EA044356B1 - VARIANT IMMUNOMODULATING PROTEINS OF ICOS LIGAND AND THEIR APPLICATION - Google Patents

Description

Cross references to related applications

The invention claims priority to US Provisional Application No. 62/323608, filed April 15, 2016, entitled Variant Immunomodulatory ICOS Ligand Proteins and Their Uses, US Provisional Application No. 62/394745, filed September 14, 2016, Variant Immunomodulatory ICOS Ligand Proteins and Their Uses , US Provisional Application No. 62/410842, filed October 20, 2016, entitled Variant ICOS Ligand Immunomodulatory Proteins and Their Uses, US Provisional Application No. 62/472568, filed March 16, 2017, entitled Variant ICOS Ligand Immunomodulatory Proteins and Their Uses and US Provisional Application No. 62/475,162, filed March 22, 2017, entitled Variant ICOS Ligand Immunomodulatory Proteins and Uses Thereof, the contents of each of which are incorporated by reference in their entirety.

Including a list of sequences with a link

The invention is submitted together with a sequence listing in electronic format. The sequence list is provided as a file called 761612000340SeqList.txt, created on April 13, 2017, which is 979474 bytes in size. Sequence listing information in electronic format is incorporated by reference in its entirety.

Field of technology to which the invention relates

The invention relates to compositions for therapeutic modulation of the immune response in the treatment of oncological and immunological diseases. In some aspects, the present invention provides specific ICOS ligand (ICOSL) variants that exhibit improved binding affinity to one or both of the cognate binding partner proteins, ICOS or CD28.

Prior Art

Modulation of the immune response by interfering with processes occurring at the immunological synapse (IS) formed through and between antigen presenting cells (APCs) or target cells and lymphocytes is of increasing interest in medicine. Mechanistically, cell surface proteins in the IS may involve coordinated and often simultaneous interactions of multiple protein targets with the single protein to which they bind. IS interaction occurs in close association with the junction of two cells, and a single protein in this structure can interact with both a protein in the same cell (cis) and a protein on the associated cell (trans), probably at the same time. Although therapeutic agents are known that can modulate IS, there is a need for improved therapeutic agents. Provided herein are immunomodulatory proteins, including soluble proteins or transmembrane immunomodulatory proteins, capable of being expressed on cells that meet such needs.

The essence of the invention

In some embodiments, what is provided herein is an ICOS ligand variant polypeptide (ICOSL) comprising an IgV domain or a specific binding fragment thereof, an IgC domain or a specific binding fragment thereof, or both of these domains, wherein the variant ICOSL polypeptide includes one or more amino acid substitutions in an unmodified ICOSL or a specific binding fragment thereof corresponding to position(s) selected from 10, 11, 13, 16, 18, 20, 25, 27, 30, 33, 37, 38, 42, 43, 47, 52, 54 , 57, 61, 62, 67, 71, 72, 74, 75, 77, 78, 80, 84, 89, 90, 92, 93, 94, 96, 97, 98, 99, 100, 102, 103, 107 , 109, 110, 111, 113, 115, 116, 117, 119, 120, 121, 122, 126, 129, 130, 132, 133, 135, 138, 139, 140, 142, 143, 144, 146, 148 , 151, 152, 153, 154, 155, 156, 158, 161, 164, 166, 168, 172, 173, 175, 190, 192, 193, 194, 198, 201, 203, 207, 208, 210, 212 , 217, 218, 220, 221, 224, 225, or 227 relative to SEQ ID NO: 32. In some embodiments, the unmodified ICOSL is a mammalian ICOSL or a specific binding fragment thereof. In some embodiments, the unmodified ICOSL is human ICOSL or a specific binding fragment thereof. In some embodiments, the unmodified ICOSL comprises (i) an amino acid sequence set forth in SEQ ID NO: 32, (ii) an amino acid sequence that has at least 95% identity with the sequence of SEQ ID NO: 32; or (iii) a portion thereof containing an IgV domain or an IgC domain or specific binding fragments thereof, or both.

In some embodiments of any of the variant ICOSL polypeptides described above, the specific binding fragment of the IgV domain or IgC domain is at least 50, 60, 70, 80, 90, 100, 110 or more amino acids in length; or the IgV domain specific binding fragment has a length that is at least 80% of the length of the IgV domain represented by amino acids 19-129 in SEQ ID NO: 5 and/or the IgC domain specific binding fragment has a length that is at least 80% of the length of the IgC domain represented by amino acids 141-227 in SEQ ID NO: 5. In some embodiments, the variant ICOSL polypeptide includes up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 amino acid modifications, optionally amino acid substitutions, insertions and/or deletions. In some embodiments, the variant ICOSL includes an amino acid sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% identity with the sequence of SEQ ID NO: 32 or a specific binding fragment thereof. In some embodiments, any of the variant ICOSL polypeptides provided herein exhibits altered binding to the ICOS, CD28, or CTLA-4 ectodomain compared to unmodified ICOSL. In some embodiments, any of the variant ICOSL polypeptides exhibit altered binding to the ICOS or CD28 ectodomain compared to unmodified ICOSL. In some embodiments, the altered binding alters binding affinity and/or alters binding selectivity.

In some embodiments of any of the ICOSL variant polypeptides described above, one or more amino acid substitutions are selected from M10V, M10I, V11E, S13G, E16V, S18R, A20V, S25G, F27S, F27C, N30D, Y33del, Q37R, K42E, T43A, Y47H , N52H, N52D, N52S, N52Q, N52Y, N52K, S54A, S54P, N57Y, N57D, R61S, R61C, Y62F, L67P, A71T, G72R, L74Q, R75Q, D77G, F78L, L80P, N84Q, D89G, E90A, K92R , F93L, H94E, H94D, L96F, L96I, V97A, L98F, S99G, Q100R, Q100K, Q100P, L102R, G103E, V107A, V107I, S109G, S109N, V110D, V110N, V110A, E111del, T1 13E, H115R, H115Q, V116A ,A117T,

N119Q, F120I, F120S, S121G, V122A, V122M, S126T, S126R, H129P, S130G, S132F, Q133H, E135K,

F138L, T139S, C140D, C140del, S142F, I143V, I143T, N144D, Y146C, V151A, Y152C, Y152H, W153R, I154F, N155H, N155Q, K156M, D158G, L161P, L161M, L1 66Q, N168Q, F172S, L173S, M175T, T190A,

T190S, S192G, V193M, N194D, C198R, N201S, L203P, L203F, N207Q, L208P, V210A, S212G, D217V,

I218T, I218N, E220G, R221G, R221I, 1224V, T225A, N227K or their conservative amino acid substitutions.

In some embodiments, one or more modifications are selected from N52Y/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52S/Y146C/Y152C, N52H/C198R, N52H/C140D/T225A, N52H/C198R/T225A, N52H/K92 R, N52H/S99G, N57Y/Q100P, N52S/G103E, N52S/S130G/Y152C, N52S/Y152C, N52S/C198R, N52Y/N57Y/Y152C, N52Y/N57Y/H129P/C198R, N52H/L161P/C1 98R, N52S/T113E, N52D/S54P, N52K/L208P, N52S/Y152H, N52D/V151A, N52H/I143T, N52S/L80P, F120S/Y152H/N201S, N52S/R75Q/L203P, N52S/D158G, N52D/Q133H,

N52S/N57Y/H94D/L96F/L98F/Q100R, N52S/N57Y/H94D/L96F/L98F/Q100R/G103E/F120S,

N52H/F78L/Q100R, N52H/N57Y/Q100R/V110D, N52H/N57Y/R75Q/Q100R/V110D, N52H/N57Y/Q100R, N52H/N57Y/L74Q/Q100R/V110D, N52H/Q100R, N52 H/S121G, A20V/ N52H/N57Y/Q100R/S109G,

N52H/N57Y/R61S/Q100R/V110D/L173S, N52H/N57Y/Q100R/V122A, N52H/N57Y/Q100R/F172S,

N52H/N57Y, N52S/F120S, N52S/V97A, N52S/G72R, N52S/A71T/A117T, N52S/E220G,

Y47H/N52S/V107A/F120S, N52H/N57Y/Q100R/V110D/S132F/M175T,

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R, Q37R/N52H/N57Y/Q100R/V110N/S142F/C198R/D217V/R221G, N52H/N57Y/Q100R/V110 D/C198R,

N52H/N57Y/Q100R/V110D/V116A/L161M/F172S/S192G/C198R, F27S/N52H/N57Y/V110N,

N52S/H94E/L96I/S109N/L166Q, S18R/N52S/F93L/I143V/R221G, A20T/N52D/Y146C/Q164L,

V11E/N30D/N52H/N57Y/H94E/L96I/L98F/N194D/V210A/I218T, N52S/H94E/L96I/V122M,

N52H/N57Y/H94E/L96I/F120I/S126T/W153R/I218N, M10V/S18R/N30D/N52S/S126R/T139S/L203F,

S25G/N30D/N52S/F120S/N227K, N30D/N52S/L67P/Q100K/D217G/R221K/T225S,

N52H/N57Y/Q100R/V110D/A117T/T190S/C198R, N52H/N57Y/Q100R/V110D/F172S/C198R,

S25G/F27C/N52H/N57Y/Q100R/V110D/E135K/L173S/C198R, N52H/N57Y/V110A/C198R/R221I,

M10I/S13G/N52H/N57Y/D77G/V110A/H129P/I143V/F172S/V193M,C198R, N52H/N57Y/R61C/Y62F/Q100R/V110N/F120S/C198R, N52H/N57Y/Q100R/V 110D/H115R/C198R,

N52H/N57Y/Q100R/V110D/N144D/F172S/C198R, N52S/H94E/L98F/Q100R, N52S/E90A,

N30D/K42E/N52S, N52S/F120S/I143V/I224V, N52H/N57Y/Q100R/V110D/C198R/S212G,

N52H/N57Y/Q100R/C198R, N52S/N194D, N52H/N57Y/Q100R/L102R/V110D/H115R/C198R, N52S/S54P, T38P/N52S/N57D, E111del, Y33del, N52H/C140del/T2 25A, N52H/F78L/ Q100R/C198R,

N52H/N57Y/R75Q/Q100P/V110D, N52H/N57Y/L74Q/V110D/S192G, N52H/S121G/C198R,

N52S/F120S/N227K, N52S/A71T/A117T/T190A/C198R, T43A/N52H/N57Y/L74Q/D89G/V110D/F172S,

N52H/N57Y/Q100R/V110D/S132F/M175T, N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G, N84Q, N119Q, N168Q, N207Q, N52Q, N52Q/N207Q, N16 8Q/N207Q, N52Q/N168Q, N84Q/N207Q, N155Q/N207Q, N119Q/N168Q, N119Q/N207Q, N119Q/N155Q, N52Q/N84Q, N52Q/N119Q, N84Q/N119Q,

N52Q/N84Q/N168Q, N52Q/N84Q/N207Q, N84Q/N155Q/N168Q, N84Q/N168Q/N207Q,

N84Q/N155H/N207Q, N155Q/N168Q/N207Q, N119Q N155Q/N168Q, N119Q/N168Q/N207Q,

N84Q/N119Q/N207Q, N119Q/N155H/N207Q, N84Q/N119Q/N155Q, N52Q/N119Q/N155Q,

N52H/N84Q/N119Q, N52H/N84Q, N52H/N84Q/N168Q, N52H/N84Q/N207Q, N52H/N84Q/N168Q/N207Q, N52Q/N84Q/N155Q, N52Q/N84Q/N168Q, N52Q/N84Q/ N155Q/N168Q, N52Q/N84Q/N119Q/N168Q,

N84Q/N119Q/N155Q/N168Q, N84Q/N155Q/N168Q/N207Q, N84Q/N119Q/N155Q/N207Q,

N52Q/N84Q/N119Q/N207Q, N52Q/N84Q/N119Q/N155Q, N52Q/N84Q/N119Q/N155Q/N207Q,

N84Q/N119Q/N155Q/N168Q/N207Q, Q100R, F138L/L203P, N52Y/F138L/L203P, N57Y/Q100R/C198R, N57Y/F138L/L203P, Q100R/F138L, L203P, N52H/N 57Y/Q100R/H115R/C198R,

N52H/N57Y/Q100R/F172S/C198R, N52H/N57Y/Q100R/H115R/F172S/C198R,

N52H/N57Y/Q100R/H115R/I143V/F172S/C198R, N52H/N57Y/Q100R/L102R/H115R/F172S/C198R,

N52H/V122A/F172S/C198R, N52H/N57Y/Q100R/H115R/F172S/N194D, N52H/N57Y/H115R/F172S/C198R, N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y /H115R, N52H/N57Y/ Q100R/H115R,

- 2 044356

N52H/N57Y/Q100R/H115R/F172S/I224V, N52H/N57Y/Q100R/H115R/F172S, N52H/N57Y/Q100R/F172S,

N52H/Q100R/H115R/I143T/F172S, N52H/N57Y/Q100P/H115R/F172S, N52Y/N57Y/Q100P/F172S,

E16V/N52H/N57Y/Q100R/V110D/H115R/C198R,

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R, N52S/E90A/H115R, N30D/K42E

N52S/H115R, N30D/K42E/N52S/H115R/C198R/R221I, N30D/K42E/N52S/H115R/C198R,

N30D/K42E/N52S/H115R/F172S/N194D, N52S/H115R/F120S/I143V/C198R, N52S/H115R/F172S/C198R, N52H/N57Y/Q100P/C198R, N52H/N57Y/Q100P H115R/F172S/C198R, N52H /N57Y/Q100P/F172S/C198R, N52H/N57Y/Q100P/H115R, N52H/N57Y/Q100P/H115R/C198R, N52H/Q100R/C198R,

N52H/Q100R/H115R/F172S, N52H/Q100R/F172S/C198R, N52H/Q100R/H115R/F172S/C198R or

N52H/N57Y/Q100R/F172S/C198R.

In some of any such embodiments, one or more amino acid modifications are at position(s) corresponding to a position selected from 52, 57, 100, 110, or 198. In some of such embodiments, one or more amino acid modifications are selected from N52H, N52D, N52S , N52K, N52Q, S54A, S54P, N57Y, Q100P, Q100R, V110A, V110D, C198R or their conservative amino acid substitution. In some embodiments, the variant ICOSL further includes one or more additional modifications, such as any described herein. In some such embodiments, the variant ICOSL polypeptide further includes one or more amino acid modifications selected from V11E, E16V, N30D, K42E, N52H, N52S, N52Y, N57Y, E90A, H94E, L96I, L98F, Q100R, Q100P, L102R, V110A, V110D, H115R, F120S, V122A, F138L, I143V, V152C, K156M, K156R, F172S, N194D, C198R, L203P, V210A, S212G, I218T, R221I, I224V or their conservative amino acid substitutions.

In some of any such embodiments, the one or more amino acid modifications are N52H/N57Y/Q100R/C198R, N52H/N57Y/Q100R/V122A, N52H/N57Y/Q100R/F172S, N52Y/N57Y/F138L/L203P, V11E/N30D/ N52H/N57Y/H94E/L96I/L98F/N194D/V210A/I218T,

N52H/N57Y/Q100R, N52H/Q100R, N52H/N57Y/Q100R/V110D/C198R/S212G,

N52H/N57Y/Q100R/L102R/V110D/H115R/C198R, E16V/N52H/N57Y/Q100R/V110D/H115R/V152C/K156M/C198R, N30D/K42E/N52S,

N52S/F120S/I143V/I224V, N52S/E90A, N52H/N57Y/V110A/C198R/R221I, N52H/N57Y/Q100P or N52S/N194D.

In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide includes an IgV domain or a specific fragment thereof and an IgC domain or a specific fragment thereof. In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide comprises the amino acid sequence set forth in any of SeQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435 -470 or a specific binding fragment or amino acid sequence thereof that has at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% sequence identity to any of SEQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470 or a specific binding fragment thereof and which includes one or more amino acid substitutions.

In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide includes an IgV domain or a specific binding fragment thereof. In some embodiments, the IgV domain or specific binding fragment thereof is the only part of the ICOSL variant ICOSL polypeptide. In some embodiments, the IgC domain or a specific binding fragment thereof is the only part of the ICOSL variant ICOSL polypeptide.

In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide comprises the amino acid sequence set forth in any of SEQ ID NOs: 197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425 -426 and 434 or a specific binding fragment thereof, an amino acid sequence that is at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% identity to any of SEQ ID NOs: 197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426 and 434 or a specific binding fragment thereof and which includes one or more amino acid substitutions.

In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS, CD28, or CTLA-4 with increased affinity compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS or CD28 with increased affinity compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide binds specifically to the ICOS ectodomain and the CD28 ectodomain, in each case with increased affinity compared to unmodified ICOSL.

In some of any such embodiments, the variant ICOSL polypeptide specifically binds to the CD28 ectodomain with increased affinity compared to unmodified ICOSL. In some such embodiments, the increased affinity for the CD28 ectodomain is increased by more than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, 50 or 60 times compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide binds specifically to the ICOS ectodomain with increased affinity compared to unmodified ICOSL.

In some of these embodiments, the increased affinity for the ICOS ectodomain is increased by more than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, 50, 60 or 70 times compared to unmodified ICOSL.

In some embodiments of any of the ICOSL variant polypeptides, one or more amino acid substitutions correspond to position(s) selected from 52, 54, or 57. In some embodiments, one or more amino acid modifications are selected from N52H, N52D, N52Q, N52S, N52Y, N52K, S54A, S54P, N57D, N57Y or their conservative amino acid substitution. In some such embodiments, one or more amino acid substitution modifications correspond to position(s) selected from 52 or 57. In some embodiments, one or more amino acid substitutions are selected from N52H, N52D, N52S, N52K, or N57Y. In some embodiments, any of the variant ICOSL polypeptides further includes yet another additional amino acid modification, such as any of those described. In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide further includes one or more amino acid modifications selected from N52H, N52D, N52S, N52Y, N52K, S54A, S54P, N57Y, R75Q, L80P, K92R, S99G, H94D , L96F, L98F, L96I, S99G, Q100R, Q100P, G103E, T113E, F120S, H129P, S130G, Q133H, F138L, C140D, C140del, I143T, Y146C, V151A, Y152C, Y152H, D158 G, L161P, C198R, N201S, L203P , L208P or T225A, or a conservative amino acid substitution thereof.

In some of any such embodiments, modifications to one or more amino acids are selected from N52Y/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52S/Y146C/Y152C, N52H/C198R,

N52H/C140del/T225A, N52H/C198R/T225A, N52H/K92R, N57Y/Q100P, N52S/C198R, N52Y/N57Y/Y152C, N52Y/N57Y/H129P/C198R, N52H/L161P/C198R, N 52S/T113E, N52S/ S54P, N52K/L208P, N52S/Y152H, N52H/I143T, N52S/R75Q/L203P, N52S/D158G, N52D/Q133H, N52H/N57Y/Q100R/V110D/C198R/S212G, N52H/N57Y/Q1 00R/C198R, N52S/ N194D, N52H/N57Y/Q100R7L102R/V110D/H115R/C198R, N52S/S54P, T38P/N52S/N57D, N52H/C140del/T225A, N52H/F78L/Q100R/C198R, N52H/N57Y/R7 5Q/Q100P/V110D, N52H/ N57Y/L74Q/V110D/S192G, N52H/S121G/C198R, N52S/F120S/N227K,

N52S/A71T/A117T/T190A/C198R, T43A/N52H/N57Y/L74Q/D89G/V110D/F172S,

N52H/N57Y/Q100R/V110D/S132F/M175T, N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G,

N52Q/N207Q, N52Q/N168Q, N52Q/N84Q, N52Q/N119Q, N52Q/N84Q/N168Q, N52Q/N84Q/N207Q, N52Q/N119Q/N155Q, N52H/N84Q/N119Q, N52H/N84Q, N52H /N84Q/N168Q, N52H/N84Q/N207Q, N52H/N84Q/N168Q/N207Q, N52Q/N84Q/N155Q, N52Q/N84Q/N168Q, N52Q/N84Q/N155Q/N168Q,

N52Q/N84Q/N119Q/N168Q, N52Q/N84Q/N119Q/N207Q, N52Q/N84Q/N119Q/N155Q,

N52Q/N84Q/N119Q/N155Q/N207Q, N52Y/F138L/L203P, N57Y/Q100R/C198R, N57Y/F138L/L203P,

N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/Q100R/F172S/C198R,

N52H/N57Y/Q100R/H115R/F172S/C198R, N52H/N57Y/Q100R/H115R/I143V/F172S/C198R,

N52H/N57Y/Q100R/L102R/H115R/F172S/C198R, N52H/V122A/F172S/C198R,

N52H/N57Y/Q100R/H115R/F172S/N194D, N52H/N57Y/H115R/F172S/C198R,

N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/H115R, N52H/N57Y/Q100R/H115R,

N52H/N57Y/Q100R/H115R/F172S/I224V, N52H/N57Y/Q100R/H115R/F172S, N52H/N57Y/Q100R/F172S, N52H/Q100R/H115R/I143T/F172S, N52H/N57Y /Q100P/H115R/F172S, N52Y/N57Y/Q100P/F172S,

E16V/N52H/N57Y/Q100R/V110D/H115R/C198R, E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R, N52S/E90A/H115R,

N30D/K42E/N52S/H115R, N30D/K42E/N52S/H115R/C198R/R221I, N30D/K42E/N52S/H115R/C198R,

N30D/K42E/N52S/H115R/F172S/N194D, N52S/H115R/F120S/I143V/C198R, N52S/H115R/F172S/C198R, N52H/N57Y/Q100P/C198R, N52H/N57Y/Q100P H115R/F172S/C198R, N52H /N57Y/Q100P/F172S/C198R, N52H/N57Y/Q100P/H115R, N52H/N57Y/Q100P/H115R/C198R, N52H/Q100R/C198R,

N52H/Q100R/H115R/F172S, N52H/Q100R/F172S/C198R, N52H/Q100R/H115R/F172S/C198R or

N52H/N57Y/Q100R/F172S/C198R.

In some of any such embodiments, one or more amino acid modifications are selected from N52Y/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52S/Y146C/Y152C, N52H/C198R,

N52H/C140del/T225A, N52H/C198R/T225A, N52H/K92R, N57Y/Q100P, N52S/C198R, N52Y/N57Y/Y152C, N52Y/N57Y/H129P/C198R, N52H/L161P/C198R, N 52S/T113E, N52S/ S54P, N52K/L208P, N52S/Y152H, N52H/I143T, N52S/R75Q/L203P, N52S/D158G or N52D/Q133H. In some such embodiments, one or more amino acid modifications are selected from N52H/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52H/K92R, N52H/C140del/T225A, N52H/C198R/T225A, N52H/K92R, N57Y/Q100P,

N52Y/N57Y/H129P/C198R, N52H/L161P/C198R, N52K/L208P or N52H/I143T.

In some of any such embodiments, the variant ICOSL polypeptide specifically binds to the CTLA-4 ectodomain with increased affinity compared to unmodified ICOSL. In some aspects, the increased affinity for the CTLA-4 ectodomain increases by more than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, 50, 60 or 70 times compared to unmodified ICOSL.

In some of any such embodiments, the variant polypeptide specifically binds to the ectodomain of ICOS, CD28, or CTLA-4 with increased selectivity compared to unmodified ICOSL. In some embodiments, the increased selectivity includes a higher ratio of binding of the variant polypeptide to one cognate binding partner selected from ICOS, CD28, and CTLA4 compared to another cognate binding partner, compared to the ratio of binding of the unmodified ICOSL polypeptide to one cognate binding partner compared to another cognate binding partner. In some cases this ratio is greater than at least, or at least about 1.5, 2.0, 3.0, 4, 5, 10, 15, 20, 30 , 40, 50 times or more.

In some embodiments of any of the variant ICOSL polypeptides, the variant ICOSL polypeptide further comprises one or more amino acids from M10V, M10I, V11E, S13G, E16V, S18R, A20V, S25G, F27S, F27C, N30D, Q37R, K42E, T43A, Y47H, N52H , N52D, N52S, N52Q, N52Y, N52K, S54A, S54P, N57D, N57Y, R61S, R61C, Y62F, L67P, A71T, G72R, L74Q, R75Q, D77G, F78L, L80P, N84Q, D89G, E90A, K92R, F93L , H94E, H94D, L96F, L96I, V97A, L98F, S99G, Q100R, Q100K, Q100P, L102R, G103E, V107A, V107I, S109G, S109N, V110D, V110N, V110A, T113E, H115R, H 115Q, V116A, A117T, F120S , F120I, S121G, V122A, V122M, S126T, S126R, H129P, S130G, S132F, Q133H, E135K, F138L, T139S, C140D, C140del, S142F, I143V, I143T, N144D, Y146C, V151A, Y152C, Y152H, W153R, I154F , N155H, N155Q, K156M, D158G, L161P, L161M, L166Q, N168Q, F172S, L173S, M175T, T190A, T190S, S192G, V193M, N194D, C198R, N201S, L203P, L203F, N207Q, L208P, V210A, S212G, D217V , I218T, I218N, E220G, R221G, R221I, 1224V, T225A, N227K or their conservative amino acid substitution. In some embodiments of any of the variant ICOSL polypeptides, the variant ICOSL polypeptide further includes one or more deletions of the amino acids corresponding to position 140 in SEQ ID NO: 32.

In some embodiments of any of the variant ICOSL polypeptides, one or more amino acid substitutions are selected from N52Y/N57Y/F138L/L203P, N52H/N57Y/Q100P,

N52S/Y146C/Y152C, N52H/C198R, N52H/C140D/T225A, N52H/C198R/T225A, N52H/K92R, N57Y/Q100P, N52S/C198R, N52Y/N57Y/Y152C, N52Y/N57Y/H1 29P/C198R, N52H/ L161P/C198R, N52S/T113E, N52D/S54P, N52K/L208P, N52S/Y152H, N52H/I143T, N52S/R75Q/L203P, N52S/D158G, N52D/Q133H, N52H/N57Y/Q100R/V1 10D/C198R/S212G, N52H/N57Y/Q100R/C198R, N52S/N194D,

N52H/N57Y/Q100R/L102R/V110D/H115R/C198R, N52S/S54P, T38P/N52S/N57D, N52H/C140del/T225A, N52H/F78L/Q100R/C198R, N52H/N57Y/R75Q/Q10 0P/V110D, N52H/ N57Y/L74Q/V110D/S192G,

N52H/S121G/C198R, N52S/F120S/N227K, N52S/A71T/A117T/T190A/C198R,

T43A/N52H/N57Y/L74Q/D89G/V110D/F172S, N52H/N57Y/Q100R/V110D/S132F/M175T or

N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G.

In some embodiments of any of the ICOSL variant polypeptides, one or more amino acid substitutions are selected from N52H/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52H/K92R, N52H/C140del/T225A, N52H/C198R/T225A, N52H/K92R, N57Y/Q100P, N52Y/N57Y/H129P/C198R, N52H/L161P/C198R, N52K/L208P, N52H/I143T, A20V/N52H/N57Y/Q100R/S109G,

N52H/N57Y/R61S/Q100R/V110D/L173S, N52H/N57Y/Q100R/V107I/V11OD/S132F/I154F/C198R/R221G,

Q37R/N52H/N57Y/Q100R/V110N/S142F/C198R/D217V/R221G, N52H/N57Y/Q100R/V110D/C198R,

F27S/N52H/N57Y/V110N, S18R/N52S/F93L/I143V/R221G, A20T/N52D/Y146C/Q164L,

N52H/N57Y/H94E/L96I/F120I/S126T/W153R/I218N, N52H/N57Y/Q100R/V110D/F172S/C198R,

S25G/F27C/N52H/N57Y/Q100R/V110D/E135K/L173S/C198R or

M10I/S13G/N52H/N57Y/D77G/V110A/H129P/I143V/F172S/V193M/C198R.

In some embodiments of any of the variant ICOSL polypeptides, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS or CD28 with increased affinity and specifically binds to the ectodomain of another of ICOS or CD28 with reduced affinity compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide binds specifically to the ICOS ectodomain with increased affinity and specifically binds to the CD28 ectodomain with reduced affinity compared to unmodified ICOSL. In some embodiments, one or more amino acid substitutions are selected from N57Y/Q100P, N52S/S130G/Y152C, N52S/Y152C, N52Y/N57Y/Y152C, N52H/L161P/C198R, N52H/L161P/C198R, N52S/L80P, A20V/N52H / N57Y/Q100R/S109G, N52H/N57Y/R61S/Q100R/V110D/L173S,

N52H/N57Y/Q100R/V107I/V110D/S132F/I154F/C198R/R221G, Q37R/N52H/N57Y/Q100R/V110N/S142F/C198R/D217V/R221G, N52H/N57Y/Q100R/V11 0D/C198R,

F27S/N52H/N57Y/V110N, S18R/N52S/F93L/I143V/R221G, A20T/N52D/Y146C/Q164L,

N52H/N57Y/H94E/L96I/F120I/S126T/W153R/I218N, N52H/N57Y/Q100R/V110D/F172S/C198R,

S25G/F27C/N52H/N57Y/Q100R/V110D/E135K/L173S/C198R or

M10I/S13G/N52H/N57Y/D77G/V110A/H129P/I143V/F172S/V193M/C198R.

In some embodiments, the variant ICOSL polypeptide binds specifically to the CD28 ectodomain with increased affinity and specifically binds to the ICOS ectodomain with reduced affinity compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes amino acid substitutions N52S/R75Q/L203P or N30D/K42E/N52S.

- 5 044356

In some embodiments of any of the variant ICOSL polypeptides, ICOS is human ICOS. In some embodiments, CD28 is human CD28.

In some embodiments of any of the variant ICOSL polypeptides, the binding is changed (increased or decreased) by more than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, at 10, at 20, at 30, at

40, 50 times compared to unmodified ICOSL.

In some embodiments of any of the variant ICOSL polypeptides, the variant ICOSL polypeptide is a soluble protein.

In some embodiments of any of the variant ICOSL polypeptides, the variant ICOSL polypeptide is linked to a multimerization domain. In some embodiments, the variant ICOSL polypeptide is a multimeric polypeptide, optionally a dimeric polypeptide, comprising a first variant ICOSL polypeptide linked to a multimerization domain and a second variant ICOSL polypeptide linked to a multimerization domain. In some embodiments, the first variant ICOSL polypeptide and the second variant ICOSL polypeptide are the same or different. In some embodiments, the multimerization domain is an Fc domain or a variant thereof with reduced effector function. In some embodiments of any of the variant ICOSL polypeptides, the variant ICOSL polypeptide is linked to a moiety that increases the biological half-life of the polypeptide.

In some embodiments, the variant ICOSL polypeptide is linked to an Fc domain or a reduced effector function variant thereof. In some embodiments, the Fc domain is mammalian, not necessarily human; or an Fc domain variant includes one or more amino acid modifications compared to an unmodified Fc domain that is mammalian, not necessarily human. In some embodiments, the Fc domain or variant thereof comprises the amino acid sequence set forth in SEQ ID NO: 226 or SEQ ID NO: 227, or an amino acid sequence that is at least 85, 86, 87, 88, 89, 90, 91 92, 93, 94, 95, 96, 97, 98, or 99% identity to SEQ ID NO: 226 or SEQ ID NO: 227.

In some embodiments, the Fc domain includes one or more amino acid modifications selected from E233P, L234A, L234V, L235A, L235E, G236del, G237A, S267K, R292C, N297G, and V302C, each according to EU numbering. In some aspects, the Fc domain includes the amino acid modification C220S in accordance with EU numbering. In some embodiments, the Fc domain comprises an amino acid sequence set forth in any of SEQ ID NOs: 474, 476, 477, 478, or an amino acid sequence that has at least 85% sequence identity with any of SEQ ID NOs: 474, 476. 477, 478 and which includes one or more amino acid modifications and/or exhibits reduced effector function.

In some embodiments, the variant ICOSL polypeptide is linked indirectly to a multimerization or Fc domain through a linker, optionally a G4S linker.

In some embodiments of any of the variant ICOSL polypeptides, the variant ICOSL polypeptide is an immunomodulatory transmembrane protein that further includes a transmembrane domain linked to an extracellular domain (ECD) or a specific binding fragment thereof of the variant ICOSL polypeptide. In some embodiments, the transmembrane domain comprises the amino acid sequence designated residues 257-277 of SEQ ID NO: 5, or a functional variant thereof that exhibits at least 85, 86, 87, 88, 89, 90, 91, 92. 93, 94, 95, 96, 97, 98, or 99% identity with residues 257-277 of SEQ ID NO: 5. In some embodiments, the variant ICOSL polypeptide further includes a cytoplasmic signaling domain associated with a transmembrane domain. In some embodiments, the cytoplasmic signaling domain comprises the amino acid sequence designated residues 278-302 of SEQ ID NO: 5, or a functional variant thereof that has at least 85, 86, 87, 88, 89, 90, 91, 92 , 93, 94, 95, 96, 97, 98 or 99% identity with residues 278-302 of SEQ ID NO: 5.

In some of any such embodiments, a variant ICOSL polypeptide comprising an amino acid sequence set forth in any of SEQ ID NOs: 494-503 or an amino acid sequence that exhibits at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% identity to any of SEQ ID NO: 494-503 and which includes one or more of the described amino acid modifications.

In some embodiments of any of the variant ICOSL polypeptides, the variant ICOSL increases IFN-gamma (interferon-gamma) expression compared to unmodified ICOSL in an in vitro assay of primary T cells. In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide reduces IFN-gamma (interferon gamma) expression relative to unmodified ICOSL in an in vitro assay of primary T cells. In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide is deglycosylated.

In some embodiments, provided herein is an immunomodulatory protein comprising a variant ICOSL polypeptide of any of the present embodiments linked to a second immunoglobulin superfamily (IgSF) domain-containing polypeptide. In some embodiments, the IgSF domain has a modified affinity and exhibits altered binding to one or more of its cognate binding partners compared to an unmodified IgSF domain or a wild-type IgSF domain. In some embodiments, the IgSF domain exhibits increased binding to one or more of its cognate binding partners compared to an unmodified IgSF domain or a wild-type IgSF domain. In some embodiments, the variant ICOSL polypeptide is a first variant ICOSL polypeptide and the IgSF domain of the second polypeptide is the IgSF domain of a second variant ICOSL polypeptide in accordance with any of the provided embodiments, wherein the first and second variant ICOSL are the same or different. In some embodiments, the variant ICOSL polypeptide is capable of specifically binding to CD28 or ICOS, and the IgSF domain of the second polypeptide is capable of binding to a cognate binding partner other than the variant ICOSL polypeptide specifically bound. In some embodiments, the IgSF domain is a member of the B7 family. In some embodiments, the IgSF domain is a tumor-localizing moiety that binds to a ligand expressed on a tumor, or is an inflammasome-localizing moiety that binds to a ligand expressed on a cell or tissue in the inflammatory environment. In some embodiments, the IgSF domain is a tumor localizing moiety that binds to a ligand expressed on the tumor. In some embodiments, the ligand is B7H6. In some embodiments, the IgSF domain is NKp30. In some embodiments, the IgSF domain is or includes an IgV domain. In some embodiments, the variant ICOSL polypeptide is or includes an IgV domain.

In some embodiments, in accordance with any one of the immunomodulatory proteins, the immunomodulatory protein includes a multimerization domain associated with one or both of the variant ICOSL polypeptides or a second polypeptide comprising an IgSF domain. In some embodiments, the multimerization domain is an Fc domain or a variant thereof with reduced effector function. In some embodiments, the immunomodulatory protein is dimeric.

In some embodiments, the immunomodulatory protein is homodimeric. In some cases, the immunomodulatory protein is heterodimeric.

In some embodiments, as provided herein, is a conjugate comprising an ICOSL variant polypeptide of any one of the embodiments or an immunomodulatory polypeptide of any of the embodiments linked to a functional moiety. In some embodiments, the functional moiety is a targeting functional moiety that specifically binds to a molecule on the surface of a cell. In some embodiments, the targeting functional moiety specifically binds to a molecule on the surface of an immune cell. In some embodiments, the immune cell is an antigen presenting cell or lymphocyte. In some embodiments, the targeting functional moiety is a tumor localizing moiety that binds to a molecule on the surface of the tumor. In some embodiments, the fragment is a protein, peptide, nucleic acid, small molecule, or nanoparticle. In some embodiments, the fragment is an antibody or antigen binding fragment. In some embodiments, the conjugate is divalent, tetravalent, hexavalent, or octavalent.

In some embodiments, what is provided herein is a nucleotide molecule encoding a variant ICOSL of any one of the provided embodiments, or an immunomodulatory polypeptide of any one of the provided embodiments. In some embodiments, the nucleotide molecule is a synthetic nucleic acid. In some embodiments, the nucleic acid is cDNA.

In some embodiments, what is provided herein is a vector containing a nucleic acid according to any of the present embodiments. In some embodiments, the vector is an expression vector. In some embodiments, the vector is a mammalian expression vector or a viral vector.

In some embodiments, what is provided herein is a cell containing a vector according to any of the provided embodiments. In some embodiments, the cell is a mammalian cell. In some embodiments, the cell is a human cell.

In some embodiments, provided herein is a method of producing a polypeptide or immunomodulatory protein, comprising introducing a nucleic acid molecule of any of the present embodiments or a vector of any of the present embodiments into a host cell under conditions for expression of the protein in the cell. In some embodiments, the method further comprises isolating or purifying the variant ICOSL polypeptide or immunomodulatory protein from a cell.

In some embodiments, provided herein is a method of constructing a cell expressing a variant ICOSL polypeptide, comprising introducing a nucleic acid molecule encoding the variant ICOSL polypeptide or an immunomodulatory polypeptide according to any of the present embodiments into a host cell under conditions in which the polypeptide expressed in the cell. In some embodiments, what is provided herein is an engineered cell expressing a variant ICOSL polypeptide, immunomodulatory polypeptide, nucleotide molecule, or vector in accordance with any of the provided embodiments.

In some embodiments, the variant ICOSL polypeptide or immunomodulatory polypeptide includes a signal peptide. In some aspects, the variant ICOSL polypeptide or immunomodulatory polypeptide does not include a transmembrane domain and/or is not expressed on the cell surface. In some cases, a variant ICOSL polypeptide or immunomodulatory polypeptide is secreted from the engineered cell. In some aspects, the engineered cell includes a variant ICOSL polypeptide that includes a transmembrane domain and/or is a transmembrane immunomodulatory protein in accordance with any of the proposed embodiments. In some embodiments, the variant ICOSL polypeptide is expressed on the surface of a cell.

In some embodiments, the engineered cell is an immune cell. In some embodiments, the immune cell is an antigen presenting cell (APC) or a lymphocyte. In some embodiments, the engineered cell is a primary cell. In some embodiments, the engineered cell is a mammalian cell. In some embodiments, the engineered cell is a human cell. In some embodiments, the lymphocyte is a T cell. In some embodiments, the APC is an artificial APC. In some embodiments, the engineered cell further includes a chimeric antigen receptor (CAR) or an engineered T-cell receptor (TCR).

Also proposed is an infectious agent comprising a nucleotide molecule encoding a variant ICOSL polypeptide or an immunomodulatory polypeptide according to any of the presented embodiments. In some embodiments, the encoded variant ICOSL polypeptide or immunomodulatory polypeptide does not include a transmembrane domain and/or is not expressed on the surface of the cell in which it is expressed. In some cases, the encoded variant ICOSL polypeptide or immunomodulatory polypeptide is secreted from the cell in which it is expressed. In some cases, the encoded variant ICOSL polypeptide includes a transmembrane domain. In some embodiments, the encoded variant ICOSL polypeptide is expressed on the surface of the cell in which it is expressed.

In some embodiments, the infectious agent is a bacterium or a virus. In some examples, the virus is an oncolytic virus. In some aspects, the oncolytic virus is adenovirus, adeno-associated virus, herpes virus, herpes simplex virus, vesicular stomatitis virus, reovirus, Newcastle disease virus, parvovirus, measles virus, vesicular stomatitis virus (VSV), coxsackie virus, or vaccinia virus. In some embodiments, the virus specifically targets dendritic cells (DC) and/or is tropic for dendritic cells. In some embodiments, the virus is a lentiviral vector that is pseudotyped with a modified Sindbis virus envelope product.

In some of any such embodiments, the infectious agent further includes a nucleotide molecule encoding an additional gene product that results in the death of a target cell, or that can enhance or activate an immune response. In some embodiments, the additional gene product is selected from an antineoplastic agent, antimetastatic agent, an antiangiogenic agent, an immunomodulatory molecule, an immune checkpoint inhibitor, an antibody, a cytokine, a growth factor, an antigen, a cytotoxic gene product, a proapoptotic gene product, an antiapoptotic gene product, a cell matrix degradation gene , a gene for tissue regeneration or reprogramming human somatic cells into pluripotent ones.

In some embodiments, as provided herein, is a pharmaceutical composition comprising an ICOSL variant polypeptide of any of the present embodiments, an immunomodulatory protein of any of the present embodiments, a conjugate of any of the present embodiments, an engineered cell of any of the present embodiments, or an infectious agent of any of the present embodiments. of the proposed embodiments. In some embodiments, the pharmaceutical composition further includes a pharmaceutically acceptable excipient. In some embodiments, the pharmaceutical composition is sterile.

In some embodiments, as provided herein is an article that includes any of the pharmaceutical compositions in an ampoule. In some embodiments, the vial is sealed. In some embodiments, a kit is also provided including any pharmaceutical compositions and instructions for use. In some aspects, a kit is provided that includes an article and instructions for use.

In some embodiments, what is provided herein relates to a method of modulating an immune response in a subject, comprising administering to the subject a pharmaceutical composition according to any of the proposed embodiments. In some embodiments, the method includes administering engineered cells, as provided in accordance with any of the proposed embodiments. In some embodiments, the engineered cells are autologous to the subject. In some embodiments, the engineered cells are allogeneic to the subject. In some embodiments, modulation of the immune response treats a disease or condition in the subject.

In some embodiments, the immune response is increased. In some embodiments, the subject is administered an immunomodulatory protein or conjugate comprising a variant ICOSL polypeptide linked to a functional tumor-localizing moiety. In some cases, the functional moiety located in the tumor is or includes a binding molecule that recognizes a tumor antigen. In some aspects, the binding molecule includes an antibody or an antigen-binding fragment thereof, or includes a wild-type IgSF domain or variant thereof.

In some of any such embodiments, a pharmaceutical composition comprising an immunomodulatory protein according to any of the provided embodiments or a conjugate according to any of the provided embodiments is administered to the subject. In some embodiments, an engineered cell containing a variant ICOSL polypeptide that is a transmembrane immunomodulatory protein and/or an engineered cell in accordance with any of the embodiments described above is administered to a subject.

In some embodiments, an infectious agent encoding a variant ICOSL polypeptide that is a transmembrane immunomodulatory protein is administered to a subject, optionally under conditions in which the infectious agent infects a tumor cell or immune cell and the transmembrane immunomodulatory protein is expressed on the surface of the infected cell. In some aspects, the transmembrane immunomodulatory protein is any provided in accordance with any of the present embodiments.

In some embodiments, the disease or condition is a tumor or malignancy. In some embodiments, the disease or condition is selected from melanoma, lung cancer, bladder cancer, hematologic malignancy, liver cancer, brain cancer, kidney cancer, breast cancer, pancreatic cancer, colorectal cancer, spleen cancer, prostate cancer, testicular cancer , ovarian cancer, uterine cancer, gastric cancer, musculoskeletal malignancy, head and neck cancer, gastrointestinal cancer, germ cell cancer or endocrine and neuroendocrine cancer.

In some embodiments, the immune response is reduced using the proposed methods of modulating the immune response. In some embodiments, a variant ICOSL polypeptide or immunomodulatory protein that is soluble is administered to a subject. In some cases, the soluble polypeptide or immunomodulatory protein is an Fc fusion protein. In some embodiments, the subject is administered a pharmaceutical composition comprising a variant ICOSL polypeptide in accordance with any of the proposed embodiments or an immunomodulatory protein in accordance with any of the proposed embodiments.

In some of any such embodiments, an engineered cell including a secreted variant ICOSL polypeptide is administered to a subject. In some embodiments, the subject is administered an engineered cell according to any of the proposed embodiments. In some embodiments, an infectious agent encoding a variant ICOSL polypeptide that is a secreted immunomodulatory protein is administered to a subject, optionally under conditions in which the infectious agent infects a tumor cell or immune cell and the secreted immunomodulatory protein is secreted from the infected cell.

In some embodiments, the disease or condition is an inflammatory or autoimmune disease or condition. In some embodiments, the disease or condition is antineutrophil cytoplasmic antibody (ANCA) associated vasculitis, vasculitis, autoimmune skin disease, transplantation, rheumatic disease, inflammatory gastrointestinal disease, inflammatory ocular disease, inflammatory neurological disease, inflammatory lung disease, inflammatory endocrine disease or autoimmune hematological disease. In some embodiments, the disease or condition is selected from inflammatory bowel disease, transplantation, Crohn's disease, ulcerative colitis, multiple sclerosis, asthma, rheumatoid arthritis, or psoriasis.

In some of any such embodiments of the ICOSL variant polypeptide, the amino acid modification is a substitution, insertion or deletion of an amino acid.

Brief description of drawings

In fig. 1 shows impedance results reflecting the cytotoxic activity of cells engineered with anti-CD19 chimeric antigen receptor (CAR) alone, or with exemplary transmembrane immunomodulatory TIPs (CD80-TIP or ICOSL-TIP) or the corresponding transmembrane protein CD80 or wild-type ICOSL after co-culture with target antigen-expressing cells. Impedance was assessed using the Acea Real-Time Cell Analyzer (RTCA), which measures changes in impedance in the culture medium of a 96-well microelectronic plate (E-plate).

In fig. Figure 2A shows that primary T cells are efficiently transduced by viruses encoding both CAR and TIP proteins. Primary human T cells were activated for 48 h with anti-CD3 plus anti-CD28 beads and then transduced with a lentivirus encoding an anti-CD19 CAR with a BFP reporter plus a second lentivirus encoding ICOSL TIP with a GFP reporter. The FACS plot shows BFP expression on the y-axis and GFP expression on the x-axis and indicates the percentage of T cells that fall into each quadrant. The results show that the cultures included only cells transduced with CAR (upper left quadrant), only cells transduced with TIP (lower right quadrant), cells transduced with both viruses (upper right quadrant), and cells that were not transduced with anything (lower left). ). In fig. 2B TIPs expressed on CAR T cells provide co-stimulation of CAR T cells. CAR-T cells with or without TIP cotransduction were labeled with Cell-Trace Far Red and incubated with the CD19+ NALM6 cell line to interact with CAR. Proliferation was assessed by the percentage of CAR-expressing cells that had diluted fluorescent dye. Cells transduced with mutant TIPs showed increased proliferation of CAR+ T cells compared to those without TIP or those transduced with wild-type ICOSL. Cells transduced with an empty control that had lost CAR expression did not proliferate in this assay.

In fig. 3A,B shows the co-stimulatory ability of wild type (WT) ICOSL or ICOSL variant when coimmobilized with anti-CD3 via cytokine release. 10 nM anti-CD3 was wet applied to the wells of 96-well flat-bottom polystyrene culture plates containing 40 nM (arrows) or 10 nM ICOSL WT or variant ICOSL. 100,000 purified CD4+ and CD8+ (pan) T cells were added and the supernatant was collected after 72 h for cytokine release analysis by ELISA. Fig. 3A shows the level of IFN-gamma protein and FIG. 3B shows the level of IL-17 protein secreted by pan T cells. The graphs are typical IFNγ and IL-17 responses to pan T cell co-stimulation.

Fig. 4A,B demonstrate, through proliferation, the co-stimulatory ability of wild type (WT) ICOSL or ICOSL variant when coimmobilized with anti-CD3. CFSE-labeled pan T cells were incubated in anti-CD3 and ICOSL-coated plates as described above for 72 h. Cells were harvested, washed, stained with fluorescently conjugated anti-CD4 or anti-CD8 antibodies, and analyzed by flow cytometry. Cytometer gates and voltages were set using unstimulated control CFSE-labeled T cells. Proliferation was determined by the dilution of CFSE relative to the control. Fig. 4A shows the percentages of total proliferating (arrows), CD4+ (solid bar), and CD8+ (shaded bar) T cells following co-stimulation with 40 nM ICOSL. Fig. 4B shows the percentage of proliferation of total pan T cells after costimulation with 10 nM ICOSL. The graphs represent the typical proliferative response to pan-T cell costimulation.

In fig. 5 depicts the function of the candidate ICOSL vIgD in the human mixed lymphocyte response (MLR). ICOSL variants and their mutations are listed on the x-axis, along with wild-type ICOSL, the negative controls PDL2-Fc and human IgG, and the positive control molecule CTLAIg belatacept. The line on the graph represents the basal amount of IFN-γ detected in the supernatants of the negative control cultures. For each candidate ICOSL or control ICOSL, three different concentrations were tested, with arrows indicating the highest protein concentration in the cultures at 40 nM. Most candidate ICOSL variants demonstrated superior antagonistic activity at all three concentrations tested compared to belatacept, which is reflected in the lower concentration of IFN-gamma in these cultures.

In fig. 6 shows the inhibition of B and T cell responses by soluble ICOSL Fc fusion proteins in a B-T cell coculture assay. In fig. 6A shows the inhibition of T cell-driven B cell proliferation by soluble ICOSL Fc fusion proteins. Purified CD4+ T cells and B cells from the same donor were labeled with CFSE and co-incubated at a 1:1 ratio in the presence or absence of the indicated mitogens with or without the indicated ICOSL Fc-fusion proteins. Cells were stimulated with Staphylococcus enterotoxin B (SEB) at 100 ng/ml, Phytolacca mitogen (PWM) at 1 mg/ml, or both. ICOSL Fc fusion proteins were included at a final concentration of 40 nM and cultures were incubated for 7 days and subjected to FACS analysis. The number of B cells that divided was determined by the number of cells in cultures that had diluted CFSE. All ICOSL Fc fusion proteins were tested, with the exception of the wild type protein, for proliferation of reduced B cells. Fig. 6B-D show that ICOSL proteins fused to Fc inhibit T cell cytokine production in B-T co-cultures. Supernatants from the above cultures were collected on day 7 and analyzed for cytokine content using the LEGENDplex Human Th Cytokine Panel (Biolegend). T cell production of IL-5 (Fig. 6B), IL-13 (Fig. 6C), and IL-21 (Fig. 6D) was attenuated by inclusion of ICOSL Fc fusion proteins.

In fig. 7A-D depict various endpoints in a mouse model of graft-versus-host disease (GVHD), where human PBMC cells were adoptively transferred into immunodeficient hosts, NSG mice. In fig. 7A shows the survival curves of treated animals. Ag

- 10 044356 a regressive course of the disease and subsequent mortality was observed in animals with control saline, while the same survival was observed in animals receiving wild-type ICOSL-Fc, as well as the variant ICOSL N52H/I143T. The N52H/N57Y/Q100P variant had improved survival rates comparable to the clinical reference belatacept. In fig. 7B shows similar trends in body weight loss with the ICOSL variant N52H/N57Y/Q100P showing similar weight maintenance as belatacept-treated animals, although all other groups experienced rapid weight loss. Fig. 7C demonstrates clinical scores on the standardized GVHD Disease Activity Index (DAI), again showing lower scores in animals treated with the variant ICOSL N52H/N57Y/Q100P, which is comparable to the clinical reference belatacept, while other groups of animals received higher DAI scores . Fig. 7D shows flow cytometric measurement of CD4 and CD8 in the blood of experimental animals performed on day 14. The percentage of CD8 cells between treatment groups was essentially the same, however, animals treated with the ICOSL variant N52H/N57Y/Q100P and belatacept had a lower percentage of CD4 cells compared with the other treatment groups.

In fig. 8 shows the co-stimulatory activity conferred by the vIgD C-L variant stack molecule, where C is the ICOSL co-stimulatory domain and L is the NKp30 localization domain. In this assay, K562 target cells expressing surface-localized B7-H6 protein were cultured in the presence of anti-CD3 with human T cells, and T cell activation was assessed by IFN-gamma levels in culture supernatants. Ahtu-CD3 alone or non-stack variant Fc molecules do not induce T cell activation. Likewise, cells cultured with only the wild-type NKp30 localizing domain or only the wild-type ICOSL co-stimulatory domain as Fc fusion proteins did not result in T cell activation. A stack domain containing the wild-type version of both the costimulatory domain and the localizing domain induced measurable IFN-gamma at the highest concentration, however, the variant localizing costimulatory stack induced more than twice as high levels of IFN-gamma at the highest concentration, and IFN levels -gamma, which were still observed as the concentration was titrated down.

In fig. Figure 9 summarizes changes in ear thickness in mice in a standard model of delayed-type hypersensitivity (DTH). Animals treated with PBS, sensitized with ovalbumin, and then ear challenged with the same protein showed the highest level of measured ear edema. In mice treated with clinical control abatacept, ear swelling was slightly reduced after the ear test. All five ICOSL treatment groups demonstrated equal or improved reduction in ear swelling compared with abatacept.

In fig. 10A-C show various exemplary configurations of a variant IgSF domain (vIgD) conjugated to an antibody (V-Mab). In fig. 10A shows various configurations in which vIgD is directly or indirectly linked to the N- and/or C-terminus of the antibody light chain. In fig. 10B shows various configurations in which vIgD is directly or indirectly linked to the N- and/or C-terminus of the antibody heavy chain. In fig. 10C shows the resulting V-Mab configurations when the light chain of FIG. 10A and the heavy chain of FIG. 10B were co-expressed in the cell.

In fig. 11A,B demonstrate the specificity of V-Mab for cognate binding partners. Binding assays were performed on Expi293 cells transiently transfected with DNA to surface express human HER2, CD28, CTLA-4, or ICOS on mammalian cells. 200,000 transfected cells were incubated with 100,000-100 pM parental antibodies (C1) or various V-Mabs (C2-9). Unbound antibody was removed, bound antibody was detected with fluorescently conjugated anti-human IgG antibody, and cells were analyzed by flow cytometry for MFI and percent positive values based on Fc controls. Fig. 11A demonstrates V-Mab transfectants binding to HER2 at levels similar to the parent antibody. Binding to cells transfected with empty control was observed with all V-Mabs, but not with WT ICOSL, due to low levels of endogenous HER2 expression on parental Expi293 cells. Fig. 11B demonstrates that binding of the parent IgSF domain (N52H/N57Y/Q100P) to its cognate partners is maintained or enhanced (C2, C3, C4, C5, C6, C8, C9) by V-Mab.

In fig. 12 demonstrates the co-stimulatory and proliferative ability of V-Mab when coimmobilized with anti-CD3. 10 nM anti-CD3 was wet coated into the wells of 96-well flat-bottom polystyrene culture plates with 30-3 nM parental antibody, V-Mab, or Fc control. CFSE-labeled T cells were added for 72 hours. IFN-γ secretion was measured by ELISA and total T cell proliferation was measured by CFSE dilution flow cytometric assay. IFN-gamma secretion and IgSF domain proliferation (N52H/N57Y/Q100P) are higher than WT ICOSL. V-Mabs exhibit increased cytokine and proliferative costimulatory capacity compared to parental IgSF.

In fig. 13A-C show various formats of the provided IgSF domain molecule variants. Fig. 13A shows soluble molecules including: (1) variant IgSF domain (vIgD) fused to an Fc chain; (2) a stack molecule containing a first variant IgSF domain (first vIgD) and a second

- 11 044356 IgSF domain, such as the second variant IgSF domain (second vIgD); (3) a tumor-targeting IgSF molecule comprising a first variant IgSF domain (vIgD) and an IgSF domain that targets a tumor antigen, such as the IgSF domain of NkP30; and (4) variant IgSF domain (vIgD) linked to antibody (V-Mab). Fig. 13B shows a transmembrane immunomodulatory protein (TIP) containing a variant IgSF domain (vIgD), such as a variant ICOSL, expressed on the cell surface. In an exemplary embodiment, the cognate binding partner of the transmembrane bound vIgD is a costimulatory receptor, such as CD28, and a TIP containing vIgD (eg, ICOSL vIgD) acts as an antagonist to the costimulatory receptor such that the TIP induces a positive signal in a cell expressing the costimulatory receptor. receptor. Fig. 13C demonstrates a secreted immunomodulatory protein (SIP) in which a variant IgSF domain (vIgD), such as a variant ICOSL, is secreted from a cell, such as a first T cell (eg, CAR T cells). In an exemplary embodiment, the cognate binding partner of the secreted vIgD is an activating receptor, such as CD28, which may be expressed in a first cell (eg, a T cell) and/or a second cell (eg, a T cell, either endogenous or engineered such as CAR T cell). When SIP binds to its cognate binding partner, signaling through the activating receptor is blocked. In all cases, vIgD can be the V domain alone (IgV), a combination of the V domain (IgV) and the C domain (IgC), including the entire extracellular domain (ECD), or any combination of Ig domains of a member of the IgSF superfamily.

In fig. 14 shows an exemplary diagram of the activity of a variant IgSF domain (vIgD) fused to an Fc (vIgD-Fc), in which vIgD is a variant of the IgSF domain from ICOSL. As shown, soluble vIgD from ICOSL interacts with its cognate binding partners to block CD80 (B7-1)/CD86 (B7-2) or ICOSL interactions with CD28 or ICOS, respectively, thereby blocking costimulation by CD28 and/or co-stimulatory receptors ICOS.

In fig. 15 shows an exemplary stacking molecule diagram for localizing a variant IgSF domain (vIgD) to a tumor cell. In this format, the stack molecule includes a first variant IgSF domain (first vIgD) and a second IgSF domain (e.g., second vIgD), wherein the second IgSF domain (e.g., second vIgD) is a tumor-targeting IgSF domain that binds to a tumor antigen . An exemplary tumor-targeting IgSF domain is the NkP30 IgSF domain, which binds to the B7-H6 tumor antigen. In this image, vIgD is a variant of the IgSF domain from ICOSL. Binding of the tumor cell surface IgSF domain is shown to localize the first vIgD to the tumor cell surface, where it can interact with one or more of its cognate binding partners (e.g., CD28 or ICOS) expressed on the surface of an adjacent immune cell (eg T cells) to stimulate the costimulatory receptor.

In fig. 16 shows various exemplary configurations of a stack molecule comprising a first variant IgSF domain (first vIgD), such as an ICOSL variant, and a second IgSF domain, such as a second variant IgSF domain (second vIgD). As shown, the first vIgD and the second IgSF domain are independently associated, directly or indirectly, with the N- or C-terminus of the Fc subunit. To produce a homodimeric Fc molecule, an Fc subunit is a subunit that is capable of forming a homodimer with the corresponding Fc subunit when the individual Fc subunits are coexpressed in a cell. To generate a heterodimeric Fc molecule, the individual Fc subunits contain mutations (eg, knob-to-groove mutations in the CH3 domain), such that heterodimer formation is favored over homodimer formation when the individual Fc subunits are co-expressed in the cell.

In fig. 17 shows an exemplary diagram of the activity of a variant IgSF domain (vIgD) conjugated to an antibody (V-Mab), in which an antibody (eg, anti-HER2 antibody) binds to an antigen on the surface of a tumor cell. In this image, vIgD is a variant of the IgSF domain from ICOSL. As shown, antibody binding to the surface of a tumor cell localizes vIgD to the surface of tumor cells, where it can interact with one or more of its cognate binding partners expressed on the surface of an adjacent immune cell (eg, T cell) to exert an agonistic effect on receptor signaling. In an illustrative embodiment, the variant IgSF domain (vIgD) is shown to be a variant of the IgSF domain from ICOSL. Binding of ICOSL vIgD to the costimulatory receptors CD28 or ICOS provides an agonistic or costimulatory signal.

Fig. 18 depicts the transcriptional signature of primary human T cells when incubated with 10 nM anti-CD3 with 40 nM control Fc protein, wild-type ICOSL-Fc, wild-type CD80-Fc, both, or variant ICOSL-Fc fusion proteins as indicated. mutations. Total RNA from the samples was obtained from the collected cells and the RNA was transferred to Nanostring, and the Cancer Immune chip was used to quantify the transcripts of 750 genes in each sample. Altered transcripts include those at higher or lower levels than diagnosed, including the flagged transcripts.

In fig. 19 shows transcript levels of exemplary transcripts during incubation as described in

- 12 044356 fig. 18 at the indicated time points in the presence of various immunomodulatory proteins.

In fig. 20A, B shows VmAb-mediated proliferation of T cells when co-cultured with HER2-expressing targets. CFSE-labeled pan T cells were activated with artificial target cells derived from K562 cells presenting anti-CD3 single-chain Fv (OKT3) and HER2 on the cell surface, in the presence of VmAb or control proteins. Proliferation was measured by flow cytometric analysis of CFSE dilution in stained CD4+ (left panel) or stained CD8+ (right panel) T cells. In fig. 20A K562 cells were titrated and plated with T cells at effector:target (E:T) ratios ranging from 40 to 1280:1. VmAb, IgSF parent domain, or WT ICOSL were added at 1000 pM. In fig. 20B K562 cells were added to T cells for an E:T ratio of 160:1. VmAb or control proteins were titrated and added at 3000 to 37 pM.

Detailed description

Provided herein are immunomodulatory proteins that are or include variants or mutants of ICOS ligand (ICOSL) or specific binding fragments thereof that have the activity of binding to at least one cognate binding partner of a target ligand (also referred to as a counterstructural protein). In some embodiments, variant ICOSL polypeptides contain one or more amino acid modifications (eg, amino acid substitutions, deletions, or additions) compared to an unmodified ICOSL polypeptide or a wild-type ICOSL polypeptide. In some embodiments, one or more amino acid modifications (eg, amino acid substitutions, deletions, or additions) are in the IgSF domain (eg, IgV) of an unmodified ICOSL polypeptide or a wild-type ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide exhibits altered, for example increased or decreased, binding activity or affinity for at least one cognate binding partner, such as at least one of ICOS, CD28, or CTLA-4. In some embodiments, the immunomodulatory proteins are soluble. In some embodiments, the immunomodulatory proteins are transmembrane immunomodulatory proteins capable of being expressed on the surface of cells. In some embodiments, also provided herein are one or more other immunomodulatory proteins that are conjugates or fusion proteins comprising a variant ICOSL polypeptide provided herein and one or more other fragments or polypeptides.

In some embodiments, variant ICOSL polypeptides and immunomodulatory proteins modulate an immune response, eg, an increased or decreased immune response. In some embodiments, the ICOSL variant polypeptides and immunomodulatory proteins provided herein can be used to treat diseases or conditions that are associated with a dysregulated immune response.

In some embodiments, the proposed ICOSL variant polypeptides modulate T cell activation through interaction with co-stimulatory signaling molecules. In general, antigen-specific T cell activation requires two distinct signals. The first signal is provided by the interaction of the T cell receptor (TCR) with antigens associated with major histocompatibility complexes (MHC) present on antigen presenting cells (APCs). The second signal is co-stimulatory for TCR engagement and is required to prevent T cell apoptosis or anergy.

In some embodiments, under normal physiological conditions, the T cell-mediated immune response is initiated by T cell receptor (TCR) antigen recognition and is regulated by a balance of co-stimulatory and inhibitory signals (ie, immune checkpoint receptors). The immune system relies on immune checkpoint receptors to prevent autoimmunity (ie, self-tolerance) and to protect tissues from excessive damage during an immune response, such as during an attack against a pathogenic infection. In some cases, however, these immunomodulatory proteins may be dysregulated in diseases and conditions, including tumors, as an immune evasion mechanism.

In some embodiments, known T cell costimulatory receptors include CD28, which is the T cell receptor for the costimulatory ligands B7-1 (CD80) and B7-2 (CD86), both of which are present on APC. These same ligands can also bind to the inhibitory T-cell receptor CTLA4 (cytotoxic T-lymphocyte protein 4) with greater affinity than CD28; binding to CTLA-4 acts to down-modulate the immune response. ICOS (inducible costimulatory receptor) is another T cell costimulatory receptor that binds to ICOS ligand (ICOSL) on the APC. In some cases, CD28 and CTLA-4 are also known to interact with ICOSL at a binding site that overlaps with ICOSL binding to the T cell costimulatory receptor ICOS (Yao et al. (2011) Immunity, 34:729-740). Although CD28 and ICOS are members of the CD28 family of activating receptors that share some common intracellular signaling motifs, the co-stimulatory effects between CD28 and ICOS differ. For example, CD28 is expressed on both unactivated and activated T cells, and its signaling pathway is important for IL-2 production and subsequent T cell effector function. ICOS is not typically expressed on the surface of T cells until T cell activation, and signaling through ICOS on activated T cells supports the specialized differentiation of a subset of T cells. Thus, in some cases, co-stimulation of CD28 and ICOS has overlapping and additive effects.

In some embodiments, the T cell costimulatory receptors CD28 and ICOS have distinct but complementary roles in modulating the immune response. Enhancing or inhibiting the activity of these receptors is of clinical importance for the treatment of inflammatory and autoimmune diseases, malignancies, and viral infections. However, in some cases, therapies to interfere with and modify the co-stimulatory effects of both receptors are limited by spatial orientation requirements as well as size restrictions imposed by the boundaries of the immunological synapse. In some aspects, existing therapeutic drugs, including antibodies, may not be able to interact simultaneously with multiple target proteins involved in modulating these interactions. In addition, in some cases, existing therapeutic drugs may have the ability to antagonize, but not agonize, the immune response. In addition, pharmacokinetic differences between drugs that independently target one or the other of these two receptors may make it difficult to properly maintain the desired blood concentrations of such drug combinations throughout the course of treatment.

In some embodiments, the proposed variant ICOSL polypeptides or immunomodulatory proteins modulate (eg, increase or decrease) the immunological activity caused by CD28 or ICOS co-stimulatory receptors. Thus, in some embodiments, the provided polypeptides overcome these limitations by providing a variant ICOSL (inducible costimulatory ligand) with altered (e.g., increased or decreased) binding affinities for both CD28 and ICOS, and in some cases CTLA-4, thereby having an agonistic or antagonistic effect on the complementary effects of receptor costimulation. Methods for creating and using these ICOSL variants are also provided.

All publications, including patents, scientific articles and patent application databases, mentioned in this specification are incorporated herein by reference in their entirety for all purposes to the same extent as if each individual publication, including a patent, patent application, scientific article or database has been specifically and individually identified for inclusion as a reference. To the extent that a definition set forth herein conflicts or is otherwise inconsistent with a definition set forth in patents, applications, published applications, and other publications that are incorporated herein by reference, the definition set forth herein shall prevail over the definition that is incorporated herein. document reference.

The section headings used herein are for organizational purposes only and should not be construed as limiting the subject matter described.

I. Definitions.

Unless otherwise indicated, all terms of the art, designations and other technical and scientific terms or terminology used herein have the same meaning as commonly understood by one skilled in the art to which the claimed subject matter pertains. In some cases, terms with generally accepted meanings are defined herein for clarity and/or as a reference, and the inclusion of such definitions herein should not necessarily be construed as materially different from what is commonly understood in the art.

The terms used in the description are defined as follows, unless otherwise limited in specific cases. As used in the English specification and the accompanying claims, the singular forms a, an and the include the plural identifiers unless the context clearly indicates otherwise. Unless otherwise specified, all technical and scientific terms, acronyms and abbreviations used herein have the same meaning as commonly understood by one skilled in the art to which the invention pertains. Unless otherwise noted, abbreviations and symbols for chemical and biochemical names follow IUPAC-IUB nomenclature. Unless otherwise noted, all numeric ranges include the values that define the range and all integer values in between.

The term modified affinity, as used in the context of an immunoglobulin superfamily domain, means a mammalian immunoglobulin superfamily (IgSF) domain having an altered amino acid sequence (relative to the corresponding parent or unmodified wild-type IgSF domain) such that it has increased or decreased binding affinity or avidity, according to to at least one of its cognate binding partners (otherwise counterstructures) compared to the parental wild-type IgSF control domain or the unmodified (ie, no affinity modified) IgSF control domain. In this context, an affinity modified ICOSL domain is included

- 14 044356

IgSF. In some embodiments, the affinity modified IgSF domain may comprise 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 or more amino acids, such as amino acid substitutions, in the wild-type IgSF domain or in the unmodified IgSF domain. Increases or decreases in binding affinity or avidity can be determined using well-known binding assays such as flow cytometry. Larsen et al., American Journal of Transplantation, Vol 5: 443-453 (2005). See also Linsley et al., Immunity, 1: 7930801 (1994). The increase in binding affinity or avidity of the protein or to its cognate partner(s) is a value of at least 10% greater than that of the wild-type IgSF domain control value, and in some embodiments, at least 20, 30, 40, 50, 100, 200, 300, 500, 1000, 5000 or 10000% greater than the wild-type IgSF domain control value. The reduction in binding affinity or avidity of a protein to at least one of its cognate binding partners is a value of no more than 90% of the control, but not less than 10% of the control value of the wild-type IgSF domain and in some embodiments no more than 80, 70, 60 , 50, 40, 30 or 20%, but not less than 10% of the wild-type IgSF domain control value. An affinity modified protein is altered in the primary amino acid sequence by substitution, addition or deletion of amino acid residues. The term affinity modified IgSF domain should not be construed as imposing any condition on any particular parent composition or method by which the IgSF affinity modified domain was created. Thus, the affinity modified IgSF domains of the present invention are not limited to wild-type IgF domains, which are then transformed into an affinity modified IgSF domain by any particular affinity modification process. An affinity-modified IgSF domain polypeptide may, for example, be derived from wild-type mammalian IgSF domain sequence information, then modeled in silico to bind to its cognate binding partner, and finally recombinantly or chemically synthesized to yield an IgSF domain composition with modified affinity of the subject matter of the invention. In one alternative example, an affinity-modified IgSF domain can be created by site-directed mutagenesis of a wild-type IgSF domain. Thus, an affinity modified IgSF domain designates a product and not necessarily a product produced by any given method. Various methods can be used, including recombinant methods, chemical synthesis, or combinations thereof.

As used herein, the term allogeneic means a cell or tissue that is removed from one organism and then introduced or adopted into a genetically dissimilar organism of the same species. In some embodiments of the invention, the species is a mouse or a human.

The term autologous as used herein means a cell or tissue that is removed from the same organism into which it is later introduced or adopted. An autologous cell or tissue can be altered, for example, by recombinant DNA techniques, so that it is no longer genetically identical to the native cell or tissue that is removed from the body. For example, a native autologous T cell can be genetically modified by recombinant DNA techniques to become an autologous engineered cell expressing a transmembrane immunomodulatory protein and/or a chimeric antigen receptor (CAR), which in some cases involves engineering a T cell or TIL (tumor infiltrating lymphocyte) ). The engineered cells are then injected into the patient from which the native T cell was isolated. In some embodiments, the organism is a human or a mouse.

The terms binding affinity and binding avidity as used herein mean the specific binding affinity and specific binding avidity, respectively, of a protein with its counterstructure under certain binding conditions. In biochemical kinetics, avidity refers to the accumulated strength of multiple affinities of individual noncovalent binding interactions, such as between ICOSL and its counterstructures ICOS and/or CD28. Thus, avidity is different from affinity, which describes the strength of a single interaction. The increase or decrease in binding affinity of an ICOSL variant containing an affinity-modified IgSF ICOSL domain to its counterstructure is determined relative to the binding affinity of an unmodified ICOSL, such as an unmodified ICOSL containing a native or wild-type IgSF domain, such as an IgV domain. Methods for determining binding affinity or avidity are known in the art. See, for example, Larsen et al., American Journal of Transplantation, Vol 5: 443-453 (2005). In some embodiments, a variant ICOSL of the invention (i.e., an ICOSL protein containing an affinity-modified IgSF domain) specifically binds to CD28 and/or ICOS as measured by flow cytometry with a binding affinity that yields a mean fluorescence intensity (MFI) value of at least 10, 20, 30, 40, 50, 60, 70, 80, 90 or 100% more than the wild-type ICOSL control in the binding assay as described in Example 6.

The term biological half-life refers to the amount of time it takes for a substance, such as an immunomodulatory polypeptide containing a variant ICOSL of the present invention, to lose half of its pharmacological or physiological activity

- 15 044356 sti or concentration. The biological half-life may be affected by the release, excretion, degradation (eg, enzymatic) of the substance, or absorption and concentration in specific organs or tissues of the body. In some embodiments, the biological half-life can be estimated by determining the time required for the plasma concentration in the blood to reach half its steady-state level (plasma half-life). Conjugates that can be used to derivatize and increase the biological half-life of the polypeptides of the invention are known in the art and include, but are not limited to, polyethylene glycol (PEG), hydroxyethyl starch (HES), XTEN (extended recombinant peptides, WO 2013130683), serum albumin human (HSA), bovine serum albumin (BSA), lipids (acylation) and poly-Pro-Ala-Ser (PAS), polyglutamic acid (glutamylation).

As used herein, the term chimeric antigen receptor or CAR refers to an artificial (ie, man-made) transmembrane protein expressed in a mammalian cell that includes at least an ectodomain, a transmembrane domain, and an endodomain. Optionally, the CAR protein includes a spacer that covalently links the ectodomain to the transmembrane domain. The spacer is often a polypeptide that links the ectodomain to the transmembrane domain via peptide bonds. CAR is typically expressed in mammalian lymphocyte. In some embodiments, the CAR is expressed in a mammalian cell, such as a T cell or tumor infiltrating lymphocyte (TIL). A CAR expressed in a T cell is referred to herein as a CAR T cell or CAR-T. In some embodiments, the CAR-T is a helper T cell, a cytotoxic T cell, a natural killer T cell, a memory T cell, a regulatory T cell, or a gamma delta T cell. When used in clinical practice, such as adoptive cell transfer, a CAR-T with antigen-binding specificity for a patient's tumor is typically engineered for expression on a native T cell derived from the patient. The engineered T cell expressing the CAR is then injected back into the patient. Thus, CART is often an autologous CAR-T, although allogeneic CAR-Ts are included within the scope of the invention. The CAR ectodomain includes an antigen-binding region, such as an antibody or an antigen-binding fragment thereof (eg, scFv), which specifically binds under physiological conditions to an antigen, such as a tumor-specific antigen. Upon specific binding, a biochemical chain of events (i.e., signal transduction) leads to modulation of CAR-T immunological activity. For example, when the antigen-binding region of a CAR-T specifically binds to its antigen, it can lead to changes in the immunological activity of T-cell activity, which is reflected in changes in cytotoxicity, proliferation or cytokine production. Signal transduction upon CAR activation is achieved in some embodiments by the CD3 zeta chain (CD3-z), which is involved in signal transduction in naïve mammalian T cells. CAR-Ts may additionally contain multiple signaling domains, such as CD28, 41BB, or OX40, to further modulate the T cell's immunomodulatory response. CD3-z includes a conserved motif known as the immunoreceptor-based tyrosine-activating motif (ITAM), which is involved in T cell receptor signal transduction.

The term collectively or cumulatively, when used in relation to cytokine production induced by the presence of two or more ICOSL variants of the invention in an in vitro assay, means the overall level of cytokine expression independent of the cytokine production induced by an individual ICOSL variant. In some embodiments, the cytokine analyzed is IFN-gamma in an in vitro primary T cell assay, such as that described in Example 7.

The terms cognate binding partner or counterstructure with respect to a polypeptide, for example with respect to the IgSF domain of an ICOSL variant, refers to at least one molecule (usually a native mammalian protein) to which the reference polypeptide binds under certain binding conditions. In some aspects, a variant ICOSL containing an affinity-modified IgSF domain specifically binds to a counterstructure of the corresponding native or wild-type ICOSL, but with increased or attenuated affinity. A species of ligand that is recognized and specifically binds to its cognate receptor under specific binding conditions is an example of a counterstructure or cognate binding partner of that receptor. A cell surface cognate binding partner is a cognate binding partner expressed on the surface of mammalian cells. The cell surface molecular junction is a cognate binding partner of immunological synapse (IS) ligands expressed on cells and by cells, such as mammalian cells, forming the immunological synapse.

As used herein, the term conjugate, conjugation, or grammatical variations thereof refers to the joining or joining together of two or more compounds, resulting in the formation of another compound, using any compounds or linking methods known in the art. It can also refer to a compound that is formed by combining or combining two or more compounds. For example, a variant ICOSL polypeptide linked directly or indirectly to one or more chemical groups or polypeptide,

- 16 044356 is an exemplary conjugate. Such conjugates include fusion proteins produced by chemical conjugates and those produced by any other means.

As used herein, the term competitive binding means that the protein is capable of specifically binding to at least two cognate binding partners, but that the specific binding of one cognate binding partner inhibits, for example, prevents or eliminates the simultaneous binding of a second cognate binding partner. Thus, in some cases, a protein can simultaneously bind two cognate binding partners. Typically, competitive binders contain the same or overlapping binding site for specific binding, but this is not a requirement. In some embodiments, competitive binding causes a measurable inhibition (partial or complete) of the specific binding of a protein to one of its cognate binding partners due to the specific binding of a second cognate binding partner. There are many known methods for quantifying competitive binding, such as ELISA (enzyme-linked immunosorbent assay) assay.

As used herein, the term conservative amino acid substitution means an amino acid substitution in which an amino acid residue is replaced by another amino acid residue having a side chain R group with similar chemical properties (eg, charge or hydrophobicity). Examples of groups of amino acids with side chains with similar chemical properties include: 1) aliphatic side chains: glycine, alanine, valine, leucine and isoleucine; 2) aliphatic-hydroxyl side chains: serine and threonine; 3) amide-containing side chains: asparagine and glutamine; 4) aromatic side chains: phenylalanine, tyrosine and tryptophan; 5) main side chains: lysine, arginine and histidine; 6) acidic side chains: aspartic acid and glutamic acid; and 7) sulfur-containing side chains: cysteine and methionine. Conservative amino acid substitution groups are: valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, glutamate-aspartate and asparagine-glutamine.

A term corresponding to protein positions, for example, an indication that nucleotide or amino acid positions correspond to nucleotide or amino acid positions in a disclosed sequence as specified in a sequence listing, refers to nucleotide or amino acid positions identified when aligned with a disclosed sequence based on structural sequence alignment or using a standard alignment algorithm such as the GAP algorithm. For example, corresponding residues can be determined by aligning the reference sequence with the sequence set forth in SEQ ID NO: 32 (ECD domain) or set forth in SEQ ID NO: 196 (IgV domain) by structural alignment methods as described herein. By aligning the sequences, one skilled in the art can identify the corresponding residues, for example, using conserved and identical amino acid residues as guides.

The terms reduce or attenuate or suppress as used herein mean a reduction by a statistically significant amount. The reduction can be at least 10, 20, 30, 40, 50, 60, 70, 80, 90 or 100%.

The terms derivative or derivatized refer to the modification of a protein by covalently linking it, directly or indirectly, to a composition so as to alter characteristics such as biological half-life, bioavailability, immunogenicity, solubility, toxicity, potency or storage efficacy, or enhance its therapeutic properties. effect. Derivatives of the immunomodulatory polypeptides of the invention are within the scope of the invention and can be prepared, for example, by glycosylation, pegylation, lipidation or Fc fusion.

As used herein, the term domain (typically a sequence of three or more, typically 5 or 7 or more amino acids, e.g., 10 to 200 amino acid residues) refers to a portion of a molecule, such as a protein or encoding nucleic acid, that is structurally and/or functionally distinct from other parts of the molecule and can be identified. For example, domains include those parts of a polypeptide chain that can form an independently complex structure within a protein consisting of one or more structural motifs and/or recognized by functional activity, such as binding activity. A protein may have one or more distinct domains. For example, a domain can be identified, defined, or separated by primary sequence or structure homology to related family members, such as homology to motifs. In another example, a domain may be distinguished by its function, such as the ability to interact with a biomolecule, such as a cognate binding partner. A domain can independently exhibit a biological function or activity, such that a domain independently or fused to another molecule can perform an activity, such as, for example, binding. A domain may be a linear amino acid sequence or a non-linear amino acid sequence. Many polypeptides contain multiple domains. Such domains are known and can be identified by those skilled in the art. Definitions are provided herein by way of example, but it will be understood by those skilled in the art to identify certain domains by name. If necessary for

- 17 044356 domain identification you can use the appropriate software.

The term ectodomain as used herein refers to the region of a membrane protein, such as a transmembrane protein, that lies outside the vesicular membrane. Ectodomains often include binding domains that specifically bind to ligands or cell surface receptors, for example, through a binding domain that specifically binds to a ligand or cell surface receptor. The ectodomain of a cellular transmembrane protein is alternately called the extracellular domain.

The terms effective amount or therapeutically effective amount refers to the amount and/or concentration of a therapeutic composition of the present invention, including a protein composition or cellular composition, which when administered ex vivo (by contact with a cell from a patient) or in vivo (by administration to a patient) either alone (i.e., as monotherapy) or in combination with additional therapeutic agents leads to a statistically significant reduction in disease progression, for example, by improving or eliminating symptoms and/or the cause of the disease. An effective amount may be an amount that relieves, reduces or alleviates at least one symptom or biological response or effect associated with a disease or disorder, prevents the progression of a disease or disorder, or improves the physical functioning of a patient. In the case of cell therapy, the effective amount is the effective dose or number of cells administered to a patient using adoptive cell therapy. In some embodiments, the patient is a mammal, such as a non-human primate, or a human patient.

As used herein, the term endodomain refers to the region found in some membrane proteins, such as transmembrane proteins, that extend into the interior space defined by the surface membrane of the cell. In mammalian cells, the endodomain is the cytoplasmic region of a membrane protein. In cells, the endodomain interacts with intracellular components and may play a role in signal transduction and therefore, in some cases, may be an intracellular signaling domain. The endodomain of a cellular transmembrane protein is alternately called the cytoplasmic domain, which in some cases may be the cytoplasmic signaling domain.

The term enhanced or increased, as used herein in the context of increasing the immunological activity of mammalian lymphocytes, means an increase in one or more activities of the lymphocytes. The increased activity may be one or more of increased cell survival, cell proliferation, cytokine production, or T cell cytotoxicity, such as a statistically significant amount. In some embodiments, reference to increased immunological activity means an increase in the production of interferon gamma (IFN-gamma), for example, a statistically significant amount. In some embodiments, immunological activity can be assessed in a mixed lymphocyte reaction (MLR) assay. Methods for performing MLR assays are known in the art. Wang et al., Cancer Immunol Res. 2014 Sep: 2(9):846-56. Methods for assessing lymphocyte activity are known in the art, including any assay as described herein. In some embodiments, the increase may be an increase of at least 10, 20, 30, 40, 50, 75, 100, 200, 300, 400, or 500% greater than the non-zero reference value.

As used herein, the term engineered cell refers to a mammalian cell that has been genetically modified through human intervention, such as recombinant DNA techniques or viral transduction. In some embodiments, the cell is an immune cell, such as a lymphocyte (eg, T cell, B cell, NK cell) or an antigen presenting cell (eg, dendritic cell). The cell may be a primary cell from a patient or may be a cell line. In some embodiments, an engineered cell of the invention includes a variant ICOSL of the invention designed to modulate the immunological activity of a T cell expressing CD28, ICOS, or CTLA-4 to which the variant ICOSL specifically binds. In some embodiments, the variant ICOSL is a transmembrane immunomodulatory protein (hereinafter referred to as a TIP) comprising an extracellular domain or a portion thereof comprising an IgV domain linked to a transmembrane domain (eg, an ICOSL transmembrane domain) and, optionally, an intracellular signaling region. In some cases, TIP is formatted as a fusion receptor containing a heterologous cytoplasmic signaling domain or endodomain. In some embodiments, the engineered cell is capable of expressing and secreting an immunomodulatory protein as described herein. Among the proposed, engineered cells are also cells further containing an engineered T-cell receptor (TCR) or chimeric antigen receptor (CAR).

As used herein, the term engineered T cell refers to a T cell, such as a helper T cell, a cytotoxic T cell (optionally a cytotoxic T lymphocyte or CTL), a natural killer T cell, a regulatory T cell, a T cell memory or gammadelta T cell that has been genetically modified by human intervention, such as recombinant DNA techniques. The engineered T cell includes a transmembrane immunomotor

- 18 044356 duplication protein (TIP) variant ICOSL of the present invention, which is expressed on a T Cell and is designed to modulate the immunological activity of the engineered T Cell itself or a mammalian cell to which the variant ICOSL expressed on the T cell specifically binds. The term engineered T cell receptor or engineered TCR refers to a T cell receptor (TCR) engineered to specifically bind with a desired affinity for a major histocompatibility complex (MHC)/peptide target antigen that is selected, cloned, and/or subsequently introduced into a population T cells often used for adoptive immunotherapy. Unlike engineered TCRs, CARs are engineered to bind target antigens in an MHC-independent manner.

As used herein, the term "expressed" is used herein to refer to a protein expressed on the surface of a cell, such as a mammalian cell. Thus, the protein is expressed as a membrane protein. In some embodiments, the expressed protein is a transmembrane protein. In some embodiments, the protein is conjugated to a small molecule moiety, such as a drug or detectable label. Proteins expressed on the cell surface may include cell surface proteins, such as cell surface receptors, which are expressed on mammalian cells.

The term half-life extending functional moiety refers to a fusion polypeptide or chemical conjugate moiety that increases the half-life of a protein circulating in mammalian serum compared to the half-life of a protein that is not conjugated to a functional moiety. In some embodiments, the half-life is increased by more than 1.2, 1.5, 2.0, 3.0, 4, 5.0, or 6.0 times. In some embodiments, the half-life is extended by more than 6, more than 12, more than 24, more than 48, more than 72, more than 96 hours, or more than 1 week after in vivo administration compared to a protein without a functional half-life extending moiety. Half-life refers to the amount of time it takes for a protein to lose half its concentration, quantity, or activity. The half-life can be determined, for example, using an ELISA or activity assay. Exemplary functional fragments that extend half-life include Fc domain, multimerization domain, polyethylene glycol (PEG), hydroxyethyl starch (HES), XTEN (extended recombinant peptides, see WO 2013130683), human serum albumin (HSA), bovine serum albumin (BSA) , lipids (acylation), poly-Pro-Ala-Ser (PAS) and polyglutamic acid (glutamylation).

As used herein, the term immunological synapse or immune synapse means the interface between a mammalian cell that expresses MHC I (major histocompatibility complex) or MHC II, such as an antigen presenting cell or a tumor cell, and a mammalian lymphocyte, such as an effector T cell or natural cell killer (NK).

The Fc (fragmented crystallizable) region or domain of an immunoglobulin molecule (also called Fc polypeptide) largely corresponds to the constant region of the immunoglobulin heavy chain and is responsible for various functions, including the effector function(s) of the antibody. The Fc region includes part or all of the hinge domain of the immunoglobulin molecule plus the CH2 and CH3 domains. The Fc domain can form a dimer of two polypeptide chains connected by one or more disulfide bonds. In some embodiments, the Fc is a variant of the Fc that exhibits reduced (eg, reduced by more than 30, 40, 50, 60, 70, 80, 90% or more) activity to facilitate effector function. In some embodiments, reference to amino acid substitutions in the Fc region is the EU numbering system, unless described by reference to a specific SEQ ID NO. The EU numbering is known and corresponds to the latest updated IMGT scientific map (IMGT®, international information system ImMunoGeneTics® http://www.imgt.org/IMGTScientificChart/Numbering/Hu_IGHGnber.html (created: May 17, 2001, last updated: January 10 2013) and the EU index presented in Kabat, E. A. et al. Sequences of Proteins of Immunological Interest. 5th ed. US Department of Health and Human Services, NIH publication No. 91-3242 (1991).

An immunoglobulin Fc fusion protein (Fc fusion protein), such as an immunomodulatory Fc fusion protein, is a molecule containing one or more polypeptides (or one or more small molecules) operably linked to the Fc region of an immunoglobulin. The Fc fusion protein may include, for example, the Fc region of an antibody (which facilitates effector functions and pharmacokinetics) and a variant ICOSL. The Fc region of an immunoglobulin may be linked indirectly or directly to one or more ICOSL variants or small molecules (fusion partners). Various linkers are known in the art and can be used to link the Fc to a fusion partner to produce an Fc fusion protein. Fc fusion proteins of identical species can be dimerized to form homodimers of Fc fusion proteins or by non-identical species to form heterodimers of Fc fusion proteins. In some embodiments, the Fc is a mammalian Fc, such as a murine or human Fc.

The term host cell refers to a cell that can be used to express proteins

- 19 044356 ka, encoded by a recombinant expression vector. The host cell may be a prokaryotic organism, such as E. coli, or it may be a eukaryotic organism, such as a single-celled eukaryotic organism (such as yeast or other fungus), a plant cell (such as a tobacco or tomato plant cell), or an animal cell (eg, a human cell, a monkey cell, a hamster cell, a rat cell, a mouse cell, or an insect cell) or a hybridoma. Examples of host cells include Chinese hamster ovary (CHO) cells or derivatives thereof, such as Veggie CHO and related cell lines that grow in serum-free medium or the CHO line DX-B11, which is DHFR deficient. In some embodiments, the host cell is a mammalian cell (eg, a human cell, a monkey cell, a hamster cell, a rat cell, a mouse cell, or an insect cell).

The term immunoglobulin (abbreviated Ig) as used herein refers to mammalian immunoglobulin protein, including any of the five human antibody classes: IgA (which includes the subclasses IGA1 and IgA2), IgD, IgE, IgG (which includes the subclasses IgG1, IgG2, IgG3 and IgG4) and IgM. The term also includes immunoglobulins that are less complete, wholly or partially synthetic (eg, produced recombinantly or by chemical synthesis) or naturally produced, such as antigen-binding fragment (Fab), variable fragment (Fv), containing _VH and _VL , single chain variable fragment (scFv) containing V _H and V _L joined together in one chain, as well as fragments of the V regions of other antibodies, such as Fab', F(ab) ₂ , F(ab') ₂ , dsFv, polypeptide fragments Fc and Fd. Bispecific antibodies, homospecific and heterobispecific, are included in the meaning of the term.

As used herein, the term immunoglobulin superfamily or IgSF refers to a group of cell surface and soluble proteins that are involved in cell recognition, binding, or adhesion. Molecules are classified as members of this superfamily based on shared structural features with immunoglobulins (i.e. antibodies); they all possess a region known as the immunoglobulin domain or immunoglobulin chain fold. Members of IgSF include cell surface antigen receptors, coreceptors and co-stimulatory molecules of the immune system, molecules involved in antigen presentation to lymphocytes, cell adhesion molecules, some cytokine receptors, and intracellular muscle proteins. They are usually associated with roles in the immune system. Proteins at the immunological synapse are often representatives of IgSF. IgSFs can also be divided into subfamilies based on common properties such as functions. Such subfamilies typically include from 4 to 30 IgSF members.

The terms IgSF domain or immunoglobulin domain or Ig domain as used herein refers to the structural domain of IgSF proteins. Ig domains are named after immunoglobulin molecules. They contain about 70-110 amino acids and are classified according to their size and function. Ig domains have a characteristic Ig chain fold that has a sandwich-like structure formed by two sheets of antiparallel beta strands. Interactions between hydrophobic amino acids on the inside of the sandwich and highly conserved disulfide bonds formed between cysteine residues in the B and F chains stabilize the Ig fold. One end of the Ig domain has a section called the complementarity determination region, which is important for the specificity of antibodies to their ligands. Ig-like domains can be classified (in classes) as: IgV, IgC1, IgC2 or IgI. Most Ig domains are either variable (IgV) or constant (IgC). IgV domains with 9 beta chains are generally larger than IgC domains with 7 beta chains. The Ig domains of some IgSF members resemble the IgV domains in amino acid sequence but are similar in size to the IgC domains. These are called IgC2 domains, whereas standard IgC domains are called IgC1 domains. T cell receptor (TCR) chains contain two Ig domains in the extracellular portion; one IgV domain at the N-terminus and one IgC1 domain, near the cell membrane. ICOSL contains two Ig domains: IgV and IgC.

As used herein, the term IgSF species means a collection of IgSF member proteins with identical or substantially identical primary amino acid sequence. Each member of the mammalian immunoglobulin superfamily (IgSF) defines the unique identity of all IgSF species that belong to that IgSF member. Thus, each member of the IgSF family is unique among other members of the IgSF family, and, accordingly, each species of a particular member of the IgSF family is unique among the species of another member of the IgSF family. However, variation between molecules of the same IgSF species can occur due to differences in post-translational modifications such as glycosylation, phosphorylation, ubiquitination, nitrosylation, methylation, acetylation and lipidation. In addition, minor sequence differences within a single IgSF species due to gene polymorphism represent another form of variation within a single IgSF species, and truncated forms of wild-type IgSF species due to, for example, proteolytic cleavage. A cell surface IgSF species is a species of IgSF expressed on the surface of a cell, typically a mammalian cell.

The term immunological activity as used herein in the context of lymphocytic

- 20 044356 mammalian activity, such as T cells, refers to one or more of cell survival, cell proliferation, cytokine production (eg, interferon-gamma) or T-cell cytotoxicity activity. In some cases, immunological activity may indicate cellular expression of cytokines such as chemokines or interleukins. Assays to determine the enhancement or suppression of immunological activity include MLR reactions (mixed lymphocyte reaction), measuring the levels of interferon-gamma cytokines in culture supernatants (Wang et al., Cancer Immunol Res. 2014 Sep: 2(9):846-56) , SEB (staphylococcal enterotoxin B) T cell stimulation assay (Wang et al., Cancer Immunol Res. 2014 Sep: 2(9):846-56) and anti-CD3 T cell stimulation assay (Li and Kurlander, J Transl Med 2010: 8: 104). Because T cell activation is associated with secretion of the cytokine IFN-gamma, determination of IFN-gamma levels in culture supernatants from these in vitro human T cell assays can be analyzed using commercial ELISA kits (Wu et al., Immunol Lett 2008 Apr 15 , 117 (1): 57-62). The induction of an immune response leads to an increase in immunological activity compared to resting lymphocytes. An immunomodulatory protein, such as a variant ICOSL polypeptide containing an affinity modified IgSF domain as provided herein, may in some embodiments increase or in alternative embodiments decrease IFN-gamma (interferon-gamma) expression in a T cell primary assay relative to an IgSF representative wild type or IgSF domain control. Experts recognize that the format of the primary T cell assay used to determine increases in IFN-gamma expression will be different from the format used to analyze decreases in IFN-gamma expression. When analyzing the ability of an immunomodulatory protein or affinity modified IgSF domain of the invention to reduce IFN-gamma expression in a primary T cell assay, a mixed lymphocyte response (MLR) assay as described in Example 6 can be used.

Conveniently, a soluble form of the affinity modified IgSF domain of the invention could be used to determine its ability to antagonize and thereby reduce IFN-gamma expression in the MLR, as described in Example 6. Otherwise, when analyzing the ability of an immunomodulatory protein or The affinity-modified IgSF domain of the invention can be used to increase the expression of IFN-gamma in the analysis of primary T cells. In a coimmobilization assay, a T cell receptor signal, represented in some embodiments by an anti-CD3 antibody, is used in combination with a coimmobilized affinity modified IgSF domain, such as a variant ICOSL, to determine the ability to increase IFN-gamma compared to a control wild-type IgSF domain. Methods for studying the immunological activity of engineered cells, including for assessing the activity of the transmembrane immunomodulatory protein variant ICOSL, are known in the field and include, without limitation, the ability to lead to the proliferation of T cells after stimulation with antigens, to maintain the proliferation of T cells in the absence of repeated stimulation and antitumor activities in relevant animal models. The assays also include assays to assess cytotoxicity, including the standard ⁵¹ Cr release assay (see, for example, Milone et al., (2009) Molecular Therapy 17: 1453-1464) or flow cytotoxicity assays or impedance-based cytotoxicity assays (Peper et al. (2014) Journal of Immunological Methods, 405:192-198).

The term immunomodulatory polypeptide is a polypeptide that modulates immunological activity. By modulating or modulating an immune response is meant that immunological activity is either increased or decreased. The immunomodulatory protein can be a single polypeptide chain or a multimer (dimers or higher order multimers), at least two polypeptide chains covalently linked to each other, for example, by disulfide bonds between the chains. Thus, higher order monomeric, dimeric and multimeric polypeptides are within the scope of the term. Multimeric polypeptides can be homomultimeric (identical polypeptide chains) or heteromultimeric (non-identical polypeptide chains). The immunomodulatory polypeptide of the invention includes a variant ICOSL.

As used herein, the term increase means an increase by a statistically significant amount. The increase may be at least 5, 10, 20, 30, 40, 50, 75, 100% or more of the non-zero control value.

The term ICOSL isoform (inducible costimulatory ligand; CD275) is one of many naturally occurring ICOSL polypeptides that differ in amino acid sequence. Isoforms may be the product of splicing variants of an RNA transcript expressed by a single gene, or the product of the expression of genes with high similarity but divergence, resulting in a functionally similar protein, as for example can occur with gene duplication. As used herein, the term ICOSL isoform also refers to the product of different alleles of the ICOSL gene (eg, ICOSLG).

As used herein, the term lymphocyte means any of the three subtypes of leukocytes in the mammalian immune system. These include natural killer cells (NK cells)

- 21 044356 (which function in cell-mediated, cytotoxic innate immunity), T cells (for cell-mediated, cytotoxic adaptive immunity) and B cells (for humoral, antibody-driven adaptive immunity). T cells include: helper T cells, cytotoxic T cells, natural killer T cells, memory T cells, regulatory T cells, or gamma delta T cells. Innate lymphoid cells (ILCs) are also included in the definition of lymphocytes.

The terms mammal, subject, or patient specifically include reference to at least one animal among a human, a chimpanzee, a rhesus monkey, a cynomolgus monkey, a dog, a cat, a mouse, or a rat.

As used herein, the term membrane protein means a protein that, under physiological conditions, is directly or indirectly attached to a lipid bilayer. The lipid bilayer that forms the membrane may be a biological membrane, such as a eukaryotic cell membrane (eg, from a mammal) or an artificial membrane (ie, human-made), such as that found on a liposome. Attachment of a membrane protein to a lipid bilayer can be accomplished by covalent fixation or by non-covalent interactions such as hydrophobic or electrostatic interactions. The membrane protein may be an integral membrane protein or a peripheral membrane protein. Membrane proteins, which are peripheral membrane proteins, are non-covalently attached to a lipid bilayer or non-covalently attached to an integral membrane protein. The peripheral membrane protein forms a transient attachment to the lipid bilayer such that, over a range of physiological conditions in a mammal, the peripheral membrane protein can associate and/or dissociate from the lipid bilayer. Unlike peripheral membrane proteins, integral membrane proteins form a substantially permanent association with the membrane lipid bilayer such that, under the range of conditions that are physiological in mammals, integral membrane proteins do not dissociate from their attachment to the lipid bilayer. A membrane protein can form an attachment to the membrane through one layer of the lipid bilayer (monotopic) or attach through both layers of the membrane (polytopic). An integral membrane protein that interacts with only one lipid bilayer is an integral monotopic protein. An integral membrane protein that interacts with both lipid bilayers is an integral polytopic protein, alternatively referred to herein as a transmembrane protein.

The terms modulating or modulating, as used herein in the context of an immune response, e.g., a mammalian immune response, refers to any change, e.g., increase or decrease, in existing or potential immune responses that occurs as a result of administration of an immunomodulatory polypeptide comprising a variant ICOSL of the present invention, or by administration of engineered cells expressing an immunomodulatory protein, such as the ICOSL transmembrane immunomodulatory protein variant of the present invention. Thus, it refers to a change, such as an increase or decrease in an immune response compared to an immune response that occurs or is present in the absence of administration of an immunomodulatory protein including a variant ICOSL or cells expressing such an immunomodulatory polypeptide. Such modulation includes any induction, activation, suppression or change in the degree or magnitude of immunological activity of an immune cell. Immune cells include B cells, T cells, NK cells (natural killer cells), NK T cells, professional antigen presenting cells (APCs) and non-professional antigen presenting cells and inflammatory cells (neutrophils, macrophages, monocytes, eosinophils and basophils). Modulation includes any changes associated with an existing immune response, a developing immune response, a potential immune response, or the ability to induce, regulate, influence, or respond to an immune response. Modulation involves any change in the expression and/or function of genes, proteins and/or other molecules in immune cells as part of an immune response. Modulation of the immune response or modulation of immunological activity includes, for example, the following: elimination, deletion or sequestration of immune cells; inducing the formation of immune cells that can modulate the functional capacity of other cells, such as autoreactive lymphocytes, antigen presenting cells or inflammatory cells; induction of a nonresponsive state in immune cells (ie anergy); enhancing or suppressing the activity or function of immune cells, including, without limitation, changing the structure of proteins expressed by these cells. Examples include altered production and/or secretion of certain classes of molecules such as cytokines, chemokines, perforins, granzymes, growth factors, transcription factors, kinases, costimulatory molecules or other cell surface receptors, or any combination of these modulatory events. Modulation can be assessed, for example, by changing IFN-gamma expression compared to wild-type control ICOSL in a primary T cell assay (see Zhao and Ji, Exp Cell Res. 2016 Jan1;340(1)132- 138). Modulation can be assessed, for example, by changing the immunological activity of the engineered cells, such as changing the cytotoxic activity of the engineered cells or changing the cytokine secretion of the engineered cells compared to cells engineered with wild-type ICOSL transmembrane protein.

As used herein, the term molecular species means a collection of proteins with identical or substantially identical primary amino acid sequence. Each member of the mammalian immunoglobulin superfamily (IgSF) defines a set of identical or substantially identical molecular species. Thus, for example, human ICOSL is a member of IgSF, and each human ICOSL molecule represents a molecular species of ICOS. Variation between molecules that belong to the same molecular species can occur due to differences in post-translational modification such as glycosylation, phosphorylation, ubiquitination, nitrosylation, methylation, acetylation and lipidation. In addition, minor sequence differences within a molecular species due to gene polymorphism represent another form of variation within a molecular species, as in the case of truncated wild-type forms of a single molecular species due to, for example, proteolytic cleavage. A cell surface molecular species is a molecular species expressed on the surface of a mammalian cell. Two or more different protein species, each present exclusively on one or exclusively the other (but not both) of the two mammalian cells forming an IS, are referred to as being in cis or cis configuration with each other. Two different kinds of protein, the first of which is exclusively present on one of the two mammalian cells forming the IS, and the second of which is exclusively present on the second of the two mammalian cells forming the IS, are said to be in trans or trans configuration. Two different protein species, each present on both mammalian cells forming the IS, are found in both cis and trans configurations on these cells.

The term multimerization domain refers to an amino acid sequence that facilitates the stable interaction of a polypeptide molecule with one or more additional polypeptide molecules, each of which includes an additional multimerization domain, which may be the same or different multimerization domain to form a stable multimer with the first domain. Typically, the polypeptide is attached directly or indirectly to a multimerization domain. Typical multimerization domains include immunoglobulin sequences or portions thereof, leucine zippers, hydrophobic regions, hydrophilic regions, and protein-protein interaction domains. For example, the multimerization domain may be an immunoglobulin constant region or domain, such as, for example, the Fc domain or portions thereof of IgG, including the IgG1, IgG2, IgG3 or IgG4 subtypes, IgA, IgE, IgD and IgM and modified forms thereof.

The terms nucleic acid and polynucleotide are used interchangeably to refer to a polymer of nucleotide residues (eg, deoxyribonucleotides or ribonucleotides) in single- or double-stranded form. Unless otherwise specified, the terms cover nucleic acids containing known analogs of natural nucleotides and which have similar binding properties to them and are metabolized in a similar way to natural nucleotides. Unless otherwise indicated, a specific nucleotide sequence also implicitly covers its conservatively modified variants (eg, degenerate codon substitutions) and complementary nucleotide sequences, as well as the explicitly specified sequence (reference sequence). In particular, degenerate codon substitutions can be obtained by generating sequences in which the third position of one or more selected (or all) codons is replaced by mixed base and/or deoxyinosine residues. The term nucleic acid or polynucleotide includes cDNA or mRNA encoded by a gene.

As used herein, the term noncompetitive binding refers to the ability of a protein to specifically bind simultaneously to at least two cognate binding partners. Thus, the protein is capable of binding to at least two different cognate binding partners simultaneously, although the binding interaction does not necessarily have to be of the same duration, so that in some cases the protein specifically binds to only one of the cognate binding partners. In some embodiments, binding occurs under certain binding conditions. In some embodiments, the simultaneous binding is such that binding of one cognate binding partner does not significantly inhibit simultaneous binding of a second cognate binding partner. In some embodiments, noncompetitive binding means that the binding of a second cognate binding partner to its binding site on the protein does not displace the binding of the first cognate binding partner to its binding site on the protein. Methods for assessing noncompetitive binding are well known in the art, for example, the method described in Perez de La Lastra et al., Immunology, 1999 Apr: 96 (4): 663-670. In some cases, in non-competitive interactions, the first cognate binding partner specifically binds to a binding site that does not overlap the interaction site of the second cognate binding partner, such that binding of the second cognate binding partner does not directly interfere with the binding of the first cognate binding partner. Thus, any influence on the binding of a cognate binding partner by binding of a second cognate binding partner occurs through a mechanism other than directly interfering with the binding of the first cognate binding partner. For example, in the context of enzyme-substrate interactions, a noncompetitive inhibitor binds to a site other than the active site of the enzyme. Noncompetitive binding includes noncompetitive binding interactions in which a second cognate binding partner specifically binds to an interaction site that does not overlap with the binding of the first cognate partner, but binds only to a second interaction site occupied by the first cognate binding partner.

The term pharmaceutical composition refers to a composition suitable for pharmaceutical use in a mammalian subject, often a human. The pharmaceutical composition typically includes an effective amount of an active agent (eg, an ICOSL variant immunomodulatory protein or engineered cells expressing the ICOSL variant transmembrane immunomodulatory protein of the present invention) and a carrier, excipient or diluent. The carrier, excipient or diluent is typically a pharmaceutically acceptable carrier, excipient or diluent, respectively.

The terms polypeptide and protein are used interchangeably herein and refer to a molecular chain of two or more amino acids linked through peptide bonds. Terms do not apply to a specific length of product. Thus, peptides and oligopeptides are included in the definition of polypeptide. The terms include post-translational modifications of the polypeptide, for example, glycosylation, acetylation, phosphorylation and the like. The terms also include molecules that include one or more amino acid analogs or non-canonical or unnatural amino acids that can be synthesized or expressed recombinantly using known protein engineering techniques. In addition, proteins can be derivatized.

As used herein, the term primary T cell assay refers to an in vitro assay for measuring interferon-gamma (IFN-gamma) expression. A variety of such primary T cell assays are known in the art, as described in Example 6. In a preferred embodiment, the assay used is an anti-CD3 coimmobilization assay. In this assay, primary T cells are stimulated with anti-CD3 immobilized with or without additional recombinant proteins. Culture supernatants are collected at time points, typically 24-72 hours. In another embodiment, the assay used is MLR. In this assay, primary T cells are stimulated with allogeneic APC. Culture supernatants are collected at time points, typically 24-72 hours. Human IFN-γ levels are measured in culture supernatants by standard ELISA methods. Commercial kits are available from suppliers and the assay is performed according to the manufacturer's recommendation.

The term purified when applied to nucleic acids, such as those encoding the immunomodulatory proteins of the present invention, generally means a nucleic acid or polypeptide that is substantially free of other components, as determined by analytical methods well known in the art (eg, a purified polypeptide or polynucleotide forms a discrete band in an electrophoretic gel, chromatographic eluate and/or density gradient centrifugation medium). For example, a nucleic acid or polypeptide that produces substantially one band on an electrophoretic gel is purified. A purified nucleic acid or protein is at least about 50% purified, typically at least about 75, 80, 85, 90, 95, 96, 99% purified, or more (e.g., percent by weight or on a molar basis).

The term recombinant indicates that the material (eg, nucleic acid or polypeptide) has been artificially (i.e., not naturally) altered by human intervention. The change can be made to a material within or by removing it from its natural environment or state. For example, a recombinant nucleic acid is a recombinant nucleic acid that is produced by recombining nucleic acids, such as during cloning, affinity modification, DNA shuffling, or other well-known molecular biology procedures. A recombinant DNA molecule consists of segments of DNA joined together using molecular biological techniques. As used herein, the term recombinant protein or recombinant polypeptide refers to a protein molecule that is expressed using a recombinant DNA molecule. A recombinant host cell is a cell that includes and/or expresses a recombinant nucleic acid or that is otherwise altered by genetic engineering, for example, by introducing into the cell a nucleic acid molecule encoding a recombinant protein, such as a transmembrane immunomodulatory protein provided herein . Transcription control signals in eukaryotes include promoter and enhancer elements. Promoters and enhancers consist of short arrays of DNA sequences that specifically interact with cellular proteins involved in transcription. Promoter and enhancer elements have been isolated from various sources.

- 24 044356 nicks of eukaryotes, including genes in yeast, insect and mammalian cells and viruses (similar control elements, i.e. promoters, are also found in prokaryotes). The choice of specific promoter and enhancer depends on what type of cell should be used to express the protein of interest. The terms functionally combined, functionally, and operably linked as used herein refer to the linking of nucleotide sequences in such a manner or orientation that a nucleotide molecule is formed capable of directing the transcription of a given gene and/or the synthesis of a protein molecule of interest.

The term recombinant expression vector, as used herein, refers to a DNA molecule containing the desired coding sequence and the corresponding nucleotide sequence necessary for expression of the operably linked coding sequence in a particular host cell. Nucleotide sequences required for expression in prokaryotes include a promoter, optionally an operator sequence, a ribosome binding site, and possibly other sequences. It is known that eukaryotic cells use promoters, enhancers, termination and polyadenylation signals. The secretory signal peptide sequence may also optionally be encoded by a recombinant expression vector operably linked to a recombinant protein coding sequence, such as a recombinant fusion protein, such that the expressed fusion protein can be secreted by the recombinant host cell to facilitate release of the fusion protein from the cell, if desired. . The term includes a vector as a self-replicating nucleotide structure, and a vector incorporated into the genome of the host cell into which it has been introduced. Among the vectors are viral vectors such as lentiviral vectors.

The term selectivity refers to the preference of a test protein or polypeptide to specifically bind to one substrate, such as one cognate binding partner, over specifically binding to another substrate, such as another cognate binding partner of the target protein. Selectivity can be reflected as the ratio of the binding activity (eg, binding affinity) of a test protein to a first substrate, such as a first cognate binding partner (eg, K _d i ) and the binding activity (eg, binding affinity) of the same test protein to a second cognate partner binding (for example, K _d2 ).

As used herein, the term sequence identity refers to sequence identity between genes or proteins at the nucleotide or amino acid level, respectively. Sequence identification is a measure of identity between proteins at the amino acid level and a measure of identity between nucleic acids at the nucleotide level. Protein sequence identification can be determined by comparing the amino acid sequence at a given position in each sequence when the sequences are aligned. Similarly, the identity of a nucleotide sequence can be determined by comparing the nucleotide sequence at a given position in each sequence when the sequences are aligned. Methods for aligning sequences for comparison are well known in the art, such methods include GAP, BESTFIT, BLAST, FASTA and TFASTA. The BLAST algorithm calculates the percentage of sequence identity and performs statistical analysis of the similarity between two sequences. Software for performing BLAST analysis is available on the National Center for Biotechnology Information (NCBI) website.

The term soluble as used herein in relation to proteins means that the protein is not a membrane protein. In general, the soluble protein comprises only the extracellular domain of the IgSF family receptor or a portion thereof containing the IgSF domain or domains or specific binding fragments thereof, but does not include the transmembrane domain. In some cases, protein solubility can be improved by binding or attaching, directly or indirectly via a linker, to the Fc domain, which in some cases can also improve the stability and/or half-life of the protein. In some aspects, the soluble protein is an Fc fusion protein.

The term species, when used herein in relation to polypeptides or nucleic acids, means an ensemble of molecules with identical or substantially identical sequences. Variation between molecules that belong to the same molecular species can occur due to differences in post-translational modification such as glycosylation, phosphorylation, ubiquitination, nitrosylation, methylation, acetylation and lipidation. Slightly truncated sequences that differ (or encode a difference) from the full-length species at the amino terminus or carboxy terminus by no more than 1, 2, or 3 amino acid residues are considered a single species. Such microheterogeneities are a common feature of manufactured proteins.

The term specific binding fragment, as used herein in relation to a full-length wild-type mammalian ICOSL polypeptide or its IgV or IgC domain, means a polypeptide having a subsequence with an IgV and/or IgC domain that specifically binds in vitro and/or in vivo to mammalian ICOS and/or mammalian CD28, such as human or murine ICOS or CD28. In some embodiments, the ICOSL IgV or ICOSL IgC specific binding fragment is at least 60, 70, 75, 80, 85, 90, 95, 96, 97, 98, or 99% of the sequence length of the full-length wild-type sequence. The specific binding moiety can be changed in sequence to form the ICOSL variant of the invention.

As used herein, the term specifically binds means the ability of a protein, under specific binding conditions, to bind to a target protein such that its affinity or avidity is at least 10 times greater, but not necessarily 50, 100, 250 or 500, or even at least at least 1000 times greater than the average affinity or avidity of the same protein for a set of random peptides or polypeptides of sufficient statistical size. A particular binding protein need not bind exclusively to one target molecule, but may specifically bind to a non-target molecule due to similarity in structural conformation between the target and non-target (eg, paralogs or orthologs). Those skilled in the art will appreciate that specific binding to a molecule having the same function in different animal species (i.e., an ortholog) or to a non-target molecule having a substantially similar epitope to the target molecule (i.e., a paralog) is possible and does not reduce binding specificity, which is determined relative to a statistically significant set of unique non-targets (for example, random polypeptides). Thus, a polypeptide of the invention may specifically bind to more than one distinct species of target molecule due to cross-reactivity. ELISA immunoassays or Biacore measurements can be used to determine the specific binding between two proteins. Typically, interactions between two binding proteins have dissociation constants (Kd) less than 1x10'5 and often up to 1x10'1 M. In some aspects of the present invention, interactions between two binding proteins have dissociation constants of ^1x10'6 , ^1x10'7 , ^1x10'8 , 1x10' ⁹ , 1x10' ¹⁰ or 1x10'11 M.

The terms surface expressed or surface expression, when applied to a mammalian cell expressing a polypeptide, means that the polypeptide is expressed as a membrane protein. In some embodiments, the membrane protein is a transmembrane protein.

As used herein, the term synthetic, when referring to, for example, a synthetic nucleic acid molecule or a synthetic gene or a synthetic peptide, refers to a nucleic acid molecule or polypeptide molecule that is produced by recombinant methods and/or chemical synthesis methods.

The term target fragment as used herein refers to a composition that is covalently or non-covalently attached to, or physically encapsulating a polypeptide containing a variant ICOSL of the present invention. The targeting functional group has a specific binding affinity for the desired counterstructure, such as a cell surface receptor (eg, PD-L1 family member B7), or a tumor antigen, such as tumor-specific antigen (TSA) or tumor-associated antigen (TAA), such as B7 -H6. As a rule, the desired counterstructure is localized in a specific tissue or cell type. Targeting functional groups include: antibodies, antigen binding fragment (Fab), variable fragment (Fv) containing VH and VL, single chain variable fragment (scFv) containing VH and VL joined together in one chain, and other fragments of the antibody V region , such as Fab', F(ab) ₂ , F(ab')2, dsFv-diabody, nanobodies, soluble receptors, receptor ligands, affinity mature receptors or ligands, and low molecular weight compositions (<500 Daltons) (e.g., specific receptor binding compositions). Targeting functional groups may also be attached covalently or non-covalently to the lipid membrane of liposomes that encapsulate the polypeptide of the present invention.

As used herein, the term transmembrane protein means a membrane protein that substantially or completely spans a lipid bilayer, such as lipid bilayers found in a biological membrane, such as a mammalian cell, or in an artificial construct such as a liposome. A transmembrane protein includes a transmembrane domain (transmembrane domain) through which it is integrated into the lipid bilayer and through which the integration is thermodynamically stable under physiological conditions. Transmembrane domains are typically predicted from their amino acid sequence by any of the commercially available bioinformatics software applications due to their increased hydrophobicity compared to regions of the protein that interact with aqueous environments (e.g., cytosol, extracellular fluid). The transmembrane domain is often a hydrophobic alpha helix that fills the membrane. A transmembrane protein can pass through both layers of the lipid bilayer one or more times. Transmembrane protein includes the provided transmembrane immunomodulatory proteins described herein. In addition to the transmembrane domain, the transmembrane immunomodulatory protein of the invention further includes an ectodomain and, in some embodiments, an endodomain.

The terms treat, cure, or therapy of a disease or disorder as used herein means the slowing, stopping, or reversing the progression of a disease or disorder, as evidenced by the subsidence, stopping, or eliminating either clinical or diagnostic symptoms, by administering a therapeutic composition (eg containing an immunomodulatory protein or engineered cells) of the invention either alone or in combination with another compound as described herein. To treat, cure, or therapy also means to reduce the severity of symptoms in an acute or chronic disease or disorder, or to reduce the frequency of relapses, such as in the case of a relapsing or remitting autoimmune disease, or to reduce inflammation, in the case of the inflammatory aspect of an autoimmune disease. When used herein in the context of a cancer, the terms treat or inhibit or inhibit a cancer refers to at least one of: a statistically significant reduction in tumor growth rate, arrest of tumor growth, or reduction in tumor size, weight, metabolic activity or volume , as measured by standard criteria such as, but not limited to, Response Evaluation Criteria for Solid Tumors (RECIST) or a statistically significant increase in progression-free survival (PFS) or overall survival (OS). Prevent, prevent or prevent a disease or disorder, as used in the context of the present invention, refers to the administration of an immunomodulatory polypeptide or engineered cells of the invention, either alone or in combination with another compound, to prevent the occurrence or onset of a disease or disorder or some or all of the symptoms of a disease or disorder or to reduce the likelihood of a disease or disorder.

The term tumor-specific antigen or TSA, as used herein, refers to an antigen that is present primarily in tumor cells of a mammalian subject, but is not typically found in normal cells of the mammalian subject. The tumor-specific antigen need not be exclusive to tumor cells, but the percentage of cells in a particular mammal that have the tumor-specific antigen is high enough, or the levels of tumor-specific antigen on the tumor surface are high enough, such that the antigen can be used as a target for anticancer therapy and provide prevention or treatment of a mammal from the effects of a tumor. In some embodiments, in a random statistical sample of tumor-bearing mammalian cells, at least 50% of the cells having TSA are malignant. In other embodiments, at least 60, 70, 80, 85, 90, 95, or 99% of the cells containing TSA are malignant.

The term variant (also modified or mutant) used in relation to a variant ICOSL means an ICOSL, for example, a mammalian (eg human or mouse) ICOSL created by human intervention. A variant ICOSL is a polypeptide having an altered amino acid sequence relative to unmodified ICOSL or wild-type ICOSL. A variant ICOSL is a polypeptide that differs from the sequence of the wild-type ICOSL isoform by one or more amino acid substitutions, deletions, additions, or combinations thereof. For purposes of the present invention, the variant ICOSL contains at least one affinity modified domain whereby one or more amino acid differences occur in the IgSF domain (eg, the IgV domain). A variant ICOSL may contain 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 or more amino acid differences, such as amino acid substitutions. The variant ICOSL polypeptide typically has at least 50, 60, 70, 80, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% or more sequence identity relative to the corresponding wild-type or unmodified ICOSL, such as the sequence of SEQ ID NO: 5, a mature sequence thereof, or a portion thereof containing an extracellular domain or an IgSF domain thereof. In some embodiments, the variant ICOSL polypeptide has at least 50, 60, 70, 80, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% or more sequence identity to the corresponding wild-type ICOSL or unmodified ICOSL containing the sequence set forth in SEQ ID NO: 32 or SEQ ID NO: 196. Unnatural amino acids as well as naturally occurring amino acids are included within the scope of permissible substitutions or additions. The variant ICOSL is not limited to any particular production method and includes, for example, de novo chemical synthesis, de novo recombinant DNA methods, or combinations thereof. The variant ICOSL of the invention specifically binds to at least one or more of mammalian CD28, ICOS or CTLA-4. In some embodiments, the altered amino acid sequence results in a change (ie, increase or decrease) in binding affinity or avidity for ICOS and/or CD28 compared to wild-type ICOSL protein. Increases or decreases in binding affinity or avidity can be determined using well-known binding assays such as flow cytometry. Larsen et al., American Journal of Transplantation, Vol 5: 443-453 (2005). See also Linsley et al., Immunity, 1: 7930801 (1994). The increase in binding affinity or avidity of the variant ICOSL to ICOS and/or CD28 is at least 5% greater than that of wild type ICOSL, and in some embodiments, at least 10, 15, 20, 30, 40 ,50, is 100% greater than the wild-type ICOSL control value. Decreased ICOSL binding affinity or avidity

- 27 044356 with ICOS and/or CD28 is no more than 95% of wild type control values, and in some embodiments no more than 80, 70, 60, 50, 40, 30, 20, 10, 5% or no detectable affinity or binding avidity of ICOS and/or wild-type CD28 controls. A variant ICOSL changes the primary amino acid sequence by substitution, addition or deletion of amino acid residues. The term variant in the context of a variant ICOSL should not be construed as imposing any condition on any particular source composition or the manner in which the variant ICOSL is created. A variant ICOSL may, for example, be generated starting from wild-type mammalian ICOSL sequence information, then modeled in silico to bind to ICOS and/or CD28, and finally recombinantly or chemically synthesized to produce the variant ICOSL of the present invention. In one alternative example, a variant ICOSL can be created by site-directed mutagenesis of wild-type ICOSL. Thus, the variant ICOSL designates a composition and not a product obtained by any method. Various methods can be used, including recombinant methods, chemical synthesis, or combinations thereof.

As used herein, the term wild type or natural or native in connection with biological materials, such as nucleotide molecules, proteins, IgSF members, host cells and the like, refers to those biological materials that are found in nature and are not modified by human intervention.

II. ICOSL variant polypeptides.

Provided herein are variant ICOSL polypeptides that exhibit altered (increased or decreased) binding activity or affinity for one or more of the ICOSL cognate binding partners. In some embodiments, the ICOSL cognate binding partner is CD28, ICOS, or CTLA-4. In some embodiments, the variant ICOSL polypeptide includes one or more amino acid modifications, such as one or more substitutions (as variant mutations or substitutions), deletions, or additions in an immunoglobulin superfamily (IgSF) domain (IgD) relative to the wild-type or unmodified ICOSL polypeptide to an ICOSL polypeptide or to a portion of wild-type or unmodified ICOSL containing an immunoglobulin superfamily (IgSF) domain or a specific binding fragment thereof. Thus, the provided variant ICOSL polypeptide is or includes a variant IgD (hereinafter referred to as vIgD) in which one or more amino acid modifications (eg, substitutions) have occurred in the IgD.

In some embodiments, the IgD includes an IgV domain or an IgC domain (eg, IgC2) or a specific binding fragment of an IgV domain or an IgC domain (eg, IgC2), or combinations thereof. In some embodiments, the IgD may be IgV alone, a combination of IgV and IgC including the entire extracellular domain (ECD), or any combination of ICOSL Ig domains. In table Figure 2 shows exemplary residues that correspond to the IgV or IgC regions of the ICOSL. In some embodiments, the variant ICOSL polypeptide includes an IgV domain or an IgC domain or specific binding fragments thereof, wherein at least one amino acid modification (eg, substitution) occurs in the IgV or IgC domain or specific binding fragment thereof. In some embodiments, due to altered binding activity or affinity, the IgV domain or IgC domain is an affinity modified IgSF domain.

In some embodiments, the variant is modified in yet another IgSF domain relative to the unmodified ICOSL sequence. In some embodiments, the unmodified ICOSL sequence is wild-type ICOSL. In some embodiments, unmodified or wild-type ICOSL has the sequence of native ICOSL or an orthologue thereof. In some embodiments, the unmodified ICOSL is or includes an extracellular domain (ECD) of ICOSL or a portion thereof containing one or more IgSF domains (see Table 2). In some embodiments, the extracellular domain of an unmodified ICOSL polypeptide or a wild-type ICOSL polypeptide includes an IgV domain and an IgC domain or domains. However, the variant ICOSL polypeptide must not include both an IgV domain and an IgC domain or domains. In some embodiments, the variant ICOSL polypeptide includes or consists primarily of an IgV domain or a specific binding fragment thereof. In some embodiments, the variant ICOSL polypeptide includes or consists primarily of an IgC domain or specific binding fragments thereof. In some embodiments, the variant ICOSL is soluble and lacks a transmembrane domain. In some embodiments, the variant ICOSL further includes a transmembrane domain, and in some cases also a cytoplasmic domain.

In some embodiments, the wild-type or unmodified ICOSL sequence is a mammalian ICOSL sequence. In some embodiments, the wild-type or unmodified ICOSL sequence may be a mammalian ICOSL, which is, but is not limited to, a human, mouse, cynomolgus, or rat. In some embodiments, the wild-type or unmodified ICOSL sequence is human.

- 28 044356

In some embodiments, the wild-type or unmodified ICOSL sequence has (i) the amino acid sequence set forth in SEQ ID NO: 5 or a mature form thereof lacking the signal sequence, (ii) an amino acid sequence that has at least 85, 86 , 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% or more sequence identity with SEQ ID NO: 5 or a mature form thereof, or (iii) is part of (i ) or (ii) containing an IgV or IgC domain or specific binding fragments thereof.

In some embodiments, the wild-type ICOSL or unmodified ICOSL sequence is or includes the extracellular domain of ICOSL or a portion thereof. In some embodiments, the unmodified ICOSL polypeptide or wild-type ICOSL polypeptide comprises the amino acid sequence set forth in SEQ ID NO: 32, or an orthologue thereof. In some cases, an unmodified or wild-type ICOSL polypeptide may comprise (i) an amino acid sequence set forth in SEQ ID NO: 32, (ii) an amino acid sequence that is at least about 85, 86, 87, 88, 89 , 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% identity with SEQ ID NO: 32 or (iii) is a specific binding fragment of sequence (i) or (ii) containing an IgV domain or IgC domain.

In some embodiments, a wild-type ICOSL polypeptide or an unmodified ICOSL polypeptide includes an IgV domain or an IgC domain, or a specific binding fragment thereof. In some embodiments, the IgV domain of a wild-type or unmodified ICOSL polypeptide comprises the amino acid sequence set forth in SEQ ID NO: 196 (corresponding to amino acid residues 19-129 of SEQ ID NO: 5), or an ortholog thereof. For example, the IgV domain of an unmodified ICOSL polypeptide or a wild-type ICOSL polypeptide may comprise (i) the amino acid sequence set forth in SEQ ID NO: 196, (ii) an amino acid sequence that is at least about 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% sequence identity with SEQ ID NO: 196, or (iii) a specific binding fragment of the amino acid sequence presented in SEQ ID NO: 196, or specific binding fragment of sequence (i) or (ii). In some embodiments, a wild-type IgV domain or an unmodified IgV domain is capable of binding one or more ICOSL cognate binding partners, such as one or more of CD28, ICOS, or CTLA-4.

In some embodiments, the IgC domain of a wild-type or unmodified ICOSL polypeptide includes the amino acid sequence represented by residues 141227 in SEQ ID NO: 5, or an orthologue thereof. For example, the IgC domain of a wild-type or unmodified ICOSL polypeptide may comprise (i) an amino acid sequence comprising residues 141-227 of SEQ ID NO: 5, (ii) an amino acid sequence that has at least about 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% sequence identity to residues 141-227 SEQ ID NO: 5 or (iii) (i) or (ii). In some embodiments, a wild-type IgV domain or an unmodified IgV domain is capable of binding one or more cognate ICOSL-binding proteins.

In some embodiments, a wild-type ICOSL polypeptide or an unmodified ICOSL polypeptide includes a specific binding fragment from ICOSL, such as an IgV domain or IgC domain specific binding fragment. In some embodiments of the invention, the specific binding moiety may bind CD28, ICOS and/or CTLA-4. In some embodiments, the specific binding fragment may have an amino acid length of at least 50 amino acids, such as at least 60, 70, 80, 90, 100 or 110 amino acids. In some embodiments, the specific IgV domain binding fragment comprises an amino acid sequence that is at least about 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% length of the IgV domain, represented as amino acids 19-129 in SEQ ID NO: 5. In some embodiments, the specific binding fragment of the IgC domain includes an amino acid sequence that is at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% of the length of the IgC domain represented as amino acids 141-227 in SEQ ID NO: 5.

In some embodiments, a variant ICOSL polypeptide includes an ECD domain or a portion thereof containing one or more affinity modified IgSF domains. In some embodiments, variant ICOSL polypeptides may comprise an IgV domain or an IgC domain, in which one or more IgSF domains (IgV or IgC) or a specific IgV domain binding fragment or a specific IgC domain binding fragment includes one or more amino acid modifications (eg, substitutions). In some embodiments, variant ICOSL polypeptides may comprise an IgV domain and an IgC domain, or a specific binding fragment of an IgV domain and a specific binding fragment of an IgC domain.

In some embodiments, the variant ICOSL polypeptide includes a full-length IgV domain. In some embodiments, the variant ICOSL polypeptide includes a full-length IgC domain. In some embodiments, the variant ICOSL polypeptide includes a specific IgV domain binding fragment. In some embodiments, the variant ICOSL polypeptide includes a specific IgC domain binding fragment. In some embodiments, the variant ICOSL polypeptide comprises sex

- 29 044356 no-size IgV domain and full-size IgC domain. In some embodiments, the variant ICOSL polypeptide includes a full-length IgV domain and a specific IgC domain binding fragment. In some embodiments, the variant ICOSL polypeptide includes a specific binding fragment of an IgV domain and a full-length IgC domain. In some embodiments, the variant ICOSL polypeptide includes an IgV domain specific binding fragment and an IgC domain specific binding fragment.

In any such embodiments, a modification of one or more amino acids (eg, substitutions) of the variant ICOSL polypeptides may be located in any one or more of the ICOSL polypeptide domains. For example, in some embodiments, one or more amino acid substitutions are in the extracellular domain of the variant ICOSL polypeptide. In some embodiments, the one or more amino acid substitutions are located in an IgV domain or a specific binding fragment of an IgV domain. In some embodiments, one or more amino acid modifications (eg, substitutions) are located in the IgC domain or a specific binding fragment of the IgC domain.

In general, each of the various attributes of polypeptides is disclosed separately below (e.g., soluble, secreted, and membrane-bound polypeptides, affinity of ICOSL for CD28, ICOS, and CTLA-4, number of variations in a polypeptide chain, number of polypeptide chains linked, number and pattern of amino acid changes per variant ICOSL, etc.). However, as one skilled in the art will appreciate, any given polypeptide may include a combination of these independent attributes. It is understood that reference to amino acids, including the specific sequence set forth as SEQ ID NO, used to describe the domain organization of the IgSF domain is for illustrative purposes and is not intended to limit the scope of the proposed embodiments. It is clear that polypeptides and the description of their domains are theoretically derived from homology analysis and alignment with similar molecules. Thus, the exact locus may vary and is not necessarily the same for every protein. Therefore, a specific IgSF domain, such as an IgV specific domain or an IgC domain, may be several amino acids (one, two, three or four) longer or shorter.

In addition, various embodiments of the invention, as discussed below, are often provided within the meaning of a certain term as described above. Therefore, the embodiments described in the specific definition are to be interpreted as being incorporated by reference when the defined term is used in discussing the various aspects and attributes described herein. Thus, the headings, the order in which the various aspects and embodiments are presented, and the separate disclosure of each independent attribute are not intended to limit the scope of the present invention.

Approximate modifications.

Provided herein are variant ICOSL polypeptides comprising at least one affinity modified IgSF domain (e.g., IgV or IgC) or a specific binding fragment thereof in an IgSF domain contained in a wild-type ICOSL polypeptide or an unmodified ICOSL polypeptide such that the variant the ICOSL polypeptide exhibits altered (increased or decreased) binding activity or affinity for one or more ligands, ICOS, CD28 or CTLA-4, relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has a binding affinity for CD28, ICOS, and/or CTLA-4 that is different from the control sequence of wild-type or unmodified ICOSL polypeptide, as determined, for example, by an enzyme-linked immunosorbent assay ELISA, flow cytometry, or Biacore assay. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28, ICOS and/or CTLA-4. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28, ICOS and/or CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. CD28, ICOS and/or CTLA-4 may be a mammalian protein, such as a human protein or a mouse protein.

The binding affinity for each of the cognate binding partners is independent; that is, in some embodiments, the variant ICOSL polypeptide has increased binding affinity to one, two, or three of CD28, ICOS, and/or CTLA-4 and decreased binding affinity to one, two, or three of CD28, ICOS, and CTLA-4, relative to the polypeptide Wild type ICOSL or unmodified ICOSL.

In some embodiments, the variant ICOSL polypeptide has increased CD28 binding affinity relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for ICOS relative to the wild type ICOSL polypeptide or the unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has a reduced affinity

- 30 044356 CD28 binding activity relative to wild type ICOSL polypeptide or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has a reduced binding affinity for ICOS relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has a reduced binding affinity for CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide.

In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28 and ICOS relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28 and decreased binding affinity for ICOS relative to a wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and ICOS relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and increased binding affinity for ICOS relative to a wild-type or unmodified ICOSL polypeptide.

In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28 and CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28 and decreased binding affinity for CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and increased binding affinity for CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide.

In some embodiments, the variant ICOSL polypeptide has increased binding affinity for ICOS and CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for ICOS and reduced binding affinity for CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for ICOS and CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for ICOS and increased binding affinity for CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide.

In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28, ICOS, and CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28 and ICOS and reduced binding affinity for CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28 and CTLA-4 and decreased binding affinity for ICOS relative to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and ICOS and increased binding affinity for CTLA-4 relative to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and increased binding affinity for ICOS and CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has increased binding affinity for CD28 and decreased binding affinity for ICOS and CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28, CTLA-4, and ICOS relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and increased binding affinity for ICOS and CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide.

In some embodiments, a variant ICOSL polypeptide with increased or greater binding affinity for CD28, ICOS and/or CTLA-4 will have an increase in binding affinity relative to a wild-type control ICOSL polypeptide or an unmodified control ICOSL polypeptide of at least about 5%, e.g. at least about 10, 15, 20, 25, 35 or 50% for CD28, ICOS and/or CTLA-4. In some embodiments, the increase in binding affinity of the relative wild-type ICOSL polypeptide or unmodified ICOSL polypeptide is greater than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40 or 50 times. In such examples, the wild-type or unmodified ICOSL polypeptide has the same sequence as the variant ICOSL polypeptide except that it does not include one or more amino acid modifications (eg, substitutions).

In some embodiments, a variant ICOSL polypeptide with reduced or decreased binding affinity for CD28, ICOS, and/or CTLA-4 will have a reduction in binding affinity relative to a control wild-type ICOSL polypeptide or a control unmodified ICOSL polypeptide of at least 5%, e.g. at least about 10, 15, 20, 30, 40, 50, 60, 70, 80, 90% or more for CD28, ICOSL and/or CTLA-4. In some embodiments, the reduction in binding affinity of the wild-type ICOSL polypeptide or unmodified ICOSL polypeptide is greater than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9 , 10, 20, 30, 40 or 50 times. In such examples, the wild-type or unmodified ICOSL polypeptide has the same sequence as the variant ICOSL polypeptide except that it does not include one or more amino acid modifications, such as substitutions.

In some embodiments, the equilibrium dissociation constant (Kd) of any of the above embodiments for CD28, ICOS and/or CTLA-4 may be less than 1x10'5, 1x10'6, ^1x10'7 , 1x10, 1x10'9, 1x10' ¹⁰ or 1x10 ^-11 , or 1x10' ¹² M.

In some embodiments, the variant ICOSL polypeptide has increased or greater binding affinity for CD28. In some embodiments, a variant ICOSL polypeptide with increased or greater CD28 binding affinity will have an increase in binding affinity relative to a wild-type or unmodified ICOSL polypeptide of at least about 25%, such as at least about 30, 40, 50 or 60% for CD28. In some embodiments, a variant ICOSL polypeptide with increased or greater binding affinity for CD28 has an equilibrium dissociation constant (Kd) of less than 200, 300, 400, 500 pM, or 600 μM for CD28. In some embodiments, the variant polypeptide specifically binds to the ectodomain of one of ICOS, CD28, or CTLA4 with increased selectivity compared to unmodified ICOSL. In some embodiments, increased selectivity for CD28. In some embodiments, the increased selectivity includes a greater ratio of binding of a variant ICOSL polypeptide to one cognate binding partner selected from ICOS, CD28, and CTLA4 relative to another cognate binding partner compared to the ratio of binding of an unmodified ICOSL polypeptide to one cognate binding partner relative to another cognate binding partner. In some embodiments, the ratio is greater than or at least about 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 30, 40, 50 times or more.

The wild-type ICOSL sequence or the unmodified ICOSL sequence need not be used as the starting composition to prepare the variant ICOSL polypeptides described herein. Therefore, the use of the term modification, such as substitution, does not mean that the present embodiments are limited to a particular method for producing variant ICOSL polypeptides. Variant ICOSL polypeptides can be produced, for example, by de novo peptide synthesis and therefore do not necessarily require modification, such as substitution, in the sense of changing the codon to encode the modification, such as substitution. This principle also applies to the terms addition and deletion of an amino acid residue, which also do not imply a specific mode of implementation. The means by which ICOSL variant polypeptides are designed or created are not limited to any particular method. In some embodiments, however, the nucleic acid encoding wild-type or unmodified ICOSL is subjected to mutagenesis from wild-type ICOSL or unmodified ICOSL genetic material and screened for the desired specific binding affinity and/or induction of IFN-gamma expression or other functional activity. In some embodiments, a variant ICOSL polypeptide is synthesized de novo using protein or nucleic acid sequences available in any quantity from publicly available databases and then screened. The National Center for Biotechnology Information provides such information and its website is publicly accessible via the Internet, as is the UniProtKB database discussed earlier.

Unless otherwise indicated as indicated throughout this specification, amino acid substitution(s) are designated by the amino acid position number corresponding to the position numbering of the unmodified ECD sequence set forth in SEQ ID NO: 32 or, where applicable, the unmodified IgV sequence set forth in SEQ ID NO : 196 (containing residues 19-129 of SEQ ID NO: 32):

- 32 044356

DTQEKEVRAMVGSDVELSCACPEGSRFDLNDVYVYWQTSESKTVVTYHIPQN SSLENVDSRYRNRALMSPAGMLRGDFSLRLFNVTPQDEQKFHCLVLSQSLGFQEVLS VEVTLHVAANFSVPVSAPHSPSQDELTFTCTSINGYPRPNVYWINKTDNSLLDQALQ NDTVFLNMRGLYDVVSVLRIARTPSVNIGCCIENV LLQQNLTVGSQTGNDIGERDKIT ENPVSTGEKNAAT (SEQ ID NO:32)

DTQEKEVRAMVGSDVELSCACPEGSRFDLNDVYVYWQTSESKTVVTYHIPQN SSLENVDSRYRNRALMSPAGMLRGDFSLRLFNVTPQDEQKFHCLVLSQSLGFQEVLS VE (SEQ ID NO: 196)

The ability to determine the appropriate position of the modification is within the skill of one of ordinary skill in the art, for example, determining the amino acid substitution in an ICOSL polypeptide, including the IgSF domain-containing portion thereof (e.g., IgV), for example, by aligning a reference sequence to SEQ ID NO: 32 or SEQ ID NO: 196. When listing modifications throughout this specification, the amino acid position is indicated in the middle with the corresponding unmodified (eg, wild type) amino acid listed before the number and the identified amino acid substitution listed after the number. If the modification is a deletion at a given position, then del is indicated, and if the modification is an insertion at a given position, then ins is indicated.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, substitutions in the wild-type ICOSL sequence or the unmodified ICOSL sequence. One or more amino acid modifications, such as substitutions, may occur in the ectodomain (extracellular domain) of the wild-type ICOSL sequence or the unmodified ICOSL sequence. In some embodiments, one or more amino acid modifications, such as substitutions, may occur in the IgV domain or a specific binding fragment thereof. In some embodiments, one or more amino acid modifications, such as substitutions, may occur in the IgC domain or a specific binding fragment thereof. In some embodiments of a variant ICOSL polypeptide, some of one or more amino acid modifications, such as substitutions, may occur in the IgV domain or a specific binding fragment thereof, and some of one or more amino acid modifications, such as substitutions, may occur in the IgC domain or its specific binding fragment.

In some embodiments, the variant ICOSL polypeptide has up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid modifications, for example replacements. The modification, eg substitution, may take place in the IgV domain or in the IgC domain. In some embodiments, the variant ICOSL polypeptide has up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acids substitutions in the IgV domain or its specific binding fragment. In some embodiments, the variant ICOSL polypeptide has up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acids substitutions in the IgC domain or its specific binding fragment. In some embodiments, the variant ICOSL polypeptide is at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% identical to the sequence of the wild-type ICOSL polypeptide or an unmodified ICOSL polypeptide or a specific binding fragment thereof, such as the amino acid sequence of SEQ ID NO: 32 or 196.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modification, for example, a substitution in the unmodified ICOSL or specific binding fragment at the corresponding position(s) 10, 11, 13, 16, 18, 20, 25, 27, 30, 33, 37 , 42, 43, 47, 52, 54, 57, 61, 62, 67, 71, 72, 74, 77, 78, 75, 80, 84, 89, 90, 92, 93, 94, 96, 97, 98 , 99, 100, 102, 103, 107, 109, 110, 111,

113, 115, 116, 117, 119, 120, 121, 122, 126, 129, 130, 132, 133, 135, 138, 139, 140, 142, 143, 144, 146, 151,

152, 153, 154, 155, 156, 158, 161, 166, 168, 172, 173, 175, 190, 192, 193, 194, 198, 201, 203, 207, 208, 210,

212, 217, 218, 220, 221, 224, 225, or 227 relative to SEQ ID NO: 32. In some embodiments, such variant ICOSL polypeptides exhibit altered binding affinity for one or more of CD28, ICOS, and/or CTLA-4 relative to wild-type ICOSL polypeptide or unmodified ICOSL polypeptide. For example, in some embodiments, the variant ICOSL polypeptide exhibits increased binding affinity for CD28, ICOS and/or CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide exhibits reduced binding affinity for CD28, ICOS, or CTLA-4 relative to the wild-type or unmodified ICOSL polypeptide.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, substitutions selected from M10V, M10I, V11E, S13G, E16V, S18R, A20V, S25G, F27S, F27C, N30D, Y33del, Q37R, K42E, Y47H, Т43А, N52H, N52D, N52Q, N52S, N52Y, N52K, S54A, S54P, N57D, N57Y, R61S, R61C, Y62F, L67P, А71Т, G72R, L74Q, R75Q, D77G, F78L, L80P, N84Q, D89G, E9 0A, K92R, F93L, H94E, H94D, L96F, L96I, V97A, L98F, S99G, Q100R, Q100K, Q100P, L102R, G103E, V107A, V107I, S109G, S109N, V110D, V110N, V110A, E111del , T113E, H115R, H115Q, V116A, A117T, N119Q, F120I, F120S, S121G, V122A, V122M, S126T, S126R, H129P, S130G, S132F, Q133H,

- 33 044356

E135K, F138L, T139S, C140D, C140del, S142F, I143V, I143T, N144D, Y146C, V151A, Y152C, Y152H, W153R, I154F, N155H, N155Q, K156M, D158G, L161P, L1 61M, L166Q, N168Q, F172S, L173S, M175T, T190S, T190A, S192G, V193M, N194D, C198R, N201S, L203P, L203F, N207Q, L208P, V210A, S212G, D217V, I218T, I218N, E220G, R221G, R221I, I2 24V, T225A, N227K or their conservative amino acid modifications , for example, replacements. A conservative amino acid modification, such as a substitution, is any amino acid that falls into the same class of amino acids as the substituted amino acids, but is different from the wild type or unmodified amino acid. The classes of amino acids are aliphatic (glycine, alanine, valine, leucine and isoleucine), hydroxyl or sulfur (serine, cysteine, threonine and methionine), cyclic (proline), aromatic (phenylalanine, tyrosine, tryptophan) basic (histidine, lysine and arginine) and acid/amide (aspartate, glutamate, asparagine and glutamine).

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, such as substitutions selected from M10V, M10I, V11E, S13G, E16V, S18R, A20V, S25G, F27S, F27C, N30D, Y33del, Q37R, K42E, T43A, Y47H , N52H, N52D, N52Q, N52S, N52Y, N52K, S54A, S54P, N57D, N57Y, R61S, R61C, Y62F, L67P, A71T, G72R, L74Q, R75Q, D77G, F78L, L80P, N84Q, D89G, E90A, K92R , F93L, H94E, H94D, L96F, L96I, V97A, L98F, S99G, Q100R, Q100K, Q100P, G103E, L102R, V107A, V107I, S109G, S109N, V110D, V110N, V110A, E111del, T1 13E, H115R, H115Q, V116A ,

A117T, N119Q, F120I, F120S, S121G, V122A, V122M, S126T, S126R, H129P, S130G, S132F, Q133H,

E135K, F138L, T139S, C140D, C140del, S142F, I143V, I143T, N144D, Y146C, V151A, Y152C, Y152H, W153R, I154F, N155H, N155Q, K156M, D158G, L161P, L1 61M, L166Q, N168Q, F172S, L173S, M175T,

Т190А, T190S, S192G, V193M, N194D, C198R, N201S, L203P, L203F, N207Q, L208P, V210A, S212G,

D217V, I218T, I218N, E220G, R221G, R2211, I224V, T225A, N227K.

In some embodiments, one or more amino acid modifications, such as substitutions, are selected from N52Y/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52S/Y146C/Y152C, N52H/C198R, N52H/C140D/T225A, N52H/C198R/T225A, N52H/K92R, N52H/S99G, N57Y/Q100P, N52S/G103E, N52S/S130G/Y152C, N52S/Y152C, N52S/C198R, N52Y/N57Y/Y152C, N52Y/N57Y/H129P/C198R,

N52H/L161P/C198R, N52S/T113E, N52D/S54P, N52K/L208P, N52S/Y152H, N52D/V151A, N52H/I143T,

N52S/L80P,

F120S/Y152H/N201S,

N52S/N57Y/H94D/L96F/L98F/Q100R,

N52S/R75Q/L203P,

N52S/D158G,

N52D/Q133H,

N52S/N57Y/H94D/L96F/L98F/Q100R/G103E/F120S,

N52H/F78L/Q100R, N52H/N57Y/Q100R/V110D, N52H/N57Y/R75Q/Q100R/V110D, N52H/N57Y/Q100R,

N52H/N57Y/L74Q/Q100R/V110D, N52H/Q100R, N52H/S121G,

N52H/N57Y/Q100P,

N52H/N57Y/R61S/Q100R/V110D/L173S,

A20V/N52H/N57Y/Q100R/S109G, N52H/N57Y/Q100R/V122A,

N52H/N57Y/Q100R/F172S, N52H/N57Y, N52S/F120S, N52S/V97A, N52S/G72R, N52S/A71T/A117T,

N52S/E220G,

Y47H/N52S/V107A/F120S,

E16V/N52H/N5 7Y/Q100R/V110D/H115R/Y152C/K15 6M/C198R, Q37R/N52H/N57Y/Q100R/V110N/S142F/C198R/D217V/R221G, N52H/N57Y/Q100R/V1 10D/V116A/L161M/ F172S/S192G/C198R,

N52H/N57Y/Q100R/V110D/S132F/M175T,

N52S/H94E/L96I/S109N/L166Q,

S18R/N52S/F93L/I143V/R221G,

V11E/N30D/N52H/N57Y/H94E/L96I/L98F/N194D/V210A/I218T,

N52H/N57Y/Q100R/V110D/C198R,

F27S/N52H/N57Y/V110N, , A20T/N52D/Y146C/Q164L,

N52S/H94E/L96I/V122M,

N52H/N57Y/H94E/L96I/F120I/S126T/W153R/I218N,

S25G/N30D/N52S/F120S/N227K,

N52H/N57Y/Q100R/V110D/A117T/T190S/C198R,

M10V/S18R/N30D/N52S/S126R/T139S/L203F, N30D/N52S/L67P/Q100K/D217G/R221 K/T225S,

N52H/N57Y/Q100R/V110D/F172S/C198R,

S25G/F27C/N52H/N57Y/Q100R/V110D/E135K/L173S/C198R, N52H/N57Y/V110A/C198R/R221I,

M10I/S13G/N52H/N57Y/D77G/V110A/H129P/I143V/F172S/V193M, C198R,

N52H/N57Y/R61C/Y62F/Q100R/V110N/F120S/C198R, N52H/N57Y/Q100R/V110D/N144D/F172S/C198R,

N52H/N57Y/Q100R/V110D/H115R/C198R,

N52S/H94E/L98F/Q100R,

N52S/E90A,

N30D/K42E/N52S,

N52S/F120S/I143V/I224V,

N52H/N57Y/Q100R/V110D/C198R/S212G,

N52H/N57Y/Q100R/C198R, N52S/N194D, N52H/N57Y/Q100R/L102R/V110D/H115R/C198R, N52S/S54P, T38P/N52S/N57D, N52H/C140del/T225A, N52H/F7 8L/Q100R/C198R, N52H/N57Y/R75Q/Q100P/V110D,

N52H/N57Y/L74Q/V110D/S192G,

N52S/A71T/A117T/T190A/C198R,

N52H/S121G/C198R,

N52S/F120S/N227K,

T43A/N52H/N57Y/L74Q/D89G/V110D/F172S,

N52H/N57Y/Q100R/V110D/S132F/M175T, N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G, Q100R,

F138L/L203P, N57Y/F138L/L203P, N57Y/Q100R/C198R, N57Y/F138L/L203P, Q100R/F138L, L203P,

N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/Q100R/H115R/F172S/C198R, N52H/N5 7Y/Q100R/L102R/H115R/F172S/C198R, N52H/N57Y/Q100R/H11 5R/F172S/N194D,

N52H/N57Y/Q100R/H115R/C198R,

N52H/N57Y/H115R,

N52H/N57Y/Q100R/F172S/C198R, N52H/N57Y/Q100R/H115R/I143V/F172S/C198R, N52H/V122A/F172S/C198R, N52H/N57Y/H115R/F172S/C198R,

N52H/N57Y/Q100R/H115R,

N52H/N57Y/Q100R/H115R/F172S/I224V, N52H/N57Y/Q100R/H115R/F172S, N52H/N57Y/Q100R/F172S,

N52H/Q100R/H115R/I143T/F172S, N52H/N57Y/Q100P/H115R/F172S, N52Y/N57Y/Q100P/F172S,

E16V/N52H/N57Y/Q100R/V110D/H115R/C198R,

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R, N52S/E90A/H115R, N30D/K42E

- 34 044356

N52S/H115R,

N30D/K42E/N52S/H115R/C198R/R221I,

N30D/K42E/N52S/H115R/C198R,

N30D/K42E/N52S/H115R/F172S/N194D, N52S/H115R/F120S/I143V/C198R, N52S/H115R/F172S/C198R, N52H/N57Y/Q100P/C198R, N52H/N57Y/Q100P H115R/F172S/C198R, N52H /N57Y/Q100P/F172S/C198R, N52H/N57Y/Q100P/H115R, N52H/N57Y/Q100P/H115R/C198R, N52H/Q100R/C198R,

N52H/Q100R/H115R/F172S, N52H/Q100R/F172S/C198R, N52H/Q100R/H115Q/F172S/C198R,

N52H/N57Y/Q100R/F172S/C198R, N52Q/N207Q, N168Q/N207Q, N52Q/N168Q, N84Q/N207Q,

N155Q/N207Q, N119Q/N168Q, N119Q/N207Q, N119Q/N155Q, N52Q/N84Q, N52Q/N119Q, N84Q/N119Q,

N52Q/N84Q/N168Q,

N84Q/N155H/N207Q,

N84Q/N119Q/N207Q,

N52Q/N84Q/N207Q, N155Q/N168Q/N207Q, N119Q/N155H/N207Q,

N84Q/N155Q/N168Q,

N119Q/N155Q/N168Q,

N84Q/N119Q/N155Q,

N84Q/N168Q/N207Q,

N119Q/N168Q/N207Q,

N52Q/N119Q/N155Q,

N52H/N84Q/N119Q, N52H/N84Q, N52H/N84Q/N168Q, N52H/N84Q/N207Q, N52H/N84Q/N168Q/N207Q, N52Q/N84Q/N155Q, N52Q/N84Q/N168Q, N52Q/N84Q/ N155Q/N168Q, N52Q/N84Q/N119Q/N168Q,

N84Q/N119Q/N155Q/N168Q, N84Q/N155Q/N168Q/N207Q, N84Q/N119Q/N155Q/N207Q,

N52Q/N84Q/N119Q/N207Q, N52Q/N84Q/N119Q/N155Q, N52Q/N84Q/N119Q/N155Q/N207Q or

N84Q/N119Q/N155Q/N168Q/N207Q.

In some embodiments, the variant ICOSL polypeptide includes any of the mutations listed in table. 1. In table. 1 also shows exemplary sequences by reference to SEQ ID NO for the wild-type ICOSL extracellular domain (ECD) or IgV domain or exemplary variant ICOSL polypeptides. As can be seen, the exact locus or residues corresponding to a given domain may vary, for example, depending on the methods used to identify or classify the domain. Additionally, in some cases, adjacent N- and/or C-terminal amino acids of a given domain (eg, IgV) may also be included in the sequence of the variant IgSF polypeptide, for example, to ensure proper folding of the domain when expressed. Thus, it is clear that the exemplification of SEQ ID NO in table. 1 should not be considered a limitation. For example, a particular domain, such as the ECD domain of an ICOSL variant polypeptide, may be several amino acids longer or shorter, particularly 1-10, e.g., 1, 2, 3, 4, 5, 6 or 7 amino acids longer or shorter, than the amino acid sequence specified in the corresponding SEQ ID NO.

In some embodiments, the variant ICOSL polypeptide includes any of the mutations listed in table. 1. In some embodiments, the variant ICOSL polypeptide includes any of the extracellular domain (ECD) sequences listed in table. 1 (i.e., any of SEQ ID NO: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470). In some embodiments, the variant ICOSL polypeptide includes a polypeptide sequence that has at least 90, at least 91, at least 92, at least 93, at least 94, at least 95, such as at least 96, 97, 98 or 99% identity with any of the extracellular domain (ECD) sequences listed in table. 1 (i.e., any of SEQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470) and includes amino acid modification(s), e.g., substitution(s) , not present in wild-type ICOSL or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes a specific binding fragment of any of the extracellular domain (ECD) sequences listed in table. 1 (i.e., any of SEQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470) and includes amino acid modification(s), such as substitution(s), not present in wild-type ICOSL or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes any of the IgV sequences listed in table. 1 (i.e., any of SEQ ID NOs: 197-199, 201-208, 210, 212, 240, 326340, 382-386, 425-426 and 434). In some embodiments, the variant ICOSL polypeptide includes a polypeptide sequence that has at least 90, at least 91, at least 92, at least 93, at least 94, at least 95, such as at least 96, 97, 98 identity or 99% identity with any of the IgV sequences listed in table. 1 (i.e., any of SEQ ID NOs: 197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426 and 434) and includes amino acid modification(s), e.g. (s) not present in wild-type ICOSL or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes a specific binding fragment of any of the IgV sequences listed in table. 1 (i.e., in any of SEQ ID NOs: 197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426 and 434) and includes amino acid substitution(s) not present in Wild type ICOSL or unmodified ICOSL. Mutations indicated by the letter X indicate that the designated position includes Q or the wild-type residue indicated at the corresponding position of SEQ ID NO: 32.

- 35 044356

Table 1

Exemplary ICOSL variant polypeptides

Mutation(s) ECD SEQ ID NO IgV SEQ ID NO Wild type 32 196 N52S 109 197 N52H 110 198 N52D 111 199 N52 Y/N57 Y/F13 8L/L203P 112 N52H/N57Y/Q100P FROM 201 N52S/Y146C/Y152C 114 N52H/C198R 115 N52H/C140D/T225A 116 N52H/C198R/T225A 117 N52H/K92R 118 202 N52H/S99G 119 203 N52Y 120 204 N57Y 121 205 N57Y/Q100P 122 206 N52S/S130G/Y152C 123 N52S/Y152C 124 N52S/C198R 125 N52Y/N57Y/Y152C 126 N52Y/N57Y/H129P/C198R 127 N52H/L161P/C198R 128 N52S/T113E 129 S54A 130 207 N52D/S54P 131 208 N52K/L208P 132 N52S/Y152H 133 N52D/V151A 134 N52H/I143T 135 N52S/L80P 136 210 F120S/Y152H/N201S 137 N52S/R75Q/L203P 138 N52S/D158G 139 N52D/Q133H 140 N52S/N57Y/H94D/L96F/L98F/Q100R 141 212 N52S/N57 Y/H94D/L96F/L98F/Q100R/G103E/F120S 142 N52S/G103E 239 240 N52H/F78L/Q100R 280 326 N52H/N57Y/Q100R/V110D 281 327 N52H/N57Y/R75Q/Q100R/V110D 282 328 N52H/N57Y/Q100R 283 329 N52H/N57Y/L74Q/Q100R/V110D 284 330 N52H/Q100R 285 331 N52H/S121G 286 А20 V/N52H/N57 Y/Q 100R/S109G 287 332 N52H/N57Y/Q100P 288 333 N52H/N57Y/R61S/Q100R/V110D/L173 S 289 N52H/N5 7 Y/Q 100R/V122 A 290 N52H/N57Y/Q100R/F172S 291 N52H/N57Y 292 334 N52S/F120S 293 N52S/V97A 294 335 N52S/G72R 295 336 N52S/A71T/A117T 296 N52S/E220G 297 Y47H/N52S/V107A/F120S 298 N52H/N57Y/Q100R/V110D/S132F/M175T 299 E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R 300 Q3 7R/N52H/N5 7 Y/Q 100R/V110N/S142F/C198R/D217 V/R221G 301 N52H/N57Y/Q100R/V110D/C198R 302 N52H/N57Y/Q100R/V110D/V116A/L161M/F172S/S192G/C198R 303 F27S/N52H/N57Y/V110N 304 337 N52S/H94E/L96I/S109N/L166Q 305

-36044356

S18R/N52S/F93L/I143V/R221G 306 A20T/N52D/Y146C/Q164L 307 VI1E/N30D/N52H/N57Y/H94E/L96I/L98F/N194D/V210A/I218T 308 N52S/H94E/L96I/V122M 309 N52H/N57Y/H94E/L96I/F120I/S126T/W153R/I218Ν 310 Μ10 V/S18R/N3 0D/N52 S/S126R/T13 9 S/L203F 311 S25 G/N3 0D/N52 S/F120 S/N227K 312 N3 0D/N52S/L67P/Q100K/D217G/R221K/T225 S 313 N52H/N57Y/Q100R/V110D/A117T/T190S/C198R 314 N52H/N57Y/Q100R/V110D/F172S/C198R 315 S25G/F27C/N52H/N57 Y/Q 100R/V11OD/E13 5K/L173 S/C198R 316 N52H/N57Y/V110A/C198R/R221I 317 Ml OI/S13G/N52H/N57Y/D77G/V110A/H129P/I143 V/F172S/V193M/C198R 318 N52H/N57Y/R61C/Y62F/Q100R/V11ON/F120S/C198R 319 N52H/N57Y/Q100R/V110D/H115R/C198R 320 N52H/N57Y/Q100R/V110D/N144D/F172S/C198R 321 N52S/H94E/L98F/Q1OOR 322 338 N52S/E90A 323 339 N30D/K42E/N52S 324 340 N52S/F120S/I143V/I224V 325 N52H/N57Y/Q100R/V110D/C198R/S212G 364 N52H/N57 Y/Q 1OOR/C198R 365 N52S/N194D 366 N52H/N57Y/Q100R/L102R/V110D/H115R/C198R 367 N52S/S54P 368 382 T38P/N52S/N57D 369 383 El Idel 370 384 Y33del 371 385 N52H/C140del/T225A 372 N52H/F78L/Q1 OOR/C 198R 373 N52H/N57Y/R75Q/Q100P/V11OD 374 386 N52H/N57Y/L74Q/V11OD/S192G 375 N52H/S121G/C198R 376 N52S/F120S/N227K 377 N52S/A71T/A117T/T190A/C198R 378 T43A/N52H/N57Y/L74Q/D89G/V110D/F172S 379 N52H/N57Y/Q100R/V110D/S132F/M175T 380 N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G 381 N84Q 387 425 N119Q 388 N168Q 389 N207Q 390 N52Q/N207X 391 N168X/N207X 392 N52Q/N168Q 393 N84Q/N207Q 394

-37 044356

N155Q/N207Q 395 N119Q/N168Q 396 N119Q/N207Q 397 N119Q/N155X 398 N52Q/N84Q 399 426 N52Q/N119Q 400 N84Q/N119Q 401 N52Q/N84Q/N168Q 402 N52Q/N84Q/N207Q 403 N84Q/N155Q/N168Q 404 N84Q/N168Q/N207Q 405 N84Q/N155H/N207Q 406 N155Q/N168Q/N207Q 407 N119Q N155Q/N168Q 408 N119Q/N168Q/N207Q 409 N84Q/N119Q/N207Q 410 N119Q/N155H/N207Q 411 N84Q/N119Q/N155Q 412 N52Q/N119Q/N155Q 413 N52H/N84Q/N119Q 414 N52H/N84Q/N168X/N207X 415 N52Q/N84Q/N155X/N168X 416 N52Q/N84Q/N119Q/N168Q 417 N84Q/N119Q/N155Q/N168Q 418 N84Q/N155Q/N168Q/N207Q 419 N84Q/N119Q/N155Q/N207Q 420 N52Q/N84Q/N119Q/N207Q 421 N52Q/N84Q/N119Q/N155Q 422 N52Q/N84Q/N119Q/N155Q/N207Q 423 N84Q/N119Q/N155Q/N168Q/N207Q 424 Q100R 427 434 F138L/L203P 428 N52Y/F138L/L203P 429 N57Y/Q100R/C198R 430 N57Y/F138L/L203P 431 Q100R/F138L 432 L203P 433 N52H/N57Y/Q100R/H115R/C198R 435 N52H/N57 Y/Q1OOR/F172S/C198R 436 N52H/N57Y/Q100R/H115R/F172S/C198R 437 N52H/N57Y/Q100R/H115R/I143 V/F172S/C198R 438 N52H/N57Y/Q100R/L102R/H115R/F172S/C198R 439 N52H/V122A/F172S/C198R 440 N52H/N57Y/Q100R/H115R/F172S/N194D 441 N52H/N57Y/H115R/F172S/C198R 442 N52H/N57Y/Q100R/H115R/C198R 443

-38 044356

N52H/N57Y/H115R 444 N52H/N57Y/Q100R/H115R 445 N52H/N57Y/Q100R/H115R/F172S/I224V 446 N52H/N57Y/Q100R/H115R/F172S 447 N52H/N57Y/Q1OOR/F172S 448 N52H/Q100R/H115R/1143 T/F172 S 449 N52H/N5 7 Y/Q 100Р/Н115R/F172 S 450 N52Y/N57Y/Q1 OOP/F 172S 451 E16V/N52H/N57Y/Q100R/V110D/H115R/C198R 452 E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R 453 N52S/E90A/H115R 454 N30D/K42E N52S/H115R 455 N30D/K42E/N52S/H115R/C198R/R2211 456 N30D/K42E/N52S/H115R/C198R 457 N30D/K42E/N52S/H115R/F172S/N194D 458 N52S/H115R/F120S/I143V/C198R 459 N52S/H115R/F172S/C198R 460 N52H/N57 Y/Q 1OOP/C198R 461 N52H/N57Y/Q1 OOP Hl 15R/F172S/C198R 462 N52H/N5 7 Y/Q 1 OOP/F 172 S/C198R 463 N52H/N57Y/Q100P/H115R 464 N52H/N57Y/Q100P/H115R/C198R 465 N52H/Q100R/C198R 466 N52H/Q100R/H115R/F172S 467 N52H/Q100R/H115X/F172S/C198R 468 N52H/Q100R/H115R/F172S/C198R 469 N52H/N57 Y/Q 100R/F172S/C198R 470

In some embodiments, the variant ICOSL polypeptide exhibits increased affinity for the CD28 ectodomain compared to a wild-type ICOSL polypeptide or an unmodified ICOSL polypeptide, for example, comprising the sequence set forth in SEQ ID NO: 32 or 196. In some embodiments, the ICOSL polypeptide exhibits increased affinity for the ICOS ectodomain compared to a wild-type ICOSL polypeptide or an unmodified ICOSL polypeptide, such as the sequence set forth in SEQ ID NO: 32 or 196. In some embodiments, the ICOSL polypeptide exhibits increased affinity for the CD28 ectodomain and the ICOS ectodomain compared to a wild-type or unmodified ICOSL polypeptide ICOSL, such as the sequence shown in SEQ ID NO: 32 or 196.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, a substitution corresponding to position(s) 52, 54, or 57. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, a substitution selected from N52H, N52D, N52Q, N52S, N52Y, N52K, S54A, S54P or N57Y or a conservative amino acid modification, such as a substitution. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, such as a substitution selected from N52H, N52D, N52S, N52K, or N57Y, or a conservative amino acid modification, such as a substitution.

In some embodiments, the variant ICOSL polypeptide may contain one or more additional amino acid modifications, e.g., a substitution in addition to the amino acid modification, e.g., a substitution at a position corresponding to position 52, 54, or 57. In some embodiments, one or more amino acid modifications, e.g., a substitution, are in the position corresponding to 10, 11, 13, 16, 18, 20, 25, 27, 30, 37, 42, 43, 47, 52, 54, 57, 61, 62, 67, 71, 72, 74, 75 , 77, 78, 80, 84, 89, 90, 92, 93, 94, 96, 97, 98, 99, 100, 102, 103, 107, 109, 110, 113, 115, 116, 117, 119, 120 , 121, 122, 126, 129, 130, 132, 133, 135, 138, 139, 140, 142, 143, 144, 146, 151, 152, 153, 154, 155, 156, 158, 161, 166, 168 , 172, 173, 175, 190, 192, 193, 194, 198, 201, 203, 207, 208, 210, 212, 217, 221, 224, 225, or 227. In some embodiments, the variant ICOSL includes one or more amino acid modifications , for example, replacement selected from M10V, M10I, V11E, S13G, E16V, S18R, A20V, S25G, F27S, F27C, N30D, Y33del, Q37R, K42E, T43A, Y47H, N52H, N52D, N52S, N52Y, N52K, N52Q, S54A, S54P, N57D, N57Y, R61S, R61C, Y62F, L67P, A71T, G72R, L74Q, R75Q, D77G, F78L, L80P, N84Q, D89G, E90A, K92R, F93L, H94E, H94D, L96F, L96I, V9 7A, L98F, S99G, Q100R, Q100K, Q100P, L102R, G103E, V107A, V107I, S109G, S109N, V110D, V110N, V110A, E111del, T113E, H115R, H115Q, V116A, A117T, N119 Q, F120I, F120S, S121G, V122A, V122M, S126T, S126R, H129P, S130G, S132F, Q133H, E135K, F138L, T139S, C140del, S142F, I143V, I143T, N144D, Y146C, V151A, Y152C, Y152H, W153R, I1 54F, K156M, D158G, L161P, L161M, L166Q, N168Q, F172S, L173S, M175T, T190A, T190S, S192G, V193M, N194D, C198R, N201S, L203P, L203F, N207Q, L208P, V210A, S212G, D217V, I218T, I2 18N, E220G, R221G, R221I, I224V, T225A, N227K or their conservative amino acid substitutions.

In some embodiments of any of the variant ICOSL polypeptides described above, the variant ICOSL polypeptide further includes one or more amino acid deletions corresponding to position 140 of SEQ ID NO: 32.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, substitutions selected from N52Y/N57Y/F138L/L203P,

N52H/N57Y/Q100P, N52S/Y146C/Y152C, N52H/C198R, N52H/C140del/T225A, N52H/C198R/T225A,

N52H/K92R, N57Y/Q100P, N52S/C198R, N52Y/N57Y/Y152C, N52Y/N57Y/H129P/C198R,

N52H/L161P/C198R, N52S/T113E, N52S/S54P, N52K/L208P, N52S/Y152H, N52H/I143T,

N52S/R75Q/L203P, N52S/D158G, N52D/Q133H, N52H/N57Y/Q100R/V110D/C198R/S212G,

N52H/N57Y/Q100R/C198R, N52S/N194D, N52H/N57Y/Q100R/L102R/V110D/H115R/C198R, N52S/S54P, T38P/N52S/N57D, N52H/C140del/T225A, N52H/F7 8L/Q100R/C198R, N52H/N57Y/R75Q/Q100P/V110D, N52H/N57Y/L74Q/V110D/S192G, N52H/S121G/C198R, N52S/F120S/N227K,

N52S/A71T/A117T/T190A/C198R, T43A/N52H/N57Y/L74Q/D89G/V110D/F172S,

N52H/N57Y/Q100R/V110D/S132F/M175T, N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G,

N52Q/N207Q, N52Q/N168Q, N52Q/N84Q, N52Q/N119Q, N52Q/N84Q/N168Q, N52Q/N84Q/N207Q, N52Q/N119Q/N155Q, N52H/N84Q/N119Q, N52H/N84Q, N52H /N84Q/N168Q, N52H/N84Q/N207Q, N52H/N84Q/N168Q/N207Q, N52Q/N84Q/N155Q, N52Q/N84Q/N168Q, N52Q/N84Q/N155Q/N168Q,

N52Q/N84Q/N119Q/N168Q, N52Q/N84Q/N119Q/N207Q, N52Q/N84Q/N119Q/N155Q,

N52Q/N84Q/N119Q/N155Q/N207Q, N52Y/F138L/L203P, N57Y/Q100R/C198R, N57Y/F138L/L203P,

N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/Q100R/F172S/C198R,

N52H/N57Y/Q100R/H115R/F172S/C198R, N52H/N57Y/Q100R/H115R/I143V/F172S/C198R,

N52H/N57Y/Q100R/L102R/H115R/F172S/C198R, N52H/V122A/F172S/C198R,

N52H/N57Y/Q100R/H115R/F172S/N194D, N52H/N57Y/H115R/F172S/C198R,

N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/H115R, N52H/N57Y/Q100R/H115R,

7 Y/Q100P/H115R/F172S , N52Y/N57Y/Q100P/F172S,

E16V/N52H/N57Y/Q100R/V110D/H115R/C198R, E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R, N52S/E90A/H115R,

N30D/K42E/N52S/H115R, N30D/K42E/N52S/H115R/C198R/R221I, N30D/K42E/N52S/H115R/C198R,

N30D/K42E/N52S/H115R/F172S/N194D, N52S/H115R/F120S/I143V/C198R, N52S/H115R/F172S/C198R, N52H/N57Y/Q100P/C198R, N52H/N57Y/Q100P H115R/F172S/C198R, N52H /N57Y/Q100P/F172S/C198R, N52H/N57Y/Q100P/H115R, N52H/N57Y/Q100P/H115R/C198R, N52H/Q100R/C198R,

N52H/Q100R/H115R/F172S, N52H/Q100R/F172S/C198R, N52H/Q100R/H115R/F172S/C198R or

N52H/N57Y/Q100R/F172S/C198R.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, e.g., substitution selected from N52H/N57Y/Q100R/C198R, N52H/N57Y/Q100R/V122A, N52H/N57Y/Q100R/F172S, N52Y/N57Y/F138L/L203P ,

V11E/N30D/N52H/N57Y/H94E/L96I/L98F/N194D/V210A/I218T, N52H/N57Y/Q100R/L102R/V110D/H115R/C198R, N52H/N57Y/Q100R, N52H/Q100R,

N52H/N57Y/Q100R/V110D/C198R/S212G, N52H/N57Y/Q100R/L102R/V110D/H115R/C198R,

E16V/N52H/N57Y/Q100R/V110D/H115R/V152C/K156M/C198R, N30D/K42E/N52S,

N52S/F120S/I143V/I224V, N52S/E90A, N52H/N57Y/V110A/C198R/R221I, N52H/N57Y/Q100P or N52S/N194D.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, substitutions selected from N52H/N57Y/Q100R/F172S, N52H/Q100R, N52H/N57Y/Q100R/C198R. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, such as substitutions selected from E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R, N52H/N57Y/Q100R, and N52H/N57Y/Q100P.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, substitutions selected from N52H/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52H/K92R, N52H/C140del/T225A, N52H/C198R/T225A, N52H /K92R, N57Y/Q100P,

N52Y/N57Y/H129P/C198R, N52H/L161P/C198R, N52K/L208P or N52H/I143T.

In some embodiments, a variant ICOSL polypeptide exhibits increased binding affinity when bound to one of the ectodomains of CD28 or ICOS and exhibits reduced binding affinity when bound to another ectodomain of CD28 or ICOS compared to a wild-type ICOSL polypeptide or an unmodified ICOSL polypeptide, e.g., comprising the sequence presented in SEQ ID NO: 32 or 196.

In some embodiments, the variant ICOSL polypeptide exhibits increased binding affinity for ICOS and exhibits reduced binding affinity for CD28. In some embodiments, one or more additional amino acid substitutions are at a position corresponding to 52, 57, 80, 100, 130, 152, 161, or 198. In some embodiments, the variant ICOSL includes one or more amino acid substitutions selected from N52S, N52H, N52Y, N52H , N57Y, L80P, Q100P Q100R, Q100K, V110D, S130G, Y152C, L161P, L161M, C198R, R221G or their conservative amino acid substitutions. In some embodiments, the variant ICOSL polypeptide has one or more amino acid substitutions selected from N57Y/Q100P, N52S/S130G/Y152C, N52S/Y152C,

N52Y/N57Y/Y152C, N52H/L161P/C198R, N52H/L161P/C198R, N52S/L80P,

A20V/N52H/N57Y/Q100R/S109G, N52H/N57Y/R61S/Q100R/V110D/L173S,

N52H/N57Y/Q100R/V107I/V110D/S132F/1154F/C198R/R221G,

Q37R/N52H/N57Y/Q100R/V110N/S142F/C198R/D217V/R221G, N52H/N57Y/Q100R/V110D/C198R,

F27S/N52H/N57Y/V110N, S18R/N52S/F93L/I143V/R221G, A20T/N52D/Y146C/Q164L,

N52H/N57Y/H94E/L96I/F120I/S126T/W153R/I218N, N52H/N57Y/Q100R/V110D/F172S/C198R,

S25G/F27C/N52H/N57Y/Q100R/V110D/E135K/L173S/C198R, M10I/S13G/N52H/N57Y/D77G/V110A/H129P/I143V/F172S/V193M/C198R.

In some embodiments, the variant ICOSL polypeptide exhibits increased binding affinity for CD28 and exhibits reduced binding affinity for ICOS. In some embodiments, the one or more amino acid substitutions are at a position corresponding to 52, 75, or 203. In some embodiments, the variant ICOSL includes one or more amino acid substitutions selected from N52S, R75Q, L203F, or L203P. In some embodiments, the variant ICOSL polypeptide has the amino acid substitutions N52S/R75Q/L203P.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, a substitution in unmodified ICOSL or a specific binding fragment thereof corresponding to position(s) 16, 30, 42, 52, 57, 90, 100, 102, 110, 115, 120, 122, 138, 143, 152, 156, 172, 194, 198, 203, 221, or 224 relative to SEQ ID NO: 32. In some embodiments, the ICOSL variant polypeptide has one or more amino acid modifications, such as substitutions selected from E16V, N30D, K42E, N52H, N52Y, N52S, N57Y, E90A, Q100R, Q100P, L102R, V110D, H115R, F120S, V122A, F138L, I143V, I143T, N152S, K156M, F172S, N194 D, C198R, L203P, R221I or I224V. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, substitutions in the unmodified ICOSL or a specific binding fragment thereof corresponding to position(s) 115, 172, or 198, relative to SEQ ID NO: 32. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example substitutions selected from H115R, F172S or C198R. In some embodiments, one or more amino acid modifications, such as substitutions, are N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/Q100R/F172S/C198R,

N52H/N57Y/Q100R/H115R/F172S/C198R, N52H/N57Y/Q100R/H115R/I143V/F172S/C198R,

N52H/N57Y/Q100R/L102R/H115R/F172S/C198R, N52H/V122A/F172S/C198R,

N52H/N57Y/Q100R/H115R/F172S/N194D, N52H/N57Y/H115R/F172S/C198R,

N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/H115R, N52H/N57Y/Q100R/H115R,

N52H/N57Y/Q100R/H115R/F172S/I224V, N52H/N57Y/Q100R/H115R/F172S, N52H/N57Y/Q100R/F172S, N52H/Q100R/H115R/I143T/F172S, N52H/N57Y /Q100P/H115R/F172S, N52Y/N57Y/Q100P/F172S,

E16V/N52H/N57Y/Q100R/V110D/H115R/C198R, E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R, N52S/E90A/H115R, N30D/K42E N 52S/H115R, N30D/K42E /N52S/H115R/C198R/R221I, N30D/K42E/N52S/H115R/C198R,

N30D/K42E/N52S/H115R/F172S/N194D, N52S/H115R/F120S/I143V/C198R, N52S/H115R/F172S/C198R, N52H/N57Y/Q100P/C198R, N52H/N57Y/Q100P H115R/F172S/C198R, N52H /N57Y/Q100P/F172S/C198R, N52H/N57Y/Q100P/H115R, N52H/N57Y/Q100P/H115R/C198R, N52H/Q100R/C198R,

N52H/Q100R/H115R/F172S, N52H/Q100R/F172S/C198R, N52H/Q100R/H115R/F172S/C198R or

N52H/N57Y/Q100R/F172S/C198R. In some embodiments, variant ICOSL polypeptides exhibit potentially increased protein solubility or increased protein expression (solubility mutations) relative to a wild-type or unmodified ICOSL polypeptide.

In some embodiments, the variant ICOSL polypeptide includes any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 435-470. In some embodiments, the variant ICOSL polypeptide includes a polypeptide sequence that has at least 90, at least 91, at least 92, at least 93, at least 94, at least 95, such as at least 96, 97, 98 or 99% identity with any of the extracellular domains (ECD) presented in SEQ ID NO: 435-470, and includes amino acid modification(s), such as substitution(s) not present in ICOSL wild type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes a specific binding fragment of any of the extracellular domain (ECD) sequences set forth in SEQ ID NO: 435470 and includes amino acid modification(s), such as substitution(s) in wild-type ICOSL or unmodified ICOSL.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modification, for example, a substitution in the unmodified ICOSL or a specific binding fragment thereof corresponding to position(s) 52, 57, 100, 138, 198, or 203 relative to SEQ ID NO: 32. B In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, substitutions selected from N52H, N52Y, N57Y, Q100R, Q100P, F138L,

- 41 044356

C198R or L203P. In some embodiments, one or more amino acid modifications, such as substitutions, are Q100R, F138L/L203P, N52Y/F138L/L203P, N57Y/Q100R/C198R,

N57Y/F138L/L203P, N52H, N57Y, N57Y/Q100P, Q100R/F138L, or L203P.

In some embodiments, the variant ICOSL polypeptide includes any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 427-433. In some embodiments, the variant ICOSL polypeptide includes a polypeptide sequence that has at least 90, at least 91, at least 92, at least 93, at least 94, at least 95, such as at least 96, 97, 98 or 99% identity with any of the extracellular domains (ECD) presented in SEQ ID NO: 427-433, and includes amino acid modification(s), such as substitution(s), that are missing in wild-type ICOSL or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes a specific binding fragment of any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 427-433 and containing amino acid modification(s), such as substitution(s) in wild-type ICOSL or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes an IgV sequence set forth in SEQ ID NO: 434. In some embodiments, the variant ICOSL polypeptide includes a polypeptide sequence that has at least 90, at least 91, at least 92, at least 93, at least 94, at least 95, such as at least 96, 97, 98 or 99% identity with the IgV sequence set forth in SEQ ID NO: 434, and includes an amino acid modification ( ii), for example, substitution(s) not present in wild-type ICOSL or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises the specific binding fragment of the IgV sequence set forth in SEQ ID NO: 434 and includes amino acid substitution(s) not present in wild-type or unmodified ICOSL.

In some embodiments, the variant ICOSL polypeptide has one or more amino acid modification, for example, a substitution in the unmodified ICOSL or a specific binding fragment thereof corresponding to position(s) 52, 84, 91, 119, 155, 168, 207 relative to SEQ ID NO: 32 In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications, for example, substitutions selected from A91S, N52H, N52Q, N84Q, N119Q, N155H, N155Q, N168Q, N207Q. In some embodiments, one or more amino acid modifications, such as substitutions, are N84Q, N119Q, N168Q, N207Q, N52Q, N52Q/N207Q, N168Q/N207Q, N52Q/N168Q, N84Q/N207Q, N155Q/N207Q, N119Q/N168Q, N119Q / N207Q, N119Q/N155Q, N52Q/N84Q, N52Q/N119Q, N84Q/N119Q, N52Q/N84Q/N168Q, N52Q/N84Q/N207Q, N84Q/N155Q/N168Q,

N84Q/N168Q/N207Q, N84Q/N155H/N207Q, N155Q/N168Q/N207Q, N119Q N155Q/N168Q,

N119Q/N168Q/N207Q, N84Q/N119Q/N207Q, N119Q/N155H/N207Q, N84Q/N119Q/N155Q,

N52Q/N119Q/N155Q, N52H/N84Q/N119Q, N52H/N84Q, N52H/N84Q/N168Q, N52H/N84Q/N207Q, N52H/N84Q/N168Q/N207Q, N52Q/N84Q/N155Q, N52Q/N84Q /N168Q, N52Q/ N84Q/N155Q/N168Q,

N52Q/N84Q/N119Q/N168Q, N84Q/N119Q/N155Q/N168Q, N84Q/N155Q/N168Q/N207Q,

N84Q/N119Q/N155Q/N207Q, N52Q/N84Q/N119Q/N207Q, N52Q/N84Q/N119Q/N155Q,

N52Q/N84Q/N119Q/N155Q/N207Q, N84Q/N119Q/N155Q/N168Q/N207Q or A91S/N119Q/N168Q/N207Q. In some embodiments, variant ICOSL polypeptides exhibit potentially reduced glycosylation relative to a wild-type or unmodified ICOSL polypeptide.

In some embodiments, the variant ICOSL polypeptide includes any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 387-424, 427-433, 435-470. In some embodiments, the variant ICOSL polypeptide includes a polypeptide sequence that has at least 90, at least 91, at least 92, at least 93, at least 94, at least 95, such as at least 96, 97, 98 or 99% identity with any of the extracellular domains (ECD) presented in SEQ ID NO: 387-424, 427-433, 435-470 and containing amino acid modification(s), e.g. , substitution(s) not present in wild-type ICOSL or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes a specific binding fragment of any extracellular domain (ECD) sequence set forth in SEQ ID NOs: 387-424, 427-433, 435-470 and containing amino acid modification(s), such as substitution(s) missing in wild-type ICOSL or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide includes any of the IgV sequences set forth in SEQ ID NOs: 425-426. In some embodiments, the variant ICOSL polypeptide includes a polypeptide sequence that has at least 90, at least 91, at least 92, at least 93, at least 94, at least 95, such as at least 96, 97, 98 or 99% identity with any of the IgV sequences specified in SEQ ID NO: 425-426, and includes amino acid modification(s), for example substitution(s), not present in the wild ICOSL type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises a specific binding fragment of any of the IgV sequences set forth in SEQ ID NO: 425-426 and includes amino acid substitution(s) not present in di-type ICOSL or unmodified ICOSL.

III. Variant polypeptide formats.

The immunomodulatory polypeptide containing the variant ICOSL provided herein, which contains vIgD, can be formatted in various forms, including as a soluble protein, a membrane-bound protein, a secreted protein, a conjugate or fusion protein, or for expression with using an infectious agent. In some embodiments, a particular format may be selected for the therapeutic use sought. In some cases, an immunomodulatory polypeptide containing a variant ICOSL polypeptide is provided in a format to antagonize or block the activity of its cognate binding partner, for example, CD28. In some embodiments, CD28 antagonism may be useful for treating inflammation or autoimmunity. In some cases, an immunomodulatory polypeptide comprising a variant ICOSL polypeptide is provided in a form to agonize or stimulate the activity of its cognate binding partner, for example, CD28. In some embodiments, CD28 agonism may be useful for the treatment of cancer. One skilled in the art can readily determine the activity of a particular format, for example, for antagonism or agonism of one or more specific cognate binding partners. Exemplary methods for assessing such activities are provided throughout this document, including examples.

In some aspects, the invention provides immunomodulatory proteins containing vIgD from ICOSL, where such proteins are soluble, for example, fused to an Fc chain. In some aspects, one or more additional IgSF domains, such as one or more additional vIgDs, may be linked to a vIgD from an ICOSL as defined herein (hereinafter referred to as a stack or stacked immunomodulatory protein). In some embodiments, the modular format of the proposed immunomodulatory proteins provides flexibility in the design or production of immunomodulatory proteins to modulate the activity of multiple counterstructures (multiple cognate binding partners). In some embodiments, such stacking molecules may be presented in a soluble format or, in some cases, may be presented as membrane-associated or secreted proteins. In some embodiments, the variant immunomodulatory protein ICOSL is provided as a conjugate wherein there is direct or indirect binding of ICOSL to a targeting agent or functional moiety, e.g., an antibody or other binding molecules that specifically bind to a ligand, e.g., an antigen, e.g., for targeting or localization of vIgD in a specific environment or cell, for example, when administered to an object. In some embodiments, a targeting agent, such as an antibody or other binding molecule, binds to a tumor antigen, thereby localizing the variant ICOSL containing vIgD to the tumor microenvironment, for example, to modulate the activity of tumor infiltrating lymphocytes (TILs) specific to the tumor microenvironment.

In some embodiments, the proposed immunomodulatory proteins are expressed in cells and are offered as part of an engineered cell therapy (ECT). In some embodiments, the variant ICOSL polypeptide is expressed in a cell, such as an immune cell (eg, a T cell or antigen presenting cell), in a membrane-bound form, thereby providing a transmembrane immunomodulatory protein (hereinafter also referred to as TIP). In some embodiments, depending on the cognate binding partner recognized by TIP, engineered cells expressing TIP can have an agonistic effect on the cognate binding partner by providing a costimulatory signal, positive or negative, to other engineered cells and/or endogenous T cells. In some aspects, the variant ICOSL polypeptide is expressed in a cell, such as an immune cell (e.g., a T cell or antigen presenting cell), in a secreted form, thereby producing a secreted or soluble form of the variant ICOSL polypeptide (hereinafter also referred to as SIP), for example , when cells are injected into an object. In some aspects, SIP may have an antagonistic effect on a cognate binding partner in the environment (eg, tumor microenvironment) in which it is secreted. In some embodiments, the variant ICOSL polypeptide is expressed in an infectious agent (e.g., a viral or bacterial agent) that, when administered to a subject, is capable of infecting a cell in vivo, such as an immune cell (e.g., a T cell or antigen presenting cell) or tumor, for delivery or expression variant polypeptide as TIP or SIP in the cell.

In some embodiments, a soluble immunomodulatory polypeptide, such as a variant ICOSL containing vIgD, can be encapsulated in a liposome, which itself can be conjugated to any one or any combination of provided conjugates (eg, a targeting functional moiety). In some embodiments, the soluble or membrane-bound immunomodulatory polypeptides of the invention are deglycosylated. In more specific embodiments, the variant ICOSL sequence is deglycosylated. In even more specific embodiments, the IgG and/or IgC (eg, IgC2) domain or domains of the ICOSL variant are deglycosylated.

- 43 044356

Non-limiting examples of proposed formats are described in FIGS. 13A-13C and further described below.

A. Soluble protein.

In some embodiments of the invention, the immunomodulatory protein comprising the ICOSL variant polypeptide is a soluble protein. Those skilled in the art will appreciate that cell surface proteins typically have an intracellular, transmembrane, and extracellular domain (ECD) and that a soluble form of such proteins can be produced using the extracellular domain or an immunologically active subsequence thereof. Thus, in some embodiments, the immunomodulatory protein comprising the ICOSL variant polypeptide lacks a transmembrane domain or a portion of a transmembrane domain. In some embodiments, the immunomodulatory protein comprising the variant ICOSL does not have an intracellular (cytoplasmic) domain or part of an intracellular domain. In some embodiments, the immunomodulatory protein comprising the variant ICOSL polypeptide comprises only the vIgD portion containing the ECD domain, or a portion thereof containing the IgV and/or IgC domain (e.g., IgC2) or domains or specific binding fragments thereof containing amino acid modification(s) .

In some embodiments, the immunomodulatory polypeptide containing a variant ICOSL may include one or more variant ICOSL polypeptides of the invention. In some embodiments, a polypeptide of the invention will contain exactly 1, 2, 3, 4, 5 variant ICOSL sequences. In some embodiments, at least two of the variant ICOSL sequences are identical variant ICOSL sequences.

In some embodiments, the present invention is an immunomodulatory polypeptide that includes two or more vIgD ICOSL sequences. The plurality of ICOSL variant polypeptides within a polypeptide chain may be identical (ie, the same) to each other, or may be non-identical (ie, different types) of ICOSL variant sequences. In addition to single polypeptide chain embodiments, in some embodiments two, three, four or more polypeptides of the invention may be covalently or non-covalently attached to each other. Thus, monomeric, dimeric, and higher order (eg, 3, 4, 5 or more) multimeric proteins are provided herein. For example, in some embodiments, exactly two polypeptides of the invention may be covalently or non-covalently attached to each other to form a dimer. In some embodiments, the coupling is accomplished through interchain disulfide bonds. Compositions containing two or more polypeptides of the invention may have identical or substantially identical polypeptide species (eg, homodimers) or non-identical polypeptide species (eg, heterodimers). A composition having a plurality of linked polypeptides of the invention may, as noted above, have one or more identical or non-identical variant ICOSL polypeptides of the invention in each polypeptide chain.

In some embodiments of the invention, the immunomodulatory protein comprises a variant ICOSL polypeptide fused to an immunoglobulin Fc (resulting in an immunomodulatory Fc Fusion protein, such as a variant ICOSL-Fc fusion protein, also referred to as an ICOSL vIgD-Fc fusion protein). In some embodiments, the ICOSL variant polypeptide attachment is at the N-terminus of the Fc. In some embodiments, the ICOSL variant polypeptide attachment is at the C-terminus of the Fc. In some embodiments, two or more variant ICOSL polypeptides (same or different) are independently linked at the N-terminus and at the C-terminus.

In some embodiments, the Fc is a murine or human Fc. In some embodiments, the Fc is a mammalian or human IgG1, IgG2, IgG3, or IgG4 Fc region. In some embodiments, the Fc is derived from IgG1, such as human IgG1. In some embodiments, the Fc comprises the amino acid sequence set forth in SEQ ID NO: 226, or an amino acid sequence that is at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96 , 97, 98, 99% or more sequence identity to SEQ ID NO: 226.

In some embodiments, the Fc region includes more than one modification to alter (eg, reduce) one or more of its normal functions. In general, the Fc region is responsible for effector functions such as complement-dependent cytotoxicity (CDC) and antibody-dependent cellular cytotoxicity (ADCC), in addition to antigen-binding capacity, which is the main function of immunoglobulins. In addition, the FcRn sequence present in the Fc region plays a role in regulating serum IgG levels by increasing the half-life in vivo through conjugation with the FcRn receptor in vivo. In some embodiments, such functions may be reduced or altered in the Fc for use with the provided Fc fusion proteins.

In some embodiments, one or more amino acid modifications may be introduced into the Fc region of the variant ICOSL-Fc fusion protein provided herein, thereby forming a variant Fc region. In some embodiments, the variant Fc region reduces effector function. There are many examples of changes or mutations in Fc sequences that can alter effector function. For example, WO 00/42072, WO 2006019447, WO 2012125850, WO

- 44 044356

2015/107026, US 2016/0017041 and Shields et al. J Biol. Chem. 9(2): 6591-6604 (2001) describe exemplary Fc variants with improved or reduced binding to FcR. The contents of these publications are expressly incorporated herein by reference.

In some embodiments of the present invention, the proposed variant ICOSL-Fc fusion proteins contain an Fc region that exhibits reduced effector functions, making them a desirable candidate for applications in which the in vivo half-life of the variant ICOSL-Fc fusion protein and certain effector functions are important. such as CDC and ADCC) are unnecessary or harmful.

In vitro and/or in vivo cytotoxicity assays can be performed to confirm reduction/depletion of CDC and/or ADCC activity. For example, Fc receptor (FcR) binding assays can be performed to ensure that the variant ICOSL-Fc fusion protein has lost its FcyR binding ability (thus likely lacking ADCC activity) but retains FcRn binding ability. The primary cells mediating ADCC, NK cells, express only FcyRIII, whereas monocytes express FcyRI, FcyRII, and FcyRIII. The expression of FcR on hematopoietic cells is shown in Table. 3 at page 464 Ravetch and Kinet, Annu. Rev. Immunol. 9:457-492 (1991). Non-limiting examples of in vitro assays for assessing the ADCC activity of a molecule of interest are described in US Pat. No. 5,500,362 (see, e.g., Hellstrom, I. et al. Proc. Nat'l Acad. Sci. USA 83:7059-7063 (1986)) and Hellstrom, I et al., Proc. Nat'l Acad. Sci. USA, 82: 1499-1502 (1985); US Patent No. 5,821,337 (see Bruggemann, M. et al, J. Exp., Med., 166: 1351-1361 (1987)). As an alternative, non-radioactive assay methods can be used (see, for example, the ACTI™ Non-Radioactive Cytotoxicity Assay for Flow Cytometry (CellTechnology, Inc. Mountain View, CA) and the CytoTox 96™ Non-Radioactive Cytotoxicity Assay (Promega, Madison, WI). Useful effector cells for such assays include peripheral blood mononuclear cells (PBMCs) and natural killer (NK) cells.Alternatively or additionally, the ADCC activity of the molecule of interest can be assessed in vivo, for example, in an animal model such as that described in dynes et al Proc Nat'l Acad Sci USA 95:652-656 (1998) C1q binding assays can also be performed to confirm that the variant ICOSL-Fc fusion protein cannot bind C1q and therefore does not have CDC activity See, for example, the ELISA to determine the binding of C1q and C3c in WO 2006/029879 and WO 2005/100402. To assess complement activation, a CDC assay can be performed (see, for example, Gazzano-Santoro et al., J. Immunol. Methods 202:163 (1996); Cragg, M. S. et al., Blood 101:1045-1052 (2003); and Cragg, M. S. and M. J. Glennie, Blood 103:2738-2743 (2004)). FcRn binding and determination of clearance/half-life in vivo can also be performed using methods known in the art (see, for example, Petkova, SB et al., Int'l. Immunol. 18(12): 1759-1769 (2006 )).

Variant ICOSL-Fc fusion proteins with reduced effector function include those with one or more residues in the Fc region of EU numbering 238, 265, 269, 270, 297, 327, and 329 replaced (US Pat. No. 6,737,056). Such Fc mutants include Fc mutants with substitutions at two or more amino acid positions from EU numbering 265, 269, 270, 297 and 327, including the so-called DANA Fc mutant with residues 265 and 297 replaced by alanine (US Pat. No. 7332581).

In some embodiments, the Fc region of variant ICOSL-Fc fusion proteins has an Fc region in which any one or more of amino acids at positions 234, 235, 236, 237, 238, 239, 270, 297, 298, 325, and 329 (denoted by EU numbering) are replaced by different amino acids compared to the native Fc region. Such Fc region changes are not limited to the changes described above and include, for example, changes such as deglycosylated chains (N297A and N297Q), IgG1-N297G, IgG1-L234A/L235A, IgG1-L234A/L235E/G237A, IgG1-A325A/A330S /P331S, IgG1-C226S/C229S, IgG1C226S/C229S/E233P/L234V/L235A, IgG1-E233P/L234V/L235A/G236del/S267K, IgG1-L234F/L235E/P331S, IgG1-S267E/L328F , IgG2-V234A/G237A , IgG2-H268Q/V309L/A330S/A331S, IgG4-L235A/G237A/E318A and IgG4-L236E, described in Current Opinion in Biotechnology (2009) 20 (6), 685-691; such as G236R/L328R, L235G/G236R, N325A/L328R and N325L/L328R described in WO 2008/092117; amino acid insertions at positions 233, 234, 235 and 237 (designated by EU numbering); and changes to the sites described in WO 2000/042072.

Several Fc variants with increased or decreased binding to FcR have been described. (see, for example, US Pat. No. 6,737,056, WO 2004/056312, WO 2006019447 and Shields et al., J. Biol. Chem. 9(2): 65916604 (2001)).

In some embodiments of the invention, a variant of the synthesis of ICOSL-Fc is provided that includes a variant Fc region containing one or more amino acid substitutions that increase half-life and/or improve binding to the neonatal Fc receptor (FcRn). Antibodies with increased half-life and improved binding to FcRn are described in US 2005/0014934 A1 (Hinton et al.) or WO 2015107026. These antibodies contain an Fc region with one or more substitutions that improve binding of the Fc region to FcRn. Such Fc variants include those that have substitutions at one or more residues of the Fc region: 238, 256, 265, 272, 286, 303, 305, 307, 311, 312,

- 45 044356

317, 340, 356, 360, 362, 376, 378, 380, 382, 413, 424 or 434 according to EU numbering, for example, substitution of residue 434 in the Fc region (US Pat. No. 7371826).

In some embodiments, the Fc region of the variant ICOSL-Fc fusion protein includes one or more amino acid substitutions E356D and M358L. In some embodiments, the Fc region of the variant ICOSL-Fc fusion protein includes one or more amino acid substitutions C220S, C226S, C229S. In some embodiments, the Fc region of the variant ICOSL fusion protein includes one or more amino acid substitutions R292C and V302C. See also Duncan & Winter, Nature 322: 738-40 (1988); Pat. US No. 5648260; Pat. US No. 5624821; and WO 94/29351 regarding other examples of Fc region variants.

In some embodiments, changes are introduced to Fc regions that result in decreased binding to C1q and/or decreased complement dependent cytotoxicity (CDC), for example, as described in US Pat. No. 6,195,551, WO 99/51642 and Idusogie et al. J. Immunol. 164: 4178-4184 (2000).

In some embodiments of the invention, a variant ICOSL-Fc fusion protein is provided, comprising a variant Fc region containing one or more amino acid modifications, wherein the variant Fc region is a derivative of IgG1, such as human IgG1. In some embodiments, the variant Fc region is derived from the amino acid sequence set forth in SEQ ID NO: 226. In some embodiments, the Fc includes at least one amino acid substitution that is N82G as numbered in SEQ ID NO: 226 (corresponding to N297G in according to EU numbering). In some embodiments, the Fc further includes at least one amino acid substitution that is R77C or V87C according to SEQ ID NO: 226 (corresponding to R292C or V302C according to EU numbering). In some embodiments, the Fc region variant further includes a C5S amino acid modification at SEQ ID NO: 226 (corresponding to EU numbering C220S). For example, in some embodiments, the Fc region variant includes the following amino acid modifications: N82G and one or more of the following amino acid modifications C5S, R77C, or V87C relative to SEQ ID NO: 226.

In some embodiments, an ICOSL-Fc synthesis variant is provided comprising a variant Fc region, wherein the Fc variant comprises an amino acid sequence set forth in any of SEQ ID NO: 474, 476, 477, 478, or 507 or an amino acid sequence that has at least At least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% or more sequence identity with any of SEQ ID NO: 474, 476, 477, 478 or 507.

In some embodiments, the Fc is derived from IgG2, such as human IgG2. In some embodiments, the Fc comprises the amino acid sequence set forth in SEQ ID NO: 227, or an amino acid sequence that is at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96 , 97, 98, 99% or more sequence identity to SEQ ID NO: 227.

In some embodiments, the Fc comprises the amino acid sequence set forth in SEQ ID NO: 505 or an amino acid sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97 98, 99% or more sequence identity to SEQ ID NO: 505. In some embodiments, the IgG4 Fc is a stabilized Fc in which the CH3 domain of human IgG4 is replaced by the CH3 domain of human IgG1 and which has inhibited aggregate formation, an antibody in which the CH3 domains and CH2 of human IgG4 are replaced by the CH3 and CH2 domains of human IgG1, respectively, or an antibody in which arginine at position 409 according to the EC index proposed by Kabat et al. for human IgG4, is replaced by lysine, and which exhibits inhibited aggregate formation (see, for example, US Pat. No. 8911726). In some embodiments, the Fc is an IgG4 containing the S228P mutation, which has been shown to prevent recombination between a therapeutic antibody and endogenous IgG4 by exchanging Fab arms (see, e.g., Labrijin et al. (2009) Nat. Biotechnol., 27 (8)767-71.) In some embodiments, the Fc comprises the amino acid sequence set forth in SEQ ID NO: 506, or an amino acid sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92 , 93, 94, 95, 96, 97, 98, 99% or greater sequence identity to SEQ ID NO: 506.

In some embodiments, the variant ICOSL polypeptide is directly linked to an Fc sequence. In some embodiments, the variant ICOSL polypeptide is indirectly linked to an Fc sequence, for example, through a linker. In some embodiments, one or more peptide linkers link the variant ICOSL polypeptide and the Fc domain. In some embodiments, the peptide linker may be a single amino acid residue or may be longer. In some embodiments, the peptide linker has at least one amino acid residue, but not more than 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2 or 1 amino acid residue. In some embodiments, the linker is (in single letter amino acid code): GGGGS (4GS) or 4GS linker multimers, such as 2, 3, 4 or 5 4GS linker repeats.

In some embodiments, the variant ICOSL-Fc fusion protein is a dimer formed by two variant ICOSL-Fc polypeptides linked to an Fc domain. In some specific embodiments, identical or substantially identical species (subject to 3 or fewer amino acid sequence differences at the N-terminus or C-terminus) of variant ICOSL-Fc fusion polypeptides will dimerize to create a homodimer. In some embodiments, the dimer

- 46 044356 is a homodimer in which the two variant ICOSL-Fc polypeptides are the same. Alternatively, different species of ICOSL-Fc fusion polypeptides can be dimerized to form a heterodimer. Thus, in some embodiments, the dimer is a heterodimer in which the two variant ICOSL-Fc polypeptides are different.

In addition, the present invention is a nucleotide molecule encoding a variant ICOSL-Fc fusion protein. In some embodiments, to produce an Fc fusion protein, a nucleotide molecule encoding an ICOSL-Fc fusion protein is inserted into an appropriate expression vector. The resulting variant ICOSL-Fc fusion protein can be expressed in host cells transformed with an expression construct where assembly between Fc domains occurs via interchain disulfide bonds formed between the Fc moieties to produce a dimer such as bivalent variant ICOSL-Fc fusion proteins .

The resulting Fc fusion proteins can be easily purified by affinity chromatography on Protein A or Protein G columns. Additional purification steps may be required to obtain heterodimers. For example, when two nucleic acids encoding different variant ICOSL polypeptides are transformed into cells, the formation of heterodimers must be achieved biochemically, since the variant ICOSL molecules bearing the Fc domain will also be expressed as disulfide homodimers. Thus, homodimers can be reduced under conditions that favor the destruction of interchain disulfides but do not affect intrachain disulfides. In some cases, different ICOSL-Fc monomers are mixed in equimolar amounts and oxidized to form a mixture of homo- and heterodimers. The components of this mixture are separated by chromatographic methods. Alternatively, the formation of this type of heterodimer can be influenced by genetic engineering and expression of Fc fusion molecules that contain the variant ICOSL polypeptide using the protrusion-in-convex methods described below.

B. Stacked molecules with additional IgSF domains.

In some embodiments, the immunomodulatory proteins may comprise any of the variant ICOSL polypeptides provided herein that is linked, directly or indirectly, to one or more other immunoglobulin superfamily (IgSF) domains (an immunomodulatory protein stack construct and also referred to as a type II immunomodulatory protein). In some aspects, this can create unique multi-domain immunomodulatory proteins that bind two or more of these, eg, three or more, cognate binding partners, thereby providing multi-targeted modulation of the immune synapse.

In some embodiments, the immunomodulatory protein includes a combination (non-wild-type combination) and/or arrangement (non-wild-type arrangement or non-wild-type permutation) of a variant ICOSL domain with one or more sequences of other affinity-modified and/or non-affinity-modified IgSF domains of another a member of the IgSF family (eg, a member of the mammalian IgSF family) that are not found in wild-type IgSF family members. In some embodiments, the immunomodulatory protein includes 2, 3, 4, 5, or 6 immunoglobulin superfamily (IgSF) domains, wherein at least one of the IgSF domains is a variant ICOSL IgSF domain (vIgD ICOSL) as provided herein.

In some embodiments, the additional IgSF domain sequences may be a modified IgSF domain that includes one or more amino acid substitution modifications, for example, compared to a wild-type IgSF domain or an unmodified IgSF domain. In some embodiments, the IgSF domain may be unaffinity modified (eg, wild type) or affinity modified. In some embodiments, the unmodified IgSF domain or wild-type IgSF domain may be derived from mouse, rat, cynomolgus or human, or combinations thereof. In some embodiments, the additional IgSF domains may be an IgSF domain of a member of the IgSF family described in Table. 2. In some embodiments, the additional IgSF domain may be an affinity modified IgSF domain containing one or more amino acid modifications, such as substitutions, compared to the IgSF domain contained in the IgSF family member described in Table. 2.

In some embodiments, the additional IgSF domain is an affinity-modified or unmodified IgSF domain contained in a member of the IgSF family selected from the signal-regulatory protein (SIRP) family, the initiating receptor-on-myeloid cell-like protein (TREML) family, the cell adhesion molecule-associated protein family carcinoembryonic antigen (CEACAM), sialic acid-binding Ig-like lectin (SIGLEC) family, butyrophilin family, B7 family, CD28 family, V-set and immunoglobulin domain-containing protein (VSIG) family, V-set transmembrane domain (VSTM) family ), major histocompatibility complex (MHC) family, lymphocyte signaling activating molecule (SLAM) family, leukocyte immunoglobulin-like receptor (LIR), nectin family (Nec), nectin-like protein family (NECL), poliovirus receptor-related protein (PVR) family natural cytotoxicity triggering receptor (NCR), T-cell immunoglobulin and mucin (TIM) family, or killer cell immunoglobulin-like receptor family

- 47 044356 (KIR). In some embodiments, the additional IgSF domains are independently derived from an IgSF protein selected from the group consisting of CD80 (B7-1), CD86 (B7-2), CD274 (PD-L1, B7-H1), PDCD1LG2 (PDL2, CD273), ICOSLG (B7RP1, CD275, ICOSL, B7-H2), CD276 (B7-I3), VTCN1 (B7-H4), CD28, CTLA4, PDCD1 (PD-1), ICOS, BTLA (CD272) CD4, CD8A (CD8- alpha), CD8B (CD8-beta), LAG3, HAVCR2 (TIM3), CEACAM1, TIGIT, PVR (CD155), PVRL2 (CD112), CD226, CD2, CD160, CD200, CD200R1 (CD200R) and NC R3 (NKp30).

In the first column of the table. 2 indicates the name and, optionally, the name of some possible synonyms for this particular IgSF member. The second column includes the protein ID in the UniProtKB database, a public database accessible online at uniprot.org, or in some cases the GenBank number. Universal Protein Resource (UniProt) is a comprehensive resource for protein sequences and annotation data. UniProt databases include UniProt Knowledgebase (UniProtKB). UniProt is a collaboration between the European Bioinformatics Institute (EMBL-EBI), the Swiss Institute of Bioinformatics SIB and Protein Information Resource (PIR) and is supported primarily by a grant from the US National Institutes of Health (NIH). GenBank is the NIH genetic sequence database, an annotated collection of all publicly available DNA sequences (Nucleic Acids Research, 2013 Jan;41(D1):D36-42). The third column includes the region in which the specified IgSF domain is located. The region is specified as a range where the domain is located between the residues that define the range. Column 3 also indicates the IgSF domain class for the specified IgSF region. Column 4 provides the region in which these additional domains are located (signal peptide, S; extracellular domain, E; transmembrane domain, T; cytoplasmic domain, C). It is understood that the description of domains may vary depending on the methods used to identify or classify the domain, and may be identified differently from different sources. Description of residues corresponding to the domain in Table. 2 is provided by way of example only and may be several amino acids (eg, one, two, three, or four) longer or shorter. Column 5 indicates, for some of the listed IgSF members, some of their cognate cell surface binding partners.

- 48 044356

table 2

Representatives of IgSF according to the present invention

Representative IgSF (Synonyms) Protein identifier UniProtKB IgSF Region & Domain Class Other domains Cognate cell surface binding partners Amino acid subsequence IgSF representative (SEQ ID NO) Precursor (mature residues) Mature ECD CD80 (B7-1) NP_00518 2.1 P33681 35-135, 35138 or 37138 IgV, 145-230 or 154-232 IgC S: 1-34, E: 35242, T: 243-263, C: 264288 CD28, CTLA4, PDL1 SEQ ID NO: 1 (35-288) SEQ ID NO: 253 SEQ ID NO: 28 CD86 (B7-2) R42081.2 33-131 IgV, 150-225 IgC2 S: 1-23, E: 24- 247, T: 248-268, C: 269329 CD28, CTLA4 SEQ ID NO: 2 (24-329) SEQ ID NO: 254 SEQ ID NO: 29 CD274 (PD-L1, B7-H1) Q9NZQ7.1 24-130 IgV, 133-225 IgC2 S: 1-18, E: 19- 238, T: 239-259, C: 260290 PD-1, B7-1 SEQ ID NO: 3 (19-290) SEQ ID NO: 255 SEQ ID NO: 30 PDCD1 LG2 (PD-L2, CD273) Q9BQ51.2 21-118 IgV, 122-203 IgC2 S: 1-19, E: 20220, T: 221-241, C: 242273 PD-1, RGMb SEQ ID NO: 4 (20-273) SEQ ID NO: 256 SEQ ID NO: 31 ICOSLG (B7RP1, CD275, ICOSL, B7-H2) 075144.2 19-129 IgV, 141-227 IgC2 S: 1-18, E: 19- 256, T: 257-277, C: 278302 ICOS, CD28, CTLA4 SEQ ID NO: 5 (19-302) SEQ ID NO: 257 SEQ ID NO: 32 CD276 (B7-NZ) Q5ZPR3.1 29-139 IgV, 145-238 IgC2, 243-357 IgV, 367- 453 IgC S: 1-28, E: 29- 466, T: 467-487, C: 488534 SEQ ID NO: 6 (29-534) SEQ ID NO: 258 SEQ ID NO: 33 VTCN1 (B7-H4) Q7Z7D3.1 35-146 IgV, 153-241 IgV S: 1-24, E: 25259, T: 260-280, C: 281282 SEQ ID NO: 7 (25-282) SEQ ID NO: 259 SEQ ID NO: 34 CD28 R10747.1 28-137 IgV S: 1-18, E: 19- B7-1, B7-2, B7RP1 SEQ ID NO: 8 SEQ ID NO: 260 SEQ ID NO: 35

- 49 044356

152, T: 153-179, C: 180- 220 (19-220) CTLA4 P16410.3 39-140 IgV S: 1-35, E: 36- 161, T: 162-182, C: 183223 B7-1, B7-2, B7RP1 SEQ ID NO: 9 (36-223) SEQ ID NO: 261 SEQ ID NO: 36 PDCD1 (PD-1) Q15116.3 35-145 IgV S: 1-20, E: 21- 170, T: 171-191, C: 192288 PD-L1, PD- L2 SEQ ID NO: 10 (21-288) SEQ ID NO: 262 SEQ ID NO: 37 ICOS Q9Y6W8. 1 30-132 IgV S: 1-20, E: 21- 140, T: 141-161, C: 162199 B7RP1 SEQ ID NO: 11 (21-199) SEQ ID NO: 263 SEQ ID NO: 38 BTLA (CD272) Q7Z6A9.3 31-132 IgV S: 1-30, E: 31- 157, T: 158-178, C: 179289 HVEM SEQ ID NO: 12 (31-289) SEQ ID NO: 264 SEQ ID NO: 39 CD4 P01730.1 26-125 IgV, 126-203 IgC2, 204317 IgC2, 317-389 IgC2 S: 1-25, E: 26- 396, T: 397-418, C: 419458 MHC class II SEQ ID NO: 13 (26-458) SEQ ID NO: 265 SEQ ID NO: 40 CD8A (CDSalpha) P01732.1 22-135 IgV S: 1-21, E: 22- 182, T: 183-203, C: 204235 MHC class I SEQ ID NO: 14 (22-235) SEQ ID NO: 266 SEQ ID NO: 41 CD8B (CDSbeta) P10966.1 22-132 IgV S: 1-21, E: 22- 170, T: 171-191, C: 192- MHC class I SEQ ID NO: 15 (22-210) SEQ ID NO: 267 SEQ ID NO: 42

- 50 044356

210 LAG3 P18627.5 37-167 IgV, 168-252 IgC2, 265-343 IgC2, 349- 419 IgC2 S: 1-28, E: 29- 450, T: 451-471, C: 472525 MNS class II SEQ ID NO: 16 (29-525) SEQ ID NO: 268 SEQ ID NO: 43 HAVCR 2 (TIM-3) Q8TDQ0.3 22-124 IgV S: 1-21, E: 22- 202, T: 203-223, C: 224301 CEACAM-1, phosphatidyl serine, G alectin-9, HMGB1 SEQ ID NO: 17 (22-301) SEQ ID NO: 269 SEQ ID NO: 44 CEACA Ml P13688.2 35-142 IgV, 145-232 IgC2, 237- 317 IgC2, 323-413 IgC S: 1-34, E: 35- 428, T: 429-452, C: 453526 TIM-3 SEQ ID NO: 18 (35-526) SEQ ID NO: 270 SEQ ID NO: 45 TIGIT Q495A1.1 22-124 IgV S: 1-21, E: 22- 141, T: 142-162, C: 163244 CD155, CD112 SEQ ID NO: 19 (22-244) SEQ ID NO: 271 SEQ ID NO: 46 PVR (CD 15 5) P15151.2 24-139 IgV, 145-237 IgC2, 244- 328 IgC2 S: 1-20, E: 21- 343, T: 344-367, C: 368417 TIGIT, CD226, CD96, poliovirus SEQ ID NO: 20 (21-417) SEQ ID NO: 272 SEQ ID NO: 47 PVRL2 (CD 112) Q92692.1 32-156 IgV, 162-256 IgC2, 261- 345 IgC2 S: 1-31, E: 32- 360, T: 361-381, C: 382538 TIGIT, CD226, CD112R SEQ ID NO: 21 (32-538) SEQ ID NO: 273 SEQ ID NO: 48 CD226 Q15762.2 19-126 IgC2, 135- 239 IgC2 S: 1-18, E: 19- 254, T: 255-275, C: 276336 CD155, CD112 SEQ ID NO: 22 (19-336) SEQ ID NO: 274 SEQ ID NO: 49 CD2 P06729.2 25-128 IgV, S: 1-24, CD58 SEQ ID NO: SEQ ID SEQ ID 129-209 IgC2 E: 25- 209, T: 210-235, C: 236351 23 (25-351) NO: 275 NO: 50 CD 160 095971.1 27-122 IgV S: 1-26 E: 27-122 HVEM, MHC family of proteins SEQ ID NO: 24 (27-159) SEQ ID NO: 276 SEQ ID NO: 51 CD200 P41217.4 31-141 IgV, 142-232 IgC2 S: 1-30, E: 31- 232, T: 233-259, C: 260278 CD200R SEQ ID NO: 25 (31-278) SEQ ID NO: 277 SEQ ID NO: 52 CD200R 1 (CD200 R) Q8TD46.2 53-139 IgV, 140-228 IgC2 S: 1-28, E: 29- 243, T: 244-264, C: 265325 CD200 SEQ ID NO: 26 (29-325) SEQ ID NO: 278 SEQ ID NO: 53 NCR3 (NKp30) 014931.1 19-126 IgClike S: 1-18, E: 19- 135, T: 136-156, C: 157201 B7-H6 SEQ ID NO:27 (19-201) SEQ ID NO: 279 SEQ ID NO: 54 VSIG8 Q5VU13 22-141 IgV 1 146-257 IgV 2 S: 1-21 E: 22-263 T: 264- 284 C: 285- 414 VISTA SEQ ID NO: 341 (22-414) SEQ ID NO: 342 SEQ ID NO: 343

- 51 044356

In some embodiments of the present invention, the provided immunomodulatory proteins, in addition to the variant ICOSL polypeptide, also contain at least 2, 3, 4, 5 or 6 additional immunoglobulin superfamily (IgSF) domains, such as an IgD domain from a member of the IgSF family indicated in table 2. In some embodiments, the provided immunomodulatory proteins comprise at least one additional IgSF domain (e.g., a second IgSF domain), wherein the at least one additional or second IgSF domain is an IgSF domain specified in a wild-type IgSF domain or an unmodified or specific binding fragment contained in the amino acid sequence specified in any of SEQ ID NOs: 1-27 and 341. In some embodiments, the wild-type or unmodified IgSF domain is an IgV domain or an IgC domain, such as an IgC1 or IgC2 domain .

In some embodiments of the present invention, the immunomodulatory proteins provided, in addition to the variant ICOSL polypeptide, also include at least one additional IgSF domain (e.g., a second IgSF domain) that is a vIgD including one or more amino acid modifications (e.g., substitutions , deletions or mutations) compared to the IgSF domain in a wild-type IgSF domain or an unmodified IgSF domain, such as the IgSF domain in a member of the IgSF family shown in Table. 2. In some embodiments, the additional or second affinity modified IgSF domain comprises at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or greater sequence identity with the wild-type or unmodified IgSF domain or specific binding fragment thereof contained in the amino acid sequence set forth in any of SEQ ID NOs: 1-27 and 341. In some embodiments, the wild-type or unmodified IgSF domain is an IgV domain or an IgC domain such as an IgC1 or IgC2 domain. In some embodiments, the additional or second IgSF domain is an affinity modified IgV or IgC domain.

In some embodiments of the present invention, the immunomodulatory protein of the invention comprises at least one additional or second IgSF domain that is a vIgD comprising one or more amino acid substitutions relative to a wild-type IgSF domain (e.g., IgV) or an unmodified IgSF domain different from ICOSL.

In some embodiments, the additional or second IgSF domain includes one or more amino acid substitutions compared to a wild-type IgSF domain or an unmodified IgSF domain, such as the IgSF domain of the IgSF family member listed in Table 1. 2. In some embodiments, the additional or second affinity modified IgSF domain is at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or greater sequence identity with the wild-type or unmodified IgSF domain or a specific binding fragment thereof contained in the amino acid sequence set forth in any of SEQ ID NO: 127. In some embodiments, the wild-type or unmodified IgSF domain is an IgV domain or an IgC domain , such as the IgC1 or IgC2 domain. In some embodiments, the additional or second IgSF domain is an affinity modified IgV or IgC domain. In table 3-5 depict exemplary polypeptides containing one or more IgSF affinity modified domains that can be used in the stacking constructs provided herein.

In some embodiments, the one or more additional IgSF domains (eg, a second IgSF) is an IgSF domain (eg, IgV) of another member of the IgSF family that binds or recognizes a tumor antigen. In such embodiments, a member of the IgSF family serves as a functional tumor localizing moiety, thereby delivering the vIgD ICOSL into close proximity to immune cells in the tumor microenvironment. In some embodiments, the additional IgSF domain (eg, a second IgSF) is an IgSF NkP30 domain that binds or recognizes B7-H6 expressed in a tumor cell. In some embodiments, at least one additional (eg, second) IgSF domain, such as NkP30, is a vIgD that includes one or more amino acid modifications (eg, substitutions, deletions, or additions). In some embodiments, one or more amino acid modifications increase the binding affinity and/or selectivity of B7-H6 relative to an unmodified IgSF domain, e.g., NkP30, e.g., by at least or at least about 1.2 to 1 ,5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40 or 50 times.

- 52 044356

Table 3

Exemplary CD 80 variant polypeptides_____________

Mutation(s) ECD SEQ ID NO IgV SEQ ID NO Wild type 28 152 L70Q/A91G 55 153 L70Q/A91G/T130A 56 L70Q/ А91G/1118 А/Т120 S/T13 0 А 57 V4M/L70Q/A91G/T120 S/T 13 0 A 58 154 L70Q/A91 G/T 120 S/T 13 0 A 59 V20L/L70Q/A91 S/T 120 S/T 13 0 A 60 155 S44P/L70Q/A91 G/T 13 0 A 61 156 L70Q/A91G/E117G/T120S/T130A 62 A91G/T120S/T130A 63 157 L70R/A91 G/T 120 S/T 13 0 A 64 158 L70Q/E81 A/A91 G/T 120 S/1127T/T13 0 A 65 159 L70Q/Y87N/A91 G/T 13 0 A 66 160 T28 S/L70Q/A91G/E95K/T120 S/T 13 0 A 67 161 N63 S/L70Q/A91 G/T 120 S/T 13 0 A 68 162 KZ 6E/I67T/L70Q/A91 G/T 120 S/T 13 0 A/N152Т 69 163 E52G/L70Q/A91 G/T 120 S/T 13 0 A 70 164 K37E/F59S/L70Q/A91G/T120S/T130A 71 165 A91G/S103P 72 K89E/T130A 73 166 A91G 74 D60 V/A91 G/T 120 S/T 13 0 A 75 167 K54M/A91G/T120S 76 168 M3 8T/L70Q/E77G/A91 G/T 120S/T130A/N152T 77 169 R29H/E52G/L70R/E88G/A91 G/T 13 0 A 78 170 Y31N/T41G/L70Q/A91 G/T 120S/T13 0 A 79 171 V68A/T110A 80 172 S66H/D90G/T110A/F116L 81 173 R29H/E52G/T120 S/T 13 0 A 82 174 A91G/L102S 83 I67T/L70Q/A91G/T120S 84 175 L70Q/A91G/T110A/T120S/T130A 85 M3 8 V/T41D/M431/W5 0G/D76G/V83 A/K89E/T120 S/T 13 0 A 86 176 V22A/L70Q/S121P 87 177 Al 2 V/S15F/Y31Н/Т41 G/T 13 OA/Р13 7L/N152Т 88 178 I67F/L70R/E88G/A91 G/T 120S/T13 0 A 89 179 E24G/L25P/L70Q/T120S 90 180 А91G/F92L/F108L/T120S 91 181 R29D/Y31L/Q3 ZN/KZ 6G/M3 8I/T41A/M43R/M47T/E81V/L85R/K89N/ A91T/F92P/K93V/R94L/H18T/N149S 92 182 R29D/Y31L/Q3 ZN/KZ 6G/M3 8I/T41A/M43R/M47T/E81V/L85R/K89N/ A91T/F92P/K93V/R94L/N144S/N149S 93 R29D/Y31L/Q3 ZN/KZ 6G/M3 8I/T41A/M42T/M43R/M47T/E81V/L85R/ K89N/A91T/F92P/K93 V/R94L/L148 S/N149 S 94 183 E24G/R29D/Y31L/Q3 ZN/KZ 6G/M3 8I/T41A/M43 R/M47T/F5 9L/E81V/ L85R/K89N/A91T/F92P/K93 V/R94L/H96R/N149S/C182S 95 184 R29D/Y31L/Q3 ZN/KZ 6G/M3 8I/T41A/M43R/M47T/E81V/L85R/K89N/ А91T/F92P/K93 V/R94L/N149S 96 R29V/M43Q/E81R/L85I/K89R/D90L/A91E/F92N/K93Q/R94G 97 185 T4H/A91G 98 186 K89R/D90K/A91G/F92Y/K93R/N122S/N177S 99 187 K89R/D90K/A91G/F92 Y/K93R 100 K36G/K37Q/M38I/F59L/E81V/L85R/K89N/A91T/F92P/K93V/R94L/E 99G/T130A/N149S 101 188 E88D/K89R/D90K/A91G/F92 Y/K93R 102 189, 543 KZ 6G/K3 7Q/M3 8I/L40M 103 190 K36G 104 191 R29H/Y31Н/Т41G/Y87N/E88G/K89E/D90N/A91G/P109S 105 192 A12T/H18L/M43V/F59L/E77K/P109S/H18T 106 193 R29V/Y31F/K36G/M38L/M43Q/E81R/V83I/L85I/K89R/D90L/A91E/F 92N/K93Q/R94G 107 194 V68M/L70P/L72P/K86E 108 195

- 53 044356

L70Q/A91G/N144D 508 L70Q/ A91G/1118 A/T120 S/T13 0 A/K169E 509 V4M/L70Q/A91G/1118 V/T 120 S/T 13 0 A/K 169E 510 L70Q/A91G/1118 V/T 120 S/T 13 0 A/K 169E 511 L70Q/A91G/1118 V/T 120 S/T 13 0 A 512 V20L/L70Q/A91S/1118 V/T 120 S/T 13 0 A 513 L70Q/A91G/E117G/1118 V/T 120S/T130 A 514 A91G/1118 V/T 120 S/T 13 0 A 515 L70R/A91G/1118 V/T 120 S/T 13 0 A/T 199 S 516 L70Q/E81 A/A91G/1118 V/T 120 S/1127T/T13 0 A 517 T28 S/L70Q/A91G/E95K/1118 V/T 120 S/1126 V/T 13 0 A/K 169E 518 N63 S/L70Q/A91G/S114T/1118 V/T 120S/T13 0 A 519 KZ 6E/I67T/L70Q/A91G/1118 V/T 120 S/T 13 0 A/N152Т 520 E52G/L70Q/A91G/D107N/1118 V/T 120 S/T 13 0 A/K 169E 521 K37E/F59S/L70Q/A91G/I118V/T120S/T130A/K185E 522 D60 V/A91G/1118 V/T 120 S/T 13 0 AK169E 523 K54M/L70Q/A91G/Y164H/T120S 524 M3 8T/L70Q/E77G/A91G/1118 V/T 120S/T13 0A/N152T 525 Y31Н/Т41G/M43L/L70Q/А91G/1118 V/T 120 S/1126 V/T 13 0 A 526 LS656H/D90G/T110A/F116L 527 R29H/E52G/D90N/I118 V/T 120 S/T 13 0 A 528 R29H/E52G/D90N/I118 V/T 120 S/T 13 0 A 529 I67T/L70Q/A91G/1118 V/T 120S 530 L70Q/A91G/T110 A/1118 V/T 120 S/T 13 0 A 531 M3 8 V/T41D/M431/W5 0G/D76G/V83 A/K89E/1118 V/T 120 S/1126 V/T 13 0A 532 Al 2 V/S15F/Y31Н/МЗ 8L/T41G/M43L/D90N/T13 0 А/Р13 7L/N149D/N1 52Т 533 I67F/L70R/E8 8G/ A91G/1118 V/T 120 S/T 13 0 A 534 E24G/L25P/L70Q/A91G/I118 VT120S/N152T 535 А91G/F92L/F108L/1118 V/T 120 S 536 E88D/K89R/D90K/A91G/F92Y/K93R/N122S/N177S 537 KZ 6G/K3 7Q/M3 8I/L40M/F59L/E81V/L85R/K89N/A91T/F92P/K93 V/R 94L/E99G/T130A/N149S 539 K36G/L40M 540 542, 544

Table 4

Exemplary NKp30 variant polypeptides

Mutation(s) ECD SEQID NO igc-like domain SEQ ID NO Wild type 54 214 L30V/A60V/S64P/S86G 143 215 L30V 144 216 A60V 145 217 S64P 146 218 S86G 147 219

Table 5

Exemplary CD 86 variant polypeptides

Mutation(s) ECD SEQ ID NO IgV SEQ ID NO Wild type 29 220 Q35H/H90L/Q102H 148 221 Q35H 149 222 H90L 150 223 Q102H 151 224

The number of such unmodified or affinity modified IgSF domains present in an immunomodulatory protein stack construct (whether wild-type or non-wild-type combinations) is at least 2, 3, 4, or 5, and in some embodiments exactly 2, 3 , 4 or 5 IgSF domains (whereby determining the number of affinity-modified IgSF domains ignores any nonspecific binding fractional sequences thereof and/or substantially immunologically inactive fractional sequences).

In some embodiments of the stacked immunomodulatory protein provided herein, the number of IgSF domains is at least 2, wherein the number of independently modified or unmodified IgSF domains is at least: 0, 1, 2, 3, 4 , 5 or 6. Thus, the number of affinity-modified IgSF domains and the number of unaffinity-modified IgSF domains, respectively (affinity-modified IgSF domain: unaffinity-modified IgSF domain), can be exactly or not less than: 2:0 ( with modified affinity: wild type), 0:2, 2:1, 1:2, 2:2, 2:3, 3:2, 2:4, 4:2, 1:1,

1:3, 3:1, 1:4, 4:1, 1:5 or 5:1.

In some embodiments of the stacked immunomodulatory protein, at least two of the unmodified and/or affinity modified domains are identical to IgSF domains.

In some embodiments, a stacked immunomodulatory protein provided herein includes at least two affinity-modified and/or unmodified IgSF domains from a single IgSF member, but in a non-wild-type arrangement (as a permutation option). One illustrative example of non-wild-type assembly or permutation is an immunomodulatory protein involving the order of non-wild-type affinity-modified and/or unmodified IgSF domain sequences relative to those found in the wild-type ICOSL, whose IgSF domain sequences served as the source of the variant IgSF domains represented in this document. Thus, in one example, the immunomodulatory protein may comprise IgV proximal and IgC distal to the transmembrane domain, albeit in an unmodified and/or affinity modified form. The presence of both non-wild type and non-wild type combinations of unmodified and/or affinity modified IgSF domains in the immunomodulatory protein provided herein is also within the scope of the invention.

In some embodiments of the stacked immunomodulatory protein, the unmodified and/or affinity modified IgSF domains are non-identical (ie, different) IgSF domains. Non-identical affinity modified IgSF domains specifically bind under specific binding conditions to different cognate binding partners and are non-identical whether or not the wild-type IgSF domains or the unmodified IgSF domains from which they were constructed were the same. Thus, for example, a non-wild type combination of at least two non-identical IgSF domains in an immunomodulatory protein may include at least one IgSF domain sequence whose origin is unique to one ICOSL, and at least one second IgSF sequence whose origin origin is unique to another member of the IgSF family that is not ICOSL, where the IgSF domains of the immunomodulatory protein are in the unmodified and/or affinity modified form. However, in alternative embodiments, two non-identical IgSF domains are derived from the same IgSF domain sequence, but at least one of them is with a modified affinity such that they specifically bind different cognate binding partners.

The plurality of unmodified and/or affinity modified IgSF domains in the polypeptide chain of a stacked immunomodulatory protein need not necessarily be directly covalently linked to each other. In some embodiments, an intervening interval of one or more amino acid residues indirectly covalently links unaffinity-modified IgSF domains and/or affinity-modified IgSF domains to each other. Communication can occur through N-terminal residues to C-terminal residues.

In some embodiments, the linkage can be made using side chains of amino acid residues that are not located at the N-terminus or C-terminus of the affinity-modified and/or unmodified IgSF domain. Thus, connections can be made through terminal or internal amino acid residues or combinations thereof.

In some embodiments, two or more IgSF domains, including vIgD from ICOSL and one or more additional IgSF domains (eg, from a second variant IgSF domain) from another IgSF family member, are covalently or non-covalently linked. In some embodiments, two or more IgSF domains are linked directly or indirectly, for example, through a linker. In some embodiments, an intermediate region of one or more amino acid residues indirectly covalently links the IgSF domains together. The connection can occur through N-terminal residues with C-terminal residues. In some embodiments, the linkage may be through side chains of amino acid residues that are not located at the N-terminus or C-terminus of the IgSF domain(s). Thus, connections can be made through terminal or internal amino acid residues or combinations thereof.

In some embodiments, one or more peptide linkers link vIgD from ICOSL and an additional IgSF domain (eg, a second variant IgSF domain). In some embodiments, the peptide linker can be a single amino acid residue or be longer. In some embodiments, the peptide linker has at least one amino acid residue, but not more than 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2 or 1 amino acid residue. In some embodiments, the linker is (single letter amino acid code): GGGGS (4GS) or 4GS linker multimers, such as 2, 3, 4 or 5 4GS linker repeats. In some embodiments, the peptide linker is (GGGGS) ₂ or (GGGGS) ₃ . In some embodiments, the linker may also include a number of alanine residues alone or in addition to another peptide linker (such as a 4GS linker or a multimer thereof). In some embodiments, the number of alanine residues in

- 55 044356 each series is: 2, 3, 4, 5 or 6 alanines.

In some embodiments, the affinity-modified and/or unmodified IgSF domains are linked by wild-type peptide linkers inserted at the N-terminus and/or C-terminus of the first and/or second affinity-modified and/or unmodified IgSF domains. In some embodiments, there is a leader peptide linker inserted at the N-terminus of the first IgSF domain and/or a first trailer sequence inserted at the C-terminus of the first affinity-modified and/or unmodified IgSF domain. In some embodiments, there is a second leader peptide linker inserted at the N-terminus of the second IgSF domain and/or a second trailer sequence inserted at the C-terminus of the second IgSF domain with modified and/or unmodified affinity. If the first and second affinity-modified and/or unmodified IgSF domains are derived from the same parent protein and linked in the same orientation, then the wild-type peptide linkers between the first and second affinity-modified and/or unmodified IgSF domains are not duplicated . For example, when the first wild-type trailer peptide linker and the second wild-type leader peptide linker are the same, the type II immunomodulatory protein includes neither the first wild-type trailer peptide linker nor the second wild-type leader peptide linker.

In some embodiments, the type II immunomodulatory proteins include a first wild-type leader peptide linker inserted at the N-terminus of a first affinity-modified and/or unmodified IgSF domain, wherein the first wild-type leader peptide linker includes at least 5 (e.g., at least about any number of 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or more) consecutive amino acids from the intervening sequence in the wild-type protein of which the first affinity-modified and/or unmodified IgSF domain derived between the parent IgSF domain and the immediately preceding domain (such as a signal peptide or IgSF domain). In some embodiments, the wild-type first leader peptide linker includes the entire intervening sequence in the wild-type protein from which the first IgSF domain is derived, with modified and/or unmodified affinity between the parent IgSF domain and the immediately preceding domain (such as a signal peptide or IgSF domain).

In some embodiments, the type II immunomodulatory proteins further comprise a first wild-type trailer peptide linker inserted at the C-terminus of the first affinity-modified and/or unmodified IgSF domain, wherein the first wild-type trailer peptide linker includes at least 5 (e.g. at least about any number of 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or more) consecutive amino acids from the intervening sequence in the wild-type protein from which the first modified IgSF domain is derived or an unmodified affinity between the parent IgSF domain and the immediately following domain (such as an IgSF domain or transmembrane domain). In some embodiments, the first wild-type trailer peptide linker includes all of the intervening sequence in the wild-type protein from which the first IgSF domain is derived, with modified and/or unmodified affinity between the parent IgSF domain and the immediately following domain (e.g., IgSF domain or transmembrane domain).

In some embodiments, the type II immunomodulatory protein further includes a second wild-type leader peptide linker inserted at the N-terminus of a second affinity-modified and/or unmodified IgSF domain, wherein the second wild-type leader peptide linker includes at least 5 (e.g. at least about any number of 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or more) consecutive amino acids from the intervening wild-type protein sequence from which the modified and/or unmodified second IgSF domain is derived affinity between the parent IgSF domain and the immediately preceding domain (such as a signal peptide or IgSF domain). In some embodiments, the wild-type second leader peptide linker includes all of the intervening sequence in the wild-type protein from which the second affinity-modified and/or unmodified IgSF domain is derived, between the parent IgSF domain and the immediately preceding domain (such as a signal peptide or IgSF domain).

In some embodiments, the type II immunomodulatory protein further includes a second longitudinal wild-type peptide linker inserted at the C-terminus of a second affinity-modified and/or unmodified IgSF domain, wherein the second wild-type peptide linker includes at least 5 (e.g., at least any number of about 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or more) consecutive amino acids from the intervening sequence in the wild-type protein from which the modified and/or unmodified second IgSF domain is derived affinity between the parent IgSF domain and the immediately following domain (such as the IgSF domain or transmembrane domain). In some embodiments, the wild-type second terminal peptide linker includes all of the intervening sequence in the wild-type protein from which a second IgSF domain is derived with modified and/or unmodified affinity between the parent IgSF domain and the immediately following domain (e.g., an IgSF domain or a transmembrane domain). domain).

An example leader sequence and trailer sequence for a type II protein containing a CD80 IgSF domain is provided in SEQ ID NO: 231 and SEQ ID NO: 232. An example leader sequence and trailer sequence for a type II protein containing an ICOSL domain IgSF is provided in SEQ ID NOs: 233 and 234. Examples of a leader sequence and a trailer sequence for a type II CD86 IgSF domain-containing protein are set forth in any one of SEQ ID NOs: 236-238. An example of a wild-type linker sequence for a type II NKp30 IgGF domain-containing protein is provided in SEQ ID NO: 235.

In some embodiments, two or more IgSF domains, including vIgD from ICOSL and one or more additional IgSF domains (eg, a second variant IgSF domain) from another IgSF family member, are linked or joined to an Fc to form a dimeric multidomain immunomodulatory protein. In some embodiments, the variant ICOSL polypeptide and the second IgSF domain are independently linked, directly or indirectly, to the N- or C-terminus of the Fc subunit. In some embodiments, the variant ICOSL polypeptide and the second IgSF domain are linked directly or indirectly, and one of the variant ICOSL or the second IgSF domain is also directly or indirectly linked to the N- or C-terminus of the Fc subunit. In some embodiments, the connection to the Fc is via a peptide linker, for example, a peptide linker such as those described above. In some embodiments, the connection between the variant ICOSL and the second IgSF domain is through a peptide linker, for example, a peptide linker such as those described above. In some embodiments, the vIgD ICOSL, one or more additional IgSF domains, and the Fc domain may be linked together in any of numerous configurations, as shown in FIG. 16. Exemplary configurations are described in the examples.

In some embodiments of the invention, the stacked immunomodulatory protein is a dimer formed by two stacked immunomodulatory Fc fusion polypeptides. Nucleic acid molecules encoding any of the stacked immunomodulatory proteins are also provided. In some embodiments, a dimeric multi-domain stacked immunomodulatory protein can be produced in cells by expressing, or in some cases co-expressing, immunomodulatory Fc-fusion polypeptides such as those described above, respectively forming dimeric Fc-fusion proteins.

In some embodiments, the dimeric multidomain stacked immunomodulatory protein is bivalent for each Fc subunit, monovalent for each subunit, or bivalent for one subunit and tetravalent for another.

In some embodiments, the dimeric multidomain stacked immunomodulatory protein is a homodimeric multidomain stacked Fc protein. In some embodiments, the dimeric multidomain stacked immunomodulatory protein includes a first stacked immunomodulatory Fc fusion polypeptide and a second stacked immunomodulatory Fc fusion polypeptide, wherein the first and second polypeptides are the same. In some embodiments, the Fc portion of the polypeptide can be any Fc described above.

In some embodiments, the multi-domain stack molecule is heterodimeric, comprising two different Fc polypeptides, wherein at least one is an Fc polypeptide containing at least one variant ICOSL polypeptide and/or at least one second IgSF domain ( for example, the second variant domain of IgSF). In some embodiments, the multi-domain stack molecule includes a first Fc polypeptide comprising a variant ICOSL and a second IgSF domain, and a second Fc polypeptide comprising a variant ICOSL and a second IgSF domain. In some embodiments, the multi-domain stack molecule includes a first Fc polypeptide comprising a variant ICOSL polypeptide and a second IgSF domain, and a second Fc polypeptide that is not associated with either the variant ICOSL polypeptide or the second IgSF domain.

In some embodiments, the multi-domain stacking molecule includes a first Fc polypeptide comprising 1, 2, 3, 4 or more variant ICOSL polypeptides and/or 1, 2, 3, 4 or more second IgSF domains, wherein the total number of IgSF domains in the first stacking polypeptide Fc greater than 2, 3, 4, 5, 6 or more. In one example of such an embodiment, the second stacked Fc polypeptide includes 1, 2, 3, 4 or more variant ICOSL polypeptides and/or 1, 2, 3, 4 or more second IgSF domains, wherein the total number of IgSF domains in the second stacked Fc polypeptide is greater than 2 , 3, 4, 5, 6 or more. In another example of such an embodiment, the second Fc polypeptide is not associated with either the variant ICOSL polypeptide or the second IgSF domain.

In some embodiments, the heterodimeric stack molecule includes a first stacked immunomodulatory Fc-fusion polypeptide and a second stacked immunomodulatory Fc-fusion polypeptide, wherein the first and second polypeptides are different. In some embodiments, the stacked heterodimeric molecule includes a first Fc subunit comprising a first variant ICOSL polypeptide and/or a second IgSF domain (e.g., a second variant IgSF domain) and a second Fc subunit comprising a different variant ICOSL polypeptide or a second IgSF domain. In some embodiments, the heterodimeric stack molecule includes a first immunomodulatory stack

- 57 044356 an Fc fusion polypeptide and a second stacked immunomodulatory Fc fusion polypeptide, wherein the first and second polypeptides are different. In some embodiments, the heterodimeric stack molecule includes a first Fc subunit comprising a first variant ICOSL polypeptide and/or a second IgSF domain (e.g., a second variant IgSF domain), and a second Fc subunit comprising both a first variant ICOSL polypeptide and a second IgSF domain (e.g. , second variant domain of IgSF), but in a different orientation or configuration from the first Fc subunit.

In some embodiments, the Fc domain of one or both of the first and second stacked immunomodulatory Fc fusion polypeptides includes modification (eg, substitution) such that the interface of the Fc molecule is modified to facilitate and/or promote heterodimerization. In some embodiments, modifications include introducing projections (bulges) into the first Fc polypeptide and cavities (holes) into the second Fc polypeptide such that the projections are positioned in the cavities to facilitate complexation of the first and second Fc-containing polypeptides. The amino acids intended to be replaced and/or modified to create protrusions or cavities in the polypeptide are typically interface amino acids that interact or contact one or more amino acids at the interface of the second polypeptide.

In some embodiments, a first polypeptide that is modified to contain an amino acid bulge comprising the replacement of a native or parent amino acid with an amino acid that has at least one side chain that protrudes from the interface of the first polypeptide and is therefore accommodated in the compensation cavity ( hole) at the adjacent interface of the second polypeptide. Most often, the replacement amino acid is one that has a larger side chain than the original amino acid residue. One skilled in the art will know how to determine and/or evaluate the properties of amino acid residues to identify those that are ideal amino acids to substitute to create a protrusion. In some embodiments, the substituting residues for forming the protrusions are naturally occurring amino acid residues and include, for example, arginine (R), phenylalanine (F), tyrosine (Y), or tryptophan (W). In some examples, the original residue identified for substitution is an amino acid residue that has a small side chain, such as, for example, alanine, asparagines, aspartic acid, glycine, serine, threonine, or valine.

In some embodiments, the second polypeptide that is modified to contain a cavity is a polypeptide that includes replacing a native or parent amino acid with an amino acid that has at least one side chain that is recessed in the interface of the second polypeptide and, thus being able to accommodate a corresponding protrusion of the interface of the first polypeptide. Most often, the replacement amino acid is one that has a smaller side chain volume than the original amino acid residue. One skilled in the art will know how to determine and/or evaluate the properties of amino acid residues to identify those that are ideal replacement residues for cavity formation. Typically, the substituting residues for cavity formation are naturally occurring amino acids and include, for example, alanine (A), serine (S), threonine (T) and valine (V). In some examples, the parent amino acid identified for substitution is an amino acid that has a large side chain, such as, for example, tyrosine, arginine, phenylalanine, or tiptophan.

The CH3 interface of human IgG1, for example, includes sixteen residues in each domain located on four antiparallel β-strands that create a 1090 A 2 depression on each surface (see, for example, Deisenhofer et al. (1981) Biochemistry, 20:2361- 2370; Miller et al., (1990) J Mol. Biol, 216, 965-973; Ridgway et al., (1996) Prot. Engin., 9: 617-621; U.S. Pat. No. 5,731,168). Modifications of the CH3 domain to create protrusions or cavities are described, for example, in US patent No. 5731168; international patent applications WO 98/50431 and WO 2005/063816; and Ridgway et al., (1996) Prot. Engin., 9: 617-621. In some examples, modifications to the CH3 domain to create protrusions or cavities typically target residues located on the two central antiparallel β-strands. The goal is to minimize the risk that the protrusions created may be accommodated by protruding into the surrounding solvent and not placed into a compensatory cavity in the CH3 partner domain.

In some embodiments, the heterodimeric molecule includes mutation T366W in the CH3 domain of the overhang chain and mutations T366S, L368A, Y407V in the CH3 domain of the cavity chain. In some cases, an additional interchain disulfide bridge between the CH3 domains can also be used (Merchant, AM, et al., Nature Biotech, 16 (1998) 677-681), for example, by introducing the Y349C mutation into the CH3 domain of the knob or hole chain and mutacite E356C or mutacy S354C to the CH3 domain of another chain. In some embodiments, the heterodimeric molecule includes mutations S354C, T366W in one of the two CH3 domains and mutations Y349C, T366S, L368A, Y407V in the other of the two CH3 domains. In some embodiments, the heterodimeric molecule includes mutations E356C, T366W in one of the two CH3 domains and mutations Y349C, T366S, L368A, Y407V in the other of the two CH3 domains. In some embodiments, the heterodimeric molecule includes mutations Y349C, T366W in one of two CH3 domains

- 58 044356 and mutations E356C, T366S, L368A, Y407V in the other of the two CH3 domains. In some embodiments, the heterodimeric molecule includes mutations Y349C, T366W in one of two CH3 domains, and mutations S354C,

T366S, L368A, Y407V in the other of the two CH3 domains. Examples of other projection-to-recess technologies are known in the art, for example as described in EP 1 870 459 A1.

In some embodiments, the Fc subunits of the heterodimeric molecule may further include one or more other Fc mutations, such as any of those described above. In some embodiments, the heterodimer molecule includes an Fc subunit with a mutation that reduces effector function.

In some embodiments, a variant Fc containing knob/cavity modifications in CH3 can be attached to the stacked immunomodulatory polypeptide at any location, but typically through its N- or C-terminus, to the N- or C-terminus of the first and/or second stacked immunomodulatory polypeptide. polypeptide, for example, to form a fusion polypeptide. The connection can be direct or indirect via a linker. Typically, a knob-hole molecule is generated by coexpression of a first stacked immunomodulatory polypeptide associated with an Fc variant containing the knob modification(s) on CH3 with a second stacked immunomodulatory polypeptide associated with the Fc variant containing the cavity modification(s) on CH3 .

C. Conjugates and fusion proteins of variant polypeptides and immunomodulatory proteins.

In some embodiments of the present invention, variant polypeptides provided herein, which are immunomodulatory proteins containing variants of the IgSF family Ig domain (vIgD), can be conjugated to or fused to a fragment, such as an effector fragment, such as another protein, directly or indirectly, with the formation of a conjugate (IgSF conjugate). In some embodiments, the compound may be covalent or non-covalent, for example, through a non-covalent biotin-streptavidin interaction. In some embodiments of the variant ICOSL-Fc fusion protein, any combination of any two or more of the above conjugates can be attached to the Fc or the variant CD80 polypeptide or both.

In some embodiments, the fragment may be a targeting moiety, a small molecule drug (a non-polypeptide drug with a molar mass of less than 500 daltons), a toxin, a cytostatic agent, a cytotoxic agent, an immunosuppressant, a radioactive agent useful for diagnostic purposes, a radioactive metal ion for therapeutic purposes, an enzyme , an activating prodrug, an agent that increases the biological half-life, or a diagnostic or detectable agent.

In some embodiments, the effector moiety is a therapeutic agent, such as an antitumor therapeutic agent, that is either cytotoxic, cytostatic, or otherwise provides some therapeutic benefit. In some embodiments, the effector moiety is a targeting moiety or agent, such as an agent that targets a cell surface antigen, such as an antigen on the surface of a tumor cell. In some embodiments, the effector portion is a label that can generate a detectable signal, directly or indirectly. In some embodiments, the effector portion is a toxin. In some embodiments, the effector moiety is a protein, peptide, nucleic acid, small molecule, or nanoparticle.

In some embodiments, 1, 2, 3, 4, 5 or more effector fragments, which may be the same or different, are conjugated, linked or fused to a variant polypeptide or protein to form an IgSF conjugate. In some embodiments, such effector fragments can be attached to a variant polypeptide or immunomodulatory protein using various molecular biological or chemical conjugation and coupling techniques known in the art and described below. In some embodiments, linkers, such as peptide linkers, cleavable linkers, non-cleavable linkers, or linkers that assist in the conjugation reaction, can be used to link or conjugate effector moieties to a variant polypeptide or immunomodulatory protein.

In some embodiments, the IgSF conjugate includes the following components: (protein or polypeptide), (L)q, and (effector moiety)n, wherein the protein or polypeptide is any of the described variant polypeptides or immunomodulatory proteins capable of binding one or more cognate counterstructural ligands as described; L represents a linker for linking a protein or polypeptide to a fragment; m is at least 1; q is 0 or more; and the resulting IgGF conjugate binds to one or more counterstructural ligands. In specific embodiments, m is from 1 to 4 and q is from 0 to 8.

In some embodiments, an IgSF conjugate containing a variant polypeptide or immunomodulatory protein provided herein is provided conjugated to a targeting agent that binds to a cell surface molecule, for example, to target the variant polypeptide or immunomodulatory protein to a specific cell. In some embodiments, the targeting agent is a molecule(s) that has the ability to localize to and bind to a molecule present in a normal cell/tissue and/or tumor cell/tumor in a subject. In other words, IgSF conjugates containing a targeting agent can bind to a ligand (directly or indirectly) that is present in a cell, such as a tumor cell. Targeting agents of the invention include antibodies, polypeptides, peptides, aptamers, other ligands, or any combination thereof that can bind a target cell component or molecule.

In some embodiments, the targeting agent binds to the tumor cell(s) or may bind in the vicinity of the tumor cell(s) (eg, tumor vasculature or tumor microenvironment) after administration to the subject. The targeting agent can bind to a receptor or ligand on the surface of a malignant tumor cell. In another aspect of the invention, a targeting agent is selected that is specific for non-malignant cells or tissue. For example, the targeting agent may be specific to a molecule typically present on a particular cell or tissue. Additionally, in some embodiments, the same molecule may be present on normal and malignant tumor cells. Various cellular components and molecules are known. For example, if the targeting agent is specific for EGFR, the resulting IgSF conjugate can target malignant tumor cells expressing EGFR as well as normal epidermal skin cells expressing EGFR. Therefore, in some embodiments, the IgSF conjugate of the invention may function through two separate mechanisms (targeting malignant tumor cells and non-malignant cells).

In various aspects of the present invention, the IgSF conjugate of the present invention disclosed herein includes a targeting agent that can bind/target a cellular component such as a tumor antigen, bacterial antigen, viral antigen, mycoplasma antigen, fungal antigen, prion antigen, parasite In some aspects, a cellular component, antigen, or molecule can be used to designate the desired target for a targeting agent. For example, in various embodiments, the targeting agent is specific for or binds to a component that includes, but is not limited to, epidermal growth factor receptor (EGFR, ErbB-1, HER1), ErbB-2 (HER2/neu), ErbB-3/HER3, ErbB-4/HER4, EGFR family of ligands; insulin-like growth factor receptor (IGFR) family, IGF-binding proteins (IGFBPs), IGFR ligand family; platelet-derived growth factor receptor (PDGFR) family, PDGFR family of ligands; fibroblast growth factor receptor (FGFR) family, FGFR ligand family, vascular endothelial growth factor receptor (VEGFR) family, VEGF family; HGF receptor family; TRK receptor family; ephrin receptor family (ERF); AXL receptor family; leukocyte tyrosine kinase (LTK) receptor family; TIE receptor family, angiopoietin 1,2; the receptor tyrosine kinase (ROR)-like orphan receptor family, such as R0R1; CD171 (L1CAM); B7-H6 (NCR3LG1); PD-L1, tumor glycosylation antigen, such as sTn or Tn, such as sTn Ag MUC1; LHR (LHCGR); phosphatidylserine, domain disc receptor (DDR) family; RET receptor family; KLG receptor family; RYK receptor family; MuSK receptor family; transforming growth factor-α (TGF-α), TGF-β receptors; cytokine receptors, class I (hematopoietin family) and family II (interferon/IL-10) receptors, tumor necrosis receptor (TNF) superfamily (TNFRSF), death receptor family; testicular cancer (CT) antigens, lineage-specific antigens, differentiation antigens, alpha-actinin-4, ARTC1, breakpoint cluster region-Abelson fusion products (Bcr-abl), B-RAF, caspase-5 (CASP-5) caspase -8 (CASP-8), β-catenin (CTNNB1), cell division cycle 27 (CDC27), cyclin-dependent kinase 4 (CdK4), CDKN2A, COA-I, dek-can fusion protein, EFTUD-2, elongation factor 2 ( ELF2), a variant Ets 6/acute myeloid leukemia 1 ETS gene fusion protein (ETC6-AML1), fibronectin (FN), e.g. fibronectin extradomain A (EDA), GPNMB, a low-density lipid receptor/GDP- fusion protein L fucose: β-D galactose fusion protein 2-a-L fucosyltransferase (LDLR/FUT), HLA-A2, arginine isoleucine substitution at residue 170 of the α-helix a2 domain in the HLA-A2 gene (HLA-A * 201-R170I), HLA-Al 1, mutated heat shock protein 70-2 (HSP70-2M), K1AAO2O5, MART2, melanoma ubiquitous, mutated 1, 2, 3 (MUM-I, 2, 3), prostatic acid phosphatase (PAP), neo -PAP, class I myosin, NFYC, OGT, OS-9, pml-RARalpha fusion protein, PRDX5, PTPRK, K-ras (KRAS2), N-ras (NRAS), HRAS, RBAF600, SIRT2, SNRPD1, SYT-SSX1 or -SSX2, triosephosphate isomerase, BAGE, BAGK-1, BAGE-2,3, 4,5, GAGE-1,2,3,4,5,6,7,8, GnT-V (aberrant Nacetylglucosanyltransferase V, MGAT5) , HERV-K-MEL, KK-LC, KM-H-1, LAGE, LAGE-I, CTL recognized antigen in melanoma (CAMEL), MAGE-A1 (MAGE-I), MAGE-A2, MAGE-A3, MAGEA4, MAGE-A5, MAGE-A6, MAGE-A8, MAGE-A9, MAGE-A10, MAGE-A11, MAGE-A12, MAGE-3, MAGE-B1, MAGE-B2, MAGE-B5, MAGE-B6, MAGE- C1, MAGE-C2, mucin 1 (MUC1), MART-1/melan-A (MLANA), gp100, gp100/Pmell7 (SILV), tyrosinase (TYR), TRP-I, HAGE, NA-88, NY-ESO -I, NY-ESO1/LAGE-2, SAGE, Spl7, SS X-1,2,3,4, TRP2-INT2, carcinoembryonic antigen (CEA), kallikrein 4, mammaglobin-A, OA1, prostate antigen (PSA), TRP-1/gp75, TRP-2, adipophilin, interferon-inducible protein in melanoma 2 (AIM-2), BING-4, CPSF, cyclin D1, epithelial cell adhesion molecule (Ep-CAM), EphA3, factor fibroblast growth 5 (FGF-5), glycoprotein 250 (gp250),

- 60 044356

EGFR (ERBB1), HER-2/neu (ERBB2), interleukin-13 receptor α2 chain (IL13Ra2), IL-6 receptor, intestinal carboxylesterase (iCE), alpha-fetobeton (AFP), M-CSF, mdm-2, MUCl, p53 (TP53), PBF, PRAME, PSMA, RAGE-I, RNF43, RU2AS, SOXlO, STEAPl, survivin (BIRC5), human telomerase reverse transcriptase (hTERT), telomerase, Wilms tumor gene (WT1) SYCPl, BRDT, SPANX, XAGE, ADAM2, PAGE-5, LIPl, CTAGE-I, CSAGE, MMA1, CAGE, BORIS, HOM-TES-85, AF15ql4, HCA661, LDHC, MORC, SGY-I, SPO1 1, TPX1, NY-SAR -35, FTHL17, NXF2, TDRD1, TEX15, FATE, TPTE, immunoglobulin idiotypes, Bence-Jones protein, estrogen receptors (ER), androgen receptors (AR), CD40, CD30, CD20, CD19, CD33, tumor antigen 72-4 (CA 72-4), tumor antigen 15-3 (CA 15-3), tumor antigen 27-29 (CA 27-29), tumor antigen 125 (CA 125), tumor antigen 19-9 (CA 199), chorionic β-human gonadotropin, β-2 microglobulin, squamous cell carcinoma antigen, neurospecific enolase, heat shock protein gp96, GM2, sargramostim, CTLA-4, 707 alanine proline (707-AP), adenocarcinoma antigen recognized by T cells 4 (ART-4 ), carcinoembryonic antigen peptide 1 (CAP-1), calcium-activated chloride channel-2 (CLCA2), cyclophilin B (Cyp-B), human signet ring cell tumor-2 (HST-2), human papillomavirus (HPV) proteins ( HPV-E6, HPV-E7, major or small capsid antigens, others), Epstein-Barr virus (EBV) proteins (latent membrane proteins EBV-LMP1, LMP2; others), hepatitis B or C virus proteins and HIV proteins.

In some embodiments, the IgSF conjugate, through its targeting agent, binds a cellular component of a tumor cell, tumor vasculature, or tumor microenvironment, thereby promoting the killing of target cells by modulating the immune response (for example, by activating co-stimulatory molecules or inhibiting negative regulatory molecules activation of immune cells), inhibition of survival signals (eg, growth factor or cytokine antagonists or hormone antagonists), activation of death signals and/or immune-mediated cytotoxicity, for example, through antibody-dependent cellular cytotoxicity. Such IgSF conjugates may function through several mechanisms to prevent, reduce or eliminate tumor cells, for example, to facilitate the delivery of conjugated effector fragments to the tumor target, for example, through receptor-mediated endocytosis of the IgSF conjugate; or such conjugates can recruit, bind and/or activate immune cells (eg, NK cells, monocytes/macrophages, dendritic cells, T cells, B cells). Moreover, in some cases, one or more of the above pathways may operate upon administration of one or more IgSF conjugates of the invention.

In some embodiments, the IgSF conjugate, through a targeting agent, will be localized, for example by binding, to a cellular component of a tumor cell, tumor vasculature, or tumor microenvironment, thereby modulating the immune response cells in the immediate vicinity of the tumor. In some embodiments, the targeting agent facilitates delivery of conjugated IgSF (e.g., vIgD) to a tumor target, e.g., to interact with its cognate binding partner to alter immune cell signaling (e.g., NK cells, monocytes/macrophages, dendritic cells, T cells , B cells) carrying a cognate binding partner. In some embodiments, localized delivery has an agonistic effect or stimulates a costimulatory receptor.

In some embodiments, the targeting agent is an immunoglobulin. As used herein, the term immunoglobulin includes natural or artificial mono- or multivalent antibodies, including, but not limited to, polyclonal, monoclonal, multispecific, human, humanized or chimeric antibodies, single chain antibodies, Fab fragments, F(ab') fragments, Fab expression library-derived fragments, single-chain Fv (scFv); anti-idiotypic (anti-Id) antibodies (including, for example, anti-Id antibodies of the invention) and epitope-binding fragments of any of the above. The term antibody as used herein refers to immunoglobulin molecules and the immunologically active portions of immunoglobulin molecules, that is, molecules that contain an antigen binding site that specifically binds to an antigen. The immunoglobulin molecules of the invention can be of any type (eg, IgG, IgE, IgM, IgD, IgA and IgY), class (eg, IgG1, IgG2, IgG3, IgG4, IgA1 and IgA2) or subclass of immunoglobulin molecule.

In some embodiments, the IgSF conjugate, through its antibody targeting moiety, will bind a cellular component of a tumor cell, tumor vasculature, or tumor microenvironment, thereby promoting apoptosis of target cells through modulation of the immune response (e.g., by activating co-stimulatory molecules or inhibiting negative regulatory molecules activation of immune cells), inhibition of survival signals (eg, growth factor receptor antagonists or cytokines or hormones), activation of death signals and/or immune-mediated cytotoxicity such as antibody-dependent cellular cytotoxicity. Such IgSF conjugates may function through several mechanisms to prevent, reduce or eliminate tumor cells, for example, to facilitate delivery of conjugated effector fragments to the tumor target, for example, through receptor-mediated endocytosis of the IgSF conjugate; or such conjugates may recruit, bind and/or activate immune cells (e.g. NK cells, monocytes/macrophages, dendritic cells,

T cells, B cells).

In some embodiments, the IgSF conjugate, through its antibody targeting moiety, will bind a cellular component of a tumor cell, tumor vasculature, or tumor microenvironment, thereby modulating the immune response (e.g., by activating co-stimulatory molecules or inhibiting negative regulation molecules of immune cell activation). In some embodiments, such conjugates can recognize, bind and/or modulate (eg, inhibit or activate) immune cells (eg, NK cells, monocytes/macrophages, dendritic cells, T cells, B cells).

Antibody targeting fragments of the invention include antibody fragments that include, but are not limited to, Fab, Fab' and F(ab')2, Fd, single chain Fv (ScFv), single chain antibodies, disulfide linked Fv (sdFv) and fragments containing either a domain VL or VH. Antigen-binding antibody fragments, including single chain antibodies, may contain variable region(s) alone or in combination with all or part of the following regions: hinge region, CH1, CH2 and CH3 domains. Also included in the invention are antigen binding fragments which also include any combination of variable region(s) with a hinge region, CH1, CH2 and CH3 domains. Also included in the invention are Fc fragments, antigen-Fc fusion proteins, and Fc-targeting fragment conjugates or fusion products (Fc peptide, Fc aptamer). Antibodies targeting fragments of the invention can be of any animal origin, including birds and mammals. In one aspect, the antibody targeting moieties are derived from human, mouse (eg, mouse and rat), donkey, sheep, rabbit, goat, guinea pig, camel, horse, or chicken. In addition, such antibodies may be humanized versions of animal antibodies. Antibodies targeting fragments of the invention may be monospecific, bispecific, trispecific, or more multispecific.

In various embodiments, the antibody/targeting moiety recruits, binds and/or activates immune cells (eg, NK cells, monocytes/macrophages, dendritic cells) through interactions between Fc (antibodies) and Fc receptors (on immune cells) and through conjugated variants polypeptides or immunomodulatory proteins presented herein. In some embodiments, the antibody/targeting moiety recognizes or binds a tumor agent through and localizes to the tumor cell conjugated variant polypeptides or immunomodulatory proteins provided herein to facilitate modulation of immune cells in the vicinity of the tumor.

Examples of antibodies that may be included in IgSF conjugates include, but are not limited to, antibodies such as cetuximab (IMC-C225; Erbitux®), trastuzumab (Herceptin®), rituximab (Rituxan®; MabThera®), bevacizumab (Avastin®) , alemtuzumab (Campath®, Campath-1H®, Mabcampath®), panitumumab (ABX-EGF, Vectibix®), ranibizumab (Lucentis®), ibritumomab, ibritumomab tiuxetan, (Zevalin®), tositumomab, iodine I131 tositumomab (BEXXAR®) , catumaxomab (Removab®), gemtuzumab, gemtuzumab ozogamicin (Mylotarg®), abatacept (CTLA4-Ig, Orencia®), belatacept (L104EA29YIg, LEA29Y, LEA), ipilimumab (MDX-010, MDX-101), tremelimumab (ticilimumab, CP675,206), PRS-010, PRS-050, aflibercept (VEGF Trap, AVE005), volociximab (M200), F200, MORAb009, SS1P (CAT-5001), cixutumumab (IMC-A12), matuzumab (EMD72000), nimotuzumab (h-R3), zalutumumab (HuMax-EGFR), necitumumab IMC-11F8, mAb806/ch806, Sym004, mAb-425, Panorex @ (17-1A) (mouse monoclonal antibody); Panorex @ (17-1A) (chimeric mouse monoclonal antibody); IDEC-Y2B8 (mouse, anti-CD20 monoclonal antibody); BEC2 (anti-idiotypic monoclonal antibody mimicking the GD epitope) (with BCG); oncolim (monoclonal antibody Lym-1); SMART MI95 Ab, humanized 13'I LYM-I (Oncolym), Ovarex (B43.13, anti-idiotypic mouse monoclonal antibody); MDX-210 (humanized bispecific antibody against HER-2); monoclonal antibody 3622W94 MAb, which binds to pancarcinoma antigens EGP40 (17-1A) on adenocarcinomas; Anti-VEGF, Zenapax (SMART Anti-Tac (IL-2 receptor), SMART MI95 Ab, humanized Ab, humanized); MDX-210 (humanized bispecific antibody against HER-2); MDX-447 (humanized bispecific antibody against EGF receptor); NovoMAb-G2 (pancrecinoma-specific antibody); TNT (chimeric monoclonal antibody-histone antigens); TNT (chimeric monoclonal antibody-histone antigens); Gliomab-H (monoclonal-humanized antibodies); GNI-250 Mab; EMD-72000 (chimeric EGF antagonist); LymphoCide (humanized LL2 antibody); and MDX-260 designed for GD-2, ANA Ab, SMART IDlO Ab, SMART ABL 364 Ab or ImmuRAIT-CEA. As illustrated by the above list, it is common to obtain antibodies to a specific target epitope.

In some embodiments of the invention, the antibody targeting fragment is a full-length antibody or an antigen binding fragment thereof containing an Fc domain. In some

- 62 044356 embodiments, the variant polypeptide or immunomodulatory protein is conjugated to the Fc portion of an antibody targeting fragment, for example, by conjugation to the N-terminus of the Fc portion of the antibody.

In some embodiments, vIgD is linked, directly or indirectly, to the N- or C-terminus of the light and/or heavy chain of the antibody. In some embodiments, the linkage may be via a peptide linker, such as any described above. Various configurations can be built. In fig. 10A-10C illustrate exemplary configurations. In some embodiments, an antibody conjugate can be produced by co-expressing an antibody heavy and light chain in a cell.

In one aspect of the present invention, the targeting agent is an aptamer molecule. For example, in some embodiments, the aptamer consists of nucleic acids that act as a targeting agent. In various embodiments, the IgSF conjugate of the invention includes an aptamer that is specific for a molecule in a tumor cell, tumor vascular network, and/or tumor microenvironment. In some embodiments, the aptamer itself may contain a biologically active sequence in addition to a target module (sequence), wherein the biologically active sequence may induce an immune response to the target cell. In other words, such an aptamer molecule is a dual-use agent. In some embodiments, the IgSF conjugate of the invention comprises conjugating an aptamer to an antibody, wherein the aptamer and antibody are specific for binding to individual molecules on a tumor cell, tumor vascular network, tumor microenvironment, and/or immune cells.

The term aptamer includes DNA, RNA or peptides that are selected based on specific binding properties to a particular molecule. For example, the aptamer(s) may be selected to bind a particular gene or gene product in a tumor cell, tumor vasculature, tumor microenvironment, and/or immune cell, as disclosed herein, where selection is accomplished by methods known in the art. in the art and familiar to one skilled in the art.

In some aspects of the present invention, the targeting agent is a peptide. For example, variant polypeptides or immunomodulatory proteins provided herein can be conjugated to a peptide that can bind to a component of malignant or tumor cells. Therefore, such IgSF conjugates of the invention include peptide-targeting agents that bind to a cellular component of a tumor cell, a tumor vascular network, and/or a component of the tumor microenvironment. In some embodiments, the targeting agent peptides may be an integrin antagonist or agonist. Integrins that include an alpha and a beta subunit include numerous types well known to one skilled in the art.

In one embodiment, the targeting agent is Vve3. The Vve3 integrin is expressed on a variety of cells and has been shown to mediate several biologically important processes, including osteoclast adhesion to bone matrix, vascular smooth muscle cell migration, and angiogenesis. Suitable target molecules for integrins include RGD peptides or peptidomimetics, as well as non-RGD peptides or peptidomimetics (see, e.g., US Pat. No. 5,767,771 and 5,780,426) for other integrins such as V4.ei (VLA-4), V4-P7 (see, for example, US Pat. No. 6365619, Chang et al., Bioorganic & Medicinal Chem Lett, 12: 159-163 (2002), Lin et al., Bioorganic & Medicinal Chem Lett, 12: 133-136 (2002)), and the like.

In some embodiments, an IgSF conjugate is provided, comprising a variant polypeptide or immunomodulatory protein provided herein, conjugated to a therapeutic agent. In some embodiments, the therapeutic agent includes, for example, daunomycin, doxorubicin, methotrexate and vindesine (Rowland et al., Cancer Immunol. Immunother. 21:183-187, 1986). In some embodiments, the therapeutic agent has intracellular activity. In some embodiments, the IgSF conjugate is internalized and the therapeutic agent is a cytotoxin that blocks protein synthesis of the cell, resulting in cell death. In some embodiments, the therapeutic agent is a cytotoxin containing a polypeptide having ribosome deactivating activity, including, for example, gelonin, buganin, saporin, ricin, ricin A chain, briodine, diphtheria toxin, restrictocin, Pseudomonas exutoxin A, and variants thereof. In some embodiments, where the therapeutic agent is a cytotoxin containing a polypeptide having ribosome deactivating activity, the IgSF conjugate must be internalized upon binding to the target cell for the protein to be cytotoxic to the cells.

In some embodiments, an IgSF conjugate is provided containing a variant polypeptide or immunomodulatory protein provided herein conjugated to a toxin. In some embodiments, the toxin includes, for example, bacterial toxins such as diphtheria toxin, plant toxins such as ricin, small molecule toxins such as geldanamycin (Mandler et al., J. Nat. Cancer Inst., 92 (19): 1573- 1581 (2000), Mandler et al., Bioorganic & Med. Chem. Letters 10: 1025-1028 (2000), Mandler et al., Bioconjugate Chem., 13: 786-791 (2002)), maytansinoids (EP 1391213, Liu et al., Proc. Natl. Acad. Sci. USA 93: 8618-8623 (1996)) and calicheamicin (Lode et al., Cancer Res., 58: 2928

- 63 044356 (1998), Hinman et al., Cancer Res., 53: 3336-3342 (1993)). Toxins may exert their cytotoxic and cytostatic effects through mechanisms including tubulin binding, DNA binding, or topoisomerase inhibition.

In some embodiments, there is provided an IgSF conjugate containing a variant polypeptide or immunomodulatory protein provided herein conjugated to a label that can generate a detectable signal, directly or indirectly. These IgSF conjugates can be used for research or diagnostic applications, such as in vivo detection of malignancy. The tag is preferably capable of producing, directly or indirectly, a detectable signal. For example, the label may be radioopaque or a radioisotope such as 3H, ^14C , ^32P , ^35S , ^123I , ^125I , ^131I ; a fluorescent (fluorophore) or chemiluminescent (chromophoric) compound such as fluorescein isothiocyanate, rhodamine or luciferin; an enzyme such as alkaline phosphatase, P-galactosidase or horseradish peroxidase; imaging agent; or a metal ion. In some embodiments, the label is a radioactive atom for scintigraphic studies, such as ⁹⁹ Tc or ¹²³ I, or a spin label for nuclear magnetic resonance (NMR) imaging (also known as magnetic resonance imaging, MRI), such as zirconium-89 , iodine-123, iodine-131, indium-111, fluorine-19, carbon-13, nitrogen-15, oxygen-17, gadolinium, manganese or iron. Zirconium-89 can be complexed with various metal chelating agents and conjugated to antibodies, for example for PET imaging (WO 2011/056983). In some embodiments, the IgSF conjugate is detected indirectly. For example, a secondary antibody specific for the IgSF conjugate and containing a detectable label can be used to detect the IgSF conjugate.

IgSF conjugates can be prepared using any methods known in the art. See, for example, WO 2009/067800, WO 2011/133886 and US Patent Application Publication No. 2014322129, incorporated herein by reference.

The variant polypeptides or immunomodulatory proteins of the IgSF conjugate can be attached to the effector moiety by any means by which the variant polypeptides or immunomodulatory proteins can be associated with, or linked to, the effector moiety. For example, variant polypeptides or immunomodulatory proteins of the IgSF conjugate can be attached to the effector moiety by chemical or recombinant means. Chemical means for producing fusion proteins or conjugates are known in the art and can be used to produce an IgSF conjugate. The method used to conjugate variant polypeptides or immunomodulatory proteins and an effector fragment must be capable of binding the variant polypeptides or immunomodulatory proteins to the effector fragment without interfering with the ability of the variant polypeptides or immunomodulatory proteins to bind to their one or more counterstructural ligands.

Variant polypeptides or immunomodulatory proteins of the IgSF conjugate may be associated indirectly with the effector fragment. For example, variant polypeptides or immunomodulatory IgSF conjugate proteins can be directly associated with a liposome containing one of several types of effector fragment. Effector fragment(s) and/or variant polypeptides or immunomodulatory proteins may also be associated with a solid surface.

In some embodiments, the variant polypeptides or immunomodulatory proteins of the IgSF conjugate and the effector fragment are both proteins and can be conjugated using methods well known in the art. There are several hundred cross-linking agents that can conjugate two proteins (see, for example, Chemistry of Protein Conjugation and Crosslinking, 1991, Shans Wong, CRC Press, Ann Arbor). The cross-linking agent is typically selected based on the reactive functional groups available or inserted on the variant polypeptides or immunomodulatory proteins and/or effector moiety. In addition, if there are no reactive groups, a photoactivatable cross-linking agent can be used. In some cases, it may be desirable to include a spacer between the variant polypeptides or immunomodulatory proteins and the effector fragment. Crosslinking agents known in the art include homobifunctional agents: glutaraldehyde, dimethyladipimidate and bis(diazobenzidine) and heterobifunctional agents: m maleimidobenzoyl-N-hydroxysuccinimide and sulfo-m maleimidobenzoyl-N-hydroxysuccinimide.

In some embodiments, the variant polypeptide or immunomodulatory IgSF conjugate proteins can be designed with specific residues for chemical attachment of the effector moiety. Specific residues used to chemically attach a molecule known in the art include lysine and cysteine. The cross-linking agent is selected based on the reactive functional groups inserted into the variant polypeptides or immunomodulatory proteins and available on the effector moiety.

The IgSF conjugate can also be produced using recombinant DNA techniques. In this case, the DNA sequence encoding variant polypeptides or immunomodulatory proteins is fused with the DNA sequence encoding the effector part, resulting in the formation

- 64 044356 hybrid DNA molecule. The fusion DNA sequence is transfected into a host cell that expresses the fusion protein. The fusion protein can be isolated from cell culture and purified using methods known in the art.

Examples of attaching an effector moiety that is a tag to variant polypeptides or immunomodulatory proteins include the methods described in Hunter, et al., Nature 144:945 (1962); David, et al., Biochemistry 13:1014 (1974); Pain, et al., J. Immunol. Meth. 40:219 (1981); Nygren, J. Histochem. and Cytochem. 30:407 (1982); Wensel and Meares, Radioimmunoimaging And Radioimmunotherapy, Elsevier, N.Y. (1983); and Colcher et al., Use Of Monoclonal Antibodies As Radiopharmaceuticals For The Localization Of Human Carcinoma Xenografts In Athymic Mice, Meth. Enzymol., 121:802-16 (1986).

Radioactive or other labels can be included in the conjugate by known methods. For example, the peptide may be biosynthesized or may be synthesized by chemical amino acid synthesis using suitable amino acid precursors including, for example, fluorine19 instead of hydrogen. Tags such as 99 Tc or ¹²³ I, ¹⁸⁶ Re, ¹⁸⁸ Re and ⁿ¹ In can be attached through a cysteine residue in the peptide. Yttrium-90 can be attached via a lysine residue. The IODOGEN method (Fraker et al., Biochem. Biophys. Res. Commun. 80:49-57 (1978)) can be used to incorporate iodine-123. Monoclonal Antibodies in Immunoscintigraphy (Chatal, CRC Press 1989) details other methods.

Conjugates of variant polypeptides or immunomodulatory proteins and a cytotoxic agent can be prepared using various bifunctional protein coupling agents such as N-succinimidyl-3-(2-pyridyldithio)propionate (SPDP), succinimidyl-4-(Mmaleimidomethyl)cyclohexane-1-carboxylate (SMCC), iminothiolane (IT), bifunctional imidoester derivatives (such as dimethyladipimidate HCl), active esters (such as disuccinimidyl suberate), aldehydes (such as glutaraldehyde), bis-azido compounds (such as bis(pazidobenzoyl)hexanediamine), bis-diazonium derivatives (such as bis-(p-diazoniumbenzoyl)ethylenediamine), diisocyanates (such as toluene-2,6-diisocyanate) and bis-active fluorine compounds (such as 1,5-difluoro-2,4-dinitrobenzene) . For example, the immunotoxin ricin can be prepared as described in Vitetta et al., Science 238: 1098 (1987). Carbon-14-labeled 1-isothiocyanatobenzyl-3-methyldiethylenetamine pentaacetic acid (MX-DTPA) is an exemplary chelating agent for radionucleotide-antibody conjugation. See, for example, WO 94/11026. The linker may be a cleavable linker that facilitates release of the cytotoxic drug into the cell. For example, an acid-labile linker, a peptidase-sensitive linker, a photolabile linker, a dimethyl linker, or a disulfide-containing linker can be used (Chari et al., Cancer Research 52: 127-131 (1992), US Pat. No. 5202020).

The IgSF conjugates of the present invention explicitly include, but are not limited to, drug conjugates prepared using cross-linking reagents: Bmps, EMCS, GMBS, HBVS, LC-SMCC, MBS, MPBH, SBAP, SIA, SIAB, SMCC, SMPB, SMPH, sulfoEMCS, sulfo-GMBS, sulfo-KMUS, sulfo-MBS, sulfo-SIAB, sulfo-SMCC and sulfo-SMPB and SVSB (succinimidyl (4-vinyl sulfone) benzoate), which are commercially available (for example, from Pierce Biotechnology , Inc., Rockford, Illinois, USA). See pp. 467-498, 2003-2004 Applications Handbook and Catalog.

D. Transmembrane and secreted immunomodulatory proteins and engineered cells.

Provided herein are engineered cells that express immunomodulatory variant ICOSL polypeptides (optionally engineered cells). In some embodiments, the expressed immunomodulatory variant ICOSL polypeptide is a transmembrane protein and is surface expressed. In some embodiments, the expressed immunomodulatory variant ICOSL polypeptide is expressed and secreted from the cell.

1. Transmembrane immunomodulatory proteins.

In some embodiments, the immunomodulatory polypeptide containing the variant ICOSL may be a membrane-bound protein. As described in more detail below, the immunomodulatory polypeptide may be a transmembrane immunomodulatory polypeptide comprising a variant ICOSL comprising: an ectodomain containing at least one affinity modified IgSF domain (IgV or IgC), a transmembrane domain, and optionally a cytoplasmic domain. In some embodiments, the transmembrane immunomodulatory protein may be expressed on the surface of an immune cell, such as a mammalian cell, including on the surface of a lymphocyte (eg, T cell or NK cell) or antigen presenting cell. In some embodiments, the transmembrane immunomodulatory protein is expressed on the surface of a mammalian T cell, including T cells such as: T helper cell, cytotoxic T cell (alternatively cytotoxic T lymphocyte or CTL), natural killer T cell, regulatory T cell, Memory T cell or gamma delta T cell. In some embodiments, the mammalian cell is an antigen presenting cell (APC). Usually, but not exclusively,

- 65 044356 ectodomain (optionally extracellular domain) includes one or more amino acid variations (eg amino acid substitutions) of the ICOSL variant of the invention. Thus, for example, in some embodiments, the transmembrane protein will comprise an ectodomain that includes one or more amino acid substitutions in a variant ICOSL of the invention.

In some embodiments, the engineered cells express variant ICOSL polypeptides that are transmembrane immunomodulatory polypeptides (TIPS), which may be a membrane protein, such as a transmembrane protein. In typical embodiments, the ectodomain of the membrane protein includes an extracellular domain or an IgSF domain thereof of a variant ICOSL provided herein that contains one or more amino acid substitutions in at least one IgSF domain as described. The transmembrane immunomodulatory proteins provided herein further comprise a transmembrane domain linked to an ectodomain. In some embodiments, the transmembrane domain provides an encoded protein for cell surface expression. In some embodiments, the transmembrane domain is linked directly to the ectodomain. In some embodiments, the transmembrane domain is linked indirectly to the ectodomain through one or more linkers or spacers. In some embodiments, the transmembrane domain comprises predominantly hydrophobic amino acid residues such as leucine and valine.

In some embodiments, a full-length transmembrane anchor domain may be used to ensure that TIPs are expressed on the surface of an engineered cell, such as an engineered T cell. Conveniently, this domain may be from a particular native protein that is affinity modified (e.g. CD80 or ICOSL or another native IgSF protein) and which is simply fused to the first membrane proximal domain sequence in a manner similar to the native IgSF protein (e.g. CD80 or ICOSL ). In some embodiments, the transmembrane immunomodulatory protein comprises the transmembrane domain of the corresponding wild type or unmodified IgSF member, such as the transmembrane domain contained in the amino acid sequence set forth in SEQ ID NO: 5 (Table 2). In some embodiments, the membrane-associated form includes the transmembrane domain of the corresponding wild-type or unmodified polypeptide, such as the corresponding residues 257-277 of SEQ ID NO: 5.

In some embodiments, the transmembrane domain is a non-native transmembrane domain that is not the transmembrane domain of the native ICOSL. In some embodiments, the transmembrane domain is derived from a transmembrane domain from another ICOSL family member polypeptide that is membrane or is a transmembrane protein. In some embodiments, a transmembrane anchor domain from another protein on T cells may be used. In some embodiments, the transmembrane domain is derived from CD8. In some embodiments, the transmembrane domain may further comprise an extracellular portion of CD8 that serves as a spacer domain. An exemplary transmembrane domain derived from CD8 is provided in SEQ ID NO: 246 or 483 or a portion thereof containing the CD8 transmembrane domain. In some embodiments, the transmembrane domain is a synthetic transmembrane domain.

In some embodiments, the transmembrane immunomodulatory protein further includes an endodomain, such as a cytoplasmic signaling domain associated with a transmembrane domain. In some embodiments, the cytoplasmic signaling domain induces cellular signaling. In some embodiments, the endodomain of the transmembrane immunomodulatory protein includes the cytoplasmic domain of a corresponding wild-type or unmodified polypeptide, such as the cytoplasmic domain contained in the amino acid sequence set forth in SEQ ID NO: 5 (see Table 2).

In some embodiments, a proposed transmembrane immunomodulatory protein that is or includes a variant ICOSL comprising an amino acid sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97 , 98 or 99% of the sequence of SEQ ID NO: 257 and includes an ectodomain containing at least one IgSF affinity modified ICOSL domain as described and a transmembrane domain. In some embodiments, the transmembrane immunomodulatory protein includes any one or more amino acid substitutions in the IgSF domain (eg, IgV domain) as described, including any listed in table. 1. In some embodiments, the transmembrane immunomodulatory protein may further comprise a cytoplasmic domain, as described. In some embodiments, the transmembrane immunomodulatory protein may further include a signal peptide. In some embodiments, the signal peptide is a native signal peptide from a wild-type IgSF representative, such as those contained in the amino acid sequence set forth in SEQ ID NO: 5 (see, for example, Table 2).

In some embodiments, the proposed are transmembrane immunomodulatory proteins containing amino acid substitutions

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R, N52H/N57Y/Q100R or

- 66 044356

N52H/N57Y/Q100P. In some embodiments, the provided transmembrane immunomodulatory protein is or includes a variant ICOSL comprising the amino acid sequence set forth in SEQ ID NO: 257 but containing amino acid substitutions

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R, N52H/N57Y/Q100R or

N52H/N57Y/Q100P at the corresponding positions in SEQ ID NO: 257 or in an amino acid sequence that is at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% identity to SEQ ID NO: 257 and includes amino acid substitutions E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R, N52H/N57Y/Q100R or

N52H/N57Y/Q100P.

In some embodiments, the present invention is a transmembrane immunomodulatory proteins containing the amino acid sequence set forth in SEQ ID NO: 496 or 497 (each including amino acid substitution N52D), SEQ ID NO: 498 or 499 (each including amino acid substitution N52H/N57Y /Q100P), SEQ ID NO: 500 or 501 (each including amino acid substitutions E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R) or SEQ ID NO: 502 or 503 (each including amino acid substitution N52H/N57Y/Q100R) or an amino acid sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% sequence identity to any of SEQ ID NO: 495-503 and which includes the specified amino acid substitutions. In some embodiments, when expressed in an engineered cell, such transmembrane immunomodulatory proteins are expressed on the surface of the cell.

The present invention also provides a nucleic acid molecule encoding such transmembrane immunomodulatory proteins. In some embodiments, the nucleotide molecule encoding a transmembrane immunomodulatory protein includes a nucleotide sequence that encodes an amino acid sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96 , 97, 98 or 99% identity to the sequence of SEQ ID NO: 257 and includes an ectodomain containing at least one affinity modified IgSF domain as described, a transmembrane domain and, optionally, a cytoplasmic domain. In some embodiments, the nucleotide molecule may further comprise a nucleotide sequence encoding a signal peptide. In some embodiments, the signal peptide is the native signal peptide of the corresponding wild-type IgSF representative (see, for example, Table 2).

An exemplary transmembrane immunomodulatory protein is an ICOSL TIP comprising i) an amino acid sequence set forth in SEQ ID NO: 383 or ii) an amino acid sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92 , 93, 94, 95, 96, 97, 98 or 99% identity with SEQ ID NO: 243 and which includes the affinity modified domain contained in SEQ ID NO: 243, or amino acid substitutions therein.

Also provided is i) a nucleotide sequence set forth in SEQ ID NO: 384, ii) a sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% of the sequence of SEQ ID NO: 244 and which encodes a TIP that includes an affinity modified domain of SEQ ID NO: 243 or iii) a sequence of i) or ii) with degenerate codons.

In some embodiments, CAR-related transmembrane immunomodulatory proteins are provided, wherein the endodomain of the transmembrane immunomodulatory protein includes a cytoplasmic signaling domain that includes at least one ITAM (immunoreceptor tyrosine activating motif)-containing signaling domain. ITAM is a conserved motif found in a number of protein signaling domains involved in immune cell signal transduction, including the CD3 zeta chain (CD3-z) involved in T cell receptor signal transduction. In some embodiments, the endodomain includes a CD3-zeta signaling domain. In some embodiments, the CD3 zeta signaling domain comprises the amino acid sequence set forth in SEQ ID NO: 243, or an amino acid sequence that is at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94 , 95, 96, 97, 98 or 99% to SEQ ID NO: 247 and retains T cell signaling activity. In some embodiments, the endodomain of the CAR-associated transmembrane immunomodulatory protein may further comprise a co-stimulatory signaling domain to further modulate T cell immunomodulatory responses. In some embodiments, the costimulatory signaling domain is CD28, ICOS, 41BB, or OX40. In some embodiments, the co-stimulatory signaling domain is derived from CD28 or 4-1BB and includes an amino acid sequence set forth in any of SEQ ID NOs: 484-487, or an amino acid sequence that is at least 85, 86, 87, 88. 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% identity with SEQ ID NO: 484-487 and retains co-stimulatory T cell signaling activity. In some embodiments, the provided CAR-associated transmembrane immunomodulatory proteins have CAR features to promote T cell trafficking upon binding of an affinity-modified IgSF domain to a cognate binding partner or counterstructure. In some embodiments, the specific binding of an IgSF affinity modified domain to its counterstructure may result in changes in the immunological activity of T cells, reflected in changes in cytotoxicity, proliferation, or cytokine production.

In some embodiments, the transmembrane immunomodulatory protein does not include an endodomain capable of mediating cytoplasmic signaling. In some embodiments, the transmembrane immunomodulatory protein has lost the signal transduction mechanism of the wild-type or unmodified polypeptide and therefore does not itself induce cellular signaling. In some embodiments, the transmembrane immunomodulatory protein does not have an intracellular (cytoplasmic) domain or part of the intracellular domain of a corresponding wild-type or unmodified polypeptide, such as a cytoplasmic signaling domain contained in the amino acid sequence set forth in SEQ ID NO: 5 (see Table 2) . In some embodiments, the transmembrane immunomodulatory protein does not include an ITIM (immunoreceptor tyrosine inhibitory motif) such as that found in certain inhibitory receptors, including the IgSF family inhibitory receptors (eg, PD-1 or TIGIT). Thus, in some embodiments, the transmembrane immunomodulatory protein includes only an ectodomain and a transmembrane domain, such as any of those described.

2. Secreted immunomodulatory proteins and engineered cells.

In some embodiments, a variant ICOSL immunomodulatory polypeptide containing one or more amino acid mutations, as described herein, is secreted, for example, when expressed from a cell. This variant immunostimulatory protein ICOSL does not include a transmembrane domain. In some embodiments, the variant immunomodulatory protein ICOSL is not conjugated to a half-life extension moiety (such as an Fc domain or multimerization domain). In some embodiments, the variant ICOSL immunomodulatory protein includes a signal peptide, such as an antibody signal peptide or other effective signal sequence for obtaining extracellular domains. When the immunomodulatory protein includes a signal peptide and is expressed by the engineered cell, the signal peptide causes the engineered cell to secrete the immunomodulatory protein. Typically, the signal peptide or part of the signal peptide is cleaved from the immunomodulatory protein upon secretion. The immunomodulatory protein may be encoded by a nucleic acid (which may be part of an expression vector). In some embodiments, the immunomodulatory protein is expressed and secreted by a cell (such as an immune cell, e.g., a primary immune cell).

Thus, in some embodiments, variant ICOSL immunomodulatory proteins are provided that further comprise a signal peptide. In some embodiments, provided herein is a nucleotide molecule encoding a variant immunomodulatory protein ICOSL operably linked to a secretion sequence encoding a signal peptide.

A signal peptide is a sequence at the N-terminus of an immunomodulatory protein that signals the secretion of the immunomodulatory protein from the cell. In some embodiments, the signal peptide is about 5 to about 40 amino acids in length (e.g., about 5 to about 7, about 7 to about 10, about 10 to about 15, about 15 to about 20, about 20 to about 25, or about 25 to about 30, about 30 to about 35, or about 35 to about 40 amino acids).

In some embodiments, the signal peptide is a native signal peptide from the corresponding wild-type ICOSL (Table 6). In some embodiments, the signal peptide is a non-native signal peptide. For example, in some embodiments, the non-native signal peptide is a mutant of the native signal peptide from the corresponding wild-type ICOSL and may include one or more (such as 2, 3, 4, 5, 6, 7, 8, 9, or 10 or more) substitutions, insertions or deletions. In some embodiments, the non-native signal peptide is a signal peptide or a mutant of a family member from the same IgSF family as the wild-type IgSF family member. In some embodiments, the non-native signal peptide is a signal peptide or mutant from an IgSF family member of a different IgSF family than the wild-type IgSF family member. In some embodiments, the signal peptide is a signal peptide or mutant thereof not from the IgSF family of proteins, such as a signal peptide from an immunoglobulin (such as IgG heavy chain or IgG kappa light chain), a cytokine (such as interleukin-2 (IL-2) or CD33), serum albumin (eg, HSA or albumin), human azurocidin precursor signal sequence, luciferase, trypsinogen (eg, chymotrypsinogen or trypsinogen), or other signal peptide capable of efficiently secreting a protein from a cell. Exemplary signal peptides include any described in table. 6.

- 68 044356

Table 6

Exemplary signal peptides

SEQ ID NO Signal peptide Peptide sequence SEQ ID NO: 346 HSA signal peptide MKWVTFISLLFLFSSAYS SEQ ID NO: 347 Ig light chain kappa MDMRAPAGIFGFFLLVLFPGYRS SEQ ID NO: 348 Signal Azurocidin preprotein sequence person MTRLT VL ALL AGLL AS SRA SEQ ID NO: 349 signal peptide heavy chain IgG MELGLSWIFLLAILKGVQC SEQ ID NO: 350 signal peptide heavy chain IgG MELGLRWVFLVAIEGVQC SEQ ID NO: 351 signal peptide heavy chain IgG MKHLWFFLLLVAAPRWVLS SEQ ID NO: 352 signal peptide heavy chain IgG MDWTWRILFLVAATGAHS SEQ ID NO: 353 signal peptide heavy chain IgG MDWTWRFLFVVAAATGVQS SEQ ID NO: 354 signal peptide heavy chain IgG MEFGLSWLFLVAILKGVQC SEQ ID NO: 355 signal peptide heavy chain IgG MEFGLSWVFLVALFRGVQC SEQ ID NO: 356 signal peptide heavy chain IgG MDLLHKNMKHLWFFLLLVAAPRWVLS SEQ ID NO: 357 IgG kappa light chain signal sequence MDMRVPAQLLGLLLLWLSGARC SEQ ID NO: 358 signal IgG kappa light chain sequence MKYLLPTAAAGLLLLLAAQPAMA SEQ ID NO: 359 Gaussia luciferase MGVKVLFALICIAVAEA SEQ ID NO: 360 Human albumin MKWVTFISLLFLFSSAYS SEQ ID NO: 361 human chymotrypsinogen MAFLWLLSCWALLGTTFG SEQ ID NO: 362 human interleukin-2 MQLLSCIALILALV SEQ ID NO: 363 Trypsinogen-2 humans MNLLLILTFVAAAVA

In some embodiments of the secreted variant immunomodulatory protein ICOSL, the immunomodulatory protein includes a signal peptide upon expression, and the signal peptide (or portion thereof) is cleaved from the immunomodulatory protein upon secretion.

In some embodiments, the engineered cells express variant ICOSL polypeptides that are secreted from the cell. In some embodiments, such a variant ICOSL polypeptide is encoded by a nucleotide molecule encoding an immunomodulatory protein under the functional control of a secretion signal sequence. In some embodiments, the encoded immunomodulatory protein is secreted when expressed from a cell. In some embodiments, the immunomodulatory protein encoded by the nucleotide molecule does not include a transmembrane domain. In some embodiments, the immunomodulatory protein encoded by the nucleotide molecule does not include a half-life extending moiety (such as an Fc domain or multimerization domain). In some embodiments, the immunomodulatory protein encoded by the nucleotide molecule includes a signal peptide. In some embodiments, the nucleic acid of the invention further includes a nucleotide sequence that encodes a secretory or signal peptide operably linked to the nucleic acid encoding an immunomodulatory protein, thereby allowing secretion of the immunomodulatory protein.

3. Cells and engineered cells.

Provided herein are engineered cells expressing any of the present immunomodulatory polypeptides. In some embodiments, the engineered cells express on their surface any of the provided transmembrane immunomodulatory polypeptides. In some embodiments, the engineered cells express and are capable of or may secrete an immunomodulatory protein from the cells under conditions suitable for protein secretion. In some embodiments, the transmembrane immunomodulatory protein is expressed on a lymphocyte, such as a tumor infiltrating lymphocyte (TIL), T cell or NK cell, or on a myeloid cell. In some embodiments, the engineered cells are antigen presenting cells (APCs). In some embodiments, the engineered cells are engineered mammalian T cells or engineered mammalian antigen presenting cells (APCs). In some embodiments, the engineered T cells or APCs are human or murine cells.

- 69 044356

In some embodiments, the engineered T cells include, but are not limited to, a helper T cell, a cytotoxic T cell (optionally a cytotoxic T lymphocyte or CTL), a natural killer T cell, a regulatory T cell, a memory T cell, or gamma delta T cell. In some embodiments, the engineered T cells are CD4+ or CD8+. In addition to the MHC signal, the engineered T cells also require a co-stimulatory signal, which in some embodiments is provided by the variant transmembrane immunomodulatory polypeptide ICOSL expressed in a membrane-bound form, as discussed previously.

In some embodiments, engineered APCs include, for example, MHC II-expressing APCs, such as macrophages, B cells, and dendritic cells, and artificial APCs (aAPCs), including both cellular and acellular (eg, biodegradable polymer microparticles) aAPCs. Artificial APCs (aAPCs) are synthetic versions of APCs that can act similarly to APCs in that they present antigens to T cells and activate them. Antigen presentation is carried out by the MHC (class I or class II). In some aspects, in engineered APCs, such as aAPC, the antigen that is loaded into the MHC is, in some embodiments, a tumor-specific antigen or a tumor-containing antigen. The antigen loaded into the MHC is recognized by the T cell receptor (TCR) of the T cell, which in some cases may express ICOS, CD28, or another molecule recognized by the variant ICOSL polypeptides provided herein. Materials that can be used to develop aAPC include: poly(glycolic acid), poly(lactic-co-glycolic acid), iron oxide, liposomes, lipid bilayers, Sepharose and polystyrene.

In some embodiments of the invention, the immunomodulatory protein, such as a transmembrane immunomodulatory protein or a secreted immunomodulatory protein provided herein, is coexpressed or engineered in a cell that expresses an antigen binding receptor, such as a recombinant receptor such as a chimeric antigen receptor (CAR) or T receptor -cells (TCR). In some embodiments, the engineered cell, such as an engineered T cell, recognizes a target antigen associated with cancer, inflammatory and autoimmune disorders, or viral infection. In specific embodiments, the antigen binding receptor includes an antigen binding moiety that specifically binds a tumor-specific antigen or tumor-associated antigen. In some embodiments, the engineered T cell is a CAR (chimeric antigen receptor) T cell that includes an antigen binding domain (eg, scFv) that specifically binds to an antigen, such as a tumor-specific antigen or tumor-associated antigen. In some embodiments, the antigen binding domain (eg, scFv) is specific for CD19. An example of a CAR is an anti-CD19 CAR, such as a CAR containing an anti-CD19 scFv set forth in SEQ ID NO: 482 or SEQ ID NO: 245. In some embodiments, the TIP protein is expressed in an engineered T cell receptor cell or in an engineered cell with chimeric antigen receptor. In such embodiments, the engineered cell co-expresses TIP and a CAR or TCR.

In some embodiments, the CAR further includes a spacer or hinge, a transmembrane domain, and an intracellular signaling domain (endodomain) comprising an ITAM signaling domain, for example, a CD3 zeta signaling domain. In some embodiments of the invention, the CAR includes a co-stimulatory signaling domain. In some embodiments, a spacer or hinge is present between the antigen binding domain and the transmembrane domain, for example, between the antigen binding domain and the plasma membrane when expressed in a cell. In some embodiments, the spacer or hinge is derived from a subclass of IgG (such as IgG1 and IgG4, IgD, or CD8 (see, e.g., Qin et al. (2017) J. Hematol. Oncol. 10:68). In some embodiments, the spacer or the hinge is derived from IgG1.

In some embodiments, the spacer and transmembrane domain are a hinge and transmembrane domain that are derived from CD8, for example, as set forth in SEQ ID NO: 246 or 483 or an amino acid sequence that is at least 85, 86, 87, 88, 89, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or greater sequence identity to SEQ ID NO: 246 or 483. In some embodiments, the endodomain includes a CD3 zeta signaling domain. In some embodiments, the CD3 zeta signaling domain comprises the amino acid sequence set forth in SEQ ID NO: 387, or an amino acid sequence that is at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94 , 95, 96, 97, 98 or 99% identity with SeQ ID NO: 387 and retains T cell signaling activity. In some embodiments, the endodomain of the CAR-associated transmembrane immunomodulatory protein may further comprise a co-stimulatory signaling domain to further modulate T cell immunomodulatory responses. In some embodiments, the costimulatory signaling domain is CD28, ICOS, 41BB, or OX40. In some embodiments, the co-stimulatory signaling domain is derived from CD28 or 4-1BB and includes an amino acid sequence set forth in any of SEQ ID NO: 484-487, or an amino acid sequence that is at least 85, 86, 87, 88, 89 , 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% identity with SEQ ID NO: 484-487 and retains co-stimulatory T cell signaling activity.

- 70 044356

In another embodiment, the engineered T cell has a TCR, including a recombinant or engineered TCR. In some embodiments, the TCR may be a native TCR. Those skilled in the art will appreciate that generally native mammalian T cell receptors contain an alpha and beta chain (or a gamma and delta chain) involved in specific antigen recognition and binding. In some embodiments, the TCR is an engineered TCR. In some embodiments, the TCR of the engineered T cell specifically binds to a tumor-associated or tumor-specific antigen represented by an APC.

In some embodiments, immunomodulatory polypeptides, such as transmembrane immunomodulatory polypeptides or secreted immunomodulatory polypeptides, can be incorporated into engineered cells, such as engineered T cells or engineered APCs, using various strategies such as those used for recombinant host cells. There are many methods known in the art for introducing DNA constructs into primary T cells. In some embodiments, viral transduction or plasmid electroporation is used. In typical embodiments, the nucleotide molecule encoding the immunomodulatory protein or expression vector includes a signal peptide that localizes the expressed transmembrane immunomodulatory proteins to the cell membrane or for secretion. In some embodiments, the nucleic acid encoding the transmembrane immunomodulatory proteins of the invention is subcloned into a viral vector, such as a retroviral vector, that allows expression in a mammalian host cell. The expression vector can be introduced into a mammalian host cell, and under culture conditions of the host cells, the immunomodulatory protein is surface expressed or secreted.

In a typical example, primary T cells can be purified ex vivo (CD4 cells or CD8 cells or both) and stimulated with an activation protocol consisting of various TCR/CD28 agonists, for example, using anti-CD3/anti-coated beads. -CD28. After a 2- or 3-day activation process, the recombinant expression vector containing the immunomodulatory polypeptide can be stably introduced into primary T cells using lentiviral or retroviral transduction protocols standard in the art or plasmid electroporation strategies. Cells can be monitored for expression of the immunomodulatory polypeptide, for example, by flow cytometry using antibodies against an epitope tag or antibodies that cross-react with the native parent molecule and polypeptides comprising the variant ICOSL. T cells that express an immunomodulatory polypeptide can be enriched by sorting with antibodies against an epitope tag or can be enriched for high or low expression depending on the application.

Upon expression of an immunomodulatory polypeptide, the engineered T cells can be analyzed for their respective function by various means. Co-expression of the engineered CAR or TCR can be confirmed to show that that portion of the engineered T cell has not significantly affected the expression of the immunomodulatory protein. Once confirmed, standard in vitro cytotoxicity, proliferation, or cytokine production assays (eg, IFNγ expression) can be used to assess the function of the engineered T cells. Typical standard endpoints are percentage of tumor line lysis, engineered T cell proliferation, or IFN-gamma protein expression in culture supernatants. An engineered construct may be selected that results in a statistically significant increase in tumor line lysis, increase in engineered T Cell proliferation, or increase in IFN-gamma expression compared to the control construct. In addition, crude cells, such as native primary or endogenous T cells, can also be included in the same in vitro assay to measure the ability of an immunomodulatory peptide expressed in engineered cells, such as engineered T cells, to modulate activity, including , in some cases, modulate the activation and generation of effector function in naïve bystander T cells. Increased expression of activation markers such as CD69, CD44 or CD62L may be monitored on endogenous T cells, and increased proliferation and/or cytokine production may indicate the desired activity of TIPs expressed on engineered T cells.

In some embodiments, similar assays can be used to compare the function of engineered T cells containing a CAR or TCR alone with those containing a CAR or TCR and a TIP construct. Typically, these in vitro assays are performed by seeding together in culture different ratios of engineered T cells and a tumor cell line containing a cognate CAR or TCR antigen. Standard endpoints include percentage of tumor line lysis, proliferation of engineered T cells, or IFN-γ production in culture supernatants. An engineered immunomodulatory protein may be selected that produces a statistically significant increase in tumor line lysis, increase in engineered T cell proliferation, or increase in IFN-gamma production compared to the same TCR or CAR construct alone. Engineered human T cells can be analyzed

- 71 044356 sed in immunocompromised mice, such as the NSG strain, which have lost murine T, NK and B cells. Engineered human T cells in which a CAR or TCR binds a target counterstructure on a xenograft and is coexpressed with an affinity-modified IgSF domain TIP can be adoptively transferred in vivo in different numbers and ratios of cells compared to the xenograft. For example, engraftment of CD19+ leukemia tumor lines containing a luciferase/GFP vector can be monitored by bioluminescence or ex vivo by flow cytometry. In a typical embodiment, the xenograft is introduced into a mouse model, followed by engineered T cells a few days later. Engineered T cells containing an immunomodulatory protein can be analyzed for increased survival, tumor clearance, or the number of engineered T cells expanded compared to engineered T cells containing only a CAR or TCR. As in the in vitro assay, endogenous, naïve (i.e., non-engineered) human T cells can be co-adoptively transferred to search for successful epitope spread in this population, resulting in better survival or tumor clearance.

E. Infectious agents expressing variant polypeptides and immunomodulatory proteins.

Also provided are infectious agents that contain nucleic acids encoding any of the variant polypeptides, such as ICOSL vIgD polypeptides, including secreted or transmembrane immunomodulatory proteins described in the present invention. In some embodiments, such infectious agents can deliver nucleic acids encoding variant immunomodulatory polypeptides described herein, such as ICOSL vIgD polypeptides, to a target cell in the subject, such as an immune cell and/or antigen presenting cell (APC) or tumor cell in the subject . Also provided are nucleic acids contained in such infectious agents and/or nucleic acids for producing or modifying such infectious agents, such as vectors and/or plasmids, and compositions containing such infectious agents.

In some embodiments, the infectious agent is a microorganism or microbe. In some embodiments, the infectious agent is a virus or bacteria. In some embodiments, the infectious agent is a virus. In some embodiments, the infectious agent is a bacterium. In some embodiments, such infectious agents can deliver nucleotide sequences encoding any of the variant polypeptides, such as ICOSL vIgD polypeptides, including the secreted or transmembrane immunomodulatory proteins described herein. Thus, in some embodiments, a cell in a subject that is infected or exposed to infectious agents can be made to express or secrete variant immunomodulatory polypeptides on the cell surface. In some embodiments, the infectious agent may also deliver one or more other therapeutic agents or nucleic acids encoding other therapeutic agents to the cell and/or environment within the entity. In some embodiments, other therapeutic agents that can be delivered by infectious agents include cytokines or other immunomodulatory molecules.

In some embodiments, the infectious agent, such as a virus or bacteria, includes nucleotide sequences that encode any of the variant polypeptides, such as ICOSL vIgD polypeptides, including secreted or transmembrane immunomodulatory proteins described herein, and by virtue of contact and/or infection of a cell in a subject, the cell expresses variant polypeptides, such as ICOSL vIgD polypeptides, including secreted or transmembrane immunomodulatory proteins encoded by nucleotide sequences contained in the infectious agent. In some embodiments, the infectious agent may be administered to a subject. In some embodiments, the infectious agent can be introduced into cells from an ex vivo site.

In some embodiments of the present invention, variant polypeptides, such as ICOSL vIgD polypeptides, including transmembrane immunomodulatory proteins expressed in a cell infected with an infectious agent, is a transmembrane protein and is expressed on the surface of the cell. In some embodiments, variant polypeptides, such as ICOSL vIgD polypeptides, including secreted immunomodulatory proteins expressed by a cell infected with an infectious agent are expressed and secreted from the cell. The transmembrane immunomodulatory protein or secreted immunomodulatory protein may be any one described herein.

In some embodiments of the present invention, the cells in the subject that are targeted by the infectious agent include a tumor cell, an immune cell, and/or an antigen presenting cell (APC). In some embodiments, the infectious agent targets a cell in the tumor microenvironment (TME). In some embodiments, the infectious agent delivers nucleic acids encoding variant polypeptides, such as ICOSL vIgD polypeptides, including secreted or transmembrane immunomodulatory proteins, into an appropriate cell (e.g., an APC, such as a cell that presents a peptide/MHC complex on its cell surface, such like dendritic

- 72 044356 cell) or tissue (for example, lymphoid tissue) that will induce and/or increase the desired effect, for example immunomodulation and/or a specific immune response to cells, for example, CD4 and/or CD8 T cell response, where CD8 T -cellular response may include a cytotoxic T-cell response (CTL). In some embodiments, the infectious agent targets an APC, such as a dendritic cell (DC). In some embodiments, the nucleotide molecule delivered by the infectious agents described herein includes the appropriate nucleotide sequences necessary for the expression of operably linked sequences encoding variant immunomodulatory polypeptides in a particular target cell, for example, regulatory elements such as promoters.

In some embodiments, an infectious agent that includes nucleotide sequences encoding immunomodulatory polypeptides may also contain nucleotide sequences that encode one or more additional gene products, such as cytokines, prodrug converting enzymes, cytotoxins, and/or detectable gene products. For example, in some embodiments, the infectious agent is an oncolytic virus, and the virus may include nucleotide sequences encoding additional therapeutic gene products (see, e.g., Kirn et al., (2009) Nat Rev Cancer 9: 64-71; Garcia-Aragoncillo et al., (2010) Curr Opin Mol Ther 12: 403-411, see US Pat. Nos. 7588767, 7588771, 7662398 and 7754221 and U.S. Pub. No. 2007/0202572, 2007/0212727, 2010/0062016 , 2009 /0098529, 2009/0053244, 2009/0155287, 2009/0117034, 2010/0233078, 2009/0162288, 2010/0196325, 2009/0136917 and 2011/0064650. In some embodiments, the additional gene product may be a therapeutic gene product that may lead to to the death of a target cell (eg, a tumor cell) or gene products that can enhance or stimulate or regulate the immune response (eg, a cytokine).Examples of gene products also include an antitumor agent, antimetastatic agent, an antiangiogenic agent, an immunomodulatory molecule, an immune checkpoint inhibitor , antibody, cytokine, growth factor, antigen, cytotoxic gene product, apoptotic gene product, anti-apoptotic gene product, cell matrix degradation gene, genes for tissue regeneration and reprogramming human somatic cells to pluripotency and other genes described herein or known to one skilled in the art. this area. In some embodiments, the additional gene product is granulocyte-macrophage colony-stimulating factor (GM-CSF).

1. Viruses.

In some embodiments, the infectious agent is a virus. In some embodiments, the infectious agent is an oncolytic virus or a virus that targets specific cells, such as immune cells. In some embodiments, the infectious agent targets a tumor cell and/or a malignant tumor cell in a subject. In some embodiments, the infectious agent targets an immune cell or antigen presenting cell (APC).

In some embodiments, the infectious agent is an oncolytic virus. Oncolytic viruses are viruses that accumulate in tumor cells and replicate in tumor cells. By virtue of replication in tumor cells and optional delivery of nucleic acids encoding variant ICOSL polypeptides or immunomodulatory polypeptides described herein, the tumor cells are lysed and the tumor shrinks and can be eliminated. Oncolytic viruses can also have a wide range of hosts and cell types. For example, oncolytic viruses can accumulate in immunoprivileged cells or immunoprivileged tissues, including tumors and/or metastases, and also including wound tissues and cells, allowing the delivery and expression of nucleic acids encoding variant immunomodulatory polypeptides described herein in a wide range of types cells. Oncolytic viruses can also replicate in a tumor cell-specific manner, resulting in tumor cell lysis and efficient tumor regression.

Exemplary oncolytic viruses include adenoviruses, adeno-associated viruses, herpes viruses, herpes simplex virus, vesicular stomatitis virus, reovirus, Newcastle disease virus, parvovirus, measles virus, vesicular stomatitis virus (VSV), coxsackie virus, and vaccinia virus. In some embodiments, oncolytic viruses can specifically colonize solid tumors without infecting other organs, and can be used as an infectious agent to deliver nucleic acids encoding the variant immunomodulatory polypeptides described herein to such solid tumors.

Oncolytic viruses for use in nucleic acid delivery encoding variant ICOSL polypeptides or immunomodulatory polypeptides described herein can be any of those known to one skilled in the art and include, for example, vesicular stomatitis virus, see, for example, US Pat. US No. 7731974, 7153510, 665.103 and publ. Pat. US No. 2010/0178684, 2010/0172877, 2010/0113567, 2007/0098743, 20050260601, 20050220818 and Pat. EP No. 1385466, 1606411 and 1520175; herpes simplex virus, see, for example, Pat. USA. No. 7 897 146, 7 731 952, 7 550 276, 7 537 924, 6 723 316, 6 428 968 and publ. Pat. US No. 2014/0154216, 2011/0177032, 2011/0158948,

- 73 044356

2010/0092515, 2009/0274728, 2009/0285860, 2009/0215147, 2009/0010889, 2007/0110720, 2006/0039894, 2004/0009604, 2004/0063 094, publ. international patent No. WO 2007/052029, WO 1999/038955; retroviruses, see, for example, Pat. US No. 6,689,871, 6,635,472, 5,851,529, 5,716,826, 5,716,613 and publ. Pat. US No. 20110212530; vaccinia viruses, see, for example, 2016/0339066, and adeno-associated viruses, see, for example, No. 8 007 780, 7 968 340, 7 943 374, 7 906 111, 7 927 585, 7 811 814, 7 662,627, 7,241,447, 7,238,526, 7,172,893, 7,033,826, 7,001,765, 6,897,045 and 6,632,670.

Oncolytic viruses also include viruses that have been genetically modified to reduce their virulence, improve their safety profile, increase their tumor specificity, and have been equipped with additional genes, such as cytotoxins, cytokines, prodrug-converting enzymes, to improve the overall effectiveness of the viruses. (see, for example, Kirn et al., (2009) Nat Rev Cancer 9: 64-71; Garcia-Aragoncillo et al., (2010) Curr Opin Mol Ther 12: 403411, see US Pat. No. 7,588,767 , 7,588,771, 7,662,398 and 7,754,221 and US Pub. Pat. 5287, 2009/0117034, 2010/0233078 , 2009/0162288, 2010/0196325, 2009/0136917 and 2011/0064650). In some embodiments, oncolytic viruses can be those that have been modified so that they selectively replicate in malignant tumor cells and are thus oncolytic. For example, an oncolytic virus is an adenovirus that has been engineered to have modified tropism for the treatment of tumors and also as vectors for gene therapy. Examples of such viruses are ONYX-015, H101 and Ad5ACR (Hallden and Portella (2012) Expert Opin Ther Targets, 16: 945-58) and TNFerade (McLoughlin et al. (2005) Ann. Surg. Oncol, 12: 825-30 ) or conditionally replicative adenovirus Oncorine®.

In some embodiments, the infectious agent is a modified herpes simplex virus. In some embodiments, the infectious agent is a modified version of Talimogene laherparepvec (also known as T-Vec, Imlygic or OncoVex GM-CSF) that is modified to include nucleic acids encoding any of the variant ICOSL polypeptides or immunomodulatory polypeptides described herein. In some embodiments, the infectious agent is a modified herpes simplex virus, which is described, for example, in WO 2007/052029, WO 1999/038955, US 2004/0063094, US 2014/0154216 or variations thereof.

In some embodiments, the infectious agent is a virus that targets a specific cell type in the subject to which the virus is administered, for example, a virus that targets immune cells or antigen presenting cells (APCs). Dendritic cells (DCs) are important APCs for initiating and controlling immune responses. DCs can capture and process antigens, migrate from the periphery to the lymphoid organ, and present antigens to resting T cells in a major histocompatibility complex (MHC)-restricted manner. In some embodiments, the infectious agent is a virus that can specifically target DCs for delivery of nucleic acids encoding variant ICOSL polypeptide or immunomodulatory polypeptides for expression in DCs. In some embodiments, the virus is a lentivirus or a variant or derivative thereof, such as an integration-deficient lentiviral vector. In some embodiments, the virus is a lentivirus that is pseudotyped to efficiently bind and productively infect cells expressing the cellular marker dendritic cell-specific non-integrin intercellular adhesion molecule-3 (DC-SIGN) molecule, such as DC. In some embodiments, the virus is a lentivirus pseudotyped with the Sindbis virus E2 glycoprotein or a modified form thereof, such as that described in WO 2013/149167. In some embodiments, the virus allows delivery and expression of a sequence of interest (eg, a nucleic acid encoding any of the variant ICOSL polypeptides or immunomodulatory polypeptides described herein) into DCs. In some embodiments, the virus includes sequences that are described in WO 2008/011636, US 2011/0064763, Tareen et al. (2014) Mol. Ther., 22: 575-587 or variations thereof. An example of a dendritic cell-tropic vector platform is ZVex™.

1. Bacteria.

In some embodiments, the infectious agent is a bacterium. For example, in some embodiments, bacteria can deliver nucleic acids encoding any of the variant immunomodulatory polypeptides described herein, e.g., an ICOSL variant polypeptide or an immunomodulatory polypeptide, to a target cell in a subject, such as a tumor cell, an immune cell, an antigen presenting cell, and /or phagocytic cell. In some embodiments, the bacterium may preferentially target a specific environment within an entity, such as a tumor microenvironment (TME), to express and/or secrete variant immunomodulatory polypeptides and/or to target specific cells in the variant immunomodulatory polypeptide expression environment.

In some embodiments, this bacterium delivers nucleic acids into cells using a tank

- 74 044356 terial-mediated transfer of plasmid DNA into mammalian cells (also called bactofection). For example, in some embodiments, delivery of genetic material is achieved by introducing the entire bacterium into target cells. In some embodiments, spontaneous or induced bacterial lysis may result in the release of the plasmid for subsequent expression in eukaryotic cells. In some embodiments, the bacterium can deliver nucleic acids to non-phagocytic mammalian cells (eg, tumor cells) and/or phagocytic cells, such as certain immune cells and/or APCs. In some embodiments, nucleic acids delivered by the bacterium can be transferred to the nucleus of a cell of the subject for expression. In some embodiments, the nucleic acids also include the corresponding nucleotide sequences necessary for the expression of operably linked sequences encoding variant immunomodulatory polypeptides in a particular host cell, for example, regulatory elements such as promoters or enhancers. In some embodiments, the infectious agent, which is a bacterium, can deliver nucleic acids encoding immunomodulatory proteins in the form of RNA, such as pre-processed translation competent RNA, delivered into the cytoplasm of the target cell for translation by the target cell machinery.

In some embodiments, the bacterium is capable of replicating and killing target cells, such as tumor cells. In some embodiments, the bacterium may contain and/or release nucleotide sequences and/or gene products into the cytoplasm of target cells, thereby killing the target cell, such as a tumor cell. In some embodiments, the infectious agent is a bacterium that can specifically replicate in a particular environment of a subject, such as a tumor microenvironment (TME). For example, in some embodiments, bacteria can replicate specifically in anaerobic or hypoxic microenvironments. In some embodiments, conditions or factors present in particular media, such as aspartate, serine, citrate, ribose or galactose, produced by cells in the TME, can act as chemoattractants to attract the bacterium to the media. In some embodiments, the bacterium may express and/or secrete the immunomodulatory proteins described herein in the environment, such as TME.

In some embodiments, the infectious agent is a bacterium that belongs to Listeria sp., Bifidobacterium sp., Escherichia sp., Closteridium sp., Salmonella sp., Shigella sp., Vibrio sp. or Yersinia sp. In some embodiments, the bacteria are selected from one or more of Listeria monocytogenes, Salmonella typhimurium, Salmonella choleraesuis, Escherichia coli, Vibrio cholera, Clostridium perfringens, Clostridium butyricum, Clostridium novyi, Clostridium acetobutylicum, Bifidobacterium infantis, Bifidobacterium longum, and Bifidobacterium adolescentis. In some embodiments, the bacterium is an engineered bacterium. In some embodiments, the bacterium is an engineered bacterium, such as described, for example, in Seow and Wood (2009) Molecular Therapy 17(5):767-777; Baban et al. (2010) Bioengineered Bugs 1:6, 385-394; Patyar et al. (2010) J Biomed Sci 17:21; Tangney et al. (2010) Bioengineered Bugs 1:4, 284-287; van Pijkeren et al. (2010) Hum Gene Ther. 21 (4): 405-416; WO 2012/149364; WO 2014/198002; US 9103831; US 9453227; US 2014/0186401; US 2004/0146488; US 2011/0293705; US 2015/0359909 and EP 3020816. The bacterium can be modified to deliver nucleotide sequences encoding any of the variant immunomodulatory polypeptides, conjugates and/or fusion proteins provided herein and/or to express such variant immunomodulatory polypeptides in the subject.

F. Nucleic acids, vectors and methods for producing polypeptides or cells.

Provided herein are isolated or recombinant nucleic acids, collectively referred to as nucleic acids, that encode any of the various embodiments of variant ICOSL polypeptides or immunomodulatory polypeptides provided herein. In some embodiments, the nucleic acids provided herein, including all those described below, are suitable for the recombinant production (eg, expression) of variant ICOSL polypeptides or immunomodulatory polypeptides provided herein. In some embodiments, the nucleic acids provided herein, including all those described below, are useful for the expression of variant ICOSL polypeptides or immunomodulatory polypeptides provided herein in cells, for example, engineered cells, such as immune cells or infectious agent cells. Nucleic acids provided herein may be in the form of RNA or in the form of DNA and include mRNA, cRNA, recombinant or synthetic RNA, and DNA and cDNA. The nucleic acids of the invention are typically DNA molecules and typically double-stranded DNA molecules. However, single-stranded DNA, single-stranded RNA, double-stranded RNA and hybrid DNA/RNA nucleic acids or combinations thereof, including any of the nucleotide sequences of the invention, are also provided.

Also provided herein are recombinant expression vectors and recombinant host cells used in the production of the variant ICOSL polypeptides or immunomodulatory polypeptides provided herein.

Proposed herein are also engineered cells such as

- 75 044356 engineered immune cells containing any of the provided nucleotide molecules or any of the variant ICOSL polypeptides or immunomodulatory polypeptides, for example, any of the transmembrane immunomodulatory polypeptides or secreted immunomodulatory polypeptides.

Also provided herein are infectious agents, such as bacterial or viral cells, containing any of the provided nucleotide molecules or any of the variant ICOSL polypeptides or immunomodulatory polypeptides, such as any of the transmembrane immunomodulatory polypeptides or secreted immunomodulatory polypeptides.

In any of the above proposed embodiments, nucleic acids encoding variant polypeptides or immunomodulatory polypeptides provided herein can be introduced into cells using recombinant DNA and cloning techniques. To do this, a recombinant DNA molecule encoding a transmembrane immunomodulatory polypeptide is obtained. Methods for producing such DNA molecules are well known in the art. For example, peptide coding sequences can be excised from DNA using suitable restriction enzymes. Alternatively, the DNA molecule can be synthesized using chemical synthesis methods such as the phosphoramidite method. Alternatively, a combination of these methods can be used. In some cases, recombinant or synthetic nucleic acid can be produced by polymerase chain reaction (PCR). In some embodiments, a DNA insert may be generated encoding one or more variant ICOSL polypeptides containing at least one IgSF affinity modified domain, and in some embodiments a signal peptide, transmembrane domain, and/or endodomain as described herein. This DNA insert can be cloned into an appropriate T cell transduction/transfection vector as known to those skilled in the art. Vectors containing nucleotide molecules are also presented.

In some embodiments, the expression vectors are capable of expressing transmembrane immunomodulatory proteins in the appropriate cell under conditions suitable for protein expression. In some aspects, the nucleotide molecule or expression vector includes a DNA molecule that encodes an immunomodulatory protein operably linked to corresponding expression control sequences. Methods for influencing this functional relationship, both before and after the DNA molecule inserted into the vector, are well known. Expression control sequences include promoters, activators, enhancers, operators, ribosomal binding sites, start signals, stop signals, cap signals, polyadenylation signals, and other signals associated with the control of transcription or translation.

In some embodiments, expression of the immunomodulatory protein is controlled by a promoter or enhancer to control or regulate expression. The promoter is operably linked to a portion of a nucleic acid molecule encoding a variant polypeptide or immunomodulatory protein. In some embodiments, the promoter is a constitutively active promoter (such as a tissue-specific constitutively active promoter or other constitutive promoter). In some embodiments, the promoter is an inducible promoter that can respond to an inducing agent (such as a T cell activation signal).

In some embodiments, the constitutive promoter is operably linked to a nucleotide molecule encoding a variant polypeptide or immunomodulatory protein. Typical constitutive promoters include the simian vacuolating virus 40 (SV40) promoter, the cytomegalovirus (CMV) promoter, the ubiquitin C (UbC) promoter, and the EF-1 alpha (EF1a) promoter. In some embodiments, the constitutive promoter is tissue specific. For example, in some embodiments, the promoter allows constitutive expression of an immunomodulatory protein in certain tissues, such as immune cells, lymphocytes, or T cells. Exemplary tissue-specific promoters described in US Pat. No. 5,998,205 include, for example, fetoprotein, DF3, tyrosinase, CEA, surfactant protein, and ErbB2 promoters.

In some embodiments, the inducible promoter is operably linked to a nucleotide molecule encoding a variant polypeptide or immunomodulatory protein such that expression of the nucleic acid is controlled by controlling the presence or absence of a corresponding transcription inducer. For example, the promoter may be a regulated promoter system and transcription factor expression system, such as published tetracycline-regulated systems or other regulated systems (see, for example, published PCT International Application No. WO 01/30843) to provide regulated expression of the encoded polypeptide. A typical regulated promoter system is the Tet-On (and Tet-Off) system available, for example, from Clontech (Palo Alto, Calif.). This promoter system allows the regulation of transgene expression controlled by tetracycline or tetracycline derivatives such as doxycycline. Other regulated promoter systems are known (see, for example, US Pub. No. 2002-0168714, entitled Regulation of Gene Expression Using Single-Chain, Monomeric, Ligand-Dependent Polypeptide Switches, which describes gene switches that contain ligand-binding domains , and transcription regulatory domains such as hormone receptors).

In some embodiments, the promoter is responsive to an element that is responsive to T cell activation signals. By way of example only, in some embodiments, the engineered T cell includes an expression vector encoding an immunomodulatory protein and a promoter operably linked to control expression of the immunomodulatory protein. The engineered T cell can be activated, for example, through transmission through the T cell receptor (TCR) or chimeric antigen receptor (CAR) and thereby initiating the expression and secretion of the immunomodulatory protein through a responsive promoter.

In some embodiments, the inducible promoter is operably linked to a nucleotide molecule encoding an immunomodulatory protein such that the immunomodulatory protein is expressed in response to nuclear factor of activated T cells (NFAT) or nuclear factor kappa light chain enhancer of activated B cells (NF-κB) . For example, in some embodiments, the inducible promoter includes a binding site for NFAT or NF-κB. For example, in some embodiments, the promoter is an NFAT or NF-κB promoter or a functional variant thereof. Thus, in some embodiments, the nucleic acids allow the expression of an immunomodulatory protein to be controlled while reducing or eliminating the toxicity of the immunomodulatory protein. In particular, engineered immune cells containing the nucleic acids of the invention express and secrete an immunomodulatory protein only when the cell (e.g., T cell receptor (TCR) or chimeric antigen receptor (CAR) expressed by the cell) is specifically stimulated by an antigen and/or cell (eg, cellular calcium signaling pathway), nonspecifically stimulated, for example, by phorbol myristate acetate (PMA)/ionomycin.

Accordingly, expression and, in some cases, secretion of an immunomodulatory protein can be controlled only when and where it is needed (eg, in the presence of an infectious agent, in a malignant tumor or at a tumor site), thereby reducing or avoiding unwanted immunomodulatory protein interactions.

In some embodiments, the nucleic acid encoding an immunomodulatory protein described herein includes a suitable nucleotide sequence that encodes an NFAT promoter, an NF-κB promoter, or a functional variant thereof. As used herein, NFAT promoter refers to one or more NFAT responsive elements associated with a minimal promoter, NF-κB promoter refers to one or more NF-κB responsive elements associated with a minimal promoter. In some embodiments, the minimal gene promoter is the human IL-2 minimal promoter or the CMV promoter. NFAT responsive elements may include, for example, NFAT1, NFAT2, NFAT3 and/or NFAT4 responsive elements. An NFAT promoter, an NF-κB promoter, or a functional variant thereof may include any number of binding motifs, such as at least two, at least three, at least four, at least five, or at least , six, at least seven, at least eight, at least nine, at least ten, at least eleven, or up to twelve binding motifs.

The resulting recombinant expression vector having a DNA molecule is used to transform the appropriate host. This transformation can be performed using methods well known in the art. In some embodiments, the nucleic acid provided herein further includes a nucleotide sequence that encodes a secretory or signal peptide operably linked to a nucleic acid encoding an immunomodulatory polypeptide such that the resulting soluble immunomodulatory polypeptide is recovered from a culture medium, a host cell, or the host cell periplasm . In other embodiments, appropriate expression control signals are selected to ensure expression of the immunomodulatory polypeptide on the membrane. In addition, commercially available kits as well as contract manufacturing companies can also be used to create the engineered cells or recombinant host cells presented herein.

In some embodiments, the resulting expression vector carrying a DNA molecule is used for transformation, such as transduction, of a suitable cell. Administration can be accomplished using methods well known in the art. Exemplary methods include those that transfer nucleic acids encoding receptors, including via viruses, such as retroviruses or lentiviruses, transduction, transposons, and electroporation. In some embodiments, the expression vector is a viral vector. In some embodiments, the nucleic acid is transferred into cells using lentiviral or retroviral transduction methods.

Any of a large number of publicly available and well-known mammalian host cells, in

- 77 044356 including mammalian T cells or APCs, can be used in the production of polypeptides or engineered cells. The choice of cell depends on a number of factors recognized in the art. These include, for example, compatibility with the chosen expression vector, toxicity of the peptides encoded by the DNA molecule, transformation rate, ease of peptide recovery, expression characteristics, biosafety and costs. The balance of these factors must be taken with the understanding that not all cells may be equally efficient at expressing a particular DNA sequence.

In some embodiments, the host cells may be various eukaryotic cells, such as yeast cells, or mammalian cells, such as Chinese hamster ovary (CHO) cells or HEK293 cells. Host cells can also be prokaryotic cells such as E. coli. The transformed recombinant host is cultured under polypeptide expression conditions and then purified to produce a soluble protein. Recombinant host cells can be cultured under conventional fermentation conditions such that the desired polypeptides are expressed. Such fermentation conditions are well known in the art. Finally, the polypeptides provided herein can be isolated and purified from recombinant cell cultures by any of a number of methods well known in the art, including ammonium sulfate or ethanol precipitation, acid extraction, anion or cation exchange chromatography, phosphocellulose chromatography, hydrophobic interaction chromatography, affinity chromatography. Protein refolding steps can be used as desired to complete the configuration of the mature protein. Finally, high performance liquid chromatography (HPLC) can be used in the final stages of purification.

In some embodiments, the cell is an immune cell, such as any of those described above in connection with the production of engineered cells. In some embodiments, such engineered cells are primary cells. In some embodiments, the engineered cells are autologous to the subject. In some embodiment, the engineered cells are allogeneic to the subject. In some embodiments, the engineered cells are obtained from the subject, for example by leukapheresis, and transformed ex vivo to express an immunomodulatory polypeptide, such as a transmembrane immunomodulatory polypeptide or a secreted immunomodulatory polypeptide.

Also provided herein are nucleic acids encoding any of the variant immunomodulatory polypeptides contained in the infectious agents described herein. In some embodiments, infectious agents deliver nucleic acids into a cell of the subject and/or enable expression of the encoded variant polypeptides in the cell. Also provided are nucleic acids that are used to obtain, produce or modify such infectious agents. For example, in some embodiments, vectors and/or plasmids are provided that contain nucleic acids encoding variant immunomodulatory polypeptides for the production of infectious agents, delivery to cells in a subject, and/or expression of variant immunomodulatory polypeptides in cells of a subject.

In some embodiments, the nucleic acids provided are recombinant viral or bacterial vectors containing nucleotide sequences encoding variant immunomodulatory polypeptides. In some embodiments, recombinant vectors can be used to produce an infectious agent that includes nucleotide sequences encoding variant immunomodulatory polypeptides and/or delivered to a target cell of the subject for expression by the target cell. In some embodiments, the recombinant vector is an expression vector. In some embodiments, the recombinant vector includes the appropriate sequences necessary for the generation and/or production of the infectious agent and expression in the target cell.

In some embodiments, the recombinant vector is a plasmid or cosmid. Plasmids and cosmids encoding the variable immunomodulatory polypeptides described herein are readily constructed using standard methods well known in the art. To produce an infectious agent, the vector or genome can be constructed in the form of a plasmid, which can then be transfected into a packaging or cell line-producing host bacterium. Recombinant vectors can be produced using any of the recombinant methods known in the art. In some embodiments, the vectors may include a prokaryotic origin of replication and/or a gene whose expression confers a detectable or selectable marker, such as a drug, for propagation and/or selection in prokaryotic systems.

In some embodiments, the recombinant vector is a viral vector. Exemplary recombinant viral vectors include a lentiviral vector genome, a poxviral vector genome, a vaccinia virus vector genome, an adenoviral vector genome, an adeno-associated virus vector genome, a herpesvirus vector genome, and an alpha viral vector genome. Viral vectors may be a live, attenuated, replication-dependent or replication-deficient vector, a non-pathogenic (defective), a replication-competent vector, and/or modified to express a heterologous gene, such as the variant immunomodulatory polypeptides provided herein . Viral production vectors can also be modified to alter viral attenuation, which includes any method of increasing or decreasing transcriptional or translational load.

Exemplary viral vectors that may be used include modified vaccinia virus viral vectors (see, e.g., Guerra et al., J. Virol. 80:985-98 (2006); Tartaglia et al., AIDS Research and Human Retroviruses 8: 1445-47 (1992); Gheradi et al., J. Gen. Virol. 86:2925-36 (2005); Mayr et al., Infection 3:6-14 (1975); Hu et al., J Virol 75: 10300-308 (2001); US Pat. No. 5,698,530, 6,998,252, 5,443,964, 7,247,615 and 7,368,116); adenovirus vector or adenovirus-associated virus vectors (see, for example, Molin et al., J. Virol. 72:8358-61 (1998); Narumi et al., Am J. Respir. Cell Mol. Biol. 19:936 -41 (1998); Mercier et al., Proc. Natl. Acad. Sci. USA 101: 6188-93 (2004); US Pat. No. 6,143,290; 6,596,535; 6,855,317; 6,936,257; 7 125 717; 7 378 087; 7 550 296); retroviral vectors, including those based on murine leukemia virus (MuLV), gibbon leukemia virus (GaLV), ecotropic retroviruses, simian immunodeficiency virus (SIV), human immunodeficiency virus (HIV) and combinations (see, for example, Buchscher et al J Virol 66:2731-39 (1992) Johann et al J Virol 66: 1635-40 (1992) Sommerfelt et al Virology 176:58-59 (1990) Wilson et al J Virol 63:2374-78 (1989) Miller et al J Virol 65:2220-24 (1991) Miller et al Mol Cell Biol 10:4239 (1990) Kolberg, NIH Res. 4:43 1992; Cornetta et al., Hum. Gene Ther. 2: 215 (1991)); lentiviral vectors, including those based on human immunodeficiency virus (HIV-1), feline immunodeficiency virus (FIV), equine infectious anemia virus, simian immunodeficiency virus (SIV) and Maedi/Visna virus (see, for example, Pfeifer et al ., Annu. Rev. Genomics Hum. Genet. 2: 177-211 (2001); Zufferey et al., J. Virol. 72: 9873, 1998; Miyoshi et al., J. Virol. 72:8150, 1998; Philpott and Thrasher, Human Gene Therapy 18:483, 2007; Engelman et al., J. Virol. 69: 2729, 1995; Nightingale et al., Mol. Therapy, 13: 1121, 2006; Brown et al., J. Virol. 73:9011 (1999); WO 2009/076524; WO 2012/141984; WO 2016/011083; McWilliams et al., J. Virol. 77: 11150, 2003; Powell et al., J. Virol. 70: 5288, 1996) or any variants thereof and/or vectors that can be used to produce any of the viruses described above. In some embodiments, the recombinant vector may include regulatory sequences, such as promoter or enhancer sequences, that can regulate the expression of the viral genome, for example, in the case of RNA viruses, in a packaging cell line (see, for example, US Pat. No. 5,385,839 and 5 168 062).

In some embodiments, the recombinant vector is an expression vector, e.g., an expression vector that allows expression of the encoded gene product when delivered to a target cell, e.g., a cell in an entity, e.g., a tumor cell, an immune cell, and/or an APC. In some embodiments, recombinant expression vectors contained in an infectious agent are capable of expressing immunomodulatory proteins in a target cell of the subject under conditions suitable for protein expression.

In some aspects, the nucleic acids or expression vector includes a nucleotide sequence that encodes an immunomodulatory protein operably linked to corresponding expression control sequences. Methods for influencing this functional relationship, either before or after the nucleotide sequence encoding the immunomodulatory protein inserted into the vector, are well known. Expression control sequences include promoters, activators, enhancers, operators, ribosomal binding sites, start signals, stop signals, cap signals, polyadenylation signals, and other signals associated with transcriptional or translational control. The promoter may be operably linked to a portion of the nucleotide sequence encoding the immunomodulatory protein. In some embodiments, the promoter is a constitutively active promoter in the target cell (such as a tissue-specific constitutively active promoter or other constitutive promoter). For example, the recombinant expression vector may also include lymphoid tissue-specific transcriptional regulatory elements (TREs), such as TREs specific for B lymphocytes, T lymphocytes, or dendritic cells. TREs specific for lymphoid tissue are known in the art (see, for example, Thompson et al., Mol. Cell. Biol. 12:1043-53 (1992); Todd et al., J. Exp. Med. 177: 1663-74 (1993); Penix et al., J. Exp. Med. 178:1483-96 (1993)). In some embodiments, the promoter is an inducible promoter that can respond to an inducing agent (such as a T cell activation signal). In some embodiments, the nucleic acids delivered to the target cell of the entity, such as a tumor cell, an immune cell, and/or an APC, may be operably linked to any of the regulatory elements described above.

In some embodiments, the vector is a bacterial vector, such as a bacterial plasmid or cosmid. In some embodiments, the bacterial vector is delivered to a target cell, such as a tumor cell, immune cells, and/or APCs, through bacteria-mediated transfer of plasmid DNA into mammalian cells (also called bactofection). In some embodiments, the delivered bacterial vector also includes appropriate expression control sequences for expression in target cells, such as a promoter sequence and/or enhancer sequences, or any of the regulatory or control sequences described above. In some embodiments, the bacterial vector includes appropriate expression control sequences for expression and/or secretion of the encoded variant polypeptides in an infectious agent, such as a bacterium.

In some embodiments, the polypeptides provided herein can also be produced by synthetic methods. Solid-phase synthesis is the preferred method for preparing individual peptides because it is the most cost-effective way to produce small peptides. For example, well-known solid-phase synthesis methods include the use of protecting groups, linkers, and solid-phase supports, as well as specific protection and deprotection reaction conditions, linker cleavage conditions, the use of scavengers, and other aspects of solid-phase peptide synthesis. The peptides can then be assembled into polypeptides as described herein.

IV. Methods for assessing the immune modulation activity of variant ICOSL polypeptides and immunomodulatory proteins.

In some embodiments, variant ICOSL polypeptides provided herein (eg, full-length and/or specific binding fragments or conjugates, stack constructs, or fusion proteins thereof) exhibit immunomodulatory activity to modulate T cell activation. In some embodiments, ICOSL polypeptides modulate IFN-gamma expression in a primary T cell assay relative to wild-type ICOSL control or unmodified ICOSL. In some cases, modulation of IFN-gamma expression may increase or decrease IFN-gamma expression compared to control. Assays to determine the specific binding and expression of IFN-gamma are well known in the art and include MLR reactions (mixed lymphocyte reactions), which measure the levels of interferon-gamma cytokines in culture supernatants (Wang et al., Cancer Immunol Res., 2014 , Sep: 2 (9): 846-56), SEB (staphylococcal enterotoxin B) T cell stimulation assay (Wang et al., Cancer Immunol Res., 2014, Sep: 2 (9): 846-56) and assays stimulation of anti-CD3 T cells (Li and Kurlander, J Transl Med., 2010: 8: 104).

In some embodiments, a variant ICOSL polypeptide may, in some embodiments, increase or, in alternative embodiments, decrease IFN-gamma (interferon-gamma) expression in a primary T cell assay relative to a wild-type ICOSL control. In some embodiments of the proposed polypeptides containing a soluble ICOSL sequence, the polypeptide can increase IFN-gamma expression, and in alternative embodiments, decrease IFN-gamma expression in a primary T cell assay relative to a wild-type ICOSL control. In some embodiments of the provided polypeptides comprising sequences from multiple variant ICOSLs, the polypeptide may increase IFN-gamma expression and, in alternative embodiments, decrease IFN-gamma expression in a primary T cell assay compared to a wild-type ICOSL control.

Those skilled in the art will appreciate that the primary T cell assay format used to determine increases in IFN-gamma expression may be different from those used to assay decreases in IFN-gamma expression. When analyzing the ability of the ICOSL variant to reduce IFN-gamma expression in a primary T cell assay, a mixed lymphocyte response (MLR) assay can be used as described in Example 6. In some cases, a soluble form of the ICOSL variant can be used to determine the ability of the ICOSL variant to exert antagonistic effects and thereby reduce the expression of IFN-gamma in the MLR, as described in example 6.

Alternatively, when testing the ability of a variant ICOSL to increase IFN-gamma expression in a primary T cell assay, the co-immobilization assay described in Example 6 can be used. In the co-immobilization assay, a TCR signal, represented in some embodiments by an anti-CD3 antibody, is used in combination with a co-immobilization assay. immobilized variant ICOSL to determine the ability to increase IFN-gamma expression compared to control ICOSL. In some cases, a soluble form of the variant ICOSL that multimerizes to the extent of allowing multivalent binding can be used to determine the ability of the variant ICOSL to exert agonistic effects and thereby increase IFN-gamma expression in the MLR, as described in Example 6.

The use of appropriate controls is known to those skilled in the art, however, in the above embodiments, controls typically involve the use of an unmodified ICOSL, e.g., a wild-type isoform of native ICOSL from the same mammalian species as the species from which the variant ICOSL was derived or developed . Regardless of whether the binding affinity to one or both of ICOS and CD28 is increased or decreased, the variant ICOSL will in some embodiments increase IFN-gamma expression and in alternative embodiments decrease IFN-gamma expression in a primary T cell assay compared to the wild ICOSL control type.

In some embodiments, the variant ICOSL increases IFN-gamma expression (i.e., protein expression) relative to control wild-type ICOSL or unmodified ICOSL by at least

- 80 044356 least, at: 5, 10, 20, 30, 40, 50, 60, 70, 80, 90% or higher. In other embodiments, the variant ICOSL reduces IFN-gamma expression (i.e., protein expression) compared to wild-type control or unmodified ICOSL by at least: 5, 10, 20, 30, 40, 50, 60, 70. 80, 90% or higher. In some embodiments, the wild-type control ICOSL is a mouse ICOSL such as is typically used for a variant ICOSL altered in sequence from the wild-type mouse ICOSL sequence. In some embodiments, the control wild-type ICOSL is a human ICOSL, such as is typically used for a variant ICOSL altered in sequence relative to a wild-type human ICOSL sequence, such as an ICOSL sequence comprising the amino acid sequence of SEQ ID NO: 32 or SEQ ID NO: 196.

V. Pharmaceutical compositions.

Provided herein are compositions containing any of the variant ICOSL polypeptides, immunomodulatory proteins, conjugates, engineered cells, or infectious agents described herein. The pharmaceutical composition may further contain a pharmaceutically acceptable excipient. For example, a pharmaceutical composition may contain one or more excipients to modify, maintain or maintain, for example, pH, osmolarity, viscosity, clarity, color, isotonicity, odor, sterility, stability, dissolution or release rate, adsorption or permeation of the composition. In some aspects, one skilled in the art will appreciate that a pharmaceutical composition containing cells may be different from a pharmaceutical composition containing protein.

In some embodiments, the pharmaceutical composition is a solid, such as a powder, capsule, or tablet. For example, the components of the pharmaceutical composition may be lyophilized. In some embodiments, the solid pharmaceutical composition is reconstituted or dissolved in a liquid prior to administration.

In some embodiments, the pharmaceutical composition is a liquid, for example, with variant ICOSL polypeptides dissolved in an aqueous solution (for example, saline or Ringer's solution). In some embodiments, the pH of the pharmaceutical composition is from about 4.0 to about 8.5 (e.g., from about 4.0 to about 5.0, from about 4.5 to about 5.5, from about 5.0 to about 6 .0, from about 5.5 to about 6.5, 6.0 and about 7.0, from about 6.5 to about 7.5, from about 7.0 to about 8.0 or from about 7.5 up to about 8.5).

In some embodiments of the invention, the pharmaceutical composition includes a pharmaceutically acceptable excipient, for example, a filler, a binder, a coating, a preservative, a lubricant, a flavoring agent, a sweetener, a coloring agent, a solvent, a buffering agent, a chelating agent, or a stabilizer. Examples of pharmaceutically acceptable excipients include cellulose, dibasic calcium phosphate, calcium carbonate, microcrystalline cellulose, sucrose, lactose, glucose, mannitol, sorbitol, maltol, pregelatinized starch, corn starch or potato starch. Examples of pharmaceutically acceptable binders include polyvinylpyrrolidone, starch, lactose, xylitol, sorbitol, maltitol, gelatin, sucrose, polyethylene glycol, methylcellulose or cellulose. Examples of pharmaceutically acceptable coatings include hydroxypropyl methylcellulose (HPMC), shellac, corn protein zein, or gelatin. Examples of pharmaceutically acceptable disintegrants include polyvinylpyrrolidone, carboxymethylcellulose or sodium starch glycolate. Examples of pharmaceutically acceptable lubricants include polyethylene glycol, magnesium stearate or stearic acid. Examples of pharmaceutically acceptable preservatives include methylparaben, ethylparaben, propylparaben, benzoic acid or sorbic acid. Examples of pharmaceutically acceptable sweeteners include sucrose, saccharin, aspartame or sorbitol. Examples of pharmaceutically acceptable buffering agents include carbonates, citrates, gluconates, acetates, phosphates or tartrates.

In some embodiments, the pharmaceutical composition further includes an agent for controlled or sustained release of a product, such as injectable microspheres, biodegradable particles, polymer compounds (polylactic acid, polyglycolic acid), granules or liposomes.

In some embodiments, the pharmaceutical composition is sterile. Sterilization can be accomplished by filtration through sterile filter membranes or by radiation. If the composition is lyophilized, sterilization using this method can be carried out either before or after lyophilization and reconstitution. The composition for parenteral administration can be stored in lyophilized form or in solution. Therapeutic GGCI compositions are typically placed in a container having a sterile entry port, such as an intravenous solution bag, or a vial having a stopper pierced by a hypodermic needle.

In some embodiments, the present invention comprises pharmaceutical compositions containing transmembrane immunomodulatory proteins, including engineered cells expressing such transmembrane immunomodulatory proteins. In some embodiments, the pharmaceutical compositions and formulations include one or more optional pharmaceutical

- 81 044356 acceptable carriers or excipients. Such compositions may include buffers, for example, neutral buffered saline, phosphate buffered saline, and the like; carbohydrates such as glucose, mannose, sucrose or dextrans, mannitol; proteins; polypeptides or amino acids such as glycine; antioxidants; chelating agents such as EDTA or glutathione; adjuvants (eg aluminum hydroxide); and preservatives. The compositions of the present invention are preferably formulated for intravenous administration.

In some embodiments of the invention, the pharmaceutical composition includes infectious agents containing nucleotide sequences encoding immunomodulatory variant polypeptides. In some embodiments, the pharmaceutical composition includes a dose of infectious agents suitable for administration to the subject in need of treatment. In some embodiments, the pharmaceutical composition includes an infectious agent that is a virus in one or more doses of about 1 x 10 ⁵ to about 1 x 10 ¹² plaque forming units (PFU), 1 x 10 ⁶ -1 x 10 ¹⁰ or 1 x 10 ⁷ -1 x 10 ¹⁰ PFU, in each case inclusive, for example, at least, or at least about or about 1 x 10 ⁶ , 1 x 10 ⁷ , 1 x 10 ⁸ , 1 x 10 ⁹ , 2 x 10 ⁹ , 3 x 10 ⁹ , 4 x 10 ⁹ , 5 x 10 ⁹ or about 1 x 10 ¹⁰ PFU. In some embodiments, the pharmaceutical composition may contain a virus concentration of from or about 10 ⁵ to about 10 ¹⁰ PFU/ml, for example, from 5x10 ⁶ to 5x10 ⁹ or from 1x10 ⁷ to 1x10 ⁹ PFU/ml, for example, at least or, at least about or about 10 ⁶ , 10 ⁷ , 10 ⁸ or 10 ⁹ PFU/ml. In some embodiments, the pharmaceutical composition includes an infectious agent that is a bacterium, in a single or multiple dose amount of about 1 x 10 ³ to about 1 x 10 ⁹ colony forming units (CFU), 1 x 10 ⁴ and 1 x 10 ⁹ CFU, or 1 x 10 ⁵ and 1 x 10 ⁷ CFU, in each case inclusive, for example, at least, or at least about or about 1x10 ⁴ , 1x10 ⁵ , 1x10 ⁶ , 1x10 ⁷ , 1x10 ⁸ or 1x10 ⁹ CFU. In some embodiments, the pharmaceutical composition may include bacteria at a concentration of from or about 103 to about 10 ⁸ CFU/ml, for example, from 5x10 ⁵ to 5x10 ⁷ or from 1x10 ⁶ to 1x10 ⁷ CFU/ml, for example, at least or, at least about or about 10 ⁵ , 10 ⁶ , 10 ⁷ or 10 ⁸ CFU/ml.

Such a composition may, for example, be in a form suitable for intravenous infusion. A pharmaceutically acceptable carrier may be a pharmaceutically acceptable material, composition or carrier that is involved in the transfer or transport of cells of interest from one tissue, organ or body part to another tissue, organ or body part. For example, the carrier may be a liquid or solid filler, diluent, excipient, solvent or encapsulating material, or a combination thereof. Each component of the carrier must be pharmaceutically acceptable because it must be compatible with the other ingredients of the composition. It must also be suitable for contact with any tissue, organ or body part it may encounter, meaning that it must not carry a risk of toxicity, irritation, allergic reaction, immunogenicity or any other complication that unduly outweighs its therapeutic benefit .

In some embodiments, the pharmaceutical composition is administered to a subject. Typically, dosages and routes of administration of a pharmaceutical composition are determined according to the size and condition of the subject, in accordance with standard pharmaceutical practice. For any composition, the therapeutically effective dose can be assessed initially either in cell culture assays or in animal models such as mice, rats, rabbits, dogs, pigs or monkeys. An animal model can also be used to determine the appropriate concentration range and route of administration. Such information can be used to determine appropriate doses and routes of administration to humans. The exact dosage will be determined in light of factors related to the subject being treated. Dosage and administration are adjusted to provide sufficient levels of the active compound or to maintain the desired effect. Factors that may be taken into account include the severity of the disease condition, the general health of the subject, the age, weight and sex of the subject, time and frequency of administration, drug combination(s), sensitivity of response, and response to therapy.

Long-acting pharmaceutical compositions may be administered every 3-4 days, every week, or every two weeks depending on the half-life and clearance rate of the particular composition. The frequency of dosing will depend on the pharmacokinetic parameters of the molecule in the formulation used. Typically, the composition is administered until a dose is reached that achieves the desired effect. Thus, the composition can be administered as a single dose or in multiple doses (at the same or different concentrations/doses) over time or as a continuous infusion. In the future, the appropriate dose will be clarified. Appropriate doses can be established using appropriate dose response data. A number of biomarkers or physiological markers for therapeutic effect can be monitored, including T cell activation or proliferation, cytokine synthesis or production (eg, TNF-α, IFN-γ, IL-2 production), induction of various activation markers (eg, CD25, receptor IL-2), inflammation, joint swelling or tenderness, serum C-reactive protein levels, production of anti-collagen antibodies and/or T-dependent antibody response.

In some embodiments, the pharmaceutical composition is administered to a subject by any route, including orally, transdermally, inhalation, intravenously, intraarterially, intramuscularly, directly applied to a wound site, applied to a surgical site, intraperitoneally, suppository, subcutaneously, intradermally, transdermally, by nebulization, intrapleural, intraventricular, intraarticular, intraocular or intravertebral.

In some embodiments of the invention, the dose of the pharmaceutical composition is a single dose or a repeat dose. In some embodiments, doses are administered to the subject once a day, twice a day, three times a day, or four or more times a day. In some embodiments, about 1 or more (eg, about 2 or more, about 3 or more, about 4 or more, about 5 or more, about 6 or more, or about 7 or more) doses are administered over the course of a week. In some embodiments, multiple doses are administered over days, weeks, months or years. In some embodiments, the course of treatment is about 1 or more doses (e.g., about 2 or more, about 3 or more, about 4 or more, about 5 or more, about 7 or more, about 10 or more, about 15 or more, about 25 or more, about 40 or more, about 50 or more, or about 100 or more doses).

In some embodiments, the administered dose of the pharmaceutical composition is about 1 μg of protein per kg of subject's body weight or more (e.g., about 2 μg of protein per kg of subject's body weight or more, about 5 or more, about 10 or more, about 25 or more , about 50 or more, about 100 or more, about 250 or more, about 500 μg of protein per kg body weight of the subject or more, about 1 mg of protein per kg of body weight or more of the subject, about 2 or more or about 5 mg of protein per kg of object's body weight or more).

In some embodiments, a therapeutic amount of the cellular composition is administered. In general, the exact amount of compositions of the present invention to be administered can be determined by the physician, taking into account individual differences in age, weight, tumor size, degree of infection or metastasis, and condition of the patient. In general, it can be stated that a pharmaceutical composition containing engineered cells, for example T cells, as described herein, can be administered at a dosage of 10 ⁴ to 10 ⁹ cells/kg body weight, for example 105 to 106 cells/ kg body weight, including all integer values within these ranges. Engineered cell compositions, such as T cell compositions, may also be administered multiple times at these dosages. The cells can be administered using infusion techniques that are widely known in immunotherapy (see, for example, Rosenberg et al., New Eng. J. Med. 319: 1676, 1988). The optimal dosage and treatment regimen for a particular patient can be easily determined by a medical professional by monitoring the patient for signs of disease and adjusting treatment accordingly.

There are many known methods for determining whether administration of a therapeutic composition of the invention sufficiently modulates immunological activity by eliminating, sequestering or inactivating immune cells mediating or capable of mediating an unwanted immune response; inducing, generating or recruiting immune cells that mediate or are capable of mediating a protective immune response; changes in the physical or functional properties of immune cells; or a combination of these effects. Examples of measurements of modulation of immunological activity include, but are not limited to, examining the presence or absence of immune cell populations (using flow cytometry, immunohistochemistry, histology, electron microscopy, polymerase chain reaction (PCR)); measurement of the functional capacity of immune cells, including the ability or resistance to proliferate or divide in response to a signal (eg, using T-cell proliferation assays and the 3H-thymidine incorporation pepscan assay after stimulation with an anti-CD3 antibody, T-cell receptor antibody , anti-CD28 antibody, calcium ionophores, PMA (phorbol 12-myristate 13-acetate), antigen presenting cells loaded with peptide or protein antigen; B cell proliferation assays); measuring the ability to kill or lyse other cells (such as cytotoxic T cell assays); measurements of cytokines, chemokines, cell surface molecules, antibodies and other cell products (eg, flow cytometry, enzyme-linked immunosorbent assays, Western blot analysis, protein microarray analysis, immunoprecipitation assay); measurement of biochemical markers of immune cell activation or signaling pathways in immune cells (eg, Western blot analysis and immunoprecipitation analysis of tyrosine, serine, or threonine phosphorylation, polypeptide cleavage, protein complex formation or dissociation, protein matrix analysis, DNA transcription, DNA array profiling, or subtractive hybridization); measurement of cell death due to apoptosis, necrosis, or other mechanisms (eg, Annexin V staining, TUNEL assays, gel electrophoresis for DNA ladder measurements, histology, fluorogenic caspase assays, Western blot analysis of caspase substrates); measurement of genes, proteins, and other molecules produced by immune cells (e.g., Northern blot, polymerase chain reaction, DNA microarray, protein microarray, two-dimensional gel electrophoresis, Western blot, enzyme-linked immunosorbent assay

- 83 044356 lysis, flow cytometry); and measuring clinical symptoms or outcomes, such as improvement in autoimmune, neurodegenerative and other diseases involving self-proteins or self-polypeptides (clinical assessments, requirements for additional treatments, functional status, imaging studies), for example, by measuring relapse rates or disease severity (using clinical parameters known to those of ordinary skill in the art) in the case of multiple sclerosis, blood glucose measurements in type I diabetes, or joint inflammation in the case of rheumatoid arthritis.

VI. Finished products and kits.

Also provided herein are finished products containing the pharmaceutical compositions described herein in appropriate packaging. Suitable packaging for the compositions (such as ophthalmic compositions) described herein is known in the art and includes, for example, vials (eg, sealed vials), vials, ampoules, bottles, jars, flexible packaging (eg, sealed mylar or plastic packages), and the like. These finished products can be further sterilized and/or sealed.

Also provided are kits containing the pharmaceutical compositions (or finished products) described herein, which may further contain instruction(s) on how to use the compositions, such as those described herein. The kits described herein may also include other materials commercially and user-desirable, including other buffers, diluents, filters, needles, syringes, and package inserts with instructions for performing any of the methods described herein.

VII. Therapeutic applications.

The pharmaceutical compositions described herein (including pharmaceutical compositions containing variant ICOSL polypeptides, immunomodulatory proteins, conjugates, engineered cells and infectious agents described herein) can be used in a variety of therapeutic applications, such as the treatment of disease. For example, in some embodiments, the pharmaceutical composition is used to treat inflammatory or autoimmune diseases, malignancies, organ transplants, viral infections and/or bacterial infections in mammals. The pharmaceutical composition can modulate (eg, increase or decrease) the immune response to treat a disease. In some embodiments, the proposed methods are useful for the therapeutic administration of variant ICOSL polypeptides, immunomodulatory proteins, conjugates, engineered cells, and infectious agents described herein. It is within the scope of one skilled in the art, given the invention provided, to select the format to be indicated depending on the type of modulation of the immune response, such as increasing or decreasing what is desired.

In some embodiments, a pharmaceutical composition as provided herein is administered to stimulate an immune response, which may be useful, for example, in the treatment of cancer, viral or bacterial infections. In some embodiments, the pharmaceutical composition includes a variant ICOSL polypeptide in a format that exhibits agonistic activity to its cognate binding partner CD28 or ICOS and/or that stimulates or initiates co-stimulatory signaling through CD28 or ICOS. Exemplary ICOSL polypeptide formats for use in such therapeutic applications include, for example, an immunomodulatory protein or stacked variant ICOSL polypeptide and an IgSF domain or variant thereof that binds a tumor antigen (e.g., Nkp30 or an affinity modified variant) (also referred to as an IgSF tumor-localizing domain ), a conjugate containing a variant ICOSL polypeptide linked to a tumor-targeting fragment (also called a tumor-localizing fragment), an engineered cell expressing a transmembrane immunomodulatory protein or an infectious agent comprising a nucleotide molecule encoding a transmembrane immunomodulatory protein, e.g. a protein in an infected cell (eg, a tumor cell or an APC, such as a dendritic cell).

In some embodiments, the pharmaceutical composition can be used to inhibit the growth of malignant tumor cells in mammals (eg, human malignant tumor cells). A method of treating cancer may include administering an effective amount of any of the pharmaceutical compositions described herein to a subject with a cancer. An effective amount of the pharmaceutical composition can be administered to inhibit, stop, or reverse the progression of cancer.

Malignancies that can be treated with the pharmaceutical compositions and methods of treatment described herein include, but are not limited to, melanoma, bladder cancer, hematologic malignancies (leukemia, lymphoma, myeloma), liver cancer, brain cancer, kidney cancer , breast cancer, pancreatic cancer (adenocarcinoma), colorectal cancer, lung cancer (small cell lung cancer and non-small cell lung cancer), cancer

- 84 044356 spleen, thymus or blood cell cancer (i.e. leukemia), prostate cancer, testicular cancer, ovarian cancer, uterine cancer, stomach cancer, musculoskeletal malignancy, head and neck cancer, gastrointestinal cancer, cancer germ cell or endocrine and neuroendocrine cancers. In some embodiments, the malignancy is Ewing's sarcoma. In some embodiments, the malignancy is selected from melanoma, lung cancer, bladder cancer, and hematologic malignancy. In some embodiments, the malignancy is lymphoma, lymphoid leukemia, myeloid leukemia, cervical cancer, neuroblastoma, or multiple myeloma.

Human malignant tumor cells can be treated in vivo or ex vivo. In ex vivo treatment of a human patient, tissue or fluids containing malignant tumor cells are processed outside the body and then the tissue or fluids are reintroduced back into the patient. In some embodiments, a cancer is treated in a human patient in vivo by administering a therapeutic composition to the patient.

In some embodiments, the pharmaceutical composition is administered as monotherapy (i.e., as a single agent) or as combination therapy (i.e., in combination with one or more additional anticancer agents, such as a chemotherapy drug, a tumor vaccine, or an immune checkpoint inhibitor In some embodiments, the pharmaceutical composition may also be administered with radiation therapy.

In some embodiments, the pharmaceutical composition suppresses the immune response, which may be useful in the treatment of inflammatory or autoimmune disorders or organ transplantation. In some embodiments, the pharmaceutical composition includes a variant ICOSL polypeptide in a format that exhibits antagonistic activity to its cognate binding partner CD28 or ICOS and/or that blocks or inhibits co-stimulatory signaling through CD28 or ICOS. Exemplary ICOSL polypeptide formats for use in such therapeutic applications include, for example, a variant ICOSL polypeptide that is soluble (e.g., an IFOSL-Fc fusion protein), an immunomodulatory protein, or a stack of a variant ICOSL polypeptide and other IgSF, including soluble forms thereof that are Fc fusion proteins, an engineered cell expressing a secreted immunomodulatory protein, or an infectious agent containing a nucleotide molecule encoding a secreted immunomodulatory protein, for example, to express and secrete a secreted immunomodulatory protein in an infected cell (e.g., a tumor cell or an APC, e.g., a dendritic cell) .

In some embodiments, the inflammatory or autoimmune disease is antineutrophil cytoplasmic antibody (ANCA) associated vasculitis, vasculitis, autoimmune skin disease, transplantation, rheumatic disease, inflammatory gastrointestinal disease, inflammatory eye disease, inflammatory neurological disease, inflammatory pulmonary disease, inflammatory endocrine disease or autoimmune hematological disease.

In some embodiments, inflammatory and autoimmune disorders that can be treated with the pharmaceutical composition described herein are Addison's disease, allergies, alopecia areata, Alzheimer's disease, antineutrophil cytoplasmic antibody-associated (ANCA) vasculitis, ankylosing spondylitis, antiphospholipid syndrome (Hughes syndrome), asthma, atherosclerosis, rheumatoid arthritis, autoimmune hemolytic anemia, autoimmune hepatitis, autoimmune inner ear syndrome, autoimmune lymphoproliferative syndrome, autoimmune myocarditis, autoimmune oophoritis, autoimmune orchitis, azoospermia, Behcet's disease, Berger's disease, bullous pemphigoid, cardiomyopathy, cardiovascular diseases, celiac sprue/celiac enteropathy, chronic fatigue immune dysfunction syndrome (CFIDS), chronic idiopathic polyneuritis, chronic inflammatory demyelination, combination of Guillain-Barré syndrome with myositis (CIPD), chronic relapsing polyneuropathy (Guillain-Barré syndrome), Churg-Strauss syndrome (CSS), cicatricial pemphigoid, cold agglutinin disease (CAD), COPD (chronic obstructive pulmonary disease), CREST syndrome, Crohn's disease, dermatitis, herpetiformis, dermatomyositis, diabetes, discoid lupus, eczema, epidermos bullosa, essential mixed cryoglobulinemia, Evan's syndrome, proptosis, fibromyalgia, Goodpasture's syndrome, Graves' disease, Hashimoto's thyroiditis, idiopathic pulmonary fibrosis, idiopathic thrombocytopenic purpura (ITP), IgA nephropathy, immunoproliferative disease or disorder, inflammatory bowel disease (IBD), interstitial lung disease, juvenile arthritis , juvenile idiopathic arthritis (JIA), Kawasaki disease, Lambert-Eaton migraine syndrome, lichen planus, lupus nephritis, lymphocytic lipophysitis, Meniere's disease, Miller-Fisher syndrome/acute disseminated encephalomyeloradiculopathy, mixed connective tissue disease, multiple sclerosis (MS), muscular rheumatism, myalgic encephalomyelitis (ME), myasthenia gravis, ocular inflammation, foliaceus, pemphigus, pemphigus vulgaris, pernicious anemia, polyarteritis nodosa, polychondria, polyglandular syndromes (Whitaker syndrome), polymyalgia rheumatica, polymyositis, primary agam

- 85 044356 Maglobulinemia, primary biliary cirrhosis/autoimmune cholangiopathy, psoriasis, psoriatic arthritis, Raynaud's phenomenon, Reiter's syndrome/reactive arthritis, restenosis, rheumatic fever, rheumatic disease, sarcoidosis, Schmidt's syndrome, scleroderma, Sjorgen's syndrome, stiff man's syndrome, systemic red lupus (SLE), systemic scleroderma, Takayusu arteritis, temporal arteritis/giant cell arteritis, thyroiditis, type 1 diabetes, ulcerative colitis, uveitis, vasculitis, vitiligo, interstitial bowel disease or Wegener's granulomatosis. In some embodiments, the inflammatory or autoimmune disease is selected from interstitial bowel disease, transplantation, Crohn's disease, ulcerative colitis, multiple sclerosis, asthma, rheumatoid arthritis, and psoriasis.

In some embodiments, the pharmaceutical composition is administered to modulate an autoimmune condition. For example, suppression of the immune response may be useful in methods of inhibiting the rejection of a tissue, cell, or organ graft from a donor by a recipient. Accordingly, in some embodiments, the pharmaceutical compositions described herein are used to limit or prevent graft-related or transplant-related diseases or disorders, such as graft-versus-host disease (GVHD). In some embodiments, pharmaceutical compositions are used to suppress autoimmune rejection of transplanted or engrafted bone marrow, organs, skin, muscle, neurons, islets, or parenchymal cells.

Pharmaceutical compositions comprising the engineered cells and methods described herein can be used in adoptive cell transfer applications. In some embodiments, cellular compositions containing engineered cells can be used in related methods, for example, to modulate immunological activity in an immunotherapeutic approach for treating, for example, mammalian cancers or in other embodiments for treating autoimmune disorders. Methods used typically include a method of contacting the TIP of the present invention with a mammalian cell under conditions that allow specific binding of the affinity modified IgSF domain and modulation of the immunological activity of the mammalian cell. In some embodiments, immune cells (such as tumor infiltrating lymphocytes (TILs), T cells (including CD8+ or CD4+ T cells), or APCs) are engineered to be expressed as a membrane protein and/or as a soluble ICOSL polypeptide, an immunomodulatory protein or a conjugate as described herein. The engineered cells can then be contacted with a mammalian cell, such as an APC, a second lymphocyte, or a tumor cell, in which modulation of immunological activity is desired and under conditions that permit specific binding of the affinity-modified IgSF domain to a counterstructure on the mammalian cell such that immunological activity can be modulated in a mammalian cell. Cells can contact in vivo or ex vivo.

In some embodiments, the engineered cells are autologous cells. In other embodiments, the cells are allogeneic. In some embodiments, the cells are autologous modified cells reintroduced into the mammal from which it was isolated. In some embodiments, the cells are allogeneic engineered cells introduced into a mammal. In some embodiments, cells are collected from the patient's blood or tumor, engineered to express a polypeptide (such as an ICOSL variant polypeptide, immunomodulatory protein, or conjugate as described herein), expanded in an in vitro culture system (e.g., by cell stimulation), and reintroduced to the patient to mediate tumor destruction. In some embodiments, the methods are carried out using adoptive cell transfer, in which TIP-expressing cells (eg, T cells) are introduced back into the patient. In some embodiments, the methods of the invention are used to treat mammalian patients with a malignancy, such as lymphoma, lymphoid leukemia, myeloid leukemia, cervical cancer, neuroblastoma, or multiple myeloma.

VIII. Exemplary implementations.

Among the proposed embodiments.

1. A variant ICOS ligand (ICOSL) polypeptide comprising an IgV domain or a specific binding fragment thereof, an IgC domain or a specific binding fragment thereof, or both, wherein the variant ICOSL polypeptide includes one or more amino acid modifications in the unmodified ICOSL or a specific binding moiety corresponding to position(s) selected from 10, 11, 13, 16, 18, 20, 25, 27, 30, 33, 37, 38, 42, 43, 47, 52, 54, 57, 61 , 62, 67, 71, 72, 74, 75, 77, 78, 80, 84, 89, 90, 92, 93, 94, 96, 97, 98, 99, 100, 102, 111, 113, 115, 116 , 117, 119, 120, 121, 122, 126, 129, 130, 132, 133, 135, 138, 139, 140, 142, 143, 144, 153, 154, 155, 156, 158, 161, 164, 166 , 168, 172, 173, 175, 190, 192, 193, 194, 198, 201, 203, 207, 208, 210, 212, 217, 220, 221, 224, 225, or 227 relative to SEQ ID NO: 32.

2. The ICOSL variant polypeptide of embodiment 1, wherein the unmodified ICOSL is mammalian ICOSL or a specific binding fragment thereof.

3. The ICOSL variant polypeptide of embodiment 2, wherein the unmodified ICOSL is

- 86 044356 is a human ICOSL or a specific binding fragment thereof.

4. The ICOSL variant polypeptide of any one of embodiments 1-3, wherein the unmodified ICOSL comprises (i) the amino acid sequence set forth in SEQ ID NO: 32, (ii) an amino acid sequence that has at least 95% sequence identity with SEQ ID NO: 32; or (iii) a portion thereof containing an IgV domain or an IgC domain or specific binding fragments thereof, or both.

5. The ICOSL variant polypeptide according to any one of embodiments 1-4, wherein:

the specific binding fragment of the IgV domain or IgC domain has a length of at least 50, 60, 70, 80, 90, 100, 110 or more amino acids; or the IgV domain specific binding fragment has a length that is at least 80% of the length of the IgV domain defined as amino acids 19-129 in SEQ ID NO: 5, and/or the IgC domain specific binding fragment has a length that is at least 80% of the length of the IgC domain, defined as amino acids 141-227 in SEQ ID NO: 5.

6. The variant ICOSL polypeptide according to any one of embodiments 1-5, where the variant ICOSL includes up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 amino acid modifications, optionally amino acid substitutions, insertions and/or deletions.

7. The variant ICOSL according to any one of embodiments 1-6, where the variant ICOSL includes an amino acid sequence that has at least 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% identity with the sequence of SEQ ID NO: 32 or a specific binding fragment thereof.

8. The variant ICOSL polypeptide of any one of embodiments 1-7, wherein the variant ICOSL polypeptide exhibits altered binding to the ICOS, CD28 or CTLA-4 ectodomain compared to unmodified ICOSL.

9. The variant ICOSL polypeptide of any one of embodiments 1-8, wherein the variant ICOSL polypeptide exhibits altered binding to the ICOS or CD28 ectodomain compared to unmodified ICOSL.

10. The ICOSL variant polypeptide of embodiment 8 or embodiment 9, wherein the altered binding is an altered binding affinity and/or an altered binding selectivity.

11. A variant ICOSL polypeptide according to any one of embodiments 1-10, where one or more amino acid modifications are selected from M10V, M101, V11E, S13G, E16V, S18R, A20V, S25G, F27S, F27C, N30D, Y33del, Q37R, K42E, T43A , Y47H, N52H, N52D, N52Q, N52S, N52Y, N52K, S54A, S54P, N57D, N57Y, R61S, R61C, Y62F, L67P, A71T, G72R, L74Q, R75Q, D77G, F78L, L80P, N84Q, E90A, K92R , F93L, H94E, H94D, L96F, L96I, V97A, L98F, S99G, Q100R, Q100K, Q100P, L102R, G103E, V107A, V107I, S109G, S109N, V110D, V110N, V110A, E111del, T1 13E, H115Q, H115R, V116A , A117T, N119Q, F120I, F120S, S121G, V122A, V122M, S126T, S126R, H129P, S130G, S132F, Q133H, E135K, F138L, T139S, C140del, S142F, I143V, I143T, N144D, Y146C, V151A, Y152C, Y152H , W153R, I154F, N155H, N155Q, K156M, D158G, L161P, L161M, L166Q, N168Q, F172S, L173S, M175T, T190A, T190S, S192G, V193M, N194D, C198R, N201S, L203P, L203F, N207Q, L208P, V210A , S212G, D217V, I218T, I218N, E220G, R221G, R221I, I224V, T225A, N227K or their conservative amino acid substitution.

12. A variant ICOSL polypeptide according to any one of embodiments 1-11, where one or more amino acid modifications are selected from N52Y/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52S/Y146C/Y152C, N52H/C198R, N52H/C140D/ T225A, N52H/C198R/T225A, N52H/K92R, N52H/S99G, N57Y/Q100P, N52S/S130G/Y152C, N52S/Y152C, N52S/C198R, N52Y/N57Y/Y152C,

N52Y/N57Y/H129P/C198R, N52H/L161P/C198R, N52S/T113E, N52D/S54P, N52K/L208P, N52S/Y152H, N52D/V151A, N52H/I143T, N52S/L80P, F120S/Y1 52H/N201S, N52S/ R75Q/L203P, N52S/D158G, N52D/Q133H, N52S/N57Y/H94D/L96F/L98F/Q100R, N52S/N57Y/H94D/L96F/L98F/Q100R/G103E/F120S, N52H/F78L/Q100R , N52H/N57Y/ Q100R/V110D, N52H/N57Y/R75Q/Q100R/V110D, N52H/N57Y/Q100R, N52H/N57Y/L74Q/Q100R/V110D, N52H/Q100R, N52H/S121G, A20V/N52H/N57Y/Q10 0R/S109G,

N52H/N57Y/R61S/Q100R/V110D/L173S, N52H/N57Y/Q100R/V122A, N52H/N57Y/Q100R/F172S,

N52H/N57Y, N52S/F120S, N52S/V97A, N52S/G72R, N52S/A71T/A117T, N52S/E220G,

Y47H/N52S/V107A/F120S, N52H/N57Y/Q100R/V110D/S132F/M175T,

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R, Q37R/N52H/N57Y/Q100R/V110N/S142F/C198R/D217V/R221G, N52H/N57Y/Q100R/V110 D/C198R,

N52H/N57Y/Q100R/V110D/V116A/L161M/F172S/S192G/C198R, F27S/N52H/N57Y/V110N,

N52S/H94E/L96I/S109N/L166Q, S18R/N52S/F93L/I143V/R221G, A20T/N52D/Y146C/Q164L,

V11E/N30D/N52H/N57Y/H94E/L96I/L98F/N194D/V210A/I218T, N52S/H94E/L96I/V122M,

N52H/N57Y/H94E/L96I/F120I/S126T/W153R/I218N, M10V/S18R/N30D/N52S/S126R/T139S/L203F,

S25G/N30D/N52S/F120S/N227K, N30D/N52S/L67P/Q100K/D217G/R221K/T225S,

N52H/N57Y/Q100R/V110D/A117T/T190S/C198R, N52H/N57Y/Q100R/V110D/F172S/C198R,

S25G/F27C/N52H/N57Y/Q100R/V110D/E135K/L173S/C198R, N52H/N57Y/V110A/C198R/R221I,

M10I/S13G/N52H/N57Y/D77G/V110A/H129P/I143V/F172S/V193M, C198R,

- 87 044356

N52H/N57Y/R61C/Y62F/Q100R/V110N/F120S/C198R, N52H/N57Y/Q100R/V110D/H115R/C198R,

N52H/N57Y/Q100R/V110D/N144D/F172S/C198R, N52S/H94E/L98F/Q100R, N52S/E90A,

N30D/K42E/N52S, N52S/F120S/I143V/I224V, N52H/N57Y/Q100R/V110D/C198R/S212G,

N52H/N57Y/Q100R/C198R, N52S/N194D, N52H/N57Y/Q100R/L102R/V110D/H115R/C198R,

N52H/N57Y/Q100R/V110D/C198R/S212G, N52H/N57Y/Q100R/C198R, N52S/N194D,

N52H/N57Y/Q100R/L102R/V110D/H115R/C198R, N52S/S54P, T38P/N52S/N57D, Ellldel, Y33del, N52H/C140del/T225A, N52H/F78L/Q100R/C198R, N52H/N57Y /R75Q/Q100P/ V110D,

N52H/N57Y/L74Q/Vll0D/Sl92G, N52H/Sl2lG/Cl98R, N52S/Fl20S/N227K,

N52S/A7lT/All7T/Tl90A/Cl98R, T43A/N52H/N57Y/L74Q/D89G/Vll0D/Fl72S,

N52H/N57Y/Q100R/V110D/S132F/M175T, N52D, N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G, N52Q/N207Q, N168Q/N207Q, N52Q/N168Q, N15 5Q/N207Q, N119Q/N168Q, N119Q/N207Q,

Nll9Q/Nl55Q, N52Q/N84Q, N52Q/Nll9Q, N84Q/Nll9Q, N52Q/N84Q/Nl68Q, N52Q/N84Q/N207Q, N84Q/Nl55Q/Nl68Q, N84Q/Nl68Q/N207Q, N84Q/Nl55H/N207 Q,Nl55Q/Nl68Q/ N207Q,

Nll9Q/Nl55Q/Nl68Q, Nll9Q/Nl68Q/N207Q, N84Q/Nll9Q/N207Q, Nll9Q/Nl55H/N207Q, N84Q/Nll9Q/Nl55Q, N52Q/Nll9Q/Nl55Q, N52H/N84Q/Nll9Q, N52H/N84 Q,N52H/N84Q/ Nl68Q, N52H/N84Q/N207Q, N52H/N84Q/Nl68Q/N207Q, N52Q/N84Q/Nl55Q, N52Q/N84Q/Nl68Q,

N52Q/N84Q/Nl55Q/Nl68Q, N52Q/N84Q/Nll9Q/Nl68Q, N84Q/Nll9Q/Nl55Q/Nl68Q,

N84Q/Nl55Q/Nl68Q/N207Q, N84Q/Nll9Q/Nl55Q/N207Q, N52Q/N84Q/Nll9Q/N207Q,

N52Q/N84Q/Nll9Q/Nl55Q, N52Q/N84Q/Nll9Q/Nl55Q/N207Q, N84Q/Nll9Q/Nl55Q/Nl68Q/N207Q,

Fl38L/L203P, N52Y/Fl38L/L203P, N57Y/Ql00R/Cl98R, N57Y/Fl38L/L203P, Ql00R/Fl38L,

N52H/N57Y/Ql00R/Hll5R/Cl98R, N52H/N57Y/Ql00R/Fl72S/Cl98R,

N52H/N57Y/Ql00R/Hll5R/Fl72S/Cl98R, N52H/N57Y/Ql00R/Hll5R/Il43V/Fl72S/Cl98R,

N52H/N57Y/Ql00R/Ll02R/Hll5R/Fl72S/Cl98R, N52H/Vl22A/Fl72S/Cl98R,

N52H/N57Y/Ql00R/Hll5R/Fl72S/Nl94D, N52H/N57Y/Hll5R/Fl72S/Cl98R,

N52H/N57Y/Ql00R/Hll5R/Cl98R, N52H/N57Y/Hll5R, N52H/N57Y/Ql00R/Hll5R,

N52H/N57Y/Ql00R/Hll5R/Fl72S/I224V, N52H/N57Y/Ql00R/Hll5R/Fl72S, N52H/N57Y/Ql00R/Fl72S, N52H/Ql00R/Hll5R/Il43T/Fl72S, N52H/N5 7Y/Q l00P/Hll5R/Fl72S , N52Y/N57Y/Ql00P/Fl72S, El6V/N52H/N57Y/Ql00R/Vll0D/Hll5R/Cl98R, El6V/N52H/N57Y/Ql00R/Vll0D/Hll5R/Yl52C/Kl56M/Fl72S/Cl98R, N52S/E90A/ Hll5R, N30D /K42E N52S/Hll5R, N30D/K42E/N52S/Hll5R/Cl98R/R22lI, N30D/K42E/N52S/Hll5R/Cl98R,

N30D/K42E/N52S/Hll5R/Fl72S/Nl94D, N52S/Hll5R/Fl20S/Il43V/Cl98R, N52S/Hll5R/Fl72S/Cl98R, N52H/N57Y/Ql00P/Cl98R, N52H/N57Y/Ql00P Hll5R/F l72S/Cl98R, N52H /N57Y/Ql00P/Fl72S/Cl98R, N52H/N57Y/Ql00P/Hll5R, N52H/N57Y/Ql00P/Hll5R/Cl98R, N52H/Ql00R/Cl98R,

N52H/Q100R/H115R/F172S, N52H/Q100R/F172S/C198R, N52H/Q100R/H115R/F172S/C198R or

N52H/N57Y/Q100R/F172S/C198R.

13. The ICOSL variant polypeptide of any one of embodiments 1-12, wherein the one or more amino acid modifications correspond to position(s) selected from 52, 57, 100, 110 or 198.

14. The variant ICOSL polypeptide according to any one of embodiments 1-13, wherein one or more amino acid modifications are selected from N52H, N52D, N52Q, N52S, N52K, S54A, S54P, N57Y, Q100P, Q100R, V110A, V110D, C198R or a conservative amino acid thereof replacements.

15. A variant ICOSL polypeptide according to any of embodiments 1-14, further containing one or more amino acid modifications selected from V11E, E16V, N30D, K42E, N52H, N52S, N52Y, N57Y, E90A, H94E, L96I, L98F, Q100R, Q100P , V110A, V110D, H115R, F120S, V122A, F138L, I143V, K156M, K156R, F172S, N194D, C198R, L203P, V210A, R221I, I224V or their conservative amino acid substitution.

16. The ICOSL variant polypeptide according to any one of embodiments 1-14, where one or more amino acid modifications are N52H/N57Y/Q100R/C198R, N52H/N57Y/Q100R/V122A,

N52H/N57Y/Q100R/F172S, N52Y/N57Y/F13 8L/L203P,

V11E/N30D/N52H/N57Y/H94E/L96I/L98F/N194D/V210A/I218Т, N52H/N57Y/Q100R/L102R/V110D/H115R/C198R, N52H/N57Y/Q100R, N52H/Q100R,

N52H/N57Y/Q100R/V110D/C198R/S212G, E16V/N52H/N57Y/Q100R/V110D/H115R/V152C/K156M/C198R, N30D/K42E/N52S,

N52S/F120S/I143V/I224V, N52S/E90A, N52H/N57Y/V110A/C198R/R221I, N52H/N57Y/Q100P or N52S/N194D.

17. The variant ICOSL polypeptide according to any one of embodiments 1-16, wherein the variant ICOSL polypeptide includes an IgV domain or a specific fragment thereof and an IgC domain or a specific fragment thereof.

18. A variant ICOSL polypeptide according to any one of embodiments 1-17, comprising the amino acid sequence presented in any of SEQ ID NO: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470 or a specific binding fragment or amino acid sequence thereof that has at least 95% sequence identity with any of SEQ ID NO: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435 -470 or its specific binding fragment-

- 88 044356 ment and which includes one or more amino acid substitutions.

19. The ICOSL variant polypeptide of any one of embodiments 1-18, wherein the variant polypeptide

The ICOSL includes an IgV domain or a specific binding fragment thereof.

20. The ICOSL variant polypeptide of any one of embodiments 1-19, wherein the IgV domain or a specific binding fragment thereof is the only ICOSL portion of the ICOSL variant polypeptide.

21. The ICOSL variant polypeptide of any one of embodiments 1-18, wherein the IgC domain or a specific binding fragment thereof is the only ICOSL portion of the ICOSL variant polypeptide.

22. A variant ICOSL polypeptide according to any one of embodiments 1-20, comprising the amino acid sequence presented in any of SEQ ID NO: 197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426 and 434 or a specific binding fragment thereof, an amino acid sequence that has at least 95% sequence identity with any of SEQ ID NO: 197-199, 201-208, 210, 212, 240, 326-340, 382-386 , 425-426 and 434 or a specific binding fragment thereof and which includes one or more amino acid substitutions.

The variant ICOSL polypeptide of any one of embodiments 1-22, wherein the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS, CD28, or CTLA-4 with increased affinity compared to unmodified ICOSL.

24. The variant ICOSL polypeptide of any one of embodiments 1-23, wherein the variant ICOSL polypeptide specifically binds to the ICOS or CD28 ectodomain with increased affinity compared to unmodified ICOSL.

25. The variant ICOSL polypeptide according to any one of embodiments 1-24, wherein the variant ICOSL polypeptide specifically binds to the ICOS ectodomain and the CD28 ectodomain each with increased affinity compared to unmodified ICOSL.

26. The variant ICOSL polypeptide of any one of embodiments 1-24, wherein the variant ICOSL polypeptide specifically binds to the CD28 ectodomain with increased affinity compared to unmodified ICOSL.

27. The ICOSL variant polypeptide of any one of embodiments 1-26, wherein the increased affinity for the CD28 ectodomain is increased by more than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, 50 or 60 times compared to unmodified ICOSL.

28. The variant ICOSL polypeptide of any one of embodiments 1-24, wherein the variant ICOSL polypeptide specifically binds to the ICOS ectodomain with increased affinity compared to unmodified ICOSL.

29. The ICOSL variant polypeptide of any one of embodiments 1-25 and 28, wherein the increased affinity for the ICOS ectodomain is increased by more than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 40, 50, 60 or 70 times compared to unmodified ICOSL.

30. The ICOSL variant polypeptide of any one of embodiments 1-29, wherein the one or more amino acid modifications correspond to position(s) selected from 52, 54 or 57.

31. The ICOSL variant polypeptide of any one of embodiments 1-30, wherein the one or more amino acid modifications are selected from N52H, N52D, N52S, N52Y, N52K, S54A, S54P, N57Y, or a conservative amino acid substitution thereof.

32. The ICOSL variant polypeptide of any one of embodiments 1-31, wherein the one or more amino acid modifications correspond to position(s) selected from 52 or 57.

33. The ICOSL variant polypeptide of any one of embodiments 1-31, wherein the one or more amino acid modifications are selected from N52H, N52D, N52S, N52K or N57Y.

34. The ICOSL variant polypeptide of any one of embodiments 25-29 and embodiment 33, further comprising one or more amino acid modifications selected from N52H, N52D, N52S, N52Y, N52K, S54A, S54P, N57Y, R75Q, L80P, K92R, S99G, H94D, L96F, L98F, S99G, Q100R, Q100P, G103E, T113E, F120S, H129P, S130G, Q133H, F138L, C140D, C140del, I143T, Y146C, V151A, Y152C, Y152H, D158G, L161P, C198R, N201S, L203P, L208P or T225A, or their conservative amino acid substitution.

35. A variant ICOSL polypeptide according to any one of embodiments 1-34, where one or more amino acid modifications are selected from N52Y/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52S/Y146C/Y152C, N52H/C198R, N52H/C140del/ T225A, N52H/C198R/T225A, N52H/K92R,

N57Y/Q100P, N52S/C198R, N52Y/N57Y/Y152C, N52Y/N57Y/H129P/C198R, N52H/L161P/C198R, N52S/T113E, N52S/S54P, N52K/L208P, N52S/Y152H, N5 2H/I143T, N52S/ R75Q/L203P, N52S/D158G, N52D/Q133H, N52H/N57Y/Q100R/V110D/C198R/S212G, N52H/N57Y/Q100R/C198R, N52S/N194D, N52H/N57Y/Q100R/L102R/ V110D/H115R/C198R, N52S/S54P, T38P/N52S/N57D, N52H/C140del/T225A, N52H/F78L/Q100R/C198R, N52H/N57Y/R75Q/Q100P/V110D, N52H/N57Y/L74Q/V110D/S192G,

N52H/S121G/C198R, N52S/F120S/N227K, N52S/A71T/A117T/T190A/C198R,

T43A/N52H/N57Y/L74Q/D89G/V110D/F172S, N52H/N57Y/Q100R/V110D/S132F/M175T,

N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G, N52Q/N207Q, N52Q/N168Q, N52Q/N84Q,

- 89 044356

N52Q/N119Q, N52Q/N84Q/N168Q, N52Q/N84Q/N207Q, N52Q/N119Q/N155Q, N52H/N84Q/N119Q,

N52H/N84Q, N52H/N84Q/N168Q, N52H/N84Q/N207Q, N52H/N84Q/N168Q/N207Q, N52Q/N84Q/N155Q,

N52Q/N84Q/N168Q, N52Q/N84Q/N155Q/N168Q, N52Q/N84Q/N119Q/N168Q, N52Q/N84Q/N119Q/N207Q,

N52Q/N84Q/N119Q/N155Q, N52Q/N84Q/N119Q/N155Q/N207Q, N52Y/F138L/L203P, N57Y/Q100R/C198R,

N57Y/F138L/L203P, N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/Q100R/F172S/C198R,

N52H/N57Y/Q100R/H115R/F172S/C198R, N52H/N57Y/Q100R/H115R/I143V/F172S/C198R,

N52H/N57Y/Q100R/L102R/H115R/F172S/C198R, N52H/V122A/F172S/C198R,

N52H/N57Y/Q100R/H115R/F172S/N194D, N52H/N57Y/H115R/F172S/C198R,

N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/H115R, N52H/N57Y/Q100R/H115R,

N52H/N57Y/Q100R/H115R/F172S/I224V, N52H/N57Y/Q100R/H115R/F172S, N52H/N57Y/Q100R/F172S, N52H/Q100R/H115R/I143T/F172S, N52H/N57Y /Q100P/H115R/F172S, N52Y/N57Y/Q100P/F172S,

E16V/N52H/N57Y/Q100R/V110D/H115R/C198R, E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R, N52S/E90A/H115R,

N30D/K42E/N52S/H115R, N30D/K42E/N52S/H115R/C198R/R221I, N30D/K42E/N52S/H115R/C198R,

N30D/K42E/N52S/H115R/F172S/N194D, N52S/H115R/F120S/I143V/C198R, N52S/H115R/F172S/C198R, N52H/N57Y/Q100P/C198R, N52H/N57Y/Q100P H115R/F172S/C198R, N52H /N57Y/Q100P/F172S/C198R, N52H/N57Y/Q100P/H115R, N52H/N57Y/Q100P/H115R/C198R, N52H/Q100R/C198R,

N52H/Q100R/H115R/F172S, N52H/Q100R/F172S/C198R, N52H/Q100R/H115R/F172S/C198R or

N52H/N57Y/Q100R/F172S/C198R.

36. The ICOSL variant polypeptide of any one of embodiments 1-35, wherein one or more amino acid modifications are selected from N52H/N57Y/F138L/L203P, N52H/N57Y/Q100P, N52H/K92R, N52H/C140del/T225A, N52H/C198R/T225A , N52H/K92R, N57Y/Q100P, N52Y/N57Y/H129P/C198R, N52H/L161P/C198R, N52K/L208P or N52H/I143T.

37. The ICOSL variant polypeptide of any one of embodiments 1-34, wherein one or more amino acid substitutions are selected from N57Y/Q100P, N52S/S130G/Y152C, N52S/Y152C, N52Y/N57Y/Y152C, N52H/L161P/C198R, N52H/L161P /C198R, N52S/L80P, A20V/N52H/N57Y/Q100R/S109G,

N52H/N57Y/R61S/Q100R/V110D/L173S, N52H/N57Y/Q100R/V107I/V110D/S132F/I154F/C198R/R221G,

Q37R/N52H/N57Y/Q100R/V110N/S142F/C198R/D217V/R221G, N52H/N57Y/Q100R/V110D/C198R,

F27S/N52H/N57Y/V110N, S18R/N52S/F93L/I143V/R221G, A20T/N52D/Y146C/Q164L,

N52H/N57Y/H94E/L96I/F120I/S126T/W153R/I218N, N52H/N57Y/Q100R/V110D/F172S/C198R,

S25G/F27C/N52H/N57Y/Q100R/V110D/E135K/L173S/C198R or

M10I/S13G/N52H/N57Y/D77G/V110A/H129P/I143V/F172S/V193M/C198R.

38. The variant ICOSL polypeptide of any one of embodiments 1-37, wherein the variant polypeptide specifically binds to the ectodomain of ICOS, CD28 or CTLA4 with increased selectivity compared to unmodified ICOSL.

39. The ICOSL variant polypeptide of embodiment 38, wherein the increased selectivity comprises a greater binding ratio of the variant polypeptide to one cognate binding partner selected from ICOS, CD28 and CTLA4 relative to another cognate binding partner compared to the binding ratio of the unmodified ICOSL polypeptide to one cognate partner relative to another cognate binding partner.

40. The variant ICOSL polypeptide of embodiment 39, wherein the ratio is greater than at least, or at least about 1.5, 2.0, 3.0, 4, 5, 10, 15, 20, 30, 40, 50 times or more.

41. The variant ICOSL polypeptide of any one of embodiments 1-40, wherein the variant ICOSL polypeptide specifically binds to an ICOS or CD28 ectodomain with increased affinity and specifically binds to an ICOS or CD28 ectodomain with reduced affinity compared to unmodified ICOSL.

42. The variant ICOSL polypeptide of embodiment 41, wherein the variant ICOSL polypeptide specifically binds to the ICOS ectodomain with increased affinity and specifically binds to the CD28 ectodomain with reduced affinity compared to unmodified ICOSL.

43. The variant ICOSL polypeptide of embodiment 41, wherein the variant ICOSL polypeptide specifically binds to the CD28 ectodomain with increased affinity and specifically binds to the ICOS ectodomain with reduced affinity compared to unmodified ICOSL.

44. A variant ICOSL polypeptide according to any one of embodiments 1-43, including amino acid substitutions N52S/R75Q/L203P or N30D/K42E/N52S.

45. The variant ICOSL polypeptide of any one of embodiments 1-40, wherein the variant ICOSL polypeptide specifically binds to the CTLA-4 ectodomain with increased affinity compared to unmodified ICOSL.

46. The ICOSL variant polypeptide of embodiment 45, wherein the increased affinity for the CTLA-4 ectodomain is increased by greater than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8 , 9, 10, 20, 40, 50, 60 or 70 times compared to unmodified ICOSL.

47. The ICOSL variant polypeptide of any one of embodiments 8-46, wherein the ICOS is human ICOS.

- 90 044356

48. The ICOSL variant polypeptide of any one of embodiments 8-47, wherein CD28 is

Human CD28.

49. A variant ICOSL polypeptide according to any one of embodiments 1-48, wherein the binding is changed (increased or decreased) by more than 1.2, 1.5, 2, 3, 4, 5, 6, 7 , 8, 9, 10, 20, 40 or 50 times compared to unmodified ICOSL.

50. The ICOSL variant polypeptide of any one of embodiments 1-49, which is a soluble protein.

51. A variant ICOSL polypeptide according to any one of embodiments 1-50, which is associated with a multimerization domain.

52. The variant ICOSL polypeptide of any one of embodiments 1-51, wherein the variant ICOSL polypeptide is a multimeric polypeptide, optionally a dimeric polypeptide, comprising a first variant ICOSL polypeptide linked to a multimerization domain and a second variant ICOSL polypeptide linked to a multimerization domain.

53. The ICOSL variant polypeptide of embodiment 52, wherein the first ICOSL variant polypeptide and the second ICOSL variant polypeptide are the same or different.

54. The ICOSL variant polypeptide of any one of embodiments 51-53, wherein the multimerization domain is an Fc domain or a reduced effector function variant thereof.

55. The ICOSL variant polypeptide of any one of embodiments 1-54, which is linked to a moiety that increases the biological half-life of the polypeptide.

56. A variant ICOSL polypeptide according to any one of embodiments 1-55, which is associated with an Fc domain or a reduced effector function variant thereof.

57. The ICOSL variant polypeptide of any one of embodiments 54-56, wherein: the Fc domain is a mammal, optionally a human; or a variant Fc domain includes one or more amino acid modifications compared to a modified Fc domain that is mammalian, not necessarily human.

58. The ICOSL variant polypeptide of any one of embodiments 54, 56, and 57, wherein the Fc domain or variant thereof comprises the amino acid sequence set forth in SEQ ID NO: 226 or SEQ ID NO: 227, or an amino acid sequence that has at least 85% sequence identity to SEQ ID NO: 226 or SEQ ID NO: 227.

59. The ICOSL variant polypeptide of any one of embodiments 54 and 56-58, wherein the Fc domain includes one or more amino acid modifications selected from E233P, L234A, L234V, L235A, L235E, G236del, G237A, S267K, R292C, N297G and V302C, each according to EU numbering.

60. The ICOSL variant polypeptide of any one of embodiments 54 and 56-59, wherein the Fc domain includes the amino acid modification C220S in accordance with EU numbering.

61. The ICOSL variant polypeptide of any one of embodiments 54 and 56-60, wherein the Fc domain comprises an amino acid sequence set forth in any of SEQ ID NO: 474, 476, 477, 478 or an amino acid sequence that exhibits at least 85% sequence identity to any of SEQ ID NOs: 474, 476, 477, 478 and exhibits reduced effector function.

62. The ICOSL variant polypeptide of any one of embodiments 51-61, wherein the ICOSL variant polypeptide is linked to the multimerization domain or Fc domain indirectly through a linker, optionally a G4S linker.

63. The variant ICOSL polypeptide of any one of embodiments 1-49, which is a transmembrane immunomodulatory protein further comprising a transmembrane domain linked to an extracellular domain (ECD) or a specific binding fragment thereof from the variant ICOSL polypeptide.

64. The ICOSL variant polypeptide of embodiment 63, wherein the transmembrane domain comprises the amino acid sequence designated residues 257-277 of SEQ ID NO: 5, or a functional variant thereof that has at least 85% sequence identity with residues 257- 277 from SEQ ID NO: 5.

65. The ICOSL variant polypeptide of embodiment 63 or embodiment 64, further comprising a cytoplasmic signaling domain linked to a transmembrane domain.

66. The variant ICOSL polypeptide of embodiment 65, wherein the cytoplasmic signaling domain comprises the amino acid sequence designated residues 278-302 of SEQ ID NO: 5, or a functional variant thereof that has at least 85% sequence identity with residues 278 -302 from SEQ ID NO: 5.

67. The ICOSL variant polypeptide of any one of embodiments 63-66, comprising an amino acid sequence set forth in any of SEQ ID NOs: 494-503, or an amino acid sequence that has at least 85% sequence identity with any of SEQ ID NOs: : 494-503.

68. The variant ICOSL polypeptide of any one of embodiments 1-67, wherein the variant ICOSL increases IFN-gamma (interferon-gamma) expression relative to unmodified ICOSL in an in vitro primary T cell assay.

- 91 044356

69. The variant ICOSL polypeptide of any one of embodiments 1-68, wherein the variant ICOSL reduces IFN-gamma (interferon-gamma) expression relative to unmodified ICOSL in an in vitro primary T cell assay.

70. The ICOSL variant polypeptide of any one of embodiments 1-51, which is deglycosylated.

71. An immunomodulatory protein comprising a variant ICOSL polypeptide of any one of embodiments 1-70 linked to a second immunoglobulin superfamily (IgSF) domain-containing polypeptide.

72. The immunomodulatory protein of embodiment 71, wherein the IgSF domain has a modified affinity and has altered binding to one or more of its cognate binding partners compared to an unmodified IgSF domain or a wild-type IgSF domain.

73. The immunomodulatory polypeptide of embodiment 72, wherein the IgSF domain exhibits increased binding to one or more of its cognate binding partners compared to an unmodified IgSF domain or a wild-type IgSF domain.

74. The immunomodulatory polypeptide of any one of embodiments 71-73, wherein the variant ICOSL polypeptide is a first variant ICOSL polypeptide and the IgSF domain of the second polypeptide is the IgSF domain of a second variant ICOSL polypeptide of any one of embodiments 1-70, wherein the first and second variants ICOSL are the same or different.

75. The immunomodulatory protein of any one of embodiments 71-74, wherein the variant ICOSL polypeptide is capable of specifically binding to CD28 or ICOS, and the IgSF domain of the second polypeptide is capable of binding to a cognate binding partner other than the binding partner specifically bound by the variant ICOSL polypeptide.

76. The immunomodulatory polypeptide of embodiment 75, wherein the IgSF domain is a member of the B7 family.

77. The immunomodulatory polypeptide of any one of embodiments 71-73 and 75, wherein the IgSF domain is a tumor localizing moiety that binds to a ligand expressed on the tumor.

78. The immunomodulatory polypeptide of embodiment 77, wherein the ligand is B7H6.

79. The immunomodulatory polypeptide of embodiment 77 or embodiment 78, wherein the IgSF domain is from NKp30.

80. The immunomodulatory polypeptide of any one of embodiments 71-79, wherein the IgSF domain is or includes an IgV domain.

81. The immunomodulatory polypeptide of any one of embodiments 71-80, wherein the variant ICOSL polypeptide is or includes an IgV domain.

82. The immunomodulatory protein of any one of embodiments 71-81, wherein the immunomodulatory protein comprises a multimerization domain linked to one or both of the variant ICOSL polypeptides or a second IgSF domain-containing polypeptide.

83. The immunomodulatory protein of embodiment 82, wherein the multimerization domain is an Fc domain or a variant thereof with reduced effector function.

84. The immunomodulatory protein according to any one of embodiments 71-83, which is dimeric.

85. The immunomodulatory protein of embodiment 84, which is homodimeric.

86. The immunomodulatory protein of embodiment 84, which is heterodimeric.

87. A conjugate comprising the ICOSL variant polypeptide of any one of embodiments 1-70 or the immunomodulatory polypeptide of any of embodiments 71-86 linked to a fragment.

88. The conjugate of embodiment 87, wherein the moiety is a targeting moiety that specifically binds to a molecule on the surface of a cell.

89. The conjugate of embodiment 88, wherein the targeting moiety specifically binds to a molecule on the surface of an immune cell.

90. The conjugate of embodiment 89, wherein the immune cell is an antigen presenting cell or lymphocyte.

91. The conjugate of embodiment 88, wherein the targeting moiety is a tumor localizing moiety that binds to a molecule on the surface of the tumor.

92. The conjugate of any one of embodiments 87-91, wherein the moiety is a protein, peptide, nucleic acid, small molecule, or nanoparticle.

93. The conjugate of any one of embodiments 87-92, wherein the fragment is an antibody or antigen binding fragment.

94. The conjugate of any one of embodiments 87-93, wherein the conjugate is divalent, tetravalent, hexavalent or octavalent.

95. Nucleic acid molecule(s) encoding the ICOSL variant polypeptide of any one of embodiments 1-70 or the immunomodulatory polypeptide of any of embodiments 71-86.

96. The nucleic acid molecule(s) of embodiment 95, which is a synthetic nucleic acid.

- 92 044356

97. The nucleic acid molecule(s) of embodiment 95 or embodiment 96, which is a cDNA.

98. A vector comprising the nucleic acid molecule(s) of any one of embodiments 95-97.

99. The vector of embodiment 98, which is an expression vector.

100. The vector of embodiment 98 or embodiment 99, wherein the vector is a mammalian expression vector or a viral vector.

101. A cell containing the vector of embodiment 102 or embodiment 103.

102. The cell of embodiment 101, which is a mammalian cell.

103. The cell of embodiment 101 or embodiment 102, which is a human cell.

104. A method of producing a variant ICOSL polypeptide or immunomodulatory protein, comprising introducing a nucleic acid molecule according to any one of embodiments 95-97 or a vector according to any one of embodiments 98-100 into a host cell under conditions of expression of the protein in the cell.

105. The method of embodiment 104, further comprising isolating or purifying the variant ICOSL polypeptide or immunomodulatory protein from a cell.

106. A method of constructing a cell expressing a variant ICOSL polypeptide, comprising introducing a nucleic acid molecule encoding the variant ICOSL polypeptide of any one of embodiments 1-70, or an immunomodulatory polypeptide of any of embodiments 71-86, into a host cell under conditions where the polypeptide is expressed in a cage.

107. An engineered cell comprising a variant ICOSL polypeptide of any of embodiments 1-70, an immunomodulatory polypeptide of any of embodiments 71-86, a nucleic acid molecule of any of embodiments 95-97, or a vector of any of embodiments 98-100.

108. The engineered cell of embodiment 107, wherein the variant ICOSL polypeptide or immunomodulatory polypeptide includes a signal peptide.

109. The engineered cell of embodiment 107 or embodiment 108, wherein the variant ICOSL polypeptide or immunomodulatory polypeptide does not include a transmembrane domain and/or is not expressed on the surface of the cell.

110. The engineered cell of any one of embodiments 107-109, wherein the variant ICOSL polypeptide or immunomodulatory polypeptide is secreted from the engineered cell.

111. The engineered cell of embodiment 107 or embodiment 108, wherein the engineered cell comprises a variant ICOSL polypeptide that includes a transmembrane domain and/or is a transmembrane immunomodulatory protein of any one of embodiments 63-67.

112. The engineered cell of embodiment 107, embodiment 108, or embodiment 111, wherein the variant ICOSL polypeptide is expressed on the surface of the cell.

113. The constructed cell of embodiment 67, wherein the cell is an immune cell.

114. The engineered cell of embodiment 113, wherein the immune cell is an antigen presenting cell (APC) or a lymphocyte.

115. The engineered cell of any one of embodiments 107-114, which is a primary cell.

116. The engineered cell of any one of embodiments 107-115, wherein the cell is a mammalian cell.

117. The engineered cell of any one of embodiments 107-116, wherein the cell is a human cell.

118. The engineered cell of any one of embodiments 107-117, wherein the lymphocyte is a T cell.

119. The engineered cell of embodiment 114, wherein the APC is an artificial APC.

120. The engineered cell of any one of embodiments 107-119, further comprising a chimeric antigen receptor (CAR) or an engineered T-cell receptor.

121. An infectious agent comprising a nucleotide molecule encoding a variant ICOSL polypeptide of any one of embodiments 1-70 or an immunomodulatory polypeptide of any of embodiments 71-86.

122. The infectious agent of embodiment 121, wherein the encoded variant ICOSL polypeptide or immunomodulatory polypeptide does not include a transmembrane domain and/or is not expressed on the surface of the cell in which it is expressed.

123. The infectious agent of embodiment 121 or embodiment 122, wherein the encoded variant ICOSL polypeptide or immunomodulatory polypeptide is secreted from a cell in which it is expressed.

124. The infectious agent of embodiment 121, wherein the encoded variant ICOSL polypeptide includes a transmembrane domain.

125. The engineered cell of embodiment 107, embodiment 108, or embodiment 111, wherein the encoded variant ICOSL polypeptide is expressed on the surface of the cell in which it is expressed.

126. The infectious agent according to any one of embodiments 121-125, where the infectious agent is

- 93 044356 is a bacterium or virus.

127. The infectious agent of embodiment 126, wherein the virus is an oncolytic virus.

128. The infectious agent of embodiment 127, wherein the oncolytic virus is adenoviruses, adeno-associated viruses, herpes viruses, herpes simplex virus, vesicular stomatitis virus, reovirus, Newcastle disease virus, parvovirus, measles virus, implanted stomatitis virus (VSV), virus Coxsackie or cowpox virus.

129. The infectious agent of embodiment 126, wherein the virus specifically targets dendritic cells (DCs) and/or is tropic for dendritic cells.

130. The infectious agent of embodiment 129, wherein the virus is a lentiviral vector that is pseudotyped with a modified Sindbis virus envelope product.

131. The infectious agent of any one of embodiments 121-130, further comprising a nucleotide molecule encoding an additional gene product that leads to the death of a target cell or that can enhance or stimulate an immune response.

132. The infectious agent of embodiment 131, wherein the additional gene product is selected from an antitumor agent, antimetastatic agent, an antiangiogenic agent, an immunomodulatory molecule, an immune checkpoint inhibitor, an antibody, a cytokine, a growth factor, an antigen, a cytotoxic gene product, a proapoptotic gene product, an antiapoptotic gene product , cell matrix degradation gene, tissue regeneration genes or reprogramming of human somatic cells into pluripotent ones.

133. A pharmaceutical composition comprising a variant ICOSL polypeptide according to any one of embodiments 1-70, an immunomodulatory protein according to any one of embodiments 71-86, a conjugate according to any one of embodiments 87-94, an engineered cell according to any one of embodiments 107-120, or an infectious agent according to any one from incarnations 121-132.

134. The pharmaceutical composition of embodiment 133, comprising a pharmaceutically acceptable excipient.

135. The pharmaceutical composition of embodiment 123 or 134, wherein the pharmaceutical composition is sterile.

136. An article containing the pharmaceutical composition according to any of embodiments 133-135 in a bottle.

137. The article of embodiment 136, wherein the bottle is sealed.

138. A kit containing the pharmaceutical composition according to any of embodiments 133-135 and instructions for use.

139. A kit containing the product of embodiments 136 and 137, and instructions for use.

140. A method of modulating an immune response in a subject, comprising administering to the subject a pharmaceutical composition according to any of embodiments 133-135.

141. A method of modulating the immune response in a subject, comprising administering engineered cells according to any one of embodiments 107-120.

142. The method of embodiment 141, wherein the engineered cells are autologous to the subject.

143. The method of embodiment 141, wherein the engineered cells are allogeneic to the subject.

144. The method of any one of embodiments 140-143, wherein modulating the immune response treats a disease or condition in the subject.

145. The method of any one of embodiments 140-144, wherein the immune response is increased.

146. The method of any one of embodiments 140, 144 and 145, wherein the subject is administered an immunomodulatory protein or conjugate comprising a variant ICOSL polypeptide linked to a tumor-locating moiety.

147. The method of embodiment 146, wherein the tumor-locating moiety is or includes a binding molecule that recognizes a tumor antigen.

148. The method of embodiment 147, wherein the binding molecule comprises an antibody or an antigen binding fragment thereof, or comprises a wild type IgSF domain or variant thereof.

149. The method of any one of embodiments 140 and 144-148, wherein the pharmaceutical composition comprising the immunomodulatory protein of any of embodiments 77-86 or the conjugate of any of embodiments 87-94 is administered to the subject.

150. The method according to any one of embodiments 140-145, wherein an engineered cell containing a variant ICOSL polypeptide, which is a transmembrane immunomodulatory protein, is administered to the subject and/or the engineered cell of 107, 108 and 111-120.

151. The method according to any one of embodiments 140, 144 and 145, wherein an infectious agent encoding a variant ICOSL polypeptide that is a transmembrane immunomodulatory protein is administered to a subject, optionally under conditions in which the infectious agent infects a tumor cell or immune cell, and the transmembrane immunomodulatory the protein is expressed on the surface of the infected cell.

152. The method according to any one of embodiments 150 or embodiment 151, wherein the transmembrane immunomo-

- 94 044356 duplicating protein according to any of embodiments 63-70.

153. The method of any one of embodiments 140-152, wherein the disease or condition is a tumor or malignancy.

154. The method of any one of embodiments 140-153, wherein the disease or condition is selected from melanoma, lung cancer, bladder cancer, hematologic malignancy, liver cancer, brain cancer, kidney cancer, breast cancer, pancreatic cancer, colorectal cancer, cancer spleen, prostate cancer, testicular cancer, ovarian cancer, uterine cancer, gastric cancer, musculoskeletal malignancy, head and neck cancer, gastrointestinal cancer, germ cell cancer or endocrine and neuroendocrine cancer.

155. The method of any one of embodiments 140-144, wherein the immune response is reduced.

156. The method of any one of embodiments 140, 144 and 155, wherein the subject is administered a variant ICOSL polypeptide or immunomodulatory protein that is soluble.

157. The method of embodiment 156, wherein the soluble polypeptide or immunomodulatory protein is an Fc fusion protein.

158. The method according to any of embodiments 140, 144 and 155-157, wherein the pharmaceutical composition comprising the ICOSL variant polypeptide according to any of embodiments 1-62 and 68-70, or the immunomodulatory protein according to any of embodiments 71-76 and 80-86, are entered into the object.

159. The method of any one of embodiments 140-144 and 155-157, wherein the engineered cell containing the secreted ICOSL polypeptide is administered to the subject.

160. The method of any one of embodiments 140-144, 155-157 and 159, wherein the engineered cell of any of embodiments 107-110 and 113-120 is administered to a subject.

161. The method of any one of embodiments 140, 144 and 155-157 and 159, wherein an infectious agent encoding a variant ICOSL polypeptide that is a secreted immunomodulatory protein is administered to a subject, optionally under conditions in which the infectious agent infects a tumor cell or immune cell and secreted immunomodulatory protein is secreted from the infected cell.

162. The method of any one of embodiments 140-144 and 155-161, wherein the disease or condition is an inflammatory or autoimmune disease or condition.

163. The method of any one of embodiments 140-144 and 155-162, wherein the disease or condition is antineutrophil cytoplasmic antibody (ANCA) associated vasculitis, vasculitis, autoimmune skin disease, transplantation, rheumatic disease, inflammatory gastrointestinal disease, inflammatory an eye disease, an inflammatory neurological disease, an inflammatory lung disease, an inflammatory endocrine disease or an autoimmune hematological disease.

164. The method of embodiment 162 or embodiment 163, wherein the disease or condition is selected from inflammatory bowel disease, transplantation, Crohn's disease, ulcerative colitis, multiple sclerosis, asthma, rheumatoid arthritis, or psoriasis.

165. The ICOSL variant polypeptide of any one of embodiments 1-70, wherein the amino acid modification is an amino acid substitution, insertion or deletion.

IX. Examples.

The following examples are included for illustrative purposes only and are not intended to limit the scope of the invention.

Example 1: Preparation of Mutant DNA Constructs of IgSF Domains Example 1 describes the preparation of mutant DNA constructs of human ICOSL IgSF domains for translation and expression on the surface of yeast as yeast display libraries.

A. Degenerate libraries.

The constructs were derived from the following wild-type human ICOSL extracellular domain (ECD) sequence shown in SEQ ID NO: 32 (containing the ECD domain corresponding to residues 19-256 set forth in UniProt Accession No. O75144.2):

DTQEKEVRAMVGSDVELSCACPEGSRFDLNDVYVYWQTSESKTVVTYHIPQ

NSSLENVDSRYRNRALMSPAGMLRGDFSLRLFNVTPQDEQKFHCLVLSQSLGFQEVL SVEVTLHVAANFSVPVVSAPHSPSQDELTFTCTSINGYPRPNVYWINKTDNSLLDQAL QNDTVFLNMRGLYDVVSVLRIARTPSVNIGCCIENVLLQQNLTVGSQTGNDIGERDKI TENPVSTGEKNAAT

For libraries that target specific residues for complete or partial randomization by degenerate codons, the coding DNA from SEQ ID NO: 32 was ordered from Integrated DNA Technologies (Coralville, Iowa) as a set of overlapping oligonucleotides up to 80 base pairs (bp). ) in length. To obtain a library of different variants of each ECD, the oligonucleotides contained the desired degenerate codons at the desired amino acid positions. Degenerate codons were generated using the algorithm at URL: rosettadesign.med.unc.edu/SwiftLib/.

In general, the positions for mutation and codon degeneration were selected from homology models

- 95 044356 (ICOSL) of target-ligand pairs of interest, to identify contact residues of the ligand, as well as residues that are at the protein interaction interface. This analysis was performed using the structure viewer available at URL: spdbv.vital-it.ch).

The next step in library development was to align the human, mouse, rat, and monkey ICOSL sequences to identify conserved residues. Based on this analysis, the conserved target residues were mutated by degenerate codons that only indicated the conserved amino acid changes plus the wild-type residue. Residues that were not conserved were mutated more aggressively but also included wild-type residues. To avoid excessive mutagenesis of the target protein, degenerate codons were placed that also encoded the wild-type residue. For the same reason, only up to 20 positions were subject to mutagenesis at a time. These residues are a combination of contact residues and non-contact residues of the interaction region.

Oligonucleotides were dissolved in sterile water, mixed in equimolar ratios, heated to 95°C for five minutes and slowly cooled to room temperature for annealing. ECD-specific oligonucleotide primers that annealed to the start and end of the ECD, respectively, were then used to generate the PCR product. ECD-specific oligonucleotides that overlap by 40-50 bp. with a modified version of the cloning vector pBYDS03 (Life Technologies USA), beyond and including the BamH1 and Kpn1 cloning sites, was then used to amplify 100 μg of PCR product from the previous step to obtain a total of 5 μg of DNA. Both PCRs were performed by polymerase chain reaction (PCR) using OneTaq 2x PCR Mastermix (New England Biolabs, USA). The second PCR product was purified using a PCR purification kit (Qiagen, Germany) and resuspended in sterile deionized water.

To prepare library inserts, a modified version of the yeast display vector pBYDS03 was digested with restriction enzymes BamH1 and Kpn1 (New England Biolabs, USA), the large vector fragment was gel purified, and dissolved in sterile deionized water. Electroporation-ready DNA for the next step was prepared by mixing 12 μg of library DNA with 4 μg of linearized vector in a total volume of 50 μl of deionized and sterile water. An alternative way to create target libraries was to perform site-directed mutagenesis (Multisite kit, Agilent, USA) of target ECDs using oligonucleotides containing degenerate codons. This approach has been used to create sublibraries designed only to mutagenesis specific regions of a target protein. In these cases, the sublibraries were mixed before proceeding to the selection stages. In general, library sizes ranged from 10E7 to 10E8 clones, with sublibraries only in the 10E4 to 10E5 range. Large libraries and sub-libraries have been created for ICOSL.

F. Random libraries.

Random libraries were also constructed to identify ECD variants from ICOSL presented in SEQ ID NO: 32 (containing the ECD domain corresponding to residues 19-256, presented under UniProt accession No. O75144.2, flanked by adjacent N- and C-terminal residues of the wild sequence type). DNA encoding wild-type ECD was cloned between the BamH1 and Kpn1 sites of the modified yeast display vector pBYDS03 and then excised using the same restriction enzymes. The excised DNA was then subjected to mutagenesis using the Genemorph II kit (Agilent, USA) to obtain an average of three to five amino acid changes per library variant. The mutagenized DNA was then amplified by two-step PCR and further processed as described above for target libraries.

Example 2: Introduction of DNA libraries into yeast.

Example 2 describes the introduction of ICOSL DNA libraries into yeast.

To introduce degenerate and random library DNA into yeast, electroporation-competent cells of the yeast strain BJ5464 (ATCC.org, ATCC number 208288) were prepared and electroporated on a Gene Pulser II (Biorad, USA) with electroporation-ready DNA essentially with stages above as described (Colby, DW et al. 2004 Methods Enzymology 388, 348-358). The only exception is that transformed cells were grown in non-inducing minimal selection medium SCD-Leu to provide the selectable marker LEU2 carried by the modified plasmid pBYDS03.

Library size was determined by plating dilutions of freshly collected cells onto SCD-Leu agar plates, then extrapolating the library size from the number of single colonies when plated to produce at least 50 colonies per plate. The remainder of the electroporated culture was grown to saturation, and cells from this culture were again subcultured in the same medium to minimize the fraction of untransformed cells. To maintain library diversity, this subculture step was performed using an inoculum that contained at least 10x more cells than the calculated library size. Cells from the second saturated culture were resuspended in fresh medium containing 25% (w/v)

- 96 044356 sterile glycerol, to a density of 10E ¹⁰ /ml and frozen and stored at -80°C (frozen library stock).

One liter of SCD-Leu medium consists of 14.7 g sodium citrate, 4.29 g citric acid monohydrate, 20 g dextrose, 6.7 g Difco Nitrogen Yeast Agar Base and 1.6 g Leucine-Free Yeast Synthetic Exclusion Supplement. Before use, the medium was sterile filtered using a 0.2 μM vacuum filter device.

Library size was determined by plating dilutions of freshly collected cells onto SCD-Leu agar plates and then extrapolating the library size from the number of individual colonies when plated to produce at least 50 colonies per plate.

To separate the plasmid from cells that contain two or more different library clones, a number of cells corresponding to 10 times the size of the library were removed from the SCD-Leu culture overnight, subcultured 1/100 in fresh SCD-Leu medium, and grown overnight . Cells from this overnight culture were resuspended in sterile 25% (w/v) glycerol to a density of 10 E10/ml and frozen and stored at -80°C (frozen library stock).

Example 3. Yeast selection.

Example 3 describes the selection of yeast expressing affinity modified ICOSL variants.

A number of cells at least 10 times the library size were thawed from individual library stocks, suspended to 0.1 x 10E6 cells/ml in SCDLeu non-inducing medium and grown overnight. The next day, a number of cells 10 times the library size were centrifuged at 2000 rpm for two minutes and resuspended to 0.5 x 10E6 cells/ml in SCDG-Leu induction medium. One liter of SCDG-Leu induction medium consists of 5.4 g Na ₂ HPO ₄ , 8.56 g NaH ₂ PO ₄ * H ₂ O, 20 g galactose, 2.0 g dextrose, 6.7 g nitrogen yeast agar base Difco and 1.6 g of yeast synthetic exclusion medium without leucine, dissolved in water, and sterilized through a 0.22 micron membrane filter device. The culture was grown for two days at 20°C to induce expression of library proteins on the surface of yeast cells.

Cells were processed using magnetic beads to reduce the number of non-binding molecules and enrich for all variant ICOSLs that have the ability to bind their exogenous recombinant counterstructural proteins. This was followed by two to three rounds of flow cytometry sorting using exogenous counterstructural protein staining to enrich for the fraction of yeast cells presenting improved binding molecules. Enrichment by magnetic beads and selection by flow cytometry are primarily described in Keith D. Miller, 1 Noah B. Pefaur, 2 and Cheryl L. Bairdl Current Protocols in Cytometry 4.7.1-4.7.30, July 2008.

Using the ICOSL library, the following target ligand proteins were obtained from R&D Systems (USA): human rCD28.Fc (i.e., recombinant CD28-Fc fusion protein), rCTLA4.Fc and rICOS.Fc. Streptavidin magnetic beads were obtained from New England Biolabs, USA. For biotinylation of the counterstructural protein, biotinylation kit No. 21955, Life Technologies, USA, was used. A Becton Dickinson FACS Aria II sorter was used for two-color, flow cytometric sorting. ICOSL presentation levels were monitored with an anti-hemagglutinin antibody labeled Alexafluor 488 (Life Technologies, USA). Ligand-binding Fc fusion proteins rCD28.Fc, rCTLA4.Fc, or rCD112R.Fc were detected using PE-conjugated goat Fabs specific for human Ig (Jackson ImmunoResearch, USA). Double yeasts were selected above the threshold using forward scatter (FSC)/side scatter (SSC) parameters, and the sorting gate was based on the higher ligand binding found in FL4, which had more limited tag expression binding in FL1.

Yeast obtained after flow cytometric sorting was analyzed for higher specific binding affinity. The sorted yield yeasts were propagated and re-induced to express specific variants of the IgSF-modified domains they encode. This population can then be compared to the parental wild-type yeast strain or to any other gate-selected yeast, such as the gate-bead yeast population, using flow cytometry.

For ICOSL, the results of the second sort (F2) were compared with the parental ICOSL yeast for binding each of the number of rICOS.Fc, rCD28.Fc and rCTLA4.Fc by double staining each population with an anti-HA (hemagglutinin) expression tag of a secondary anti-human Fc antibody for detection of ligand binding.

For variant ICOSLs in yeast selected for binding to ICOS, F2 sort results yielded a mean fluorescence intensity (MFI) of 997 when stained with 5.6 nM rICOS.Fc, whereas the parent strain ICOSL had an MFI of 397 when stained with the same the rICOS.Fc concentration itself. This represents an approximately threefold improvement in average binding in this selected pool of F2 clones, and individual clones from this pool are predicted to have

- 97 044356 significantly improved MFI/affinity in individual testing.

For variant ICOSLs in yeast selected for binding to CD28, F2 sort results yielded MFI values of 640 when stained with 100 nM rCD28.Fc, whereas the MFI of the parent strain ICOSL was 29 when stained with the same concentration of rCD28.Fc (an improvement of 22 times). For the ICOSL yeast variants selected for binding to CTLA4, the F2 sort results yielded MFI values of 949 when stained with 100 nM rCTLA4.Fc, whereas the MFI of the parent strain ICOSL was measured at 29 when stained with the same concentration of rCTLA4.Fc (32-fold improvement ).

It is important to note that the MFI values of all F2 results described above, when measured with an antibody against the HA tag on FL1, were not increased and were sometimes decreased compared to wild-type strains, indicating that increased binding was not a function of increased expression of selected variants to the surface of yeast, and a validated selection strategy above a threshold only selects for medium and low expressors with high ligand binding.

Selected variant ICOSL ECD domains were further formatted as fusion proteins and tested for binding and functional activity as described below.

Example 4. Reformatting of selection results in the form of Fc fusion proteins and in various types of immunomodulatory proteins.

Example 4 describes the reformatting of the selection results into immunomodulatory proteins containing a modified (variant) extracellular domain (ECD) of ICOSL fused to an Fc molecule (variant ECD-Fc fusion protein molecules).

Cells obtained from the final flow cytometric sorting ICOSL were grown to maximum density in SCD-Leu medium. Plasmid DNA for each result was isolated using a yeast plasmid DNA isolation kit (Zymoresearch, USA). For Fc fusion, PCR primers with added restriction sites suitable for cloning into the selected Fc fusion vector were used to periodically amplify from preparations of plasmid DNA encoding the mutant ECD target DNA. After restriction digestion, the PNR products were ligated into the appropriate Fc vector, followed by chemical transformation into Blue E. Coli strain XL1 (Agilent, USA) or NEB5alpha (New England Biolabs), as specified by the supplier. An example of an Fc fusion vector is pFUSE-hIgG1-Fc2 (Invivogen, USA).

Dilutions of transformation reactions were plated on LB agar containing 100 μg/ml carbenicillin (Teknova, USA) to obtain individual colonies. Up to 96 colonies from each transformation were then grown in 96-well plates to saturation overnight at 37°C in LB broth (Teknova cat #L8112) and a small aliquot from each well was submitted for DNA sequencing of the ECD insert in order to identify the mutation ( i) in all clones. Sample preparation for DNA sequencing was performed using protocols provided by the service provider (Genewiz, Cayc Plainfield, NJ). After removing the DNA sequencing sample, glycerol was added to the remaining cultures to obtain a final glycerol content of 25% and the plates were stored at -20°C for future use as master plates (see below). Otherwise, samples for DNA sequencing were obtained by replication subculture from growing liquid cultures on solid agar plates using a disposable 96-well replicator (VWR, USA). These plates were incubated overnight to obtain growth patches, and the plates were provided by Genewiz, as directed by Genewiz. In some cases, resequencing was performed to check for mutations.

After identifying clones of interest from analysis of DNA sequencing data obtained from Genewiz, clones of interest were recovered from master plates and individually grown to density in 5 ml liquid LB broth containing 100 μg/ml carbenicillin (Teknova, USA) and 2 ml of each culture was then used to prepare about 10 μg of miniprep plasmid DNA of each clone using a standard kit such as the Pureyield kit (Promega). Identification of clones of interest typically involves the following steps. First, DNA sequence data files were downloaded from the Genewiz website. All sequences were then manually processed to start at the beginning of the ECD coding region. Validated sequences were then batch translated using a suitable program available at URL: www.ebi.ac.uk/Tools/st/emboss_transeq/. The translated sequences were then aligned using a suitable program available at URL: multalin.toulouse.inra.fr/multalin/multalin.html.

Clones of interest were then identified using the following criteria: 1) an identical clone occurs at least twice in the alignment and 2) a mutation occurs at least twice in the alignment and preferably in separate clones. Clones that met at least one of these criteria were clones that were enriched by our sorting process due to improved binding.

To produce immunomodulatory proteins that are Fc fusion proteins containing the ECD from ICOSL with at least one affinity modified domain (e.g., variant ICOSL ECD-Fc), an encoding nucleic acid molecule was prepared to encode

- 98 044356 design of a protein designed as follows: a signal peptide followed by a variant (mutant) ECD, followed by a three-alanine (AAA) linker, followed by a human IgG1 Fc containing the N82G mutation, relative to the wild-type human IgG1 Fc, presented in SEQ ID NO: 226 (which corresponds to N297G by EU numbering). This exemplary Fc also contained stabilizing cysteine mutations R77C and V87C and a replacement of a cysteine residue with a serine residue at position 5 (C5S), each relative to the wild-type human IgG1 Fc shown in SEQ ID NO: 226 (corresponding to R292C, V302C and C220S, respectively, according to EU numbering). In some cases, the NotI cloning site that contributes to the AAA linker sequence was removed to create a direct fusion of the ICOSL ECD and the start of the Fc. Since the construct does not include any antibody light chains that can form a covalent bond with cysteine, the human IgG1 Fc also contained a replacement of cysteine residues with a serine residue at position 5 (C5S) compared to the wild-type or unmodified Fc presented in SEQ ID NO: 226.

Example 5 Expression and Purification of Fc Fusion Proteins.

Example 5 describes the large-scale expression and purification of Fc fusion proteins containing the ICOSL variant ECD.

Recombinant variant Fc fusion proteins were produced using the Expi293 expression system (Invitrogen, USA). 4 μg of each plasmid DNA from the previous step was added to 200 μl of Opti-MEM (Invitrogen, USA) while separately adding 10.8 μl of ExpiFectamine to another 200 μl of Opti-MEM. After 5 min, 200 μl of plasmid DNA was mixed with 200 μl of ExpiFectamine and further incubated for an additional 20 min before adding the mixture to the cells. Ten million Expi293 cells were dispensed into individual sterile 10 ml wells of a 24-well conical bottom deep plate (Thomson Instrument Company, USA) in a volume of 3.4 ml Expi293 medium (Invitrogen, USA). The plates were shaken for 5 days at 120 rpm in a mammalian cell culture incubator at 95% humidity and 8% CO ₂ . After 5 days of incubation, cells were pelleted and culture supernatants were removed.

Protein was purified from supernatants using a high-throughput protein A purification kit in 96-well format according to the manufacturer's protocol (cat. no. 45202, Life Technologies, USA). The resulting eluted fractions were buffer exchanged to PBS using a Zeba 96 spin desalting plate (cat. no. 89807, Life Technologies, USA) according to the manufacturer's protocol. Purified protein was quantified at an optical density of 280 nm measured by a Nanodrop instrument (Thermo Fisher Scientific, USA), and protein purity was assessed by loading 5 μg of protein onto NUPAGE precast polyacrylamide gels (Life Technologies, USA) under denaturation and reduction conditions and then gelling. -electrophoresis. Proteins were visualized in the gel using standard Coomassie staining.

Example 6: Assessment of binding and activity of molecules containing an affinity mature IgSF domain.

A. Binding to counterstructures expressed by cells.

This example describes binding studies of Fc fusion proteins to demonstrate the specificity and affinity of variant ICOSL domain immunomodulatory proteins for cognate binding partners.

To obtain cells expressing cognate binding partners, constructs based on full-length mammalian surface expression sequences for each of human CD28 and ICOS were designed in the pcDNA3.1 expression vector (Life Technologies) and generated using GenScript, USA. Binding studies were performed using the Expi293F transient transfection system (Life Technologies, USA). The number of cells required for the experiment was determined and performed at the appropriate 30 ml scale using the manufacturer's recommended protocol. For each 30-ml CD28, ICOS, or negative control transfection, 75 million Expi293F cells were incubated with 30 μg of expression construct DNA and 1.5 ml of diluted ExpiFectamine 293 reagent for 48 h, after which the cells were collected for staining.

For flow cytometry staining, 200,000 cells from the corresponding transient transfection or negative control were seeded in 96-well round-bottom plates. Cells were centrifuged and resuspended in staining buffer (PBS, 1% BSA, and 0.1% sodium azide) for 20 min to block nonspecific binding. The cells were then centrifuged again and resuspended in 50 μl of staining buffer containing 100 nM-1 nM variant immunomodulatory protein, depending on the experiment with each of the candidate variant CD80 Fc, variant ICOSL Fc, or stacked variant IgSF Fc fusion protein. Primary staining was performed on ice for 45 min before washing the cells twice in staining buffer. PE-conjugated antibodies against human Fc (Jackson ImmunoResearch, USA) were diluted 1:150 in 50 μl of staining buffer and added to the cells and incubated for another 30 min on ice. The secondary antibody was washed out twice, cells were fixed in 4% formaldehyde/PBS, and samples were analyzed on a flow cytometer.

- 99 044356

FACScan (Becton Dickinson, USA) or Hypercyt flow cytometer (Intellicyte, USA)

Mean fluorescence intensity (MFI) was calculated for each transfectant and negative control parental line using Cell Quest Pro software (Becton

Dickinson, USA), or a Hypercyt flow cytometer (Intellicyte, USA).

B. Characterization of bioactivity.

This example further describes the characterization of the bioactivity of a variant Fc fusion protein in in vitro assays of primary human T cells.

1. Mixed lymphocyte reaction (MLR).

The bioactivity of soluble rICOSL.Fc was tested in the human mixed lymphocyte reaction (MLR). Primary human dendritic cells (DC) were prepared by culturing monocytes isolated from PBMCs (BenTech Bio, USA) in vitro for 7 days with 500 U/ml rIL-4 (R&D Systems, USA) and 250 U/ml rGM-CSF (R&D Systems, USA) in Ex-Vivo 15 environment (Lonza, Switzerland). 10,000 mature DCs and 100,000 purified allogeneic CD4+ T cells (BenTech Bio, USA) were co-cultured with ICOSL fusion protein, CF80 Fc fusion protein or control in 96-well round-bottom plates in 200 μl final volume of Ex-Vivo 15 medium. On day 5, IFNgamma secretion in culture supernatants was analyzed using the Human IFNgamma Duoset ELISA kit (R&D Systems, USA). Optical density was measured using a VMax ELISA microplate reader (Molecular Devices, USA) and determined against the rIFN-gamma titrated standard included in the IFN-gamma Duo kit kit (R&D Systems, USA). The second MLR protocol consisted of human primary dendritic cells (DCs) obtained by culturing monocytes isolated from PBMCs (BenTech Bio, USA) in vitro for 7 days with 50 ng/ml rIL-4 (R&D Systems, USA) and 80 ng/ml rGM-CSF (R&D Systems, USA) in Ex-Vivo 15 medium (Lonza, Switzerland). On days 3 and 5, half of the medium was removed and replaced with fresh medium containing 50 ng/ml rIL-4 and 80 ng/ml rGM-CSF. To fully induce DC maturation, lipopolysaccharide (LPS) (InvivoGen Corp., USA) at 100 ng/ml was added to DC cultures on day 6 and cells were incubated for an additional 24 h. About 10,000 mature DCs and 100,000 purified allogeneic CD3+ T cells (BenTech Bio, USA) were co-cultured with IFOSL Fc fusion proteins and control in 96-well round-bottom plates in 200 μl of a final volume of Ex-Vivo 15 medium. On day 5, IFN-gamma secretion in culture supernatants was analyzed with using the Human IFN-gamma Duoset ELISA kit (R&D Systems, USA). Optical density was measured on a BioTek Cytation Multimode Microplate Reader (BioTek Corp., USA) and determined using the rIFN-gamma titrated standard included in the IFN-gamma Duo Kit (R&D Systems, USA).

1. Coimmobilization assay with anti-CD3.

The costimulatory biological activity of the ICOSL fusion variants was determined in coimmobilization assays with anti-CD3. 1 or 10 nM mouse anti-human CD3 (OKT3, Biolegends, USA) was diluted in PBS with 1 nM-80 nM RICOSL.Fc variant proteins. This mixture was added to treated 96-well flat-bottom tissue culture plates (Corning, USA) overnight to facilitate adhesion of the stimulating proteins to the wells of the plate. The next day, unbound protein was washed off the plates and 100,000 purified human pan T cells (BenTech Bio, USA) or human BC3 T cell clones (Astarte Biologics, USA) were added to each well in a final volume of 200 μl Ex-Vivo 15 medium ( Lonza, Switzerland). In some cases, human pan-T cells were labeled with 0.25 μM carboxyfluorescein succinimidyl ester (CFSE, ThermoFisher Scientific, USA). Cells were cultured 3 days before collecting culture supernatants and measuring human IFN-gamma levels using a Duoset ELISA kit (R&D Systems, USA) as mentioned above. Cell proliferation was determined by the percentage of original cells that entered division by measuring by CFSE dilution on cells stained with fluorescently conjugated anti-CD4, anti-CD8 antibodies (BD, USA) or total T cells by flow cytometric analysis on LSR II ( BD, USA).

C. Results.

The results of binding and activity studies for exemplary tested variants are shown in table. 7, which identifies the amino acid substitution(s) in the exemplary IgSF domain of the ICOSL ECD selected by affinity mature screening against the corresponding ICOS and CD28 cognate structures. In the tables, exemplary amino acid substitutions are designated by the number of the amino acid position similar to the corresponding original unmodified ECD sequence, as follows. For example, the reference unmodified ECD sequence in Table. 7 (WT ICOSL) is the unmodified ICOSL ECD sequence shown in SEQ ID NO: 32. The amino acid position is indicated in the middle with the corresponding unmodified (eg, wild type) amino acid listed before the number and the identified amino acid substitution listed after the number. Column 2 lists the ECD variant SEQ ID NO for each variant ECD-Fc fusion molecule.

Also shown is the binding activity, as measured by the mean fluorescence intensity (MFI) value for binding of each variant Fc fusion molecule to cells engineered

- 100 044356 mi for expression of a cognate counterstructural ligand, and the MFI ratio compared to the binding of a corresponding unmodified ECD-Fc fusion protein molecule containing no amino acid substitution(s) to the same cellularly expressed counterstructural ligand. The functional activity of the variant Fc fusion molecules in modulating T cell activity is also shown based on calculated IFN-gamma levels in culture supernatants (pg/ml) produced with either i) the indicated variant ECD-Fc fusion protein molecule coimmobilized with anti-CD3 or ii) with the specified variant ECD-Fc fusion protein molecule in an MLR assay. The table also shows the ratio of IFN-gamma produced by each variant ECD-Fc compared to the corresponding unmodified ECD-Fc (wild type) in both functional assays.

As shown, screens have led to the identification of a number of variant ICOSL IgSF domains that have been modified in affinity, demonstrating increased binding to at least one, and in some cases more than one, cognate counterstructure ligand. In addition, the results showed that affinity modification of variant molecules also exhibited improved activities for both increasing and decreasing immunological activity depending on the molecule format. For example, ligand coimmobilization likely provides a multivalent interaction with the cell to cluster or increase avidity, which promotes agonist activity and increases T cell activation compared to an unmodified (e.g., wild-type) ECD-Fc molecule lacking the amino acid substitution(s). . However, when the molecule is presented as a divalent Fc molecule in solution, the same variant IgSF domains exhibit antagonistic activity in reducing T cell activation compared to an unmodified ECD-Fv molecule (eg, wild type) lacking the amino acid substitution(s).

Table 7

ICOSL variants selected against CD28 or ICOS. Molecular sequences, binding data and costimulatory bioactivity data

Mutation(s) ICOSL SEQ ID NO (ECD) Binding Coimmobilization with aHTH-CD3 MLR ICOS OD (original ratio) CD28 MFI (original ratio) IFN-gamma pg/ml (initial ratio) IFN- levels gamma pg/ml (original ratio) N52S 109 1.33 (1.55) 162 (9.00) 1334 (1.93) 300 (0.44) N52H 110 1.30 (1.51) 368 (20.44) 1268 (1.83) 39 (0.06) N52D 111 1.59 (1.85) 130 (7.22) 1943 (2.80) 190 (0.28) N52Y/N57Y/F138L/ L203P 112 1.02 (1.19) 398 (22.11) 510* (1.47*) 18 (0.03) N52H/N57Y/Q100P 113 1.57 (1.83) 447 (24.83) 2199 (3.18) 25 (0.04) N52S/Y146C/Y152C 114 1.26 (1.47) 39 (2.17) 1647 (2.38) 152 (0.22) N52H/C198R 115 1.16 (1.35) 363 (20.17) 744* (2.15*) ND (ND) N52H/C140del/T225 A 372 ND (ND) 154 (8.56) 522* (1.51*) ND (ND) N52H/C198R/T225A 117 1.41 (1.64) 344 (19.11) 778* (2.25*) 0 (0)

- 101 044356

N52H/K92R 118 1.48 (1.72) 347 (19.28) 288* (0.83*) 89 (0.13) N52H/S99G 119 0.09 (0.10) 29 (1.61) 184* (0.53*) 421 (0.61) N52Y 120 0.08 (0.09) 18 (1.00) 184* (0.53*) 568 (0.83) N57Y 121 1.40 (1.63) 101 (5.61) 580* (1.68*) 176 (0.26) N57Y/Q100P 122 0.62 (0.72) 285 (15.83) 301* (0.87*) 177 (0.26) N52S/S130G/Y152C 123 0.16 (0.19) 24 (1.33) 266* (0.77*) 1617 (2.35) N52S/Y152C 124 0.18 (0.21) 29 (1.61) 238* (0.69*) 363 (0.53) N52S/C198R 125 1.80 (2.09) 82 (4.56) 1427 (2.06) 201 (0.29) N52Y/N57Y/Y152C 126 0.08 (0.09) 56 (3.11) 377* (1.09*) 439 (0.64) N52Y/N57Y/H129P/ C198R 127 ND (ND) 449 (24.94) 1192 (1.72) ND (ND) N52H/L161P/C198R 128 0.18 (0.21) 343 (19.05) 643* (1.86*) 447 (0.65) N52S/T113E 129 1.51 (1.76) 54 (3.00) 451* (1.30*) 345 (0.50) S54A 130 1.62 (1.88) 48 (2.67) 386* (1.12*) 771 (1.12) N52D/S54P 368 1.50 (1.74) 38 (2.11) 476* (1.38*) 227 (0.33) N52K/L208P 132 1.91 (2.22) 291 (16.17) 1509 (2.18) 137 (0.20) N52S/Y152H 133 0.85 (0.99) 68 (3.78) 2158 (3.12) 221 (0.32) N52D/V151A 134 0.90 (1.05) 19 (1.06) 341* (0.99*) 450 (0.66) N52H/I143T 135 1.83 (2.13) 350 (19.44) 2216 (3.20) 112 (0.16) N52S/L80P 136 0.09 (0.10) 22 (1.22) 192* (0.55*) 340 (0.49) F120S/Y152H/N201 S 137 0.63 (0.73) 16 (0.89) 351* (1.01*) 712 (1.04) N52S/R75Q/L203P 138 1.71 (1.99) 12 (0.67) 1996 (2.88) 136 (0.20) N52S/D158G 139 1.33 (1.55) 39 (2.17) 325* (0.94*) 277 (0.40) N52D/Q133H 140 1.53 (1.78) 104 (5.78) 365* (1.05*) 178 (0.26) WT ICOSL 32 0.86 (1.00) 18 (1.00) 692 /346* (1.00) 687 (1.00)

*Initial ratio calculated using IFN-gamma 346 pg/ml for WT ICOSL.

Binding assays were repeated essentially as described above, except that binding was also assessed against cells expressing full-length human CTLA4. ICOSL Fc fusion protein variants were also further evaluated in an anti-CD3 coimmobilization assay as described above. The results confirmed the identification of a number of IgSF variant ICOSL domains that exhibited increased binding affinity to at least one, and in some cases more than one, cognate ligand. In addition, the results showed that affinity modification of the variant molecules also exhibited improved activities in the coimmobilization assay.

Example 7: Additional Affinity Modified IgSF Domains.

These examples describe the design, generation, and screening of additional sensitized modified CD80 (B7-1), CD86 (B7-2), and NKM30 immunomodulatory proteins, which are other components of the immune synapse (IS) that have demonstrated dual roles in both immune activation and and in inhibition. These examples demonstrate that affinity modification of IgSF domains produces proteins that can act to either enhance or reduce immunological activity. This work also describes various combinations of such domains fused in pairs (i.e., stacked) with an affinity-modified variant ICOSL to form a type II immunomodulatory protein to produce immunomodulatory activity.

Mutant DNA constructs of the human CD80, CD86, and NKp30 IgSF domains for translation and expression as yeast display libraries were prepared essentially as described in Example 1. For libraries that target specific residues of the target protein, either complete or partial randomization with degenerate codons encoding DNA for the extracellular domains (ECD) of human CD80 (SEQ ID NO: 28) and NKp30 (SEQ ID NO: 54) were ordered from Integrated DNA Tech- 102 044356 nologies (Coralville, Iowa) as a set of overlapping oligonucleotides up to 80 base pairs (bp) long. Otherwise, the remaining residues were mutated by site-directed mutagenesis essentially as described in Example 1. Otherwise, random libraries were constructed to identify ECD variants CD80 (SEQ ID NO: 28), CD86 (SEQ ID NO: 29) and NKp30 (SEQ ID NO: 54) essentially as described in Example 1.

The target DNA and random library DNA were introduced into yeast essentially as described in Example 2 to produce a yeast library. The libraries were used to select yeast expressing affinity modified CD80, CD86 and NKp30 variants essentially as described in Example 3. Cells were treated to reduce the number of non-interacting molecules and enriched for CD80, CD86 or NKp30 variants to bind their exogenous recombinant counterstructure proteins essentially as described in Example 3. For example, yeast presented targeted or random CD80 libraries for each of CD28, CTL-4 and PD-L1 separately. This was followed by two to three rounds of flow cytometric sorting using exogenous counterstructure protein staining to enrich for the fraction of yeast cells presenting improved binding molecules. Enrichment by magnetic beads and selection by flow cytometry are primarily described in Keith D. Miller, 1 Noah B. Pefaur, 2 and Cheryl L. Baird1 Current Protocols in Cytometry 4.7.1-4.7.30, July 2008.

Using the CD80, CD86, ICOSL and NKp30 libraries at R&D Systems (USA), the following target protein ligands were obtained: human rCD28.Fc (i.e. recombinant CD28-Fc fusion protein), rPDL1.Fc, rCTLA4.Fc and rB7H6.Fc. Two-color flow cytometry was performed essentially as described in Example 3. Yeast resulting from flow cytometry sorting was analyzed for higher specific binding affinity. The resulting sorted yeasts were propagated and re-induced to express specific affinity-modified variants of the IgSF domains that they encode. This population could then be compared to the parental wild-type yeast strain or to any other yeast, such as a bead-selected yeast population, using flow cytometry.

In the case of yeast with NKp30 variants selected for binding to B7-H6, F2 sort yields gave MFI values of 533 when stained with 16.6 nM rB7H6. Fc, whereas the parent strain NKp30 gave an MFI value of 90 when stained with the same concentration of rB7H6.Fc (6-fold improvement).

Among the NKp30 variants that were identified was a variant that contained the L30V/A60V/S64P/S86G mutation relative to positions in the extracellular domain of NKp30 corresponding to the positions presented in SEQ ID NO: 54. Among the CD86 variants identified was a variant that contained the Q35H mutations /H9OL/Q1O2H relative to positions in the extracellular domain of CD86 corresponding to the positions presented in SEQ ID NO: 29. Among the CD80 variants identified were the variants shown in table. 8 and further described below.

It is important to note that the MFI values for all F2 yields described above, when measured with an antibody against the HA tag on FL1, were not increased and sometimes decreased compared to wild-type strains, indicating that the increased binding was not a function of increased expression of selected variants on the surface of yeast, and validated selection strategies above the threshold only select for medium and low expressors with high ligand binding.

The exemplary selection results were reformatted as immunomodulatory proteins containing an affinity-modified (variant) extracellular domain (ECD) from CD80, combined with an Fc molecule (variant ECD-Fc fusion protein molecules) essentially as described in Example 4, and fused to Fc protein was expressed and purified essentially as described in Example 5.

The binding of exemplary variant CD80-Fc fusion proteins to cell-expressed counterstructures was then assessed, as described in Example 6. To obtain cells expressing cognate binding partners, surface expression constructs of full-length mammalian proteins were generated for each of CD28, CTLA4, and PD- human L1 as described in Example 6. Binding studies and flow cytometry were performed essentially as described in Example 6. In addition, the biological activity of the variant Fc fusion protein was characterized by either mixed lymphocyte reaction (MLR) or coimmobilization with anti- CD3 essentially as described in Example 6.

The results of binding and activity studies for exemplary tested variants are shown in table. 8 and 9. In particular, in table. 8 shows exemplary amino acid substitutions (substitutions) of the IgSF domain in the ECD from CD80, selected by affinity maturation screening against the corresponding cognate structure of CD28. Table 9 shows exemplary IgSF domain amino acid substitutions in ECDs from CD80 selected by affinity maturation screening against the corresponding PD-L1 cognate structure. As above, for each table, exemplary amino acid substitutions are indicated by the number of the amino acid position similar to the corresponding one in the reference unmodified ECD sequence. For example, the reference unmodified ECD sequence in Table. 8 and 9 represent the unmodified ECD sequence

CD80, shown in SEQ ID NO: 28. The amino acid position is indicated in the middle with the corresponding unmodified (eg, wild type) amino acid listed before the number and the identified amino acid substitution listed after the number. Column 2 indicates the variant ECD SEQ ID NO for each variant ECD-Fc fusion protein molecule.

Also shown is the binding activity, as measured by the mean fluorescence intensity (MFI) value for the binding of each variant Fc fusion protein molecule to cells engineered to express the cognate counterstructure ligand, and the ratio of the MFI compared to the binding of the corresponding unmodified ECD-Fc fusion protein molecule, not containing amino acid substitution(s) to the same cell-expressed counterstructural ligand. The functional activity of the variant Fc fusion molecules to modulate T cell activity is also shown based on calculated IFN-gamma levels in culture supernatants (pg/ml) produced with either i) the indicated variant ECD-Fc fusion protein molecule coimmobilized with anti-CD3 or ii) with the specified variant ECD-Fc fusion protein molecule in an MLR assay. The tables also show the IFN-gamma ratio obtained with each ECD-Fc variant compared to the corresponding unmodified ECD-Fc in both functional assays.

As shown, the screens led to the identification of a number of affinity modified CD80 IgSF domain variants that exhibited increased binding to at least one, and in some cases more than one, cognate counterstructure ligand. In addition, the results showed that affinity modification of variant molecules also exhibited improved activities for both increasing and decreasing immunological activity depending on the molecule format. For example, ligand coimmobilization likely provides a multivalent interaction with the cell to cluster or increase avidity, which promotes agonist activity and increases T cell activation compared to an unmodified (e.g., wild-type) ECD-Fc molecule lacking the amino acid substitution(s). . However, when the molecule is presented as a divalent Fc molecule in solution, the same variant IgSF domains exhibit antagonistic activity to reduce T cell activation compared to an unmodified (eg, wild type) ECD-Fv molecule lacking the amino acid substitution(s).

Table 8

CD80 variants selected against CD28. Molecular sequences, binding data and costimulatory bioactivity data

EQ Binding Coimmobilization with anti-POP IFN-gamma MLR levels CD28 CTLA- PD-L1 Mutation(s) CD 80 ID NO MFI (original 4 MFI (Exodus- MFI (original pg/ml (initial IFN- gamma new new new proportionally pg/ml (ECD) proportionally proportionally proportionally sewing) (Exodus sewing) sewing) sewing) new proportionally sewing) L70Q/A91G/N144D 508 125 (1.31) 283 (1.36) 6 (0.08) 93 (1.12) 716 (0.83) L70Q/A91G/T130A 56 96 (1.01) 234 (1.13) 7 (0.10) 99 (1.19) 752 (0.87) L70Q/A91G/I118A/ 123 226 7 86 741 T120S/T130A/K169E 59 (1.29) (1.09) (0.10) (1.03) (0.86) V4M/L70Q/A91G/1118 V/T 120 S/T 1 89 263 6 139 991 30A/K169E 510 (0.94) (1.26) (0.09) (1.67) (1.14) L70Q/A91G/I118V/T120S/T130A/K 106 263 6 104 741 169E 59 (1.12) (1.26) (0.09) (1.25) (0.86) V20L/L70Q/A91S/1118 V/T 120 S/T 1 105 200 9 195 710 30A 513 (1.11) (0.96) (0.13) (2.34) (0.82) S44P/L70Q/A91G/T130A 61 88 (0.92) 134 (0.64) 5 (0.07) 142 (1.71) 854 (0.99) L70Q/A91G/E117G/1118 V/T 120 S/T 120 (1.27) 193 6 98 736 130A 514 (0.93) (0.08) (1.05) (0.85) A91G/1118 V/T 120 S/T 13 0 A 515 84 (0.89) 231 (1.11) 44 (0.62) 276 (3.33) 714 (0.82)

- 104 044356

L70R/ A91G/1118 V/T 120 S/T 13 0 A/T 199S 516 125 (1.32) 227 (1.09) 6 (0.09) 105 (1.26) 702 (0.81) L70Q/E81 A/A91G/1118 V/T 12 0 S/11 140 185 18 98 772 27T/T130A 517 (1.48) (0.89) (0.25) (1.18) (0.89) L70Q/Y87N/A91 G/T 13 0 A 66 108 (1.13) 181 (0.87) 6 (0.08) 136 (1.63) 769 (0.89) T28 S/L70Q/A91G/1118 V/E95K/T12 32 65 6 120 834 0S/I126V/T130A/K169E 518 (0.34) (0.31) (0.08) (1.44) (0.96) N63 S/L70Q/A91G/S114T/1118 V 124 165 6 116 705 /T120S/T130A 519 (1.30) (0.79) (0.08) (1.39) (0.81) KZ 6E/I67T/L70Q/A91G/1118 V/T 12 8 21 5 53 852 0S/T130A/N152T 520 (0.09) (0.10) (0.08) (0.63) (0.98) E52G/L70Q/A91G/D107N/1118 V/T 113 245 6 94 874 120S/T130A K169E 521 (1.19) (1.18) (0.08) (1.13) (1.01) K37E/F59S/L70Q/A91G/I118V/T12 20 74 6 109 863 0S/T130A/K185E 522 (0.21) (0.36) (0.08) (1.31) (1.00) A91G/S103P 72 39 (0.41) 56 (0.27) 9 (0.13) 124 (1.49) 670 (0.77) K89E/T130A 73 90 (0.95) 148 (0.71) 75 (1.07) 204 (2.45) 761 (0.88) A91G 96 200 85 220 877 74 (1.01) (0.96) (1.21) (2.65) (1.01) D60 V/A91G/1118 V/T 120 S/T 13 0 A/ 111 222 12 120 744 K169E 523 (1.17) (1.07) (0.18) (1.44) (0.86) K54M/L70Q/A91 G/Yl 64H 524 68 (0.71) 131 (0.63) 5 (0.08) 152 (1.83) 685 (0.79) M3 8T/L70Q/E77G/A91G/1118 V/T 1 61 102 5 119 796 20S/T130A/N152T 525 (0.64) (0.49) (0.07) (1.43) (0.92) R29H/E52G/L70R/E88G/A91 G/T 13 100 119 5 200 740 0A 78 (1.05) (0.57) (0.08) (2.41) (0.85) Y31H/T41G/M43L/L70Q/A91G/I11 85 85 6 288 782 8V/T120S/I126V/T130A 526 (0.89) (0.41) (0.08) (3.47) (0.90) V68A/T110A 103 233 48 163 861 80 (1.08) (1.12) (0.68) (1.96) (0.99) L65H/D90G/T110A/F116L 527 33 (0.35) 121 (0.58) 11 (0.15) 129 (1.55) 758 (0.88) R29H/E52G/D90N/I118V/T120S/T1 141 eleven 124 30A 66 800 82 (0.69) (0.68) (0.15) (1.49) (0.92) A91G/L102S 83 6 (0.06) 6 (0.03) 5 (0.08) 75 (0.90) 698 (0.81) I67T/L70Q/A91G/I118V T120S 530 98 (1.03) 160 (0.77) 5 (0.08) 1751 (21.1) 794 (0.92) L70Q/A91 G/T 110 A/1118 V/T 120 S/T 8 14 5 77 656 130A 531 (0.09) (0.07) (0.07) (0.93) (0.76) M3 8 V/T41D/M43I/W50G/D76G/V8 82 (0.99) 671 (0.78) 3 A/K89E/1118 V/T 120 S/1126 V/T 13 0 A 532 5 (0.06) 8 (0.04) 8 (0.11) V22A/L70Q/S121P 87 5 (0.06) 7 (0.04) 5 (0.07) 105 (1.27) 976 (1.13) Al 2 V/S15F/Y31H/M3 8L/T41G/M4 104 (1.25) 711 (0.82) 3L/D90N/T13 0 A/P13 7L/N149D N152T 533 6 (0.06) 6 (0.03) 5 (0.08) I67F/L70R/E88G/A91G/I118V/T120 5 6 6 62 1003 S/T130A 534 (0.05) (0.03) (0.08) (0.74) (1.16) E24G/L25P/L70Q/A91G/1118 V/T 12 26 38 8 101 969 0S/N152T 535 (0.27) (0.18) (0.11) (1.21) (1.12) A91G/F92L/F108L/1118 V/T 120 S 536 50 (0.53) 128 (0.61) 16 (0.11) 59 (0.71) 665 (0.77) WT CD80 95 208 70 83 866 28 (1.00) (1.00) (1.00) (1.00) (1.00)

- 105 044356

Table 9

CD80 variants selected against PD-L1. Molecular sequences, ________binding data and costimulatory bioactivity data _______

Mutation(s) CD 80 EQ ID NO (ECD) Binding Coimmobilization with anti-POPs MLR CD28 MFI (original ratio) CTLA-4 MFI (original ratio) PD-L1 MFI (original ratio) IFN-gamma pg/ml (initial ratio) levels IFN- gamma pg/ml (original ratio) R29D/Y31L/Q33H/ K36G/M38I/ Т41А/ 92 1071 (0.08) 1089 (0.02) 37245 (2.09) 387 (0.76) 5028 (0.26) M43R/M47T/E81V/ L85R/K89N/A91T/ F92P/K93 V/ R94L/ I118T/N149S R29D/Y31L/Q33H/ K36G/M38I/T41A/ M43R/M47T/E81V/ L85R/K89N/A91T/ F92P/K93V/R94L/ N144S/N149S 93 1065 (0.08) 956 (0.02) 30713 (1.72) 400 (0.79) 7943 (0.41) R29D/Y31L/Q33H/ K36G/M38I/T41A/ M42T/M43R/M47T/ E81V/L85R/K89N/ A91T/F92P/K93V/ R94L/L148S/N149S 94 926 (0.07) 954 (0.02) 47072 (2.64) 464 (0.91) 17387 (0.91) E24G/R29D/Y31L/ Q33H/K36G/M38I/ T41A/M43R/M47T/ F59L/E81V/L85R/ K89N/A91T/F92P/ K93V/R94L/H96R 95 1074 (0.08) 1022 (0.02) 1121 (0.06) 406 (0.80) 13146 (0.69) R29D/Y31L/Q33H/ K36G/M38I/T41A/ M43R/M47T/E81V/ L85R/K89N/A91T/ F92P/K93V/R94L/ N149S 96 1018 (0.08) 974 (0.02) 25434 (1.43) 405 (0.80) 24029 (1.25) R29V/M43Q/E81R/ L85I/K89R/D90L/ A91E/F92N/K93Q/ R94G 97 1029 (0.08) 996 (0.02) 1575 (0.09) 342 (0.67) 11695 (0.61) T41I/A91G 98 17890 (1.35) 50624 (1.01) 12562 (0.70) 433 (0.85) 26052 (1.36) E88D/K89R/D90K/A91 G/ F92Y/K93R/N122S/ N178S 537 41687 (3.15) 49429 (0.99) 20140 (1.13) 773 (1.52) 6345 (0.33) E88D/K89R/D90K/A91 G/F92Y/K93R 538 51663 (3.91) 72214 (1.44) 26405 (1.48) 1125 (2.21) 9356 (0.49) K36G/K37Q/M38I/ L40M/F59L/E81V/L85R /K89N/A91T/F92P/ K93V/R94L/E99G/ T130A/N149S 539 1298 (0.10) 1271 (0.03) 3126 (0.18) 507 (1.00) 3095 (0.16) AE88D/K89R/D90K/ A91G/F92Y/K93R 102 31535 (2.38) 50868 (1.02) 29077 (1.63) 944 (1.85) 5922 (0.31) K36G/K37Q/M38I/ L40M 103 1170 (0.09) 1405 (0.03) 959 (0.05) 427 (0.84) 811 (0.04) K36G/L40M 540 29766 (2.25) 58889 (1.18) 20143 (1.13) 699 (1.37) 30558 (1.59) WTCD80 28 13224 (1.00) 50101 (1.00) 17846 (1.00) 509 (1.00) 19211 (1.00)

Example 8: Preparation and Evaluation of Stacked Molecules Containing Various Affinity Modified Domains.

This example further describes immunomodulatory proteins that have been engineered as stacked constructs containing at least two different affinity modified domains from identified ICOSL variant polypeptides and one additional

- 106 044356 variant of the molecules CD80, CD86, ICOSL and NKp30, interconnected and fused with Fc.

Selected variant molecules described above that were modified in affinity for one or more counterstructure ligands were used to prepare a stack molecule (ie, type II immunomodulatory protein) containing two or more affinity-modified IgSF domains. Stack constructs were generated as gene blocks (Integrated DNA Technologies, Coralville, Iowa) that encode the stack in a format that allows for fusion to Fc by standard Gibson assembly using the Gibson assembly kit (New England Biolabs).

The encoding nucleic acid molecule of all stacks was prepared to encode a protein constructed as follows: a signal peptide followed by the first IgV variant, followed by a 15 amino acid linker that consists of three GGGGS (G4S) motifs (SEQ ID NO: 228), followed by followed by a second IgV, followed by two GGGGS linkers (SEQ ID NO: 229), followed by three alanines (AAA), followed by a human IgG1 Fc, as described above. To maximize the likelihood of correct folding of the IgV domains in each stack, the first IgV was preceded by all residues that are normally found in the wild-type protein between that IgV and the signal peptide (leader sequence). Likewise, the first IgV was followed by all the residues that normally bind it in the wild-type protein, either to the next Ig domain (usually the IgC domain) or, if such a second IgV domain is absent, to the residues that connect it to the transmembrane domain (trailer sequence ). The same design principle was applied to the second IgV domain, except that when both IgV domains were derived from the same parent protein (e.g., IgV CD80 stacked with another IgV CD80), the linker between them was not duplicated.

In table Figure 11 shows the design of exemplary stack structures. Exemplary stacking molecules shown in Table. 10 contain Ig domains (eg, IgV domain) as indicated, and additionally trailer sequences as described above. The table shows the following components: signal peptide (SP, SEQ ID NO: 225), Ig domain 1 (for example, Ig1), trailer sequence 1 (TS1), linker 1 (LR1, SEQ ID NO: 228), Ig domain 2 (Ig2), trailer sequence 2 (TS2), linker 2 (LR2, SEQ ID NO: 230) and Fc domain (SEQ ID NO: 226, containing the amino acid substitution C5S/R77C/N82G/V87C). In some cases, leader sequence 1 (LS1) is present between the signal peptide and IgV1, and in some cases, leader sequence 2 (LS2) is present between the linker and IgV2.

- 107 044356

Table 10

Amino acid sequence (SEQ ID NO) of components of exemplary stacking constructs

S P First domain LR 1 Second domain LR2 Fc LSI igi TSI LS2 Ig2 TS2 Domain 1: NKp30 WT Domain 2: ICOSL WT + - 214 235 + - 196 233 + + Domain 1: NKp30 L30V/A60V/S6 4P/S86G Domain 2: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R + - 215 235 + - 212 233 + + Domain 1: NKp30 L30V/A60V/S6 4P/S86G) Domain 2: ICOSL N52D + - 215 235 + - 199 233 + + Domain 1: NKp30 L30V/A60V/S6 4P/S86G Domain 2: ICOSL N52H/N57Y/Q1 OOP + - 215 235 + - 201 233 + + Domain 1: ICOSL WT Domain 2: Nkp30 WT + - 196 233 + - 214 235 + + Domain 1: ICOSL N52D Domain 2: NKp30 L30V/A60V/S6 + - 199 233 + - 215 235 + +

- 108 044356

4P/S86G Domain 1: ICOSL N52H/N57Y/Q1 OOP Domain 2: NKp30 L30V/A60V/S6 4P/S86G + - 201 233 + - 215 235 + + Domain 1: CD80 WT Domain 2: ICOSL WT + - 152 471 + - 196 233 + + Domain 1: CD80 E88D/K89R/D9 0K/A91G/F92Y/ K93R Domain 2: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S + - 189 471 + - 213 233 + + Domain 1: CD80 A12T/H18L/M4 3V/F59L/E77K/ P109S/I118T Domain 2: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S + - 193 471 + - 213 233 + + Domain 1: CD80 A12T/H18L/M4 3V/F59L/E77K/ P109S/I118T Domain 2: ICOSL N52D + - 193 471 + - 199 233 + + Domain 1: CD80 E88D/K89R/D9 0K/A91G/F92Y/ K93R Domain 2: ICOSL N52H/N57Y/Q1 OOP + - 189 471 + - 201 233 + + Domain 1: CD80 A12T/H18L/M4 3V/F59L/E77K/ P109S/I118T + - 193 471 + - 201 233 + +

- 109 044356

Domain 2: ICOSL N52H/N57Y/Q1 OOP Domain 1: ICOSL WT Domain 2: CD80 WT + - 196 233 + - 152 471 + + Domain 1: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S Domain 2: CD80 E88D/K89R/D9 0K/A91G/F92Y/ K93R + - 213 233 + - 189 471 + + Domain 1: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S Domain 2: CD80 A12T/H18L/M4 3V/F59L/E77K/ P109S/I118T + - 213 233 + - 193 471 + + Domain 1: ICOSL N52D Domain 2: CD80 E88D/K89R/D9 0K/A91G/F92Y/ K93R + - 199 233 + - 189 471 + + Domain 1: ICOSL N52D Domain 2: CD80 A12T/H18L/M4 3V/F59L/E77K/ P109S/I118T + - 199 233 + - 193 471 + + Domain 1: ICOSL N52H/N57Y/Q1 OOP Domain 2: CD80 E88D/K89R/D9 0K/A91G/F92Y/ K93R + - 201 233 + - 189 471 + + Domain 1: ICOSL + - 201 233 + - 193 471 + +

- 110 044356

N52H/N57Y/Q1 OOP Domain 2: CD80 A12T/H18L/M4 3V/F59L/E77K/ P109S/I118T Domain 1: CD86 WT Domain 2: ICOSL WT + 236 220 237 + - 196 233 + + Domain 1: CD80 R29H/Y31H/T4 1G/Y87N/E88G/ K89E/D90N/A9 1G/P109S Domain 2: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S + - 192 471 + - 213 233 + + Domain 1: CD80 I67T/L70Q/A91 G/T120S Domain 2: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S + - 175 471 + - 213 233 + + Domain 1: CD80 R29H/Y31H/T4 1G/Y87N/E88G/ K89E/D90N/A9 1G/P109S Domain 2: ICOSL N52D + - 192 471 + - 199 233 + + Domain 1: CD80 I67T/L70Q/A91 G/T120S Domain 2: ICOSL N52D + - 175 471 + - 199 233 + + Domain 1: CD80 R29H/Y31H/T4 1G/Y87N/E88G/ K89E/D90N/A9 1G/P109S Domain 2: ICOSL N52H/N57Y/Q1 + - 192 471 + - 201 233 + +

- 111 044356

OOP Domain 1: CD80 I67T/L70Q/A91 G/T120S Domain 2: ICOSL N52H/N57Y/Q1 OOP + - 175 471 + - 201 233 + + Domain 1: CD86 Q35H/H90L/Q1 02H Domain 2: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S + 236 221 237 + - 213 233 + + Domain 1: CD86 Q35H/H90L/Q1 02H Domain 2: ICOSL N52D + 236 221 237 + - 199 233 + + Domain 1: CD86 Q35H/H90L/Q1 02H Domain 2: ICOSL N52H/N57Y/Q1 OOP + 236 221 237 + - 201 233 + + Domain 1: ICOSL WT Domain 2: CD86 WT + - 196 233 + 236 220 237 + + Domain 1: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S Domain 2: CD80 R29H/Y31H/T4 1G/Y87N/E88G/ K89E/D90N/A9 1G/P109S + - 213 233 + - 192 471 + + Domain 1: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S Domain 2: CD80 + - 213 233 + - 175 471 + +

- 112044356

I67T/L70Q/A91 G/T120S Domain 1: ICOSL N52D Domain 2: CD80 R29H/Y31H/T4 1G/Y87N/E88G/ K89E/D90N/A9 1G/P109S + - 199 233 + - 192 471 + + Domain 1: ICOSL N52D Domain 2: CD80 I67T/L70Q/A91 G/T120S + - 199 233 + - 175 471 + + Domain 1: ICOSL N52H/N57Y/Q1 OOP Domain 2: CD80 R29H/Y31H/T4 1G/Y87N/E88G/ K89E/D90N/A9 1G/P109S + - 201 233 + - 192 471 + + Domain 1: ICOSL N52S/N57Y/H9 4D/L96F/L98F/ Q100R/G103E/ F120S Domain 2: CD86 Q35H/H90L/Q1 02H + - 213 233 + 236 221 237 + + Domain 1: ICOSL N52D Domain 2: CD86 Q35H/H90L/Q1 02H + - 199 233 + 236 221 237 + + Domain 1: ICOSL N52H/N57Y/Q1 OOP Domain 2: CD86 Q35H/H90L/Q1 02H + - 201 233 + 236 221 237 + +

Large-scale expression and purification of variant IgV stacked-Fc fusion proteins containing various combinations of variant IgV domains from CD80, CD86, ICOSL, or Nkp30 containing at least one affinity-modified IgV domain was performed as described in Example 5 Binding of variant IgV-stack-Fc fusion proteins to the corresponding counterstructures and functional activity using an anti-CD3 co-immobilization assay was also assessed as described in Example 6. For example, the co-stimulatory bioactivity of IgSF Fc stack fusion proteins was determined in a similar immobilization assay. anti-CD3 as described above. In this case, 4 nM bhtu-CD3 (OKT3, Biolegend, USA) was coimmobilized with 4-120 nM human rB7-H6.Fc (R&D Systems, USA) or human rPD-L1.Fc (R&D Systems, USA) overnight in 96-well treated tissue culture plates (Corning, USA). The next day, unbound protein was washed away with PBS and 100,000 purified pan-T cells were added to each well in 100 μl of Ex-Vivo 15 medium (Lonza, Switzerland). IgSF stack domains were then added at concentrations ranging from 8 to 40 nM in a volume of 100 μL for a total volume of 200 μL. Cells were cultured 3 days before collecting culture supernatants and measuring human IFN-gamma levels using a Duoset ELISA kit (R&D Systems, USA) as mentioned above.

The results are shown in table. 11-13. In particular, in table. Figure 11 shows the binding and functional activity results of variant IgV-Fc stack protein fusion molecules containing the IgV NKp30 domain and the IgV ICOSL domain. In table 12 and 13 show binding and functional activity results for variant IgV-stack-Fc fusion proteins containing a variant ICOSL domain of IgV and variant Ig80 IgV or CD86 IgV domains.

For each of the tables. 11-13 Column 1 shows the structural organization and orientation of domains in a stack of affinity-modified or wild-type (WT) domains, with the amino-terminal (N-terminal) domain followed by the middle WT or affinity-modified domain located before the C-terminal domains human IgG1 Fc. Column 2 shows the SEQ ID NO for the sequence of each IgV domain contained in the corresponding molecule in the stack. Column 3 shows the binding partners against which the indicated affinity-modified stacking domains from column 1 were selected.

Also shown is the binding activity, as measured by the mean fluorescence intensity (MFI) value for binding of each stack molecule, to cells engineered to express various counter-structural ligands and the ratio of MFI compared to the binding of the corresponding stack molecule containing unmodified IgV domains that do not include an amino acid substitution. (s), with the same counterstructural ligand expressed by the cells. The functional activity of the variant stack molecules to modulate T cell activity is also shown based on the calculated IFN-gamma levels in culture supernatants (pg/ml) obtained with the indicated variant stack molecule and the corresponding ligand coimmobilized with anti-CD3, as described in Example 6. The tables also display the ratio of IFNγ produced by each variant stack molecule compared to the corresponding unmodified stack molecule in the coimmobilization assay.

As shown, the results demonstrated that it is possible to create stack molecules containing at least one variant IgSF domains that have modified, increased binding affinity activity to at least one cognate counterstructure ligand compared to a corresponding stack molecule containing the same unmodified IgV domain (eg wild type). In some cases, a stacked molecule, either from one or a combination of both variant IgSF domains in the molecule, exhibits increased binding to more than one cognate counterstructure ligand. The results also showed that the order of IgV domains in stacked molecules may, in some cases, alter the extent of increased binding activity. In some cases, the activity of functional T cells was also altered when assessed in a targeted co-immobilization assay.

Table 11

Stacked variant IgV Fc fusion proteins comprising the IgV NKp30 domain and the ICOSL domain of IgV

Domain structure from N-terminus to C-terminus: domain 1/domain 2/Fc SEQ ID NO (ig domain) Counterstructure selected against Binding activity Analysis of co-immobilization with anti-POPs pg/ml IFN-gamma (ratio of IFN-gamma to original wild type) B7H6 MFI (ratio of MFI to original wild type) ICOS MFI (ratio of MFI to original wild type) CD28 MFI (ratio MFI to original wild type) Domain l:NKp30WT Domain 2: ICOSL WT 214 196 64538 (1.00) 26235 (1.00) 6337 (1.00) 235 (1.00) Domain 1: NKp30 L30V/A60V/S64P/S8 6G Domain 2: ICOSL N52S N57Y H94D L96F L98FQ100R 215 212 B7-H6 ICOSCD28 59684 (0.92) 12762 (0.49) 9775 (1.54) 214 (0.91) Domain 1: NKp30 L30V/A60V/S64P/S8 6G Domain 2: ICOSL N52D 215 199 B7-H6 ICOSCD28 65470 (1.01) 30272 (1.15) 9505 (1.50) 219 (0.93) Domain 1: NKp30 L30V/A60V/S64P/S8 6G Domain 2: ICOSL N52H N57YQ100P 215 201 B7-H6 ICOSCD28 38153 (0.59) 27903 (1.06) 11300 (1.78) 189 (0.80) Domain 1: ICOSL WT Domain 2: Nkp30 WT 196 214 117853 (1.0) 70320 (1.0) 7916 (1.0) 231 (1.0) Domain LICOSL N52D Domain 2: NKp30 L30V/A60V/S64P/S8 6G 199 215 ICOSCD28 B7-H6 100396 (0.85) 83912 (1.19) 20778 (2.62) 228 (0.98) Domain 1: ICOSL N52H/N57Y/Q100P Domain 2: NKp30 L30V/A60V/S64P/S8 6G 201 215 ICOSCD28 B7-H6 82792 (0.70) 68874 (0.98) 72269 (9.12) 561 (2.43)

- 114 044356

Table 12

Stacked variant IgV Fc fusion proteins comprising IgV CD80 domain and IgV ICOSL domain

Domain structure from N-terminus to C-terminus: domain 1/domain 2/Fc SEQ ID NO (ig domain) Counterstructure selected against Binding activity Analysis of co-immobilization with anti-POPs pg/ml IFN-gamma (ratio of IFN-gamma to original wild type) CD28 MFI (ratio of MFI to original wild type) PD-L1 MFI (ratio of MFI to original wild type) ICOS MFI (ratio of MFI to original wild type) Domain 1: CD80WT Domain 2: ICOSL WT 152 196 1230 (1.00) 2657 (1.00) 11122 (1.00) 69 (1.00) Domain 1: CD80 E88D/K89R/D90K/A 91G/F92Y/K93R Domain 2: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S 189 213 PD-L1 ICOS/CD 28 3383 (2.75) 4515 (1.70) 5158 (0.46) 90 (1.30) Domain 1: CD80 A12T/H18L/M43V/F5 9L/E77K/P109S/I118 T Domain 2: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 193 213 PD-L1 ICOS/CD 28 2230 (1.81) 2148 (0.81) 3860 (0.35) 112 (1.62)

- 115 044356

E/F120S Domain 1: CD80 A12T/H18L/M43V/F5 9L/E77K/P109S/I118 T Domain 2: ICOSL N52D 193 199 PD-L1 ICOS/CD 28 5665 (4.61) 6446 (2.43) 15730 (1.41) 126 (1.83) Domain 1: CD80 E88D/K89R/D90K/A 91G/F92Y/K93R Domain 2: ICOSL N52H/N57Y/Q100P 189 201 PD-L1 ICOS/CD 28 6260 (5.09) 4543 (1.71) 11995 (1.08) 269 (3.90) Domain 1: CD80 A12T/H18L/M43V/F5 9L/E77K/P109S/I118 T Domain 2: ICOSL N52H/N57Y/Q100P 193 201 PD-L1 ICOS/CD 28 3359 (2.73) 3874 (1.46) 8541 (0.77) 97 (1.41) Domain 1: ICOSL WT Domain 2: CD80 WT 196 152 3000 (1.00) 2966 (1.00) 14366 (1.00) 101 (1.00) Domain 1: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S Domain 2: CD80 E88D/K89R/D90K/A 91G/F92Y/K93R 213 189 ICOS/CD 28 PD-L1 3634 (1.21) 4893 (1.65) 6403 (0.45) 123 (1.22) Domain 1: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S Domain 2: CD80 A12T/H18L/M43V/F5 9L/E77K/P109S/I118T 213 193 ICOS/CD 28 PD-L1 1095 (0.37) 5929 (2.0) 7923 (0.55) 127 (1.26) Domain 1: ICOSL N52D Domain 2: CD80 E88D/K89R/D90K/A 91G/F92Y/K93R 199 189 ICOS/CD 28 PD-L1 2023 (0.67) 5093 (1.72) 16987 (1.18) 125 (1.24) Domain 1: ICOSL N52D Domain 2: CD80 A12T/H18L/M43V/F5 9L/E77K/P109S/I118 T 199 193 ICOS/CD 28 PD-L1 3441 (1.15) 3414 (1.15) 20889 (1.45) 165 (1.63) Domain 1: ICOSL N52H/N57Y/Q100P Domain 2: CD80 E88D/K89R/D90K/A 91G/F92Y/K93R 201 189 ICOS/CD 28 PD-L1 7835 (2.61) 6634 (2.24) 20779 (1.45) 95 (0.94) Domain 1: ICOSL N52H/N57Y/Q100P Domain 2: CD80 A12T/H18L/M43V/F5 9L/E77K/P109S/I118 T 201 193 ICOS/CD 28 PD-L1 8472 (2.82) 3789 (1.28) 13974 (0.97) 106 (1.05)

- 116044356

Table 13

Stacked variant IgV Fc fusion proteins comprising an IgV CD80 domain or CD86 domain and an IgV ICOSL domain

Domain structure from N-terminus to C-terminus: domain 1/domain 2/Fc SEQ ID NO (domain Ig) Counterstructure selected against Binding activity Functional Activity MLR IFN-gamma pg/ml PD-L1 MFI (ratio of MFI to original wild type) CTLA-4 MFI (ratio of MFI to original wild type) Domain 1: CD80WT Domain 2: ICOSL WT 152 196 1230 (1.00) 11122 (1.00) 1756 (1.00) Domain 1: CD86WT Domain 2: ICOSL WT 220 196 29343 (1.00) 55193 (1.00) 6305 (1.00) Domain 1: CD80 R29H/Y31H/T41G/Y 87N/E88G/K89E/D90 N/A91G/P109S Domain 2: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S 192 213 CD28 ICOS/CD 28 2280 (1.85) 3181 (0.29) 2281 (1.30) Domain 1: CD80 I67T/L70Q/A91G/T12 0S Domain 2: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S 175 213 CD28 ICOS/CD 28 2309 (1.88) 26982 (2.43) 1561 (0.89) Domain 1: CD80 R29H/Y31H/T41G/Y 87N/E88G/K89E/D90 N/A91G/P109S Domain 2: ICOSL N52D 192 199 CD28 ICOS/CD 28 4285 (3.48) 22744 (2.04) 1612 (0.92) Domain 1: CD80 I67T/L70Q/A91G/T12 0S Domain 2: ICOSL N52D 175 199 CD28 ICOS/CD 28 3024 (2.46) 16916 (1.52) 3857 (2.20) Domain 1: CD80 R29H/Y31H/T41G/Y 87N/E88G/K89E/D90 N/A91G/P109S Domain 2: ICOSL N52H/N57Y/Q100P 192 201 CD28 ICOS/CD 28 6503 (5.29) 7240 (0.65) 6886 (3.92)

- 117 044356

Domain 1: CD80 I67T/L70Q/A91G/T12 0S Domain 2: ICOSL N52H/N57Y/Q100P 175 201 CD28 ICOS/CD 28 3110 (2.53) 4848 (0.44) 3393 (1.93) Domain 1: CD86 Q35H/H90L/Q102H Domain 2: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S 221 213 CD28 ICOS/CD 28 11662 (0.40) 21165 (0.38) 880 (0.14) Domain 1: CD86 Q35H/H90L/Q102H Domain 2: ICOSL N52D 221 199 CD28 ICOS/CD 28 24230 (0.83) 73287 (1.33) 1110 (0.18) Domain 1: CD86 Q35H/H90L/Q102H Domain 2: ICOSL N52H/N57Y/Q100P 221 201 CD28 ICOS/CD 28 1962 (0.07) 1630 (0.03) 587 (0.09) Domain 1: ICOSL WT Domain 2: CD80 WT 196 152 3000 (1.00) 14366 (1.00) 4113 (1.00) Domain 1: ICOSL WT Domain 2: CD86 WT 196 220 18005 (1.00) 53602 (1.00) 18393 (1.00) Domain 1: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S Domain 2: CD80 R29H/Y31H/T41G/Y 87N/E88G/K89E/D90 N/A91G/P109S 213 192 ICOS/CD 28 CD28 10426 (3.48) 51286 (3.57) 18680 (4.54) Domain 1: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S Domain 2: CD80 I67T/L70Q/A91G/T12 OS 213 175 ICOS/CD 28 CD28 17751 (5.92) 29790 (2.07) 10637 (2.59) Domain 1: ICOSL N52D Domain 2: CD80 R29H/Y31H/T41G/Y 87N/E88G/K89E/D90 N/A91G/P109S 199 192 ICOS/CD 28 CD28 2788 (0.93) 25870 (1.80) 6205 (1.51) Domain 1: ICOSL N52D Domain 2: CD80 I67T/L70Q/A91G/T12 0S 199 175 ICOS/CD 28 CD28 2522 (0.84) 13569 (0.94) 5447 (1.32) Domain 1: ICOSL N52H/N57Y/Q100P Domain 2: CD80 201 192 ICOS/CD 28 9701 (3.23) 9187 (0.64) 5690 (1.38) R29H/Y31H/T41G/Y 87N/E88G/K89E/D90 N/A91G/P109S CD28 Domain 1: ICOSL N52S/N57Y/H94D/L9 6F/L98F/Q100R/G103 E/F120S Domain 2: CD86 Q35H/H90L/Q102H 213 221 ICOS/CD 28 CD28 27050 (1.50) 21257 (0.40) 8131 (0.44) Domain 1: ICOSL N52D Domain 2: CD86 Q35H/H90L/Q102H 199 221 ICOS/CD 28 CD28 34803 (1.93) 80210 (1.50) 6747 (0.37) Domain 1: ICOSL N52H/N57Y/Q100P Domain 2: CD86 Q35H/H90L/Q102H 201 221 ICOS/CD 28 CD28 5948 (0.33) 4268 (0.08) 26219 (1.43)

Example 9. Preparation and evaluation of engineered cells expressing a transmembrane immunomodulatory protein.

T cells were engineered in which transmembrane immunomodulatory protein (TIP),

- 118 044356 including an extracellular domain (ECD) containing either a CD80 variant as described above or an affinity-modified IgSF domain from ICOSL were co-expressed with a chimeric antigen receptor (CAR). TIP also contained a transmembrane domain and a cytoplasmic domain corresponding to the wild-type CD80 or ICOSL transmembrane protein sequence. The immunomodulatory activity of the engineered cells was compared with cells that expressed CAR alone or cells that coexpressed the corresponding transmembrane protein CD80 or wild-type ICOSL with CAR.

An exemplary CD80-TIP is a variant CD80 having an affinity modified IgSF domain containing amino acid mutations in the IgV and IgC domains corresponding to I67T/L70Q/A91G/T120S, relative to positions in the CD80 extracellular domain shown in SEQ ID NO: 28 and transmembrane and a cytoplasmic domain corresponding to residues 243-288 in SEQ ID NO: 1. The amino acid sequence of the exemplary CD80-TIP is set forth in SEQ ID NO: 241 and is encoded by the nucleotide sequence shown in SEQ ID NO: 242. The corresponding wild-type CD80 transmembrane protein had the amino acid sequence shown as amino acid residues 35-288 in SEQ ID NO: 1 and encoded by the amino acid sequence shown in SEQ ID NO: 251.

An exemplary ICOSL-TIP is a variant ICOSL having an affinity modified IgSF domain containing amino acid mutations in the IgV domain corresponding to N52H/I143T, relative to positions in the extracellular domain of ICOSL specified in SEQ ID NO: 32, and in the transmembrane and cytoplasmic domain, corresponding to residues 257-302 in SEQ ID NO: 5. The amino acid sequence of the exemplary ICOSL-TIP is shown in SEQ ID NO: 243 and is encoded by the nucleotide sequence shown in SEQ ID NO: 244. The corresponding wild-type ICOSL transmembrane protein had the amino acid sequence designated as amino acid residues 19-302 in SEQ ID NO: 5 and is encoded by the amino acid sequence shown in SEQ ID NO: 252.

TIP containing an affinity-modified domain or a wild-type transmembrane protein containing the corresponding unmodified IgSF domain was co-expressed in T cells with a 1st generation chimeric antigen receptor (CAR) containing the CD3 zeta intracellular signaling domain. The 1st generation CAR included ScFv specific for CD19 (SEQ ID NO: 245), a hinge and transmembrane domain derived from CD8 (SEQ ID NO: 246), and an intracellular signaling domain derived from CD3 zeta (from SEQ ID NO: 47). The nucleotide sequence encoding the CD19 scFv-CD3 zeta CAR is presented in SEQ ID NO: 248, and the amino acid sequence of the CAR19 scFv-CD3 zeta CD19 is presented in SEQ ID NO: 479.

Nucleotide molecules were prepared encoding the CAR alone, or also encoding one of the exemplary TIP or wild-type transmembrane proteins separated from the CAR by the self-cleaving sequence T2A (SEQ ID NO: 250 and encoded by the nucleotide sequence presented in SEQ ID NO: 249). Examples of constructs contain the nucleotide sequences given in table. 14. Also, as a control, a nucleotide construct encoding a 2nd generation CAR was created, additionally containing the costimulatory domain of CD28 (CD19 scFvCD28-CD3 zeta).

Table 14

Nucleotide constructs

CAR (SEQ ID NO) Linker T2A (SEQ ID NO) TIP (SEQ ID NO) CD19 scFv - CD3 zeta + (248) - - CD 19 scFv - CD3 zeta - Т2А-В7-1 + (248) + (249) CD80 wild type (251) CD 19 scFv - CD3 zeta - T2A-B7-1TIP + (248) + (249) CD80 TIP (242) CD 19 scFv - CD3 zeta T2A - ICOSL + (248) + (249) Wild type ICOSL (252) CD 19 scFv - CD3 zeta - T2A-ICOSL TIP + (248) + (249) ICOSL TIP (244)

The nucleotide molecules were individually cloned into a lentiviral vector, which was used to transduce T cells isolated from human PBMC samples obtained from three different healthy donors. Lentiviral particles containing nucleotide sequences were obtained after co-transfection of HEK293 cells with vectors and lentiviral packaging constructs. Lentiviral particles were collected from the culture medium by ultracentrifugation and titrated by qRT-PCR. Human peripheral blood mononuclear cells (PBMCs) were isolated from three normal blood donors using density gradient sedimentation. PBMC were cultured overnight with anti-CD3 and anti-CD28 antibodies and IL-2, then transduced with lentivirus preparations at a multiplicity of infection of 5:1. Lentiviral vectors encoding the 2nd generation control CAR were used only to transduce cells from a single donor.

After two weeks (14 days) of culture, cells were assayed for cytotoxicity after co-culture with target antigen-expressing cells using an Acea Real-Time Cell Analyzer (RTCA), which measures impedance variations in the culture medium of a 96-well microelectronic plate (E-plate). plate) and shows changes in cell number and cell morphology in a real-time graph. CD19-expressing HeLa target cells (HeLa-CD19) were seeded in a 96-well E plate, and the impedance of each monolayer was monitored for 24 h using an RTCA system. The engineered T cells were added to the wells at an effector-to-target ratio of 10:1, and the wells were monitored for an additional 48 hours. The results were displayed and recorded as a Cell Index value (CI) derived from the change in measured electrical impedance and then converted by dividing the CI readings of all wells at all time points by the CI readings of individual wells at the same time (baseline time point) to obtain a normalized cell index value representing the percentage of the value at the baseline time point (see Zhang et al. Introduction to the Data Analysis of the Roche xCELLigence®System with RTCA Package. Bioconductor. May, 3, 2016, bioconductor.org/packages/devel/bioc/vignettes/RTCA/inst/doc/aboutRTCA.pdf. Accessed September 9, 2016) . In this assay, the decrease in monolayer impedance reflects the killing of target cells by the transduced cells.

The results showed that a decrease in impedance was observed in cells expressing 1st generation CAR compared to non-transduced T cells, although the extent of the decrease in impedance for cells expressing 1st generation CAR was less than for cells expressing 2nd generation CAR. The decrease in impedance in cells expressing 1st generation CARs continued until 8 h of analysis, whereas the impedance of 2nd generation CAR-expressing cells continued to decrease thereafter.

As shown in FIG. 2, in a single donor, each of the cells co-expressing TIP or the corresponding wild-type transmembrane protein with the 1st generation CAR exhibited a greater decrease in impedance, indicating greater cytotoxic activity, compared to cells expressing only the 1st generation CAR generations. In addition, the results showed that the cytotoxic activity was greater in CAR-expressing cells that co-expressed CD80TIP or ICOSL-TIP relative to CAR-expressing cells that co-expressed the corresponding wild-type CD80 or ICOSL transmembrane proteins containing an unaffinity-modified IgSF domain. The observed results of these engineered TIP cells showed that cytotoxic activity in cells coexpressing CD80-TIP or ICOSL-TIP with CAR exhibited enhanced activity to modulate the cytotoxic immune response of antigen-specific T cells, such as CAR-expressing T cells.

In the other two donors, cells expressing CD80-TIP did not result in much lower impedance compared with cells expressing the corresponding wild-type CD80 transmembrane protein. One donor did not have enough cells to be transduced with the wild-type transmembrane protein construct, although in this donor ICOS-L TIP provided better cytotoxicity compared with other constructs tested. In another donor, cells expressing ICOS-L-TIP did not result in more reduced impedance compared with cells expressing the corresponding wild-type ICOS-L transmembrane protein. In the cells tested, all cells coexpressing either CD80-TIP, ICOSL-TIP, or the corresponding wild-type transmembrane protein with the CAR exhibited greater cytotoxic activity than cells expressing only the first generation CAR. Differences in results observed among donors may be due to differences in T cells among donors, differences in the expression levels of various engineered proteins on the cell surface, the specific conditions used in this example assay to assess cell cytolysis (e.g., when assessing transformed cells 14th day, when assessing the single effector:target cell ratio) or other factors.

Example 10. Assessment of binding and activity of ICOSL IgSF domain variants.

Additional ECD ICOSL variants were identified using a yeast selection method essentially as described above and used to generate ECD-Fc fusion proteins as described in Example 5. Binding studies were conducted to evaluate the specificity and affinity of immunomodulatory proteins with the ICOSL domain for cognate binding partners essentially as described in Example 6.

A. Binding and functional characterization.

Binding was assessed to cells expressing full-length cognate binding partners CD28, ICOS and CTLA-4 essentially as described in Example 6. The bioactivity of ECD ICOSL variants was also assessed in a co-immobilization assay with anti-CD3 or human mixed lymphocyte reaction (MLR) essentially , as described in Example 6, except that for the co-immobilization assay, co-stimulatory activity was determined by culture of human T cells with a mixture of 10 nM

- 120 044356 plate-bound anti-CD3 antibodies and 40 nM ICOSL Fc.

In table 15 shows exemplary results for additional IgSF ICOSL domain variants for binding to cell-expressed counterstructures and bioactivity in an anti-CD3 coimmobilization or MLR assay. Approximate amino acid substitutions shown in table. 15 are indicated by the amino acid position number corresponding to the analogous reference unmodified ICOSL ECD sequence shown in SEQ ID NO: 32. The amino acid position is indicated in the middle with the corresponding unmodified (e.g., wild type) amino acid listed before the number and the identified amino acid substitution listed after the number . Column 2 lists the ECD variant SEQ ID NO for each variant ECD-Fc fusion protein molecule.

The results presented in table. 15 show binding activity measured by the mean fluorescence intensity (MFI) value for binding of each variant Fc fusion protein molecule to cells engineered to express a cognate counterstructure ligand and the ratio of MFI compared to binding of the corresponding unmodified ECD-Fc fusion protein molecule. containing no amino acid substitutions, with the same counterstructural ligand expressed by the cells. The functional activity of the variant Fc fusion molecules to modulate T cell activity is also shown based on the calculated levels of IFN-gamma in culture supernatants (pg/ml) produced with either i) the indicated variant ECD-Fc fusion protein molecule coimmobilized with anti- CD3 or ii) with the specified variant ECD-Fc fusion protein molecule in an MLR assay. The Table also shows the IFN-gamma ratio produced by each ECD-Fc variant compared to the corresponding unmodified (parent) ECDFc in both functional assays.

The results indicate altered, including increased, binding affinity of ICOSL IgSF domain variants with modified affinity to at least one cognate counterstructural ligand, and/or improved immunological activity. In particular, similar to the original variants obtained in Example 6, the selection resulted in the identification of a number of additional affinity modified ICOSL IgSF domain variants that exhibit increased binding to at least one, and in some cases more than one cognate counterstructural ligand. In addition, the results showed that affinity modification of the variant molecules also exhibited improved activities for both increasing and decreasing immunological activity depending on the molecule format, as described in Example 6.

Table 15

ICOSL variants: binding data and costimulatory activity data

ICOSL mutation(s) SEQ ID NO (ECD) ICOS tfxn MFI (original ratio) CD28 tfxn MFI (original ratio) CTLA-4 tfxn MFI (original ratio) IFNgamma AntiPOP co-immobilization assay and pg/ml (initial ratio) MLR IFNgamma pg/ml (initial ratio) N52H, F78L, Q100R, C198R 373 9568 (0.12) 1966 (0.24) 1454 (0.12) 130(0.31) 5927 (1.84) N52H, N57Y, Q100R, V110D, C198R, S212G 364 9418 (1.16) 136665 (16.55) 115352 (9.59) 944 (2.21) 821 (0.25) N52H, N57Y, R75Q, Q100P, V110D 374 5558 (0.07) 7465 (0.90) 4689 (0.39) 122 (0.28) 1136 (0.35) N52H, N57Y, Q100R, C198R 365 9148 (1.13) 134923 (16.33) 83241 (6.92) 1060 (2.48) 375 (0.12) N52H, N57Y, L74Q, 375 9448 128342 123510 1137 889 (0.28)

- 121 044356

V11OD, S192G (1.17) (15.54) (10.26) (2.66) N52H,Q100R 285 9478 (1.17) 151977 (18.40) 133929 (11.13) 972 (2.28) 794 (0.25) N52H, S121G, C198R 376 9128 (1.13) 124732 (15.10) 182607 (15.18) 827 (1.94) 1257 (0.39) A20V, N52H, N57Y, Q100R, S109G 287 5828 (0.72) 76973 (9.32) 73640 (6.12) 447 (1.05) 2283 (0.71) N52H, N57Y, QI OOP, C198R 461 9548 (1.18) 130676 (15.82) 81966 (6.81) 1125 (2.64) 643 (0.20) N52H, N57Y, R61S, Q100R, VI10D, L173S 289 1018 (0.13) 9129 (1.11) 5790 (0.48) 109 (0.25) 5094 (1.58) N52H, N57Y, Q100R, V122A 290 9978 (1.23) 137372 (16.63) 70764 (5.88) 1316 (3.08) 473 (0.15) N52H, N57Y, Q100R, F172S 291 1028 (1.27) 135821 (16.44) 73320 (6.09) 1561 (3.66) 486 (0.15) N52H, N57Y, Q100R 283 9858 (1.22) 140612 (17.02) 75106 (6.24) 1648 (3.86) 778 (0.24) N52S, F120S, N227K 377 9438 (1.17) 67796 (8.21) 82370 (6.85) 1157 (2.71) 1626 (0.50) N52S, N194D 366 9798 (1.21) 59431 (7.19) 74502 (6.19) 1671 (3.91) 1690 (0.52) N52S, V97A 294 3138 (0.04) 1733 (0.21) 1541 (0.13) 84 (0.20) 3858 (1.20) N52S, F120S 293 9068 (1.12) 67233 (8.14) 97880 (8.13) 1178 (2.76) 2814 (0.87) N52S, G72R 295 9288 (1.15) 51638 (6.25) 62339 (5.18) 1161 (2.72) 2947 (0.91) N52S, A71T, A117T, T190A, C198R 378 8918 (1.10) 44044 (5.33) 56646 (4.71) 1076 (2.52) 4031 (1.25) N52S, E220G 297 3878 (0.05) 2047 (0.25) 1796 (0.15) 122 (0.29) 1927 (0.60) Y47H, N52S, V107A, F120S 298 3268 (0.04) 2562 (0.31) 2104 (0.17) 334 (0.78) 4390 (1.36) WT ICOSL 32 8088 (1.00) 8260 (1.00) 12033 (1.00) 427 (1.00) 3226 (1.00) T43A, N52H, N57Y, L74Q, D89G, V110D, F172S 379 2821 (0.02) 2180 (0.49) 2051 (0.12) 184 (0.75) N52H, N57Y, Q100R, 381 174586 122383 76202 985 (4.01) 1037

- 122 044356

VI071, V110D, S132F, I154F, C198R, R221G (0.97) (27.24) (4.31) (0.36) E16V, N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, C198R 300 190765 (1.05) 129070 (28.73) 68488 (3.87) 4288 (17.46) 1225 (0.43) Q37R, N52H, N57Y, Q100R, VI ION, S142F, C198R, D217V, R221G 301 148638 (0.82) 91104 (20.28) 13498 (0.76) 62 (0.25) 7643 (2.68) N52H, N57Y, Q100R, V110D, C198R 302 179194 (0.99) 123312 (27.45) 84136 (4.76) 762 (3.10) 1342 (0.47) N52H, N57Y, Q100R, V110D, V116A, L161M, F172S, S192G, C198R 303 5236 (0.03) 4160 (0.93) 3305 (0.19) 49 (0.20) 2039 (0.72) F27S, N52H, N57Y, VI ION 304 20154 (0.11) 8613 (1.92) 3903 (0.22) 83 (0.34) 7522 (2.64) F27S, N52H, N57Y, VI ION 304 5236 (0.03) 4160 (0.93) 2957 (0.17) 40 (0.16) - N52S, H94E, L96I, S109N, L166Q, 305 198604 (1.10) 100361 (22.34) 102892 (5.82) 1253 (5.10) 5645 (1.98) S18R, N52S, F93L, I143V, R221G 306 154561 (0.85) 7625 (1.70) 4254 (0.24) 203 (0.83) 5239 (1.84) A20T, N52D, Y146C, Q164L 307 149661 (0.83) 9073 (2.02) 6901 (0.39) 287 (1.17) 4829 (1.69) VI IE, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 180016 (1.00) 120230 (26.76) 62809 (3.55) 2218 (9.03) 7283 (2.56) N52S, H94E, L96I, V122M 309 198717 (1.10) 88901 (19.79) 94231 (5.33) 590 (2.40) 618 (0.22) N52H, N57Y, H94E, L96I, Fl 201, S126T, W153R, I218N 310 87711 (0.48) 42035 (9.36) 31798 (1.80) 67 (0.27) 2500 (0.88) M10V, S18R, N30D, N52S, S126R, T139S, L203F 311 180665 (1.00) 64929 (14.45) 48362 (2.73) 1193 (4.86) 13647 (4.79) S25G, N30D, N52S, F120S, N227K 312 178834 (0.99) 66127 (14.72) 46631 (2.64) 1246 (5.07) 2202 (0.77) N30D, N52S, L67P, Q100K, D217G, R221K, T225S 313 18630 (0.10) 1986 (0.44) 1940 (0.11) 54 (0.22) 2752 (0.97)

- 123 044356

WT ICOSL 32 180900 (1.00) 4493 (1.00) 17685 (1.00) 246 (1.00) 2850 (1.00) N52H, N57Y, Q100R, V110D, A117T, T190S, C198R 314 2831 (0.04) 2881 (0.57) 2464 (0.23) 59 (0.08) - N52H, N57Y, Q100R, V110D, F172S, C198R 315 58478 (0.79) 74031 (14.75) 56850 (5.33) 712 (0.96) 1093 (0.23) S25G, F27C, N52H, N57Y, Q100R, V110D, E135K, L173S, C198R 316 22514 (0.30) 21320 (4.25) 20450 (1.92) 353 (0.48) 5765 (1.21) N52H, N57Y, V110A, C198R, R221I 317 84236 (1.14) 81842 (16.31) 121519 (11.39) 4593 (6.18) 1137 (0.24) M10I, S13G, N52H, N57Y, D77G, V110A, Н129Р, I143V, F172S, V193M, C198R 318 6362 (0.09) 6001 (1.20) 4834 (0.45) 141 (0.19) 4326 (0.91) N52H, N57Y, R61C, Y62F, Q100R, VI ION, F120S, C198R 319 4355 (0.06) 4316 (0.86) 3430 (0.32) 110(0.15) 6854 (1.44) N52H, N57Y, Q100R, L102R, V110D, H115R, C198R 367 96736 (1.31) 77881 (15.52) 148012 (13.88) 8765 (11.79) 630 (0.13) N52H, N57Y, Q100R, V110D, N144D, F172S, C198R 321 67578 (0.91) 64953 (12.94) 95731 (8.98) 1672 (2.52) 1490 (0.31) N52S, H94E, L98F, Q100R, 322 80690 (1.09) 78750 (15.69) 148160 (13.89) 3564 (4.80) 1497 (0.32) N52S, E90A 323 108908 (1.47) 31086 (6.19) 108866 (10.21) 4564 (6.14) 3927 (0.83) N30D, K42E, N52S 324 85726 (1.16) 4293 (0.86) 10755 (1.01) 5211 (7.01) 5656 (1.19) N52S, F120S, I143V, I224V 325 90862 (1.23) 28443 (5.67) 105229 (9.87) 4803 (6.46) 4357 (0.92) WT ICOSL 32 73964 (1.00) 5018 (1.00) 10665 (1.00) 743 (1.00) 4748 (1.00)

C. Cytokine production in anti-POP costimulation assays.

The exemplary variant ECD ICOSL Fc fusion molecules described above were further assessed for IL-17 cytokine stimulation in the anti-POP co-stimulatory bioactivity assay described above. A mixture of 10 nM plate-bound anti-POP and 40 nM ICOSL-Fc variant polypeptides were cultured with human T cells. Supernatants were collected and IL-17 levels were determined by ELISA. The amount of IL-17 in culture supernatants (pg/ml) produced with the indicated variant ECD-Fc fusion proteins and the corresponding unmodified (parental) ECD-Fc coimmobilized with anti-POP was measured. For comparison, also in this table, the results for the production of IFN-gamma in the same analysis as shown in the table are shown. 15 for approximate options.

The results are presented in table. 16, which shows the pg/ml IL-17 measured in the supernatant, as well as the ratio (fold increase) of IL-17 produced by each ECD-Fc variant compared to the corresponding unmodified ECD-Fc (wild type). Similar results are shown for IFN-gamma. The percentage of total IL-17 or IFN-gamma cytokines produced by the cells is also shown. The results showed that in a co-stimulation assay, affinity modification of variant molecules demonstrated altered T cell functional activity to increase IL-17 in addition to IFN-gamma.

- 124044356

Table 16

Costimulatory bioactivity data for ICOSL IgSF domain variants

ICOSL mutation(s) SEQ ID NO (ECD) IL-17A [pg/ml] IL-17A multiplicity μντ IFN-g [pg/ml] IFN-g multiplicity μντ Total multiplicity ΐ WT % of total cytokine produced % General IL-17+ IFN-g % IL-17 % IFN-g N52H, N57Y, Q100R, C198R 365 617 7.93 1060 2.48 10.42 5.51 0.77 6.28 N52H, N57Y, Q100R, V122A 290 647 8.33 1316 3.08 11.41 5.79 0.96 6.75 N52H, N57Y, Q100R, F172S 291 549 7.06 1561 3.66 10.72 4.91 1.14 6.05 N52Y, N57Y, F138L, L203P 112 90 1.05 1999 2.69 3.74 0.81 2.91 3.72 Vl IE, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 319 3.16 2218 9.03 12.19 2.85 3.23 6.08 N52H, N57Y, Q100R, L102R, V110D, H115R, C198R 367 510 5.90 8765 11.79 17.70 4.56 12.78 17.33 N52H, N57Y, Q100R 283 473 6.08 1648 3.86 9.94 4.23 1.20 5.43 N52H,Q100R 285 358 4.60 972 7.01 11.62 3.20 0.71 3.91 N52H, N57Y, Q100R, V110D, C198R, S212G 364 124 1.60 944 2.21 3.81 1.11 0.69 1.80 N52H, N57Y, Q100P 113 127 1.47 4922 6.62 8.09 1.14 7.17 8.31 E16V, N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, C198R 300 22 7.11 130 17.46 24.57 6.41 6.25 12.66 N30D, K42E, N52S 324 349 4.04 5211 7.01 11.05 3.12 7.60 10 .71 N52S, F120S, I143V, I224V 325 292 3.39 4803 6.46 9.85 2.61 7.00 9.62 N52S, E90A 323 306 3.54 4564 6.14 9.68 2.73 6.65 9.39 N52H, N57Y, V110A, C198R, R221I 317 290 3.35 4593 6.18 9.53 2.59 6.69 9.28 N52S, N194D 366 428 5.50 1671 3.90 9.4 1.52 5.19 5.40 N52H, Π43Τ 135 84 - 1727 - 3.30 0.75 2.52 3.27 N52D 111 126 - 1447 - 3.41 1.13 2.11 3.23

Example 11: Generation of an additional engineered T cell expressing a transmembrane immunomodulatory protein and assessment of proliferation.

This example describes the generation of additional engineered T cells in which a transmembrane immunomodulatory protein (TIP) containing an extracellular domain (ECD) including an affinity-modified ICOSL domain of IgSF was co-expressed with a chimeric antigen receptor (CAR). Specifically, a TIP was generated to include the ECD of an exemplary variant ICOSL containing the amino acid mutations N52D, N52H/N57Y/Q100P,

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R or N52H/N57Y/Q100R relative to positions in the ICOSL extracellular domain shown in SEQ ID NO: 32. TIP also contained a transmembrane domain and a cytoplasmic domain of the corresponding transmembrane protein sequence Wild type ICOSL corresponding to residues 257-302 of SEQ ID NO: 5. The sequence of TIP with and without signal peptide is as follows: N52D (SEQ ID NO: 496 and 497); N52H/N57Y/Q100P (SEQ ID NO: 498 and 499);

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R (SEQ ID NO: 500 and 501) and N52H/N57Y/Q100R (SEQ ID NO: 502 and 503). For comparison, full-length wild-type transmembrane ICOSL (amino acid residues 19-302 SEQ ID NO: 5) was also expressed in cells. The sequence of wild-type TIP with and without signal peptide is set forth in SEQ ID NOs: 494 and 495. The nucleic acid encoding TIP also included a sequence encoding green fluorescent protein (GFP) separated from TIP by a T2A self-cleaving sequence.

TIP containing an affinity-modified domain or a wild-type transmembrane protein containing the corresponding unmodified IgSF domain was co-expressed in chimeric antigen receptor (CAR) T cells. The nucleotide sequence encoding the CAR encodes, in order: CD8 signal sequence (SEQ ID NO: 481), anti-CD19 scFv (SEQ ID NO: 482), CD8-derived hinge/transmembrane region (SEQ ID NO: 483), a 4-1BB-derived co-stimulatory signaling domain (SEQ ID NO: 484) and a CD3zeta signaling domain (SEQ ID NO: 247). The resulting anti-CD19 CAR has the amino acid sequence shown in SEQ ID NO: 490. The nucleic acid encoding the CAR also included

- 125 044356 sequence encoding blue fluorescent protein (BFP, SEQ ID NO: 489), separated from

CAR using the T2A self-cleaving sequence (set forth in SEQ ID NO: 488).

Separately, viral vector constructs were produced in which either a nucleic acid molecule encoding the CAR alone or a nucleotide molecule encoding one of the exemplified wild-type TIPs or ICOSLs were cloned. A viral vector encoding CAR and a viral vector encoding wild-type TIP or ICOSL were cotransformed into T cells. For transduction, primary T cells were activated with anti-CD3 and anti-CD28 beads (Dynal) at a bead ratio of 1:1 and incubated in the presence of 100 IU/ml IL-2 at 37°C for 2 days. T cells were then collected and transduced with 400 μl CAR viral supernatant and 400 μl TIP viral supernatant in the presence of 8 μg/ml polybrene. Cells were centrifugally inoculated at 1000 g for 30 min at 30°C. Cells were then transferred and incubated overnight at 37°C. After incubation, cells were harvested and viral supernatant was removed. Cells were resuspended with complete medium and 50 IU/ml IL-2. Cells were expanded and replenished with IL-2 and medium every two days for 6 days. The beads were removed from the cells using a magnet and counted before evaluation in the proliferation assay. An exemplary TIP and CAR expression profile in exemplary transduced T cells is shown in FIG. 2A.

To assess proliferation of CAR T cells and CAR-TIP T cells in response to antigen, cells were labeled using trace amounts of far-red dye. CD19-expressing Nalm6 target cells were titrated starting at a target:T cell ratio of 1.5:1 and a 1:2 dilution with an 8-point dilution. CAR T-labeled cells or CAR-TIP T cells were added to Nalm6 cells and the culture was incubated for 4 days before the cells were analyzed by flow cytometry. The supernatant was collected and further evaluated in a cytokine release assay.

As shown in FIG. 2B, CAR+ primary T cells proliferate in a dose-dependent manner into CD19+ NALM6 cells. Compared to CAR-only T cells, T cells coexpressing CAR and wild-type ICOSL, or one of the exemplary ICOSL TIPs, showed increased proliferation compared to CAR-only T cells. Co-expression of CAR and TIP containing either N52H/N57Y/Q100P variants

E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R, or the N52H/N57Y/Q100R ICOSLECD variant, showed greater proliferation than T cells coexpressing wild-type CAR and ICOSL, indicating that TIP, expressed on primary T cells provide an enhanced co-stimulatory signal to enhance T cell proliferation.

Example 12 Purification and Evaluation of Purified ICOSL Domain Variants of IgSF.

The purification strategy was used for the exemplary selected candidates described in Examples 6 and 10. Human cells derived from cell line 293 (Expi293) were transiently transfected with an expression construct and the cells expressed the ECD ICOSL Fc fusion protein molecule. Fc fusion proteins were then purified from protein A supernatants using affinity chromatography (MabSelect SuRe). This initial purification step was followed by size exclusion chromatography (SEC) to further purify the proteins (Superdex200 16x60). Samples from both purification steps were stored and compared using analytical SEC. Protein concentration was determined after purification of Protein A. The resulting purified proteins were also analyzed by high performance liquid chromatography (HPLC) analytical SEC to assess purity.

The percentage of the major peak in the purified samples was determined and compared to the protein purified initially in the Protein A step (% of the major peak of the prot. A pool) compared to the protein purified using Protein A followed by preparative SEC (% of the major peak of the SEC T pool =D0). As shown in table. 17, the additional SEC step significantly increases the protein purity of the purified proteins. To further evaluate protein stability, preparative SEC-purified proteins were left at room temperature for 24 h and then scored and compared by HPLC major peak (% SEC pool major peak T=D24) and compared to sample D0. The change in % of the main peak in D0 compared to D24 (A% of the main peak of the SEC pool) was determined. As shown in table. 17, most of the exemplary variant ECD ICOSL Fc fusion protein molecules tested showed little change in the % of the main peak, indicating minimal aggregation of the protein variants.

- 126 044356

Table 17

Purification of ICOSL protein variants

ICOSL mutation(s) SEQ ID NO (ECD) Expi293 cont. Prot. A mg/l % main peak pool Prot. A % of the main peak of the SEC pool T=D0 % of the main peak of the SEC pool T=D24 A% of the main peak of the pool SEC N52S, N194D 366 120 87.9 93.5 92 1.5 N52H, N57Y, Q100R, F172S 291 217 86.9 97.4 95.6 1.8 N52S, E90A 323 128 86.5 89.5 88.3 1.2 N52H,Q100R 285 176 85.9 97.5 96.1 1.4 N52H, N57Y, Q100R 283 186 85.1 97.6 95.7 1.9 N52S, F120S, I143V, I224V 325 87 83.2 88.9 88.3 0.6 N52H, N57Y, Q100R, C198R 365 204 82.9 95.8 92.3 3.5 N52H, N57Y, Q100P 113 63 80.5 94.5 88.5 6 N30D, K42E, N52S 324 81 80 95.4 91.3 4.1 N52H, N57Y, Q100R, L102R, V110D, H115R, C198R 367 141 78.9 96 92.9 3.1 N52H, N57Y, Q100R, V122A 290 260 77.6 96.4 95.2 1.2 N52Y/N57Y/F13 8L/L203P 112 40 75.6 96.8 94.8 2 E16V, N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, C198R 300 60 73.8 97.1 95.8 1.3 N52H, N57Y, V110A, C198R, R221I 317 95 65.4 90.9 86 4.9 N52H, N57Y, Q100R, V110D, C198R, S212G 364 73 50.6 87.9 78.6 9.3 VI IE, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 58 - - - - N52H, P43T 135 134 93.2 96 92.7 3.3 N52D 111 136 90.4 95.5 93.3 2.2

Example 13: Evaluation of co-stimulatory bioactivity of purified selected variants of ICOSL IgSF domains.

Exemplary ECD ICOSL Fc fusion protein molecules, purified as described in Example 12, were assessed for their bioactivity by MLR essentially as described in Example 6. A mixture of 10 nM or 40 nM variant ICOSL Fc proteins were coupled overnight in 96-well plates in the presence of 10 nM anti-CD3. The plates were washed and 100,000 CFSE-labeled pan T cells were added for 96 h. Supernatants were collected, and IFN-γ and IL-17 levels were measured by ELISA.

Results for cytokine secretion induced by anti-CD3 costimulation with exemplary variants tested (10 and 40 nM ICOSL Fc) are shown in FIG. 3A and 3B, which depict exemplary amino acid substitutions (substitutions) of IgSF domains in the ECD from ICOSL. The histograms in Fig. 3A and 3B show the amount of IFN-gamma and IL-17 secreted, respectively, by ELISA in the supernatants (pg/ml). The level of cytokine release induced by anti-CD3 costimulation with the tested variants compared to the level induced by anti-CD3 costimulation with WT ICOSL is indicated by the horizontal line. The results showed that affinity modification of the variant molecules was active in modulating the functional activity of T cells, including a significant increase in the secretion of IFN-gamma and IL-17 in a costimulation assay. Increased immunological activity has been observed with some variants.

Example 14. Assessment of proliferation of purified selected variants of the ICOSL domain of IgSF.

Exemplary ECD ICOSL Fc fusion protein variants purified as described in Example 12 were evaluated for their ability to co-stimulate anti-CD3-induced T cell proliferation. Primary T cells were labeled with carboxyfluorescein succinmidyl ester (CFSE). A mixture of 10 nM or 40 nM ECD variant ICOSL Fc protein or wild-type ICOSL protein was bound overnight to 96-well plates in the presence of 10 nM anti-CD3 and then T cells were plated and incubated for 3 days. As a control, proliferation was also assessed in the presence of bound antiCD3 and IgG or IgG alone. Cells were stained for CD4 or CD8 surface markers, and T cell, CD4+ T cell, or CD8+ T cell proliferation was determined by assessing CFSE dilution using flow cytometry.

The results are shown in Fig. 4A and 4B for exemplary embodiments tested at 40 nM and 10 nM ICOSL, respectively. As shown in FIG. 4A, almost all variant ECD ICOSL Fc fusion protein molecules tested induced proliferation above that of the WT control. As shown in FIG. 4B, differences in proliferation were more apparent at 10 nM with some variants providing maximum proliferation even at this lower concentration.

Example 15. Evaluation of binding and activity of purified selected domain variants of ICOSL IgSF.

- 127 044356

Exemplary variant ECD ICOSL Fc fusion protein molecules, purified as described in Example 12, were assessed for binding and functional activities using methods substantially as described in Example 6 or Example 10.

A. Flow cytometric binding assays.

Human cells derived from cell line 293 (Expi293) were transfected with CD28, CTLA-4, ICOS, or mock transfected. Cells were then incubated with wild-type ECD ICOSL Fc or ECD ICOSL-Fc fusion protein molecules that were titrated from 100,000 to 46 pM, and binding was observed using a PE-conjugated anti-human Fc antibody as described in Example 6. Binding was assessed by flow assay cytometry and mean fluorescence intensity (MFI), and the percentage (%) of cells positive for signal was determined using Cell Quest Pro software (Becton Dickinson, USA). The ICOSL-Fc concentration that produced the half-maximal MFI response (MFI EC50) or % positive cells (%(+) _EC50 ) was determined.

In table 18 presents the results. ICOSL amino acid substitutions shown in table. 18 are designated by the amino acid position number corresponding to the similar unmodified ICOSL ECD sequence shown in SEQ ID NO: 32. For some values (eg, WT binding to CD28), the EC ₅₀ cannot be obtained, so a value of 1,000,000 pM was arbitrarily chosen for data formatting purposes. Similar to the results obtained in the previous binding assays described in Example 10 above, altered binding affinity of the variant ICOSL ECD-Fc fusion protein molecule was observed for at least one cognate counterstructure ligand.

Table 18

EC50 by flow cytometry for ICOSL variants

ICOSL mutation(s) SEQ ID NO (ECD) CD28 MFI EC50 [nM] CD28% (+) EC50 [pm] CTLA-4 MFI EC50 [nM] CTLA-4% (+) EC50 [nM] ICOSMFI EC50 [nM] ICOS % (+) EC50 [nM] WT ICOSL 32 1000000 1000000 1000000 1000000 10543 762 N52H, I143T 135 19147 567 20259 1891 2666 286 N52H, N57Y, Q100R C198R 365 950 159 73548 422 1032 179 N52H, N57Y, Q100R VI22 A 290 29701 152 1008 293 302 64 N52H, N57Y, Q100R,F172S 291 1006 231 1332 396 779 130 N52Y/N57Y/F13 8L/L203P 112 7844 386 7457 994 3104 408 VI IE, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 5961 595 6909 1026 5514 852 N52H, N57Y, Q100R L102R, V110D, H115R C198R 367 1034 307 23328 579 3172 347 N52H, N57Y, Q100R 283 1665 238 11002 533 383 131 N52H,Q100R 285 1305 274 8593 1997 702 167 N52H, N57Y, Q100R V110D, C198R S212G 364 4987 594 30382 922 50219 814 N52H, N57Y, Q100P 113 21137 402 22651 758 4090 320 E16V, N52H, N57Y, Q100R V110D, H115R, Y152C, K156M, C198R 300 2508 387 5399 806 2381 421 N30D, K42E, N52S 324 - 3683800 8593 1997 3251 558 N52S, F120S, I143V, I224V 325 902400 9060 28126 2948 4366 245 N52S, E90A 323 1339700 31302 31419 5828 5225 473 N52H, N57Y, V110A, C198R R221I 317 1809 426 7201 841 1293 433 N52S, N194D 366 944669 11876 1254880 5170 473 206 N52D 111 288617 17793 396841 3891 2642 137

B. ForteBio binding assay.

Protein-protein interactions between receptors and ICOSL variant domain immunomodulatory proteins were further assessed using Fortebio binding assays. ICOS, CD28, and CTLA-4 were loaded individually onto anti-human capture sensors (ForteBio Octet AHC) and unmodified wild-type ICOSL ECD-Fc molecule, wild-type PD-L2 EDFc fusion protein molecule, or variant ICOSL Fc fusion protein molecules were bound to the receptors in 4-point titrations . Each titration was globally fitted to calculate the association (ko _n ) and dissociation (K _dis ) of each protein. The response of the anti-human uptake sensors of each receptor tested to a variant ICOSL ECD-Fc fusion protein molecule was determined. The dissociation constant (KD) was calculated and compared with the wild type to determine the fold improvement (fold improvement).

- 128 044356

The results of binding to ICOS are given in table. 19, with CD28 - in table. 20, and with CTLA-4 - in table. 21.

Approximate amino acid substitutions shown in table. 19-21, are designated by the number of the amino acid position corresponding to the same original unmodified sequence

ICOSL ECD shown in SEQ ID NO: 32.

Table 19

ForteBio ICOS Binding Assay

ICOSL mutation(s) SEQ ID NO (ECD) Answer KD(M) Kon (1/Mc) Kdi _S (l/c) full R ² Multiple improvement WT ICOSL 32 0.73 8.83E-10 l,78E+05 l,58E-04 0.9908 - N52H, I143T 135 0.87 3.32E-10 ZDZE+05 l,04E-04 0.9683 2.7 N52H, N57Y, Q100R, C198R 365 0.74 4.92E-10 3.85E+05 l,89E-04 0.9882 1.8 N52H, N57Y, Q100R V122A 290 0.67 4.72E-10 3.77E+05 l,78E-04 0.9775 1.9 N52H, N57Y, Q100R, F172S 291 0.68 4.20E-10 4.34E+05 l,82E-04 0.9545 2.1 N52Y/N57Y/F13 8L/L203P 112 0.64 7.69E-10 2.22E+05 l,71E-04 0.9782 1.1 VUE, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 0.67 3.62E-10 3.55E+05 l,29E-04 0.9687 2.4 N52H, N57Y, Q100R, L102R, V110D, H115R, C198R 367 0.76 4.77E-10 3.29E+05 l,57E-04 0.9616 1.9 N52H, N57Y, Q100R 283 0.74 3.69E-10 2.87E+05 l,06E-04 0.9817 2.4 N52H,Q100R 285 0.79 3.73E-10 4.45E+05 l,66E-04 0.968 2.4 N52H, N57Y, Q100R, V110D, C198R, S212G 364 0.60 l,29E-09 l,66E+05 2D5E-04 0.9846 0.7 N52H, N57Y, Q100P 113 0.73 3.82E-10 3.71 E+05 l,42E-04 0.9729 2.3 E16V, N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, C198R 300 0.75 5.43E-10 2.65E+05 l,44E-04 0.9848 1.6 N30D, K42E, N52S 324 0.80 3.71E-10 4.48E+05 l,66E-04 0.9651 2.4 N52S, F120S, I143V, I224V 325 0.80 3.llE-10 5.03E+05 l,56E-04 0.9673 2.8 N52S, E90A 323 0.88 3.40E-10 4.85E+05 l,65E-04 0.9792 2.6 N52H, N57Y, V110A, C198R, R221I 317 0.68 4.77E-10 ZD5E+05 l,50E-04 0.976 1.9 N52S, N194D 366 0.88 3.37E-10 3.38E+05 1D4E-04 0.9723 2.6 N52D 111 0.87 3.38E-10 3.91E+ 05 l,32E-04 0.9792 2.6 Wildtype PD-L2 ED-Fc - 0.03

Table 20

ForteBio CD28 Binding Assay

ICOSL mutation(s) SEQ ID NO (ECD) Answer KD(M) Kon (1/Mc) K _dis (l/c) full R ² Multiple improvement WT ICOSL 32 0.33 l,39E-08 6.69E+04 9.29E-04 0.9715 - N52H, I143T 135 0.95 5.25E-10 4.27E+05 2.24E-04 0.9877 26.5 N52H, N57Y, Q100R C198R 365 1.14 4.47E-10 4D2E+05 l,84E-04 0.9877 31.0 N52H, N57Y, Q100R, V122A 290 1.04 3.90E-10 4.07E+05 l,59E-04 0.9878 35.6 N52H, N57Y, Q100R, F172S 291 1.06 2.93E-10 4.26E+05 l,25E-04 0.9836 47.3 N52Y/N57Y/F13 8L/L203P 112 0.86 7.83E-10 l,79E+05 l,40E-04 0.993 17.7 V11E, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 0.92 5.53E-10 2.54E+05 l,40E-04 0.9906 25.1 N52H, N57Y, Q100R, L102R, V110D, H115R, C198R 367 1.10 3.66E-10 3.41 E+05 l,25E-04 0.986 37.9 N52H, N57Y, Q100R 283 1.04 3.68E-10 3.72E+05 l,37E-04 0.983 37.7 N52H,Q100R 285 1.09 4.01E-10 5.0SE+05 2.02E-04 0.9938 34.7 N52H, N57Y, Q100R, V110D, C198R, S212G 364 0.94 8.96E-10 l,78E+05 l,60E-04 0.9961 15.5 N52H, N57Y, Q100P 113 0.99 4.36E-10 3.29E+05 l,43E-04 0.9835 31.8 E16V, N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, C198R 300 1.06 5.03E-10 3.06E+05 l,54E-04 0.9872 27.6 N30D, K42E, N52S 324 0.54 l,95E-09 2.74E+05 5.33E-04 0.9772 7.1 N52S, F120S, I143V, I224V 325 0.84 9.10E-10 4.51 E+05 4.10E-04 0.9742 15.3 N52S, E90A 323 0.94 9.69E-10 4.74E+05 4.59E-04 0.978 14.3

- 129 044356

N52H, N57Y, V110A, C198R, R221I 317 0.94 5.63E-10 2.63E+05 1.48E-04 0.9781 24.7 N52S, N194D 366 0.82 1.04E-09 3.53E+05 3.68E-04 0.9887 13.3 N52D 111 0.86 1D6E-09 3.36E+05 3.90E-04 0.989 11.9 wildtype PD-L2 ED-Fc - -0.04

Table 21

ForteBio CTLA-4 Binding Assay

ICOSL mutation(s) SEQ ID NO (ECD) Answer KD(M) Kon (1/Mc) K _dis (l/c) Full R ² Multiple improvement WT ICOSL 32 0.21 7.71E-08 l,92E+04 l,48E-03 0.8919 - N52H, I143T 135 0.96 6.78E-10 7.26E+05 4.92E-04 0.9641 113.8 N52H, N57Y, Q100R, C198R 365 1.57 6.45E-10 4.79E+05 3.09E-04 0.9875 119.6 N52H, N57Y, Q100R VI22 A 290 1.43 5.76E-10 4.73E+05 2.72E-04 0.9926 133.9 N52H, N57Y, Q100R F172S 291 1.47 5.36E-10 5DZE+05 2.75E-04 0.9924 144.0 N52Y/N57Y/F13 8L/L203P 112 1.33 8.33E-10 3.45E+05 2.87E-04 0.9943 92.6 VI IE, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 1.50 6.48E-10 ZD2E+05 2.02E-04 0.9943 119.0 N52H, N57Y, Q100R, L102R, V110D, H115R, C198R 367 1.60 8.64E-10 4.79E+05 4.14E-04 0.9825 89.2 N52H, N57Y, Q100R 283 1.65 7.19E-10 4.28E+05 3.08E-04 0.9895 107.2 N52H,Q100R 285 1.17 5.92E-10 8.37E+05 4.96E-04 0.9629 130.3 N52H, N57Y, Q100R, V110D, C198R, S212G 364 1.32 l,47E-09 2.34E+05 3.44E-04 0.9937 52.6 N52H, N57Y, Q100P 113 1.51 6.47E-10 3.61 E+05 2.33E-04 0.9911 119.2 E16V, N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, C198R 300 1.58 l,06E-09 4.24E+05 4.49E-04 0.9779 72.8 N30D, K42E, N52S 324 0.42 2.81E-09 2.42E+05 6.81E-04 0.9676 27.4 N52S, F120S, I143V, I224V 325 0.58 l,20E-09 ZD0E+05 3.72E-04 0.9283 64.3 N52S, E90A 323 0.64 1D2E-09 3.28E+05 3.68E-04 0.9184 68.7 N52H, N57Y, V110A, C198R R221I 317 1.44 l,07E-09 4.0SE+05 4.32E-04 0.9811 72.3 N52S, N194D 366 0.59 2.52E-09 2.66E+05 6.69E-04 0.9643 30.6 N52D 111 0.62 l,52E-09 4D6E+05 6.32E-04 0.9234 50.7 wildtype PD-L2 ED-Fc - 0.00

D. Coimmobilization analysis.

The costimulatory bioactivity of the ICOSL fusion protein variants was determined in co-immobilization assays with anti-POPs essentially as described in Example 6. Approximately 0.37, 1.3 or 10 nM mouse anti-human CD3 antibody (OCT3, Biolegends, USA) was diluted in PBS using 10 or 40 nM variant ICOSL ECD Fc or wild-type ICOSL ECD-Fc. This mixture was added to treated 96-well flat-bottom tissue culture plates overnight to facilitate adhesion of stimulatory proteins to the wells of the plate. The next day, unbound protein was washed from the plates and 100,000 purified human pan T cells were added to each well. Cells were cultured for 3 days before collecting culture supernatants and measuring human IFN-gamma levels using an ELISA kit.

In table 22 shows the amount of IFN-gamma (pg/ml) produced by cells under various conditions in an anti-POP co-immobilization assay. The table shows the amino acid substitutions of exemplary variant ICOSL-Fc fusion proteins, which are designated by the amino acid position number corresponding to the unmodified ICOSL ECD sequence presented in SEQ ID NO: 32, and the corresponding SEQ ID NO for the variant ECD for each variant ICOSL fusion protein molecule ECD-Fc. Shown is the IFN-gamma ratio obtained in the presence of each ICOSL ECD-Fc variant in the functional assay compared to the presence of the corresponding unmodified ECD-Fc (wild type) (fold 7 WT). As shown, co-stimulatory signaling of variant ICOSL-ECD-Fc molecules was significantly higher compared to wild-type ICOSL.

- 130044356

Table 22

Evaluation of IFN-gamma responses in costimulation assays

ICOSL mutation(s) SEQ ID NO (ECD) IFN-gamma [pg/ml 40 nM ligand 10 nM ligand 40 nM ligand 10 mM OKT3 Multiple of TWT 10 mM OKT3 Multiple of TWT 10 mM OKT3 1.1 nM OKT3 0.37 nM OKT3 E16V, N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, C198R 300 14372 17.3 4903 29.9 8379.2 7422.8 2893.7 N52H, N57Y, Q100R, VI22 A 290 10640 12.8 6456 39.4 5636.2 4724.2 2246.3 N52H, N57Y, Q100R, F172S 291 10379 12.5 3741 22.8 3979.7 4067.7 1415.5 N52H, N57Y, Q100R, C198R 365 9590 11.5 4048 24.7 4215.8 2787.1 1072.4 N52H, N57Y, Q100R 283 9568 11.5 3270 19.9 4412.3 3862.0 1820.0 N52H, N57Y, Q100R, L102R, V110D, H115R, C198R 367 6939 8.4 3234 19.7 5495.2 4081.6 1442.8 N52S, F120S, I143V, I224V 325 6567 7.9 717 4.4 2145.4 2185.7 646.1 N52S, N194D 366 5690 6.8 272 1.7 2315.1 1485.0 1140.6 VI IE, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 5345 6.4 1152 7.0 2747.0 3383.4 1701.2 N52S, E90A 323 5097 6.1 706 4.3 5019.8 3036.4 1482.4 N52H, N57Y, V110A, C198R R221I 317 4737 5.7 520 3.2 2501.5 1632.1 937.5 N52H,Q100R 285 4122 5.0 1466 8.9 5782.1 2861.4 967.5 N30D, K42E, N52S 324 4080 4.9 273 1.7 1336.8 1260.7 541.1 N52H, N57Y, Q100P 113 3344 4.0 229 1.4 2525.4 2439.5 1233.9 N52H, N57Y, Q100R, V110D, C198R S212G 364 3064 3.7 1471 9.0 2699.5 2629.9 678.2 N52Y, N57Y, F138L, L203P 112 2177 2.6 200 1.2 1889.5 1757.9 808.8 N52H, I143T 135 1906 2.3 138 0.8 1417.1 1367.9 275.2 WT ICOSL 32 831 1.0 164 1.0 558.8 377.7 152.0 N52D 111 88 0.1 231 1.4 1288.9 1737.9 289.0

E. Mixed lymphocyte reaction to assess suppression of bioreactivity.

Modulation of T cell activity by fusion protein variants was determined in a mixed lymphocyte reaction (MLR) essentially as described in Example 6. Human monocytes were incubated for 6 days in the presence of IL-4 and GM-CSF and matured into dendritic cells upon addition of LPS During the last 24 hours, 1 x 10 ⁴ dendritic cells and 1 x 10 ⁵ CFSE-labeled human T cells were seeded per well and incubated for 4 days in the presence of three different concentrations (40, 13.3 or 4.4 nM) of wt. type or recombinant variant molecule ICOSL ECD-Fc, diluted in PBS. The same concentrations of human IgG, PD-L2-Fc, or belatacept (CTLA4-Fc containing the L104E and A29Y mutations) were used as controls. Supernatants were collected and IFNγ responses were characterized by ELISA.

In fig. 5 depicts the production of IFN-gamma under various conditions. The levels of IFN-gamma produced by cells in the presence of wild-type ICOSL are represented by the horizontal line. There was no suppression of IFN-γ production in the presence of the negative control protein PD-L2-Fc. In contrast, most ICOSL variants tested showed some degree of inhibition of IFN-γ production in the MLR. In some embodiments, significant inhibition of IFN-gamma was observed, with very low to undetectable IFN-gamma production obtained in cultures, even at the lowest concentration tested of 4.4 nM. The percentage of MLR suppression in the presence of 4.4 nM ECD variant ICOSL-Fc is given in table. 23. In the table, negative values indicate an inflammatory effect in the assay.

- 131 044356

Table 23

Costimulatory bioactivity data for ICOSL in MLR

ICOSL mutation(s) SEQID NO (ECD) % MLR suppression (4.4 nM) N52H, N57Y, Q100R, C198R 365 93.6 N52H, N57Y, Q100R, V122A 290 94.4 N52H, N57Y, Q100R, F172S 291 100.0 N52 Y/N57 Y/F13 8L/L203P 112 100.0 Vl IE, N30D, N52H, N57Y, H94E, L96I, L98F, N194D, V210A, I218T 308 100.0 N52H, N57Y, Q100R, L102R, V110D, H115R, C198R 367 100.0 N52H, N57Y, Q100R 283 98.2 N52H,Q100R 285 97.5 N52H, N57Y, Q100R, V110D, C198R, S212G 364 90.4 N52H, N57Y, QI OOP 113 100.0 E16V, N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, C198R 300 100.0 N30D, K42E, N52S 324 -38.8 N52S, F120S, I143V, I224V 325 -44.2 N52S, E90A 323 -30.4 N52H, N57Y, V110A, C198R, R221I 317 100.0 N52S, N194D 366 -22.3 N52H, I143T 135 -78.0 N52D 111 0.5

F. Assessment of proliferation and intracellular cytokine markers by flow cytometry.

Carboxyfluorescein succinimidyl ester (CFSE)-labeled pan T cells from MLR studies as described above, which were incubated for 4 days in the presence of wild-type ICOSL ECD-Fc molecules or recombinant variant ICOSL ECD-Fc molecules, were further tested for cytokine levels by restimulation with phorbol acetate myristate (PMA)/ionomycin for 6 h in the presence of a Golgi inhibitor (Golgi/Block/Plug). T cells from the MLR study that were incubated with human IgG, anti-CD28, anti-ICOSL, PD-L2-Fc, or belatacept (CTLA4-Fc containing the L104E and A29Y mutations) were also restimulated. Tcells were stained for surface markers CD4 or CD8, fixed, permeabilized, and intracellularly stained for various cytokines as indicated in Table 1 . 24 and 25.

The percentage (%) of CD4+ and CD8+ T cells that were positive for specific intracellular cytokines are presented in Table 1. 24 respectively. The results showed that a number of variant ICOSL ECD-Fc molecules are capable of inhibiting one or more cytokines, including in some cases a majority of cytokines. The total score and mean score were calculated to evaluate the cumulative effects of the individual molecules tested across the parameters considered in this analysis. Proliferation and the percentage of cells that entered division were also assessed, as determined by CFSE dilution, which is also shown in Table. 24 and 25. From the results presented, it is clear that certain variants demonstrate comparable or better activity than belatacept, especially on CD8+ cells.

- 132 044356

Table 24

Assessment of proliferation and intracellular cytokine levels in CD4+ T cells

% cytokine + Variant SEQ ID NO (ECD) or Protein About the bodice %IFNg+ %IL4 + %IL21+ %IL22+ %TNF + %IL2 + %IL10 + %IL17A+ Total score Average score 308 h,o 9.9 3.0 1.0 1.3 34.9 31.3 0.1 0.2 41.0 4.6 300 2.7 π,ι 3.4 1.1 1.4 38.0 35.0 0.0 0.1 63.0 7.0 317 2.9 10.9 3.3 1.1 1.5 37.3 34.1 0.1 0.2 65.0 7.2 291 3.2 8.1 2.5 0.6 1.1 27.9 24.1 0.9 1.3 66.0 7.3 283 z, z 9.2 3.0 0.8 1.4 31.1 26.3 0.8 1.3 70.0 7.8 364 3.4 10.9 3.3 1.0 1.6 36.5 32.3 0.5 0.9 89.0 9.9 390 3.6 9.5 3.1 0.9 1.5 33.8 29.4 0.8 1.4 92.0 10.2 367 2.8 12.0 3.5 1.1 1.6 40.9 38.5 0.1 0.3 92.0 10.2 CTLA-4Ig: L104E, A29Y (Belatacept) 10.7 10.5 2.7 2.4 2.0 24.5 19.6 0.6 1.4 99.0 11.0 112 3.5 12.0 3.8 1.3 1.6 41.3 36.7 0.2 0.4 109.0 12.1 285 4.4 9.8 3.2 1.2 1.7 32.7 29.3 0.9 1.4 114.0 12.7 113 h,o 13.2 4.1 1.2 1.8 43.2 39.8 0.1 0.2 115.0 12.8 365 3.6 11.3 3.9 1.0 2.0 39.1 34.6 0.8 1.3 118.0 13.1 WT ICOSL 12.7 16.7 5.8 4.9 4.7 36.2 29.2 0.2 0.4 127.0 14.1 113 3.5 12.3 3.9 1.1 1.8 39.2 37.5 0.4 0.6 127.0 14.1 366 10.9 16.0 7.2 4.9 5.4 40.6 33.0 0.1 0.2 135.0 15.0 321 10.6 16.7 6.0 4.3 5.0 41.3 37.3 0.2 0.3 146.0 16.2 mlgGctl 15.5 15.6 5.9 5.2 4.6 31.7 26.0 0.4 1.7 146.0 16.2 PDL2 12.3 17.9 5.7 4.7 5.2 41.4 36.0 0.3 0.6 163.0 18.1 323 11.9 17.7 6.2 5.2 5.3 42.4 37.5 0.2 0.4 167.0 18.6 WT ICOSL 12.8 17.4 6.3 5.4 5.6 38.9 32.1 0.3 0.6 167.0 18.6 Anti- ICOSL 15.5 16.3 6.5 5.2 35.1 29.1 0.7 2.0 168.0 18.7 135 12.6 17.4 6.5 5.4 4.9 44.3 37.2 0.4 0.5 179.0 19.9 HuIgG 12.7 17.1 6.3 5.9 4.9 41.1 32.4 0.7 1.3 181.0 20.1 AntiCD28 88.2 42.9 5.0 5.8 5.4 43.5 25.5 0.4 1.4 183.0 20.3 325 12.7 18.7 6.5 5.4 5.6 44.2 40.0 0.1 0.3 186.0 20.7 111 13.7 18.3 6.8 5.9 6.1 42.1 35.2 0.3 0.5 194.0 21.6

Table 25

Assessment of proliferation and intracellular cytokine levels in CD8+ T cells

% Cytokine + Variant SEQ ID NO (ECD) or Protein About the bodice %IFNg+ %IL4 + %IL21+ %IL22+ %TNF+ %IL2 + %IL10 + % IL17A + Total score Average score 308 4.2 8.4 1.6 2.2 47.2 7.6 8.0 0.2 0.1 67.0 7.4

- 133 044356

300 3.8 8.5 2.0 2.2 47.2 8.1 9.1 0.1 0.0 69.0 7.7 317 3.9 8.8 2.0 1.7 45.6 8.4 9.0 0.1 0.1 64.0 7.1 291 3.8 7.0 1.4 1.8 46.6 5.9 5.9 1.1 1.1 78.0 8.7 283 4.3 8.4 1.9 1.7 50.2 7.1 6.6 1.1 1.1 98.0 10.9 364 4.1 9.0 1.8 1.8 46.4 8.2 8.5 0.7 0.8 87.0 9.7 390 4.1 7.9 1.8 1.8 49.8 7.3 7.2 1.0 1.2 93.0 10.3 367 3.5 9.0 1.7 2.0 47.3 8.6 10.4 0.2 0.2 78.0 8.7 CTLA-41g ^: L104E, A29Y (Belatacept) 12.3 14.5 2.0 3.2 38.1 11.2 8.8 0.9 1.6 121.0 13.4 112 4.2 9.6 1.9 1.8 40.4 9.4 9.9 0.3 0.3 81.0 9.0 285 5.4 9.5 2.0 2.7 44.2 7.8 8.4 1.2 1.3 112.0 12.4 113 3.7 9.5 1.9 1.3 44.4 9.4 10.5 0.1 0.1 62.0 6.9 365 4.1 9.6 2.3 1.8 46.8 9.0 9.3 0.9 1.0 122.0 13.6 ICOSL 17.2 22.3 4.9 6.4 46.4 22.0 15.7 0.4 0.7 181.0 20.1 113 4.2 9.5 2.0 2.1 45.6 8.9 10.7 0.5 0.5 110.0 12.2 366 14.5 19.4 5.6 4.8 48.4 19.2 13.8 0.1 0.2 142.0 15.8 321 13.4 18.9 4.3 5.0 46.3 18.4 15.4 0.3 0.6 138.0 15.3 mlgGctl 20.2 25.0 4.4 4.1 35.5 24.6 15.8 0.7 1.8 174.0 19.3 PDL2 15.6 21.2 4.1 4.8 44.1 20.7 15.6 0.5 0.8 147.0 16.3 323 15.4 20.8 4.7 5.6 44.9 20.6 17.1 0.2 0.5 149.0 16.6 WT ICOSL 17.5 22.1 5.5 4.6 45.0 21.0 12.8 0.2 0.4 148.0 16.4 Anti- ICOSL 21.5 26.4 4.8 5.3 33.4 26.9 19.0 1.0 2.3 198.0 22.0 135 17.2 22.2 4.7 6.7 39.4 22.5 18.1 0.8 0.9 178.0 19.8 HuIgG 15.9 21.5 4.4 6.4 41.6 21.4 15.1 1.4 1.7 179.0 19.9 AntiCD28 60.6 44.3 3.5 2.1 38.6 32.5 16.0 1.2 1.6 182.0 20.2 325 16.5 22.0 4.9 5.4 44.0 21.9 18.5 0.2 0.6 161.0 17.9 111 17.7 22.8 5.3 6.1 45.4 22.9 16.7 0.4 0.6 183.0 20.3

Example 16. Evaluation of cytokine production in a co-culture of B-T cells.

B cells and CD4+ T cells were purified from the same donor, labeled with CSFE, and seeded in 96-well plates at a 1:1 cell ratio of 5 x 10 ⁴ cells per well each. ICOSL ECD-Fc fusion protein molecules or belatacept were added at a final concentration of 40 nM per well. Cells were either unstimulated or incubated with 100 ng/ml Staphylococcus enterotoxin B (SEB), 100 μg/ml phytolacca mitogen (PWM), or both for 7 days at 37°C in a final volume of 200 μg/well.

Cells were collected and their surface stained with the following markers of B and T cell lines (IgM, IgD, CD38, CD138, CD27, CD19, CD4, CD3). Proliferation was assessed by flow cytometry and culture supernatants were assayed for the cytokines IL-5, IL-13 or IL-21 using the LEGENDplex Human Cytokine Detection Kit (Biolegend, USA).

As shown in FIG. 6A, the number of dividing B cells was reduced in B/T cell co-cultures when incubated in the presence of variant ICOSL ECD Fc fusion protein molecules compared to controls without protein. The magnitude of antagonistic effect for the exemplary variants tested was similar to that of CTLA-4-Ig belatacept (L104E, A29Y). Likewise, as shown in FIG. 6B-D, variant ICOSL ECD Fc fusion protein molecules inhibited cytokine production in co-cultures of primary human B/T cells in vitro compared to controls without the protein, as well as to cultures containing wild-type control ICOSL. Compared to belatacept, exemplary variant ICOSL ECD Fc fusion protein molecules tested were more effective at blocking cytokine production in some cases.

Example 17: Assessment of survival and disease activity in a graft-versus-host disease (GvHD) model.

An exemplary variant protein, ICOSL ECD-Fc, was evaluated for activity in a graft-versus-host disease (GVHD) model. Female NGS mice (n=10 per group) were irradiated (100 rad) and given 10 mg gamma globulin subcutaneously on day −1. On day 0, mice received 10 million human peripheral blood mononuclear cells (PBMCs) and an intraperitoneal injection of either 100 μg of WTICOSL ECD Fc, ICOSL ECD Fc variant molecule N52H/I143T (ECD provided in SEQ ID NO: 135), ICOSL ECD Fc variant molecule N52H /N57Y/Q100P (ECD shown in SEQ ID NO: 113), 75 μg belatacept (CTLA-4-Ig L104E/A29Y, US Patent Application Publication No. US 20160271218) or saline as control. On day 15, engrafted human CD45+ cells were phenotyped by flow cytometry. After study completion at day 35, endpoints of survival, weight loss, and disease activity were assessed.

In fig. 7A shows the survival of GVHD mice treated with saline, WT ICOSL-ECD Fc, ICOSL ECD-Fc variant molecule, or belatacept. Observed significant times

- 134 044356 difference in survival of mice treated with variant ICOSL ECD-Fc N52H/N57Y/Q100P (ECD represented in SEQ ID NO: 113) compared to mice treated with saline or WT ICSOL ECD-Fc (p< 0.0001 by Cox-Mantel test and Gehan-Breslow-Wilcoxon test). In fig. 7B shows similar differences between the body weight loss of mice treated with saline, WT ICOSL ECD-Fc, ICOSL ECD-Fc variant molecules, or belatacept over the course of the study.

Disease Activity Index (DAI) was determined three times per week during the study and assessed for weight loss, posture, activity, and coat and skin appearance of the mice. The gradation of the disease during the study is shown in Fig. 7C. Treatment groups that received the variant ICOSL Fc N52H/N57Y/Q100P (ECD set forth in SEQ ID NO: 113) or belatacept showed significantly improved DAI scores. The percentage of human T cells in peripheral blood on study day 14 was also assessed by flow cytometry. Measurements were averaged across treatment groups, and error bars represented standard error of the mean (SEM). Fig. 7D shows the percentage of living CD3+/CD4+ or CD3+/CD8+ cells in the blood. Treatment groups that received the variant ICOSL ECD Fc with N52H/N57Y/Q100P (ECD set forth in SEQ ID NO: 113) or belatacept showed significantly different levels of CD4+ T cells compared to the saline treatment group (p = 0.008 and 0.006, respectively, unpaired Student's t test). This study demonstrates the therapeutic effect of a variant ICOSL Fc protein in primary human T cells and GVHD in an in vivo model.

Example 18. Evaluation of activation by stacking molecules.

Variant IGV Fc stack fusion proteins containing the IGV NKp30 domain as the localization domain (designated L) and the ICOSL domain of IgV as the costimulatory domain (designated C) were prepared and evaluated essentially as described in Example 8. Specifically, the constructs , tested in this experiment include: (1) a stack construct with a variant IgV Fc fusion protein (vIgD CL) containing NKp30, consisting of Ig from the NKp30 consensus variant (SEQ ID NO: 493) and the IgV variant ICOSL domain N52H/N57Y/ Q100P (ECD represented in SEQ ID NO: 113 and IgV represented in SEQ ID NO: 201); (2) a stack construct with a wild-type NKp30 Ig domain (SEQ ID NO: 214) and a wild-type ICOSL V domain (WT CL) (SEQ ID NO: 196); (3) a wild-type ICOSL domain construct of IgV (WT C domain, SEQ ID NO: 196) and (3) a wild-type NKp30 domain construct (WT L domain, SEQ ID NO: 214).

CD32+ K562 cells were engineered to stably express B7-H6 on the cell surface. Cells were then treated with mitomycin and plated with pan T cells in the presence/absence of 10 nM anti-CD3 and stack variants or control domains at 100, 33, 11, or 3.7 nM. Cells were cultured for 3 days before collecting culture supernatants and measuring human IFN-gamma levels using ELISA.

As shown in FIG. 8, costimulatory-localizing domain stacks, both variant and wild type, were able to localize to engineered K562 cells and deliver a costimulatory signal to pan-T cells to induce IFN-gamma secretion. A stacked construct with a variant IgV Fc fusion protein (vIgD C-L) resulted in increased functional activity at all concentrations tested, while the individual domain components had no effect when not stacked together.

Example 19: Assessment of delayed-type hypersensitivity in vivo.

Variant ICOSL ECD-Fc fusion protein molecules were evaluated for anti-inflammatory activity in vivo in a mouse model of delayed-type hypersensitivity (DTH). Delayed hypersensitivity immune responses were detected in mice sensitized with ovalbumin (OVA) and the response was assessed after challenge. The variant ICOSL ECD-Fc fusion protein molecules tested contained an ECD variant with the following amino acid substitutions: N52H/N57Y/Q100P (ECD represented in SEQ ID NO: 113), N52H/Q100R (ECD represented in SEQ ID NO: 285), or N52H/N57Y/Q100R/F172S (ECD shown in SEQ ID NO: 291). The variants were fused to either an Fc backbone containing the EU numbering mutations C220S/L234A/L235E/G237A (designated Fc#1), which is represented in SEQ ID NO: 477, or an Fc backbone containing the mutations C220S/E233P/L234V/ L235A/G236del/S267K by EU numbering (designated Fc#2), which is presented in SEQ ID NO: 478), either with or without the G4S linker (GGGGS). In table 26 shows the tested constructs.

- 135 044356

Table 26

ICOSL ECD-Fc fusion constructs

ICOSL ECD (SEQ ID NO) Linker G4S Fc (SEQ ID NO) N52H/N57Y/Q100P (G4S)- FC #1 113 + 477 N52H/Q100R (G4S)-Fc #1 285 + 477 N5 2H/N5 7 Y/Q100R/F172 S (G4S)-Fc 291 + 478 N52H/N57Y/Q100R/F172S (G4S)-Fc 291 + 477 N52H/N57Y/Q100R/F172SFc 291 - 477

For sensitization, 8-week-old female BALB/C mice were injected subcutaneously with 100 μg of OVA emulsified in Sigma adjuvant (100 μl; catalog no. S6322-1VL) at the base of the tail on day 0. On days 1 and 4, mice were treated with variant ICOSL ECD-Fc fusion proteins, 75 μg CTLA-4 Fc (abatacept), or PBS as a negative control by intraperitoneal injection. On day 7, two to three hours before the OVA challenge, mice were additionally administered PBS control, 75 μg of CTLA-4 Fc (abatocept from Orencia), or the indicated ICOSL variant polypeptide by intraperitoneal injection. Abatacept and various ICOSL-Fc fusion protein molecules were dosed in molar equivalents.

For a challenge dose of OVA, an intradermal injection of 10 μg of OVA was performed in the left ear in a volume of 10 μl of PBS under gas isoflurane anesthesia 2-3 hours before therapeutic treatment. Baseline ear thickness was measured before the challenge dose of OVA.

On day 8, ear thickness was measured under isoflurane anesthesia using a Mitutoyo caliper, and the change in ear thickness before and after the challenge dose of OVA was determined. As shown in FIG. 9, mice treated with the indicated variant ICOSL ECD-Fc fusion protein molecules showed significantly less OVA-induced ear swelling compared to PBS controls (<0.0001 by one-way ANOVA). There were no significant differences between ear thicknesses for mice treated with abatacept compared with any of the indicated variant ICOSL ECD-Fc fusion proteins or between variant ICOSL treatments. These results indicate that variant ICOSL molecules can reduce immune responses in vivo.

Example 20. Preparation and evaluation of binding and activity of variant ICOSL IgSF domain-containing molecules.

Additional variant ICOSL IgSF (eg, ECD) domain-containing molecules were prepared as described below. In each of the tables below, the amino acid substitutions in the ICOSL variant ECD are indicated by the amino acid position number corresponding to the amino acid positions in the corresponding original unmodified ICOSL extracellular domain (ECD) sequence shown in SEQ ID NO: 32. In some cases, deletion of the AAA linker sequence variant ICOSL ECD-Fc, designated AAAA. Column 2 indicates the SEQ ID NO for each variant ECD domain contained in the variant ECD-Fc fusion protein molecule.

A. Getting additional options 1.

Solubility options.

From the collection of mutants containing the mutations presented below, including E16V, N30D, K42E, N52H, N52Y, N52S, N57Y, E90A, Q100R, Q100P, L102R, V110D, H115R, F120S, V122A, F138L, I143V, I143T, H152C, Yu56 M, F172S, N194D, C198R, L203P, R221I, 1224V, H115R, F172S and C198R mutations have been identified as mutations that can potentially increase protein solubility or enhance protein expression (solubility mutations). These three mutations (H115R, F172S, and C198R) were randomly introduced by site-directed mutagenesis into the same set of clones to generate a set of derivatives that contain one or more of these solubility mutations. Because the site-directed mutagenesis reaction was performed with pooled mutagenic oligonucleotides that reacted with pooled parental clones in one reaction, some of the clones also contain some mutations from other parental clones. The resulting variants contain from 3 to 10 different amino acid mutations in various combinations described in table. 27A.

- 136 044356

Table 27A

Exemplary ICOSL variant polypeptides

Mutation(s) ECD SEQ ID NO Wild type 32 N52H/N57Y/Q100R/H115R/C198R 435 N52H/N57Y/Q100R/F172S/C198R 436 N52H/N57Y/Q100R/H115R/F172S/C198R 437 N52H/N57Y/Q100R/H115R/I143 V/F172S/C198R 438 N52H/N57Y/Q100R/L102R/H115R/F172S/C198R 439 N52H/V122A/F172S/C198R 440 N52H/N57Y/Q100R/H115R/F172S/N194D 441 N52H/N57Y/H115R/F172S/C198R 442 N52H/N57Y/Q100R/H115R/C198R 443 N52H/N57Y/H115R 444 N52H/N57Y/Q100R/H115R 445 N52H/N57Y/Q100R/H115R/F172S/I224V 446 N52H/N5 7 Y/Q100R/H115R/F172 S 447 N52H/N57Y/Q100R/F172S 448 N52H/Q100R/H115R/1143 T/F172 S 449 N52H/N57Y/Q100Р/Н115R/F172S 450 N52Y/N57Y/Q100P/F172S 451 E16V/N52H/N57Y/Q100R/V110D/H115R/C198R 452 E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R 453 N52S/E90A/H115R 454 N30D/K42E N52S/H115R 455 N30D/K42E/N52S/H115R/C198R/R221I 456 N30D/K42E/N52S/H115R/C198R 457 N30D/K42E/N52S/H115R/F172S/N194D 458 N52S/H115R/F120S/I143V/C198R 459 N52S/H115R/F172S/C198R 460 N52H/N57 Y/Q100Р/С198R 461 N52H/N57Y/Q1 OOP Hl 15R/F172S/C198R 462 N52H/N5 7 Y/Q 100P/F172 S/C198R 463 N52H/N57Y/Q100P/H115R 464 N52H/N57Y/Q100Р/Н115R/C198R 465 N52H/Q100R/C198R 466 N52H/Q100R/H115R/F172S 467 N52H/Q100R/H115X/F172S/C198R 468 N52H/Q100R/H115R/F172S/C198R 469 N52H/N57Y/Q100R/F172S/C198R 470

2. Reverse options.

Specific examples of mutations, including N52H, N52Y, N57Y, Q100R, Q100P, F138L, C198R, L203P, identified in the selected variants described in Example 6 were further combined into the wild-type ICOSL ECD framework relative to the positions presented in SEQ ID NO: 32 to create additional combination options. The resulting variants, containing from 1 to 3 different amino acid mutations in various combinations, are presented in table. 27V.

Table 27B

Exemplary ICOSL variant polypeptides Mutation(s) ECD SEQ ID NO: ID NO Wild type 32 Q100R 427 F138L/L203P 428 N52Y/F138L/L203P 429 N57Y/Q100R/C198R 430 N57Y/F138L/L203P 431 N52H 110 N57Y 121 N57Y/Q100P 122 Q100R/F138L 432 L203P 433

3. Glycosylation options.

Exemplary glycosylation mutations selected from N52H, N52Q, N84Q, N119Q, N155H, N155Q, N168Q, N207Q were combined in various combinations in the wild-type ICOSL ECD framework relative to the positions presented in SEQ ID NO: 32 to obtain additional combination variants. The resulting variants contain from 1 to 5 different amino acid mutations in various combinations presented in table. 27C. Mutations denoted by X indicate either N,

- 137 044356 or Q in the indicated position.

Table 27C (glyc.)

Exemplary ICOSL variant polypeptides

Mutation(s) ECD SEQ ID NO Wild type 32 N84Q 387 N119Q 388 N168Q 389 N207Q 390 N52Q/N207X 391 N168X/N207X 392 N52Q/N168Q 393 N84Q/N207Q 394 N155Q/N207Q 395 N119Q/N168Q 396 N119Q/N207Q 397 N119Q/N155X 398 N52Q/N84Q 399 N52Q/N119Q 400 N84Q/N119Q 401 N52Q/N84Q/N168Q 402 N52Q/N84Q/N207Q 403 N84Q/N155Q/N168Q 404 N84Q/N168Q/N207Q 405 N84Q/N155H/N207Q 406 N155Q/N168Q/N207Q 407 N119QN155Q/N168Q 408 N119Q/N168Q/N207Q 409 N84Q/N119Q/N207Q 410 N119Q/N155H/N207Q 411 N84Q/N119Q/N155Q 412 N52Q/N119Q/N155Q 413 N52H/N84Q/N119Q 414 N52H/N84Q/N168X/N207X 415 N52Q/N84Q/N155X/N168X 416 N52Q/N84Q/N119Q/N168Q 417 N84Q/N119Q/N155Q/N168Q 418 N84Q/N155Q/N168Q/N207Q 419 N84Q/N119Q/N155Q/N207Q 420 N52Q/N84Q/N119Q/N207Q 421 N52Q/N84Q/N119Q/N155Q 422 N52Q/N84Q/N119Q/N155Q/N207Q 423 N84Q/N119Q/N155Q/N168Q/N207Q 424

B. Binding to counterstructures expressed by cells.

Additional variants were formatted as the Fc fusion proteins described in Example 4. The variant Fc fusion molecules were evaluated in binding studies to assess the specificity and affinity of the variant ICOSL domain immunomodulatory proteins for cognate binding partners. The binding studies described in Example 6 used Expi293 cells transfected with cognate binding partners of human CD28, ICOS and CTLA4. MFI was determined for each transfectant and compared with the corresponding unmodified (parental) ECD-Fc.

Binding results for exemplary variant ICOSL ECD-Fc fusion protein molecules are shown in Table. 28A-S. As shown in table. 28A-C, IgSF ICOSL domain variants (eg, ECD) generated with various combinations of specific mutations exhibit increased binding to at least one, and in some cases more than one, cognate counterstructure ligand.

C. Characterization of bioactivity using anti-CD3 coimmobilization assay.

The costimulatory bioactivity of the resulting variant Fc fusion molecules was also assessed in the anti-CD3 coimmobilization assays described in Example 6. 28A-C show the IFN-gamma ratio produced by each ECD-Fc variant in the assay compared to the corresponding unmodified ICOSL ECD-Fc (wild type). Mutations indicated by X indicate a Q or wild-type residue corresponding to the indicated position of SEQ ID NO: 32 at the indicated position. As shown, the resulting variant Fc fusion molecules exhibit improved immunological enhancing activities.

- 138 044356

Table 28A

Molecular sequences, binding data and co-stimulatory bioactivity data of variant ICOSLECD-Fc molecules including selected mutations

Mutations ICOSL SEQ ID NO (ECD) Binding Coimmobilization with anti-POPs IFN-gamma pg/ml (initial ratio) ICOS MFI (original ratio) CD28 MFI (original ratio) CTLA-4 MFI (original ratio) N52H, 436 N57Y, Q100R, F172S, 178790 C198R 118145 (1.33) 59651 (29.60) (41.12) 5059 (37.90) N52H, 437 N57Y, Q100R, H115R, F172S, 211000 C198R 125341 (1.41) 51604 (25.60) (48.53) 8218 (61.57) N52Y, 451 174224 N57Y, 121280 (1.37) 63663 (31.59) (40.07) 8123 (60.86) Q100P, F172S E16V, 453 N52H, N57Y, Q100R, V110D, H115R, Y152C, K156M, F172S, 170080 C198R 107819 (1.22) 68883 (34.18) (39.12) 8936 (66.95) N52S, 459 H115R, F120S, I143V, C198R 116235 (1.31) 25582 (12.69) 22483 (5.17) 125 (0.93) N52H, 461 N57Y, Q100P, 172319 C198R 107164(1.21) 56103 (27.84) (39.63) 1258 (9.43) N52H, 462 N57Y, Q100P, H115R, F172S, 176637 C198R 120864 (1.36) 54586 (27.08) (40.63) 5507 (41.26) N52H, 463 N57Y, Q100P, F172S, 151265 C198R 117954 (1.33) 59376 (29.46) (34.79) 3884 (29.10) N52H, 464 N57Y, Q100P, 178812 H115R 126221 (1.42) 53321 (26.46) (41.13) 4154 (31.13) N52H, 465 N57Y, Q100P, H115R, 148417 C198R 137004 (1.55) 55454 (27.51) (34.14) 5069 (37.98) N52H, 466 Q100R, 116111 C198R 111428 (1.26) 58608 (29.08) (26.71) 3729 (27.94) N52H, 467 Q100R, H115R, 106295 F172S 105532 (1.19) 58287 (28.92) (24.45) 5294 (39.67) N52H, 468 Q100R, N115X, 171815 F172S, 106555 (1.20) 73397 (36.42) (39.52) 6961 (52.16)

- 139 044356

C198R N52H, Q100R, H115R, F172S, C198R 469 114223 (1.29) 66686 (33.09) 157154 (36.15) 7592 (56.88) N52H, N57Y, Q100R, F172S, C198R 470 99350 (1.12) 61292 (30.41) 182288 (41.93) 9167 (68.68) N52H, N57Y, Q100R, H115R, F172S, C198R 437 114057 (1.29) 52011 (25.81) 146471 (33.69) 6545 (49.04) N52H, N57Y, Q100R, H115R, F172S 447 136143 (1.54) 66516 (33.00) 177376 (40.80) 8527 (63.89) N52H, N57Y, Q100R, H115R, F172S, C198R 437 132970 (1.50) 59633 (29.59) 133247 (30.65) 5999 (44.95) Q100R 427 62064 (8.31) 16740 (8.31) 29654 (8.31) 35 (0.26) Q100R ΔΑΑΑ 427 1594 (8.20) 16535 (8.20) 33457 (8.20) 87 (0.65) F138L L203P 428 53804 (0.75) 1510(0.75) 2151 (0.75) 35 (0.26) F138L L203P A AAA 428 53044 (0.93) 1882 (0.93) 1623 (0.93) 35 (0.26) N52Y F138L L203P 429 99761 (23.50) 47369 (23.50) 67300 (23.50) 1489 (11.16) N52Y F138L L203P AAAA 429 59576 (26.23) 52865 (26.23) 66553 (26.23) 997 (7.47) N57Y Q100R C198R 430 58706 (28.65) 57739 (28.65) 99426 (28.65) 9962 (74.64) N57Y Q100R C198R A AAA 430 98514 (28.63) 57694 (28.63) 131458 (28.63) 6763 (50.67) N57Y F138L L203P N57Y F138L L203P A AAA N52H N57Y 431 431 110 121 109472 (20.98) 97777 (22.29) 91598 (28.91) 109031 (21.71) 42276 (20.98) 44924 (22.29) 58264 (28.91) 43754 (21.71) 64477 (20.98) 64742 (22.29) 103025 (28.91) 50683 (21.71) 4979 (37.30) 6507 (48.75) 3393 (25.42) 4881 (36.57) N57Y, QI OOP 122 72480 (29.85) 60161 (29.85) 109522 (29.85) 2797 (20.95) Q100R, F138L 432 65974 (2.23) 4485 (2.23) 8136 (2.23) 685 (5.13) L203P 433 61554 (0.76) 1533 (0.76) 2031 (0.76) 2434 (18.24) ICOSL wild type ECD 32 88625 (1.00) 2015 (1.00) 4348 (1.00) 133 (1.00)

- 140 044356

Table 28B

Molecular sequences, binding data and co-stimulatory bioactivity data of variant ICOSLECD-Fc molecules including selected mutations

Mutations ICOSL SEQ ID NO (ECD) Binding Coimmobilization with anti-POPs ICOS MFI (original ratio) CD28 MFI (original ratio) CTLA-4 MFI (original ratio) IFN-gamma pg/ml (initial ratio) N52H, N57Y, Q100R, H115R 445 165027 (1.97) 51666 (9.89) 287581 (60.27) 5858 (20.36) N52H, N57Y, Q100R, F172S 448 184449 (2.20) 51394 (9.84) 182109 (38.16) 3449 (11.99) N52H, N57Y, Q100R, H115R, F172S, I224V 446 165120 (1.97) 46636 (8.93) 274026 (57.43) 2053 (7.13) N52H, N57Y, Q100R, H115R, F172S 447 164750 (1.97) 40046 (7.67) 259351 (54.35) 3722 (12.93) N52H, N57Y, Q100R, H115R, C198R 435 186017 (2.22) 39073 (7.48) 200505 (42.02) 3909 (13.58) N52H, N57Y, Q100R, F172S, C198R 436 181118 (2.16) 38233 (7.32) 210709 (44.16) 1199 (4.17) N52H, N57Y, Q100R, H115R, F172S, C198R 437 155392 (1.85) 28828 (5.52) 169736 (35.57) 3449 (11.99)

- 141 044356

N52H, N57Y, Q100R, H115R, I143V, F172S, C198R 438 139977 (1.67) 31459 (6.02) 179089 (37.53) 1620 (5.63) N52H, N57Y, Q100R, L102R H115R, F172S, C198R 439 146799 (1.75) 29636 (5.68) 200000 (41.91) 2712 (9.43) N52H, N57Y, Q100R, H115R F172S, N194D 441 150863 (1.80) 31304 (5.99) 167783 (35.16) 15607 (54.24) N52H, N57Y, H115R, F172S, C198R 442 126909 (1.51) 35803 (6.86) 152858 (32.03) 5374 (18.67) N52H, N57Y, Q100R, H115R, C198R 443 131730 (1.57) 37595 (7.20) 139041 (29.14) 9306 (32.34) N52H, N57Y, H115R 444 162632 (1.94) 49847 (9.55) 266878 (55.93) 2918 (10.14) N52H, Q100R, H115R, I143T F172S 449 132873 (1.59) 52058 (9.97) 186366 (39.06) 3086 (10.72) N52H, N57Y, QI OOP, H115R, F172S 450 148160 (1.77) 46851 (8.97) 246636 (51.69) 4987 (17.33) E16V, N52H, N57Y, Q100R, V110D, H115R, C198R 452 154036 (1.84) 48674 (9.32) 212905 (44.62) 5095 (17.71) N52S, E90A, H115R 454 142963 (1.71) 3597 (0.69) 3772 (0.79) 2241 (7.79) N30D, K42E, N52S, H115R, C198R R221I 456 124095 (1.48) 8066 (1.54) 7751 (1.62) 417(1.45) N30D, K42E, N52S, H115R, C198R 457 161734 (1.93) 2791 (0.53) 2919 (0.61) 841 (2.92) N30D, K42E, N52S, H115R, F172S, N194D 458 117880 (1.41) 4395 (0.84) 4941 (1.04) 2904 (10.09) N30D, K42E, N52S, H115R, 455 114107 (1.36) 2935 (0.56) 2748 (0.58) 549 (1.91) N52S, E90A, H115R, 454 120450 (1.44) 12768 (2.45) 23282 (4.88) 2890 (10.04) N30D, K42E, N52S, H115R 455 115273 (1.38) 11964 (2.29) 22779 (4.77) 2241 (7.79) N52S, H115R, F172S, C198R 460 95537 (1.14) 7614 (1.46) 21701 (4.55) 1458 (5.07) Wild type 32 83813 (1.00) 5222 (1.00) 4772 (1.00) 288 (1.00)

- 142 044356

Table 28C

Molecular sequences, binding data and co-stimulatory bioactivity data of variant ICOSLECD-Fc molecules including glycosylation mutations

ICOSL mutation(s) SSEQ ID NO (ECD) Binding Coimmobilization with anti-POPs ICOS MFI (original ratio) CD28 MFI (original ratio) CTLA-4 MFI (original ratio) IFN-gamma pg/ml (initial ratio) N84Q 387 34426 (0.94) 1755 (1.16) 5757 (1.51) 100 (2.03) N119Q 388 30806 (0.84) 4102 (2.70) 19836 (5.21) 81 (1.66) N168Q 389 27041 (0.74) 1410(0.93) 18641 (4.90) 67 (1.36) N207Q 390 36516(1.00) 11923 (7.86) 25701 (6.76) 206 (4.20) N52Q, N207X 391 30216 (0.83) 12086 (7.97) 27952 (7.35) 77 (1.56) N168X, N207X 392 37191 (1.02) 5787 (3.81) 12280 (3.23) 104 (2.12) N52Q, N168Q 393 32576 (0.89) 12638 (8.33) 27167 (7.14) 101 (2.06) N84Q, N207Q 394 37176 (1.02) 5292 (3.49) 3153 (0.83) 31 (0.63) N155Q, N207Q 395 34884 (0.95) 1489 (0.98) 987 (0.26) 73 (1.48) N119Q, N168Q 396 29099 (0.80) 2534 (1.67) 11289 (2.97) 51 (1.05) N119Q, N207Q 397 32603 (0.89) 1861 (1.23) 6795 (1.79) 153 (3.12) N119Q N155X 398 38516 (1.05) 15318 (10.10) 27498 (7.23) 173 (3.52) N52Q, N84Q 399 33988 (0.93) 1675 (1.10) 3525 (0.93) 39 (0.80) N52Q,N119Q 400 35729 (0.98) 11040 (7.28) 26139 (6.87) 51 (1.03) N84Q, N119Q 401 34777 (0.95) 1493 (0.98) 2877 (0.76) 39 (0.80) N52Q, N84Q, N168Q 402 27021 (0.74) 1584 (1.04) 958 (0.25) 38 (0.78) N52Q, N84Q, N207Q 403 39942 (1.09) 13396 (8.83) 26360 (6.93) 37 (0.76) N84Q, N155Q, N168Q 404 27812 (0.76) 357 (0.24) 466 (0.12) 30 (0.61) N84Q, N168Q, N207Q 405 30659 (0.84) 737 (0.49) 861 (0.23) 25 (0.52) N84Q, N155H, N207Q 406 13557 (0.37) 685 (0.45) 607 (0.16) 29 (0.59) N155Q, N168Q, 407 13999 (0.38) 277 (0.18) 317(0.08) 40 (0.82)

- 143 044356

N207Q N119Q, N155Q, N168Q 408 36896 (1.01) 4094 (2.70) 2179 (0.57) 50 (1.02) N119Q, N168Q, N207Q 409 29543 (0.81) 921 (0.61) 3744 (0.98) 72 (1.47) N84Q, N119Q, N207Q 410 21357 (0.58) 569 (0.38) 640 (0.17) 59 (1.20) N119Q, N155H, N207Q 411 37310 (1.02) 614 (0.40) 931 (0.24) 86 (1.75) N84Q, N119Q, N155Q 412 2675 (0.07) 262 (0.17) 291 (0.08) 34 (0.70) N52Q, N119Q, N155Q 413 27853 (0.76) 552 (0.36) 772 (0.20) 42 (0.87) N52H, N84Q, N119Q 414 40700 (1.11) 4580 (3.02) 4601 (1.21) 39 (0.80) N52H, N84Q, N168X, N207X 415 8796 (0.24) 587 (0.39) 481 (0.13) 32 (0.66) N52Q, N84Q, N155X, N168X 416 43521 (1.19) 6605 (4.35) 4811 (1.26) 32 (0.66) N52Q, N84Q, N119Q, N168Q 417 39342 (1.07) 4519 (2.98) 3300 (0.87) 37 (0.76) N52Q, N84Q, N119Q, N207Q 421 7011 (0.19) 602 (0.40) 433 (0.11) 37 (0.75) ICOSL wild type ECD 32 36602 (1.00) 1517(1.00) 3804 (1.00) 49 (1.00)

Example 21: Generation and Evaluation of Protein Fusion Molecules Using an Antibody Targeting HER2.

This example describes the preparation and evaluation of variant ICOSL ECD-Fc protein fusion molecules conjugated to a tumor-targeting agent to form a conjugate (vIgD conjugate).

The V domain only from ICOSL vIgD (N52H/N57Y/Q100P, shown in SEQ ID NO: 201) was combined with the amino and carboxyl termini of the light chain (Fig. 10A) and heavy chain (Fig. 10B) of the HER2-targeting antibody , with intermediate linkers GGGSGGGS. Exemplary vIgD conjugate configurations are shown in FIG. 10C.

To assess binding to HER2, Expi293 transfectants with HER2 DNA or empty control were stained with titrated amounts of HER2-targeting antibody that contains a variant ICOSL conjugate (vIgD conjugate N52H/N57Y/Q100P) at concentrations ranging from 100 pM to 100 nM. Control proteins were also tested, including a wild-type ICOSL ECD-Fc fusion protein, a wild-type PD-L2 IgV-Fc fusion protein, and a variant ICOSL ECD-Fc fusion protein with mutations at N52H/N57Y/Q100P. For each transfectant, the mean fluorescence intensity (MFI) or percentage of positive cells was determined as described in Example 6. All IgSF conjugates prepared as shown in FIG. 11A,B retained binding to HER2 compared to the endogenous level of HER2 expression observed in Expi293 cells. Likewise, vIgD conjugates also exhibited binding to cognate binding partners of ICOSL, including CD28, CTLA-4, and ICOS.

Protein biological activity and proliferation of human primary T cells in in vitro assays were also characterized as described in Example 6. vIgD conjugates were coupled overnight to 96-well plates at 30-0.1 nM in the presence of 10 nM anti-CD3. The plates were washed and 100,000 CFSE-labeled pan T cells were added to the plates and incubated for 72 hours. IFN-γ levels in the supernatant were analyzed by ELISA. As shown in FIG. 12, vIgD conjugates with these configurations showed greater IFN-γ secretion and proliferation compared to the parental wild-type ICOSL ECD-Fc fusion protein conjugate.

Example 22 Nanostring transcriptional signature of primary human T cells.

Tissue culture plates were coated with 10 nM anti-CD3 with 40 nM control Fc protein, wild-type ICOSL-Fc, wild-type CD80-Fc, both, or variant ICOSL-Fc fusion proteins with the indicated mutations. Purified human T cells were then seeded onto protein-coated plates and incubated at 37°C. Cultures from each treatment group described above were harvested at 24, 48, and 72 h, and total RNA was isolated from each cell sample. RNA was transferred to Nanostring and the Cancer Immune chip was used to quantify 750 transcripts

- 144 -

Claims (69)

genes in each sample. Transcript values were normalized using proprietary Nanostring software allowing comparison of transcript levels between treatment groups and over different time periods. As shown in FIG. 18 and 19, the ICOSL ECD-Fc variant polypeptides tested showed altered inflammatory activity compared to wild-type CD80 ECD-Fc, wild-type ICOSL ECD-Fc, or a combination thereof. Example 23: Generation and Evaluation of Protein Fusion Molecules Using an Antibody Targeting HER2. The proliferation of human T cells cocultured with VmAb and HER2-expressing target cells was also characterized. CFSE-labeled pan T cells were stimulated for 72 h with artificial target cells derived from K562 cells presenting cell surface anti-CD3 single-chain Fv (OKT3) and HER2 in the presence of VmAb or control proteins. Proliferation was measured using CFSE dilution flow cytometric analysis on stained CD4+ or CD8+ T cells. Vmab was analyzed by varying either the number of target cells or the concentration of VmAb used. In the first assay, K562 target cells were titrated from 2500 to 78 cells/well and added to 100,000 T cells for an effector:target (E:T) range of 40 to 1280:1. VmAb, IgSF parent domain, or WT ICOSL were added at a concentration of 1000 pM. In the second assay, K562 target cells were added at levels ranging from 625 cells/well to 100,000 T cells for an effector:target ratio of 160:1. VmAb or control proteins were titrated and added at 3000 to 37 pM. As shown in FIG. 20A and 20B, both assay configurations demonstrate that VmAbs containing the vIgD conjugate provide superior proliferation compared to the parent antibody, parent IgSF domain, or WT ICOSL. In addition, vIgD conjugates mediate proliferation at low E:T ratios (1280:1) or at low protein concentrations (37 pM). The present invention is not to be limited in scope by the specific embodiments disclosed, which are provided, for example, to illustrate various aspects of the present invention. Various modifications to the described compositions and methods will become apparent from the description and teachings of the present invention. Such variations may be practiced without departing from the true scope and spirit of the invention and are intended to fall within the scope of the present invention. CLAIM 1. Variant ICOSL polypeptide including an IgV domain: wherein the variant ICOSL polypeptide contains the amino acid substitution N52H in the unmodified ICOSL with reference to SEQ ID NO: 32 or a portion thereof containing an IgV domain, wherein the IgV domain is represented by amino acids 19-129 in SEQ ID NO: 5, where: the variant ICOSL polypeptide contains an amino acid sequence that exhibits at least 95% sequence identity to SEQ ID NO: 32 or a portion of SEQ ID NO: 32 containing the specified IgV domain; and the variant ICOSL polypeptide specifically binds to the ectodomain of human CD28 with increased affinity compared to unmodified ICOSL. 2. Variant ICOSL polypeptide according to claim 1, characterized in that the variant ICOSL includes up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 amino acid substitutions. 3. The variant ICOSL according to claim 1 or 2, wherein the variant ICOSL includes an amino acid sequence that has at least 96, 97, 98 or 99% identity with the sequence of SEQ ID NO: 32 or a portion thereof containing an IgV domain. 4. Variant ICOSL polypeptide according to any one of claims 1-3, characterized in that the variant ICOSL contains amino acid substitutions selected from N52H/N57Y/Q100P, N52H/C198R, N52H/C140del/T225A, N52H/C198R/T225A, N52H/K92R, N52H/S99G, N52H/L161P/C198R, N52H/I143T, N52H/N57Y/Q100R, N52H/Q100R, N52H/N57Y/Q10 0R/V122A, N52H/ N57Y/Q100R/F172S, E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/C198R, N52H/N57Y/Q100R/V110D/C198R, N52H/N57Y/V110A/C198R/R221I, M10I/S13G/N52H/N57Y/D77G/V110A/H129P/I143V/F172S/V193M/C198R, N52H/N57Y/Q100R/V110D/C198R/S212G , N52H/N57Y/Q100R/ L102R/V110D/H115R/C198R, N52H/N57Y/Q100R/C198R, N52H/N57Y/L74Q/V110D/S192G, N52H/S121G/C198R, N52H/N84Q/N119Q, N52H/N57Y/Q100R/H115R/C198R, N52H/N57Y/Q1 00R/H115R/F172S/ C198R, N52H/N57Y/Q100R/H115R/I143V/F172S/C198R, N52H/N57Y/Q100R/L102R/H115R/F172S/C198R, N52H/N57Y/Q100R/H115R/F172S/N194D, N52H/N57Y/H115R/F172S/C198R, N52H/N57Y/H115R, N52H/N57Y/Q100R/H115R, N52H/N57Y/Q100R/H115R/F172S/I224V, N52H/N57Y/Q100R/H115R/F172S, N52H/Q100R/H115R/I143T/F172S, N52H/N57Y /Q100P/H115R/F172S, E16V/N52H/N57Y/Q100R/V110D/H115R/C198R, E16V/N52H/N57Y/Q100R/V110D/H115R/Y152C/K156M/F172S/C198R, N52H/N57Y/Q100P/C198R, - 145 044356 N52H/N57Y/Q100P/F172S/C198R, N52H/N57Y/Q100P/H115R, N52H/N57Y/Q100P/H115R/C198R, N52H/Q100R/C198R, N52H/Q100R/H115R/F172S, N52H/Q100R/H115R/F172S/C198R, N52H/N57Y/Q100R/F172S/C198R, N52H/N57Y/Q100P/H115R/F172S/C198R, N52H/Q100R/H115X/F172S/C198R, N52H/N57Y/Q100R/V107I/V110D/S132F/I154F/C198R/R221G, Q37R/N52H/N57Y/Q100R/V110N/S142F/C198R/D217V/R221G, F27S/N52H/N57Y/V110N, V11E/N30D/N52H/N57Y/H94E/L96I/L98F/N194D/V210A/I218T, N52H/N57Y/H94E/L96I/F120I/S126TAV153R/I218N, N52H/N57Y/Q100R/V110D/F172 S/C198R, S25G/F27C/N52H/N57Y/Q100R/V110D/E135K/L173S/C198R, N52H/N57Y/Q100R/V110D/N144D/F172S/C198R, A20V/N52H/N57Y/Q100R/S109G, N52H/N57Y/R61S/Q100R/V110D/L173S, N52H/S121G, N52H/N84Q, N52H/F78L/Q100R, N52H/N57Y/Q100R/V110D/S132F/M175T, N52H/N57Y/Q100R/V107I/V110D/I154F/C198R/R221G, N52H/Q100R/F172S/C198R. 5. The variant ICOSL polypeptide according to any one of claims 1-4, where the variant ICOSL contains amino acid substitutions N52H/N57Y/Q100R. 6. A variant ICOSL polypeptide according to any one of claims 1-5, including the amino acid sequence presented in any of SEQ ID NO: 110, 113, 115, 117-119, 128, 135, 280, 283, 285, 287, 289- 292, 300-302, 304, 308, 310, 315-318, 321, 364, 365, 367, 372, 375, 376, 381, 435-439, 441-450, 452, 453, 461-470 or amino acid sequence which has at least 95% sequence identity with any of SEQ ID NO: 110, 113, 115, 117-119, 128, 135, 280, 283, 285, 287, 289-292, 300-302, 304 , 308, 310, 315-318, 321, 364, 365, 367, 372, 375, 376, 381, 435-439, 441-450, 452, 453, 461-470. 7. A variant ICOSL polypeptide according to any one of claims 1-6, containing the amino acid sequence presented in any of SEQ ID NO: 110, 113, 115, 117-119, 128, 135, 280, 283, 285, 287, 289- 292, 300-302, 304, 308, 310, 315-318, 321, 364, 365, 367, 372, 375, 376, 381, 435-439, 441-450, 452, 453, 461-470. 8. A variant ICOSL polypeptide according to any one of claims 1-7, containing the amino acid sequence presented in any of SEQ ID NO: 198, 201-203, 326-334, 337, 386. 9. The variant ICOSL polypeptide according to any one of claims 1 to 8, characterized in that the IgV domain is the only domain of the immunoglobulin superfamily of the variant ICOSL polypeptide. 10. A variant ICOSL polypeptide according to any one of claims 1 to 9, comprising an amino acid sequence presented in any of SEQ ID NO: 198, 201-203, 326-334, 337, 386, or an amino acid sequence that has at least , 95% sequence identity with any of SEQ ID NO: 198, 201-203, 326-334, 337, 386 or a specific binding fragment thereof, and includes one or more amino acid substitutions. 11. The variant ICOSL polypeptide according to any one of claims 1 to 10, wherein the variant ICOSL polypeptide specifically binds to each of the ectodomain of human ICOS and the ectodomain of human CD28 with increased affinity compared to the binding of unmodified ICOSL to the ectodomain of human ICOS and human CD28. 12. Variant ICOSL polypeptide according to any one of claims 1 to 11, characterized in that the increased affinity for the CD28 ectodomain increases by more than 1.2, 1.5, 2, 3, 4, 5, 6 , 7, 8, 9, 10, 20, 30, 40, 50 or 60 times compared to unmodified ICOSL. 13. Variant ICOSL polypeptide according to claim 11, characterized in that the increased affinity for the ICOS ectodomain increases by more than 1.2, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, 50, 60 or 70 times compared to unmodified ICOSL. 14. The ICOSL variant polypeptide according to any one of claims 1 to 13, which is a soluble protein. 15. A variant ICOSL polypeptide according to any one of claims 1 to 14, which is associated with a multimerization domain. 16. The variant ICOSL polypeptide according to any one of claims 1 to 15, wherein the variant ICOSL polypeptide is a transmembrane immunomodulatory protein comprising a transmembrane domain linked to the extracellular domain (ECD) of the variant ICOSL polypeptide and also contains a cytoplasmic signaling domain linked to the transmembrane domain. 17. A variant ICOSL-Fc fusion protein comprising a variant ICOSL polypeptide according to any one of claims 1 to 16 and an Fc domain. 18. The variant ICOSL-Fc fusion protein of claim 17, wherein the Fc domain is a variant Fc domain with reduced effector function. 19. A variant ICOSL-Fc fusion protein comprising a variant ICOSL polypeptide fused via a linker to an Fc domain, where: (i) the variant ICOSL polypeptide contains amino acid substitutions N52H/N57Y/Q100P in the amino acid sequence represented in SEQ ID NO:32 or in a portion thereof containing an IgV domain, where the IgV domain is represented by amino acids 19-129 in SEQ ID NO:5; and (ii) the Fc domain is a variant human IgG1 Fc with reduced effector function. 20. A variant ICOSL-Fc fusion protein comprising a variant ICOSL polypeptide fused via a linker to an Fc domain, where: - 146 044356 (i) the ICOSL variant polypeptide contains the amino acid substitutions N52H/N57Y/Q100R in the amino acid sequence represented in SEQ ID NO:32 or in the portion thereof containing the IgV domain, where the IgV domain is represented by amino acids 19-129 in SEQ ID NO: 5; and (ii) the Fc domain is a variant human IgG1 Fc with reduced effector function. 21. A variant ICOSL-Fc fusion protein comprising a variant ICOSL polypeptide fused via a linker to an Fc domain, where: (i) the variant ICOSL polypeptide contains amino acid substitutions N52H/Q100R in the amino acid sequence represented in SEQ ID NO:32 or in a portion thereof containing an IgV domain, where the IgV domain is represented by amino acids 19-129 in SEQ ID NO: 5; and (ii) the Fc domain is a variant human IgG1 Fc with reduced effector function. 22. The variant ICOSL-Fc fusion protein of any one of claims 16 to 21, wherein the Fc domain comprises an amino acid sequence set forth in SEQ ID NO:226 or SEQ ID NO:227, or an amino acid sequence that exhibits at least 85% identity sequences with SEQ ID NO: 226 or SEQ ID NO: 227. 23. The variant ICOSL-Fc fusion protein according to any one of claims 17 to 22, wherein the Fc domain contains one or more amino acid modifications selected from E233P, L234A, L234V, L235A, L235E, G236del, G237A, S267K, R292C, N297G and V302C , each according to EU numbering. 24. The variant ICOSL-Fc fusion protein according to any one of claims 17 to 23, wherein the Fc domain contains the amino acid modification C220S according to the EU numbering. 25. The variant ICOSL-Fc fusion protein of any one of claims 17 to 24, wherein the Fc domain comprises an amino acid sequence set forth in SEQ ID NO: 474, 476, 477, 478 or an amino acid sequence that exhibits at least 85% sequence identity with SEQ ID NO: 474, 476, 477, 478 and exhibits reduced effector function. 26. The variant ICOSL-Fc fusion protein according to any one of claims 17 to 25, wherein the variant ICOSL polypeptide is linked to the Fc domain via a G4S linker. 27. A variant ICOSL-Fc fusion protein according to any one of claims 17 to 26, which is a dimer. 28. The variant ICOSL-Fc fusion protein of claim 27, which is a homodimer. 29. A conjugate comprising a variant ICOSL polypeptide according to any one of claims 1 to 16, linked to a moiety, wherein the moiety is a targeting moiety that specifically binds to a cell surface molecule. 30. A conjugate comprising the variant ICOSL-Fc fusion protein of any one of claims 17 to 28, linked to a moiety, wherein the moiety is a targeting moiety that specifically binds to a molecule on the surface of a cell. 31. The conjugate according to claim 29 or 30, characterized in that the targeting moiety specifically binds to a molecule on the surface of the immune cell. 32. The conjugate according to claim 29 or 30, characterized in that the targeting moiety is a tumor-localizing moiety that binds to a molecule on the surface of the tumor. 33. The conjugate according to any one of claims 29-32, characterized in that the fragment is an antibody or an antigen-binding fragment. 34. The conjugate according to any one of claims 29-33, which is a fusion protein. 35. A nucleic acid molecule encoding a variant ICOSL polypeptide according to any one of claims 1 to 16. 36. A nucleic acid molecule encoding a variant ICOSL-Fc fusion polypeptide according to any one of claims 17 to 28. 37. A vector containing a nucleic acid molecule according to claim 35. 38. A vector containing a nucleic acid molecule according to claim 36. 39. The vector according to claim 37 or 38, which is an expression vector. 40. The vector according to any one of claims 37 to 39, characterized in that the vector is a mammalian expression vector or a viral vector. 41. A cell comprising a nucleic acid molecule according to claim 35 or 36. 42. A cell comprising a vector according to any one of claims 37-40. 43. The cell of claim 41 or 42, which is a mammalian cell. 44. The cell according to any one of claims 41-43, which is a human cell. 45. A method for producing a variant ICOSL polypeptide, comprising introducing a nucleic acid molecule according to claim 35 or 36 or a vector according to any one of claims 37-40 into a host cell under conditions for expression of the protein in the cell. 46. A method for producing a variant ICOSL-Fc fusion protein, comprising introducing a nucleic acid molecule according to claim 35 or 36 or a vector according to any one of claims 37-40 into a host cell under conditions for expression of the protein in the cell. 47. The method of claim 45 or 46, further comprising isolating or purifying the variant ICOSL polypeptide or variant ICOSL-Fc fusion protein from the cell. 48. A method for constructing a cell in vitro expressing a variant ICOSL polypeptide, - 147 044356 comprising introducing a nucleic acid molecule encoding a variant ICOSL polypeptide according to any one of claims 1 to 16 into a host cell under conditions in which the polypeptide is expressed in the cell. 49. A method of constructing a cell in vitro expressing a variant ICOSL-Fc fusion protein, comprising introducing a nucleic acid molecule encoding the variant ICOSLFc fusion protein of any one of claims 17 to 28 into a host cell under conditions in which the polypeptide is expressed in the cell. 50. An engineered cell comprising a variant ICOSL polypeptide according to any one of claims 1 to 16, or an ICOSL-Fc fusion protein according to any one of claims 17 to 28, a nucleic acid molecule according to claim 35 or 36, or a vector according to any one of claims 37 -40. 51. The engineered cell according to claim 50, characterized in that the cell is an immune cell. 52. The engineered cell of claim 50 or 51, further comprising a chimeric antigen receptor (CAR) or an engineered T-cell receptor. 53. An infectious agent comprising a nucleotide molecule encoding a variant ICOSL polypeptide according to any one of claims 1 to 16 or a variant ICOSL-Fc fusion protein according to any one of claims 17 to 28, wherein the infectious agent is a bacterium or a virus. 54. A pharmaceutical composition comprising a variant ICOSL polypeptide according to any one of claims 1 to 16 and a pharmaceutically acceptable excipient. 55. A pharmaceutical composition comprising the variant ICOSL-Fc fusion protein according to any one of claims 17 to 28 and a pharmaceutically acceptable excipient. 56. A pharmaceutical composition comprising the conjugate according to any one of claims 29 to 34 and a pharmaceutically acceptable excipient. 57. A pharmaceutical composition comprising an engineered cell according to any one of claims 50-52 and a pharmaceutically acceptable excipient. 58. A pharmaceutical composition comprising an infectious agent according to claim 53 and a pharmaceutically acceptable excipient. 59. A bottle comprising a pharmaceutical composition according to any one of claims 54-58. 60. A kit for use in the treatment of an autoimmune or inflammatory disease in a subject, comprising a pharmaceutical composition according to any one of claims 54-58 or a bottle according to claim 59 and instructions for use. 61. Use of a pharmaceutical composition according to any one of claims 54 to 58 for the preparation of a medicament for use in modulating an immune response in a mammal, wherein modulating the immune response treats an autoimmune or inflammatory disease in the mammal. 62. Use of a pharmaceutical composition according to any one of claims 54 to 58 for modulating an immune response in a mammal, wherein modulating the immune response treats an autoimmune or inflammatory disease in the mammal. 63. A method of modulating an immune response in a subject, comprising administering a pharmaceutical composition according to any one of claims 54 to 58 to the subject. 64. The method of claim 58, wherein modulating the immune response treats a disease or condition in the subject. 65. Method according to claim 63 or 64, characterized in that the immune response is reduced. 66. A method of treating a disease or condition in a subject, comprising administering a pharmaceutical composition according to any one of claims 54 to 58 to the subject. 67. The method according to any one of claims 64-66, characterized in that the disease or condition is an inflammatory or autoimmune disease or condition. 68. The method according to any one of claims 64 to 67, wherein the disease or condition is antineutrophil cytoplasmic antibody (ANCA) associated vasculitis, vasculitis, autoimmune skin disease, transplantation, rheumatic disease, inflammatory gastrointestinal disease, inflammatory an eye disease, an inflammatory neurological disease, an inflammatory lung disease, an inflammatory endocrine disease or an autoimmune hematological disease. 69. The method according to any one of claims 64-68, characterized in that the disease or condition is selected from inflammatory bowel disease, transplantation, Crohn's disease, ulcerative colitis, multiple sclerosis, asthma, rheumatoid arthritis, psoriasis, graft-versus-host disease (GVHD) or systemic lupus erythematosus (SLE). - 148 -