EA041306B1 - AGONISTIC ANTIBODIES AGAINST ICOS AND THEIR USE - Google Patents

Description

The field of technology to which the invention belongs

The present invention relates to anti-inducible costimulatory T cell (ICOS) agonist antibodies and pharmaceutical compositions thereof, as well as methods of using such antibodies, for example, for the treatment of cancer by administering anti-ICOS agonist antibodies and pharmaceutical compositions.

Related Applications

This application claims the priority of U.S. Provisional Application Nos. 62/483158 (filed April 7, 2017), 62/514151 (filed June 2, 2017), 62/545732 (filed August 15, 2017) and 62/581412 (filed November 3, 2017). The entire content of the above applications is incorporated herein by reference.

Sequence listing

This application contains the Sequence Listing, which has been submitted electronically in ASCII format and incorporated herein by reference in its entirety. The mentioned ASCII copy, created on March 30, 2018, is named MXI-556PC_SL.txt and is 321751 bytes in size.

Prerequisites for the creation of the invention

There is a need to confront the global cancer epidemic. Cancer is one of the leading causes of disease and the second leading cause of death worldwide. In 2015, cancer accounted for 8.8 million deaths. Globally, almost one in six deaths is related to cancer. In the United States, an estimated 1,735,350 new diagnosed cases of cancer and 609,640 cancer-related deaths are expected in 2018. In 2012, there were 3.5 million new cancer cases and 1.9 million cancer deaths in Europe. According to World Health Organization estimates in 2018, new cancer cases are expected to increase by approximately 70% over the next two decades.

Conventional cancer treatments include surgery, radiation therapy, and chemotherapy, among other treatments. In recent years, immuno-oncology has emerged as a new method of cancer treatment. Immuno-oncology differs from conventional cancer treatments, which, for example, attempt to target tumors directly or cut off the tumor's blood supply. Instead, immuno-oncology involves using the patient's own immune response to treat cancer. Understanding how the immune system affects the development of cancer and how it can be used to treat cancer is a complex, challenging task. For example, patients may not respond to certain immuno-oncology drugs, and some develop mechanisms of resistance, such as T cell depletion, which occurs when a T cell, a specific type of white blood cell, no longer functions properly. (Dempke et al., Eur. J. of Cancer, 74 55-72 (2017))

An important role of the immune system lies in its ability to distinguish between normal cells and foreign cells. Thus, the immune system can attack foreign cells and leave only normal cells. To do this, the immune system uses checkpoints, which are molecules on certain immune cells that must be activated or inactivated in order to initiate an immune response. Tumor cells can sometimes use these checkpoints to avoid being attacked by the immune system. Some immuno-oncology drugs target these checkpoints by acting as checkpoint inhibitors. Programmed cell death protein 1 (PD-1) is a checkpoint inhibitor that normally acts as a brake to prevent T cells from attacking other cells in the body. PD-1 does this when it binds to programmed cell death ligand 1 (PD-L1), a protein on some normal (and cancerous) cells. When PD-1 binds to PD-L1, this interaction indicates that the T cell is not attacking other cells. Some cancer cells have large amounts of PD-L1, which helps them evade immune attack. Therapeutic agents such as monoclonal antibodies that target this PD-1/PD-L1 interaction, such as nivolumab (Opdivo®), can block PD-1/PD-L1 binding to increase the body's immune response against tumor cells.

There is a need for drugs that target different mechanisms of action that work alone or in combination with checkpoint inhibitors for the safe and effective treatment of cancer and other diseases or conditions. T cell activation and function are regulated by the innate immune system through costimulatory molecules in the CD28 superfamily (eg, positive and negative costimulatory molecules that stimulate or inhibit T cell receptor signal activation, respectively). The inducible costimulatory molecule (ICOS), also known as CD278, is an immune checkpoint protein that is a member of this CD28 superfamily. ICOS is a 55-60 kDa type I transmembrane protein that is expressed on T cells after T cell activation and co-stimulates T cell activation after binding of its ligand, ICOS-L (B7H2). ICOS is expressed by CD4+ cells, CD8+ cells and regulatory T cells (Treg). ICOS has also been shown to play a key role in follicular helper T cell (Tfhs) function and the humoral immune response.

- 1 041306

The magnitude and quality of the T cell immune response depends in part on a complex balance between co-stimulatory and inhibitory signals transmitted to T cells. There is a need for new immuno-oncological therapies to improve patient response rates after immunotherapy and to overcome drug resistance.

Brief description of the invention

The present invention provides isolated monoclonal antibodies (eg, humanized and human monoclonal antibodies) that bind to human ICOS (SEQ ID NO: 1), ie. anti-huICOS antibodies, and exhibit therapeutically desirable functional properties. The antibodies of the invention can be used as an agonist to stimulate or enhance an immune response in a subject, for example to stimulate human ICOS activity and/or to elicit antigen-specific T cell responses against a tumor or viral antigen. The antibodies of the invention may also be used in combination with other antibodies (eg, anti-PD-1, PD-L1 and/or CTLA-4 antibodies) to treat various conditions, such as cancer. Accordingly, the antibodies described herein, alone or in combination with other agents, can be used to treat a variety of conditions or diseases, including cancer. In other embodiments, the antibodies disclosed herein can be used in methods for detecting an ICOS protein.

In one aspect, an isolated antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g., human ICOS-L) with human ICOS, and (a) induces proliferation and interferon-gamma (IFN-γ) production in CD25-CD4+ T cells with an EC50 value of about 0.01-0.16 nM in an in vitro CHO-OKT3-CD32A co-culture assay; and/or (b) induces the production of IFN-γ in CD25-CD4+ T cells with an EC50 value of about 0.002-0.4 nM in staphylococcal enterotoxin B in a co-culture assay of CD25-CD4+ T cells and B cells.

In another embodiment, the antibody (or antigen-binding portion thereof) exhibits one or more of the following hallmarks:

(a) binds to human T cells with an EC50 value of about 0.7 nM and cynomolgus monkey T cells with an EC50 value of about 0.3 nM;

(b) binds to human activated CD4+ T cells;

(c) does not bind to human CD28 or human CTLA-4;

(d) activates at least one primary T lymphocyte, such as a CD4+ effector T cell (Teff), a follicular T helper cell (Tfh), and a regulatory T cell (Treg);

(e) induces protein kinase B (pAkt) phosphorylation in an in vitro primary T cell signaling assay with an EC50 value of about 30 nM;

(f) induces the production of interleukin-10 (IL-10) in response to staphylococcal enterotoxin B in a co-culture assay of Tfh and naïve B cells;

(g) induces a greater increase in CD3-stimulated Teff proliferation compared to CD45RA+ Tregs and CD45RO+ Tregs in an in vitro assay;

(h) reduces Teff suppression by Treg cells;

(i) does not increase cytokine production in whole blood assay at 10 µg/mL;

(j) increases secretion of at least one of IL-10 and IFN-g by Tfh cells in vitro;

(k) stimulates ICOS-mediated signaling;

(l) has increased affinity for CD32B and/or CD32A and/or (m) has reduced affinity for CD16.

In another embodiment, the isolated antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g., human ICOS-L) with human ICOS, and induces proliferation and production of interferon -gamma (IFN-γ) in CD25-CD4+ T cells with an EC50 value of approximately 0.083 nM in an in vitro CHO-OKT3-CD32A co-culture assay. In another embodiment, the isolated antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g., human ICOS-L) with human human ICOS, and induces proliferation and production interferon-gamma (IFN-γ) in CD25-CD4+ T cells with an EC50 value of approximately 0.01-0.1 nM in an in vitro CHO-OKT3-CD32A co-culture assay.

In one aspect, an isolated antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g., human ICOS-L) with human ICOS and induces the production of IFN-γ in CD25 - CD4+ T cells with an EC50 value of approximately 0.2 nM in staphylococcal enterotoxin B in a co-culture assay of CD25-CD4+ T cells and B cells. In another aspect, the

A humanized antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g. human ICOS-L) with human ICOS and induces the production of IFN-γ in CD25 - CD4+ T cells with an EC50 value of approximately 0.01-0.1 nM in staphylococcal enterotoxin A in a co-culture assay of CD25-CD4+ T cells and B cells.

In another embodiment, the antibody (or antigen-binding portion thereof) binds to human, cynomolgus monkey, mouse, and rat ICOS.

In another aspect, the isolated antibody binds to a human inducible costimulatory (ICOS) molecule and contains:

(a) a heavy chain variable domain comprising the CDR1, CDR2 and CDR3 regions having the amino acid sequences set forth in SEQ ID NOs: 9, 10 and 11, respectively, and a light chain variable domain containing the CDR1, CDR2 and CDR3 regions having the amino acid sequences indicated in SEQ ID NO: 12, 14 and 15, respectively;

(b) a heavy chain variable domain containing CDR1, CDR2 and CDR3 regions having the amino acid sequences shown in SEQ ID NOs: 18, 19 and 20, respectively, and a light chain variable domain containing CDR1 regions,

CDR2 and CDR3 having the amino acid sequences shown in SEQ ID NOs: 21, 22 and 23, respectively;

(c) a heavy chain variable domain comprising the CDR1, CDR2 and CDR3 regions having the amino acid sequences set forth in SEQ ID NOs: 26, 27 and 28, respectively, and a light chain variable domain containing the CDR1, CDR2 and CDR3 regions having the amino acid sequences indicated in SEQ ID NO: 29, 30 and 31, respectively;

(d) a heavy chain variable domain comprising the CDR1, CDR2 and CDR3 regions having the amino acid sequences set forth in SEQ ID NOs: 34, 35 and 36, respectively, and a light chain variable domain containing the CDR1, CDR2 and CDR3 regions having the amino acid sequences indicated in SEQ ID NO: 37, 38 and 39, respectively;

(e) a heavy chain variable domain comprising the CDR1, CDR2 and CDR3 regions having the amino acid sequences set forth in SEQ ID NOs: 42, 43 and 44, respectively, and a light chain variable domain containing the CDR1, CDR2 and CDR3 regions having the amino acid sequences indicated in SEQ ID NO: 45, 46 and 47, respectively;

(f) a heavy chain variable domain containing the CDR1, CDR2 and CDR3 regions having the amino acid sequences set forth in SEQ ID NOs: 42, 43 and 44, respectively, and a light chain variable domain containing the CDR1, CDR2 and CDR3 regions having the amino acid sequences specified in SEQ ID NOs: 49, 50, and 51, respectively, or (g) a heavy chain variable domain containing the CDR1, CDR2, and CDR3 regions having the amino acid sequences specified in SEQ ID NOs: 191, 192, and 193, respectively, and the variable a light chain domain containing CDR1, CDR2 and CDR3 regions having the amino acid sequences shown in SEQ ID NOs: 194, 195 and 196, respectively.

In another aspect, the isolated antibody binds to a human inducible costimulatory (ICOS) molecule and the heavy and light chain variable regions comprise:

in SEQ ID NO in SEQ ID NO in SEQ ID NO in SEQ ID NO in SEQ ID NO (a) amino acid sequences indicated (b) amino acid sequences indicated (c) amino acid sequences indicated (d) amino acid sequences indicated (e ) amino acid sequences indicated and 6, respectively;

and 176 respectively;

and 25, respectively;

and 33, respectively;

and 41, respectively;

(f) the amino acid sequences shown in SEQ ID NOs: 40 and 48, respectively; or (g) the amino acid sequences shown in SEQ ID NOs: 186 and 189, respectively.

In another aspect, an isolated, full length, humanized monoclonal antibody that binds to a human inducible costimulatory (ICOS) molecule contains heavy chains that contain the amino acid sequence set forth in SEQ ID NO: 7 and light chains that contain the amino acid sequence set forth in SEQ ID NO: 7. SEQ ID NO: 8.

In one embodiment, the isolated antibody competes for binding to ICOS or binds to the same epitope as an antibody that blocks the interaction of human ICOS with human ICOS-L. In another embodiment, the isolated antibody specifically binds to one or more SIFDPPPFKVTL (SEQ ID NO: 203) residues of human ICOS. In another embodiment, the ICOS epitope contains the amino acid residues SIFDPPPFKVTL (SEQ ID NO: 203) of human ICOS.

In one embodiment, the antibodies of the invention are full length antibodies, for example an IgG1, IgG2, IgG2a or IgG4 isotype. In another embodiment, the antibodies are binding fragments, such as Fab, Fab' or (Fab')2 fragments, or single chain antibodies.

In one aspect, anti-ICOS antibodies, or antigen-binding portions thereof, bind to Fc

- 3 041306 receptors, such as one or more activating Fc gamma receptors (FcyRs). In some embodiments, the antibody contains at least one amino acid substitution in the Fco region relative to the human IgG1 sequence (SEQ ID NO: 206) that increases the affinity of the antibody for an FcyR, such as FcyRIIb, such as one or more amino acid substitutions at a position including at least one of 234, 235, 236, 237, 239, 266, 267, 268, 325, 326, 327, 328 and/or 332, according to the EU index, for example, 234D, 234E, 234F, 234W, 235D, 235F, 235R, 235Y, 236D, 236N, 237D, 237N, 239D, 239E, 266M, 267D, 267E, 268D, 268E, 327D, 327E, 328F, 328W, 328Y and/or 332E. In other embodiments, the Fc region contains at least two substitutions 235Y-267E, 236D-267E, 239D-268D, 239D-267E, 267E-268D, 267E-268E, and/or 267E-328F compared to a human IgG1 sequence (SEQ ID NO: 206). In another embodiment, the amino acid substitution in the Fc region is S267E compared to the human IgG1 sequence shown in SEQ ID NO: 206.

In another aspect, the invention provides immunoconjugates comprising an antibody of the invention, or an antigen-binding portion thereof, associated with a therapeutic agent, such as a cytotoxic agent or a radioactive isotope, as well as bispecific molecules containing an antibody of the invention, or an antigen-binding portion thereof, associated with a second functional fragment, having a binding specificity that is different from the binding specificity of said antibody or antigen-binding portion thereof.

Also provided are compositions (eg, pharmaceutical compositions) comprising an antibody, or an antigen-binding portion thereof, or an immunoconjugate, or a bispecific molecule of the invention, and a pharmaceutically acceptable carrier. In another aspect, the composition further comprises a soluble, neutrally active glycoprotein hyaluronidase.

Also provided are nucleic acid molecules encoding antibodies (e.g., cDNA) or antigen-binding portions (e.g., variable regions and/or CDRs) of the invention, as well as expression vectors containing such nucleic acids, and host cells containing such expression vectors. . Also provided are methods for producing anti-ICOS antibodies by expressing the antibody in such host cells and isolating the antibody from the host cell.

In one aspect, the isolated antibody has reduced antibody-dependent cell-mediated cytotoxic (ADCC) activity compared to a control IgG1 antibody.

In another aspect, the invention provides methods for stimulating immune responses using anti-ICOS antibodies or antigen-binding portions thereof according to the invention. In one embodiment, the method includes stimulating an antigen-specific T cell response by contacting T cells with an antibody or antigen-binding portion of the invention such that an antigen-specific T cell response is stimulated. In another embodiment, the production of interleukin-2 by an antigen-specific T cell is stimulated. In yet another embodiment, the subject has a tumor(s) and an immune response against the tumor is being stimulated. In another embodiment, the subject has a virus and an immune response against the virus is being stimulated.

In yet another aspect, the invention provides a method for inhibiting tumor cell growth in a subject, comprising administering to the subject an antibody of the invention, or an antigen-binding portion thereof, such that tumor growth in the subject is inhibited. In another aspect, the invention provides a method of treating a viral infection in a subject, comprising administering to the subject an antibody of the invention, or an antigen-binding portion thereof, such that the subject's viral infection is treated. Such methods include administering an antibody or antigen-binding portion, composition, bispecific molecule, or immunoconjugate of the invention.

In yet another aspect, the invention provides a method for stimulating an immune response in a subject, comprising administering to the subject an antibody of the invention, or an antigen-binding portion thereof, for example, in combination with at least one additional therapeutic agent such as an anti-PD-1 antibody, anti-PD- An L1 antibody and/or an anti-CTLA-4 antibody, such that an immune response is stimulated in the subject, for example to inhibit tumor growth or to stimulate an antiviral response. In one embodiment, the additional immunostimulatory antibody is an anti-PD-1 antibody. In another embodiment, the additional immunostimulatory agent is an anti-PD-L1 antibody. In yet another embodiment, the additional immunostimulatory agent is an anti-CTLA-4 antibody. In yet another embodiment, an antibody of the invention, or an antigen-binding portion thereof, is administered with a cytokine (e.g., IL-2 modified with IL-2 and/or IL-21) or a co-stimulatory antibody (e.g., an anti-CD137 and/or anti-GITR antibody) . In some embodiments, the antibodies are, for example, human, chimeric, or humanized antibodies.

In one embodiment, the isolated antibody is administered with one or more additional therapeutic agents to a human subject. In another embodiment, the additional therapeutic agent is a chemotherapeutic agent.

Also provided herein are methods for treating cancer in a subject (e.g., a patient

- 4 041306 human), comprising administering to the patient an anti-ICOS antibody or a combination of an anti-ICOS antibody and at least one additional antibody (for example, an anti-PD-1 antibody, an anti-PD-Ll antibody and/or an anti-CTLA-4 antibodies), wherein the anti-ICOS antibody or antibody combination is administered according to a specific dosing regimen (i.e., at a specific dose size and according to a specific dosing regimen). In one aspect, the method includes at least one cycle of administration, and for each of at least one cycle, at least one dose of the antibody is administered at a dose of about 375 mg. In another aspect, the antibody is administered in an amount or frequency sufficient to achieve and/or maintain receptor occupancy of less than about 80%. In another embodiment, the method includes administering at intervals of once a week, once every two weeks, once every three weeks, once every four weeks, once every five weeks, once every six weeks, once every seven weeks, once every eight weeks, once every nine weeks, once every ten weeks, once every eleven weeks, or once every twelve weeks.

The methods disclosed herein include the treatment of cancer such as colorectal cancer (CRC), head and neck squamous cell carcinoma (HNSCC), non-small cell lung cancer (NSCLC), prostate cancer (PRC), urothelial carcinoma (UCC), urinary tract cancer. bladder cancer, breast cancer, uterine/cervical cancer, ovarian cancer, testicular cancer, esophageal cancer, gastrointestinal cancer, pancreatic cancer, colon cancer, kidney cancer, gastric cancer, germ cell cancer, bone cancer, liver cancer, thyroid cancer glands, skin cancer, neoplasms of the central nervous system, lymphoma, leukemia, myeloma, sarcoma or virus-associated cancer.

In yet another embodiment, the antibodies are formulated for intravenous administration. In another embodiment, the antibodies are formulated for subcutaneous administration. In another embodiment, the antibodies are administered simultaneously (eg, in the same formulation or simultaneously in separate formulations). Alternatively, in another embodiment, the antibodies are administered sequentially (eg, as separate formulations).

The effectiveness of the treatments provided herein can be evaluated using any suitable means. In some embodiments, the treatment reduces tumor size, reduces the number of metastatic lesions over time, induces a complete response, induces partial results, and/or results in stable disease.

In another aspect, the invention provides anti-ICOS antibodies, or antigen-binding portions thereof, and compositions of the invention for use in the above methods, or for the manufacture of a medicament for use in the above methods (eg, for the treatment of various conditions).

Also provided are kits that include a pharmaceutical composition containing an anti-ICOS antibody in a therapeutically effective amount adapted for use in the methods described herein. In another embodiment, the kit comprises an anti-ICOS antibody and another antibody (e.g., an anti-PD-1 antibody, an anti-PD-L1 antibody, and/or an anti-CTLA-4 antibody) in therapeutically effective amounts adapted for use in methods described in this document. For example, the set contains:

(a) a dose of anti-ICOS antibody;

(b) a dose of anti-PD-1 antibody, anti-PD-L1 antibody, and/or anti-CTLA-4 antibody; and (c) instructions for using the antibody in the method of the invention.

In another aspect, an anti-ICOS antibody is provided for administration (or co-administration with another antibody, e.g., an anti-PD-1 antibody, an anti-PD-Ll antibody, and/or an anti-CTLA-4 antibody) according to the methods described in this document.

Other features and advantages of the present invention will be apparent from the following detailed description and examples, which should not be construed as limiting. The contents of all references, GenBank entries, patents, and published patent applications cited throughout this application are expressly incorporated herein by reference.

Brief description of the drawings

In FIG. 1 shows the sequence of human ICOS (SEQ ID NO: 1). The results of epitope binding analysis for ICOS.4 using hydrogen-deuterium exchange mass spectrometry (HDX-MS) are shown by bolding and underlining the ICOS.4 epitope.

In FIG. 2 shows a portion of the human IgG1f constant domain sequence (SEQ ID NO: 52, renumbered as residues 118-446) that can be used in the Fc sequence variants disclosed herein. The residues in bold are illustrative residues that have been modified. The modified amino acid is shown in bold under the specific residue. The D270E replacement is underlined. The C-terminal lysine residue (K) is omitted from SEQ ID NO: 52, but is present in some embodiments. Similarly, in some embodiments, nucleic acids encoding these embodiments include nucleotides encoding an additional lysine at the 3' end of the nucleic acid.

- 5 041306

In FIG. 3 shows the sequence alignment of the human germline heavy and light chains used to humanize the parental hamster antibody (C398.4). VH3-15 was chosen for the heavy chain and VKI O18 was chosen for the light chain based on framework sequence homology. The human germline FW4, JK3 was also selected for the light chain based on sequence homology. The human germline FW4, JH4 was selected for the heavy chain based on sequence similarity and did not contain residues that could pose a potential risk of exposure. Asterisks and underscores indicate amino acid residues that differ between the germline sequences and the sequence of the parental bat antibody (C398.4).

In FIG. 4 shows the sequences of the heavy and light chain variable regions of the anti-ICOS antibody ICOS.33 IgG1f S267E. The heavy and light chain variable regions CDR1, 2 and 3 are in bold, underlined and labelled.

In FIG. 5A and 5B are graphs that show interferon-gamma (IFN-γ) production and cell proliferation induced by ICOS.33 IgG1f S267E in co-cultures of CD25CD4+ T cells and CHO-OKT3-CD32A cells.

In FIG. 6 is a graph that illustrates the induction of IFN-γ by anti-ICOS antibodies in a SEB-stimulated co-culture assay of CD25-CD4+ T cells and B cells.

In FIG. 7A and 7B are graphs that show the induction of IL-10 and IFN-γ in a SEB-stimulated co-culture of Tfh and naive B cells. Average 4.4x induction.

In FIG. 8A and 8B are graphs that show the elimination of Teff suppression by costimulation of Treg with an anti-ICOS antibody.

In FIG. 9A and 9B are graphs that compare the ability of ICOS.33 IgG1f S267E and ICOS.33 IgG1 to induce ADCC using cells from two different donors (Donors 9 and 12).

In FIG. 10 is a graph of ELISA results comparing the ability of ICOS.33 IgG1f S267E and ICOS.33 IgG1 to bind to the C1q component of human complement.

In FIG. 11A-C are graphs that show the antitumor activity of ICOS Fc variants, ICOS IgG1 SE (ICOS hg1 SE) and ICOS IgG1 (ICOS hg1) antibodies, and IgG1 isotype control antibody (hIgG1) in the MC38 tumor model.

In FIG. 12A-E are graphs that show tumor growth curves by treatment group. Mice were treated with mG1 isotype control, ICOS.1 mg1 D265A, ICOS.4 mg1, ICOS.4 hg1 or ICOS.4 mg2a on days 7, 10 and 14 after Sa1N cell implantation.

In FIG. 13A and 13B are graphs that illustrate mean and median tumor growth curves by treatment group. Mice were treated with mG1 isotype control, ICOS.1 mg1 D265A, ICOS.4 mg1, ICOS.4 hg1 or ICOS.4 mg2a on days 7, 10 and 14 after Sa1N cell implantation.

In FIG. 14A-D are graphs showing the percentage of Foxp3+ Treg cells, CD4+ Teff cells, and CD8+ T cells in tumors at Day 15. Mice were treated with isotype control mG1, ICOS.1 mg1 D265A, ICOS.4 mg1, ICOS.4 hg1 or ICOS.4 on days 7, 10 and 14 after Sa1N cell implantation.

In FIG. 15A-J are graphs that show tumor growth curves for individual mice by treatment group: mIgG1 isotype control, mIgG1 anti-PD-1 antibody D265A (PD1), and/or mIgG1 anti-ICOS.4 antibody (ICOS.4mg1).

In FIG. 16A and 16B are graphs that show mean and median tumor growth curves by treatment group: mIgG1 isotype control, mg1 anti-PD-1 antibody, and/or mIgG1 anti-ICOS.4 antibody (ICOS.4 mg1).

In FIG. 17A-D are graphs that show the mean percentage (SEM) of Foxp3+, CD8+, Ki-67, and Granzyme B in tumors. Mice were treated with mIgG1 isotype control, mg1 anti-PD-1 antibody, and/or mIgG1 anti-ICOS.4 antibody (ICOS.4 mg1).

In FIG. 18A-I are graphs that show tumor growth curves for individual mice by treatment group: mIgG1 isotype control, mIgG1 anti-PD-1 antibody D265A (PD1), and/or mIgG1 anti-ICOS.4 antibody (ICOS).

In FIG. 19A and 19B are graphs that show mean and median tumor growth curves by treatment group: mIgG1 isotype control, mIgG1 anti-PD-1 antibody D265A (PD-1), and/or mIgG1 anti-ICOS.4 antibody (ICOS).

In FIG. 20A-D are graphs that show the percentage of Foxp3+ Treg cells, CD4+ Teff cells, and CD8+ T cells in tumors. Mice were treated with mIgG1 isotype control, anti-PD-1 mIgG1 antibody D265A (PD-1) and/or anti-ICOS.4 mIgG1 antibody (ICOS).

In FIG. 21A-C are graphs that show the mean percentage of Ki-67 in tumors. Mice were treated with mIgG1 isotype control, anti-PD-1 mIgG1 antibody D265A (PD-1) and/or anti-ICOS.4 mIgG1 antibody (ICOS).

- 6 041306

In FIG. 22A-D are graphs that show ICOS-L expression on B cells in the spleen and PBMC. Mice were treated with mIgG1 isotype control, anti-PD-1 mIgG1 antibody D265A (PD-1) and/or anti-ICOS.4 mIgG1 antibody (ICOS).

In FIG. 23A-F are graphs that show tumor growth curves for individual mice by treatment group: mIgG1 isotype control, mIgG2b anti-CTLA-4 antibody (CTLA-4 mg2b), mIgG1 anti-ICOS.4 antibody (ICOS.4 mg1), and /or anti-ICOS.4 antibody mIgG2a (ICOS.4 mg2a).

In FIG. 24A and 24B are graphs that show mean and median tumor growth curves by treatment group: mIgG1 isotype control, mIgG2b anti-CTLA-4 antibody (CTLA-4 mg2b), mIgG1 anti-ICOS.4 antibody (ICOS.4 mg1), and /or anti-ICOS.4 antibody mIgG2a (ICOS.4 mg2a).

In FIG. 25A and 25B are graphs that show ICOS.33 IgG1f S267E binding to human, cynomolgus, rat and mouse T cells as measured using FACS.

In FIG. 26A and 26B are graphs that show that the ICOS.33 IgG1f antibody has higher binding avidity to CD4+ T cells, as calculated from EC50 values, compared to two competing anti-ICOS antibodies.

In FIG. 27 is a diagram illustrating a dose escalation clinical trial using an anti-ICOS antibody in combination with an anti-PD-1 antibody and/or an anti-CTLA-4 antibody.

In FIG. 28 is a graph showing the effects of increasing doses of an anti-ICOS antibody, ICOS.33 IgG1f S267E, in combination with an anti-PD1 antibody and the effect on tumor growth inhibition in a mouse model.

Detailed description of the invention

The present invention provides isolated antibodies, such as monoclonal antibodies, eg humanized or human monoclonal antibodies, that specifically bind to human ICOS (huICOS) and have agonistic activity to stimulate an immune response. In some embodiments, the antibodies described herein contain specific structural features, such as CDR regions containing specific amino acid sequences. In other embodiments, the antibodies compete for binding to the human ICOS protein or bind to the same epitope as the antibodies of the present invention.

Also provided herein are methods for preparing such antibodies, immunoconjugates and bispecific molecules containing such antibodies or antigen-binding fragments thereof, and pharmaceutical compositions formulated to contain the antibodies or antibody fragments. Also provided herein are methods of using antibodies, either alone or in combination with other agents, eg, other immunostimulatory agents (eg, antibodies), to enhance an immune response, eg, to treat cancer and/or infections. Thus, the anti-huICOS antibodies described herein can be used to treat a variety of conditions, including, for example, tumor growth inhibition.

Definitions

For easier understanding of the present description, some terms are first defined. Additional definitions are provided in the detailed description. The headings provided herein are not intended to limit the various aspects of the invention, which may be understood by referring to the description as a whole. Accordingly, the terms defined immediately below are defined in more detail by reference to the description of the invention in its entirety.

As used herein, ICOS refers to an inducible T cell co-stimulator protein, which in humans is encoded by the ICOS gene. ICOS is also known as an inducible costimulator, activation-induced immunomodulatory lymphocytic molecule, AILIM, CVID1 and CD278. Human ICOS is further described in GENE ID NO: 29851 and MIM database (Mendelian Inheritance in Man): 604558. The human ICOS sequence (NP_036224.1) including the 20 amino acid signal sequence is given as SEQ ID NO: 1 and shown in Fig. 1.

Below are the amino acid sequences of two isoforms of human ICOS.

Isoform 1 (Q9Y6W8)

MKSGLWYFFL FCLRIKVLTG EINGSANYEM FIFHNGGVQI LCKYPDIVQQ

FKMQLLKGGQ ILCDLTKTKG SGNTVSIKSL KFCHSQLSNN SVSFFLYNLD

HSHANYYFCN LSIFDPPPFK VTLTGGYLHI YESQLCCQLK FWLPIGCAAF

VVVCILGCIL ICWLTKKKYS SSVHDPNGEY MFMRAVNTAK KSRLTDVTL (SEQ W NO: 1)

Isoform 2 (Q9Y6W8-2)

- 7 041306

MKSGLWYFFLFCLRIKVLTGEINGSANYEMFIFHNGGVOILCKYPDIVOOFKMOLLK GGQILCDLTKTKGSGNTVSIKSLKFCHSQLSNNSVSFFLYNLDHSHANYYFCNLSIFDP PPFKVTLTGGYLHIYESQLCCQLKFWLPIGCAAFVVVCILGCILICWLTKKM (SEQ ID NO: 205)

The signal sequence of isoforms 1 and 2 corresponds to amino acids 1-20 (underlined above). Thus, mature isoforms 1 and 2 consist of amino acids 21-199 of SEQ ID NO: 1 and amino acids 21-158 of SEQ ID NO: 205.

ICOS interacts with the ICOS ligand (ICOS-L), which is also known as ICOSL, ICOS-LG, LICOS, B7H2, B7-H2, B7RP1, B7RP-1, CD275 and GL50. Human ICOS-L is further described in GENE ID NO: 23308 and MIM: 605717. The sequence of human ICOS-L (NP_001269979.1), including the 18 amino acid signal sequence, is shown in SEQ ID NO: 2. Thus, mature the ICOS-L form consists of amino acids 19-302 of SEQ ID NO: 2.

The terms antibody or immunoglobulin, which are used interchangeably herein, refer to a protein containing at least two heavy (H) chains and two light (L) chains linked together by disulfide bonds. Each heavy chain consists of a heavy chain variable region (abbreviated herein as V _H ) and a heavy chain constant region (abbreviated herein as CH). In some antibodies, such as naturally occurring IgG antibodies, the heavy chain constant region consists of a hinge and three domains, CH1, CH2 and CH3. In some antibodies, such as naturally occurring IgG antibodies, each light chain consists of a light chain variable region (abbreviated herein as V _L ) and a light chain constant region. The light chain constant region consists of a single domain (abbreviated herein as CL). The V _H and V _L regions can be further subdivided into regions of hypervariability called complementarity determining regions (CDRs) interspersed with more conservative regions called framework regions (FRs). Each VH and V _L region consists of three CDRs and four FRs, arranged from amino to carboxy in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The variable regions of the heavy and light chains contain a binding domain that interacts with the antigen. Antibody constant regions can mediate immunoglobulin binding to host tissues or factors, including various cells of the immune system (eg, effector cells) and the first component (C1q) of the classical complement system. The heavy chain may or may not have a C-terminal lysine. Unless otherwise indicated, amino acids in the variable regions are numbered using the Kabat numbering system and amino acids in the constant regions are numbered using the EU system. The immunoglobulin may be of any known isotype, including IgA, secretory IgA, IgD, IgE, IgG, and IgM. The IgG isotype is divided into subclasses in some species: IgG1, IgG2, IgG3 and IgG4 in humans, and IgG1, IgG2a, IgG2b and IgG3 in mice. In some embodiments, the anti-ICOS antibodies described herein are of the IgG1 subtype. Immunoglobulins, such as IgG1, exist in several allotypes that differ from each other by no more than a few amino acids. By way of example, an antibody includes both naturally occurring and non-naturally occurring antibodies; monoclonal and polyclonal antibodies; chimeric and humanized antibodies; human and non-human antibodies and fully synthetic antibodies.

As used herein, an IgG antibody has the structure of a natural IgG antibody, that is, it has the same number of heavy and light chains and disulfide bonds as a naturally occurring IgG antibody of the same subclass. For example, an IgG1, IgG2, IgG3, or IgG4 anti-ICOS antibody consists of two heavy chains (HC) and two light chains (LC), with the two heavy chains and light chains linked by the same number and arrangement of disulfide bridges that occur in natural antibodies. IgG1, IgG2, IgG3 and IgG4, respectively (unless the antibody has been mutated to modify disulfide bonds).

An antigen is a molecule or substance that triggers an immune response and to which an antibody binds. Antibodies typically bind specifically to their recognized antigen with high affinity, as reflected by a dissociation constant (K _D ) of 10 ^-7 -10 ^-11 M or less. Any value of KD greater than about 10 ^-6 M, as a rule, indicates non-specific binding. As used herein, an antibody that specifically binds to an antigen refers to an antibody that binds to an antigen, and in some cases substantially identical antigens, with high affinity, i.e., has a KD of 10 ^-7 M or less, 10 ^-8 M or less, 5x10 ^-9 M or less, or 10 ^-8 M to 10 ^-10 M or less, but does not bind with high affinity to unrelated antigens. An antigen is substantially identical to a given antigen if it has a high degree of sequence identity with the given antigen, for example, if it exhibits at least 80%, at least 90%, at least 95%, at least 97%, or at least least 99% sequence identity with the sequence of a given antigen. As an example, an antibody that specifically binds to human ICOS, in some embodiments, also cross-reacts with ICOS antigens from certain primate species other than

- 8 041306 human (eg cynomolgus monkeys) but does not cross-react with ICOS from other species or with an antigen other than ICOS.

As used herein, the term antigen-binding portion or antigen-binding fragment of an antibody refers to one or more portions of an antibody that retain the ability to specifically bind to an antigen (eg, human ICOS). It has been shown that the antigen-binding function of an antibody can be performed by fragments or portions of a full length antibody. Examples of binding fragments encompassed by the term antigen-binding portion or antigen-binding fragment of an antibody, such as an anti-ICOS antibody described herein, include:

(1) Fab-fragment (fragment resulting from cleavage with papain) or similar monovalent fragment, consisting of V _L , VH, LC and CH1 domains;

(2) an F(ab')2 fragment (a fragment resulting from cleavage with pepsin) or a similar bivalent fragment containing two Fab fragments linked by a disulfide bridge in the hinge region;

(3) Fd fragment consisting of VH and CH1 domains;

(4) Fv fragment, consisting of V _L and VH domains of one arm of the antibody;

(5) a single domain antibody (dAb) fragment (Ward et al., (1989) Nature 341:544-46) which consists of a V _H domain;

(6) a dedicated complementarity determining region (CDR); and (7) a combination of two or more isolated CDRs, which may optionally be joined with a synthetic linker. In addition, although the two domains of the Fv fragment, V _L and V _H , are encoded by separate genes, they can be connected using recombinant methods using a synthetic linker, which allows them to be obtained as a single protein chain in which the VL and VH regions pair to form monovalent molecules (known as single-stranded Fv (scFv); see, for example, Bird et al. (1988) Science 242:423-426; and Huston et al. (1988) Proc. Natl. Acad. Sci. USA 85:5879-5883). It is contemplated that such single chain antibodies are also encompassed by the term antigen-binding portion or antigen-binding fragment of an antibody. These antibody fragments are prepared using conventional techniques known to those skilled in the art and the fragments are screened for applicability in the same manner as intact antibodies. Antigen-binding portions can be obtained by methods using recombinant DNA or by enzymatic or chemical cleavage of intact immunoglobulins.

The term acceptor human framework refers to a framework containing the amino acid sequence of a light chain variable domain (VL) framework or a heavy chain variable domain (VH) framework derived from a human immunoglobulin framework or from a human consensus framework. An acceptor human framework derived from a human immunoglobulin framework or a human consensus framework may have the same amino acid sequence as a natural human immunoglobulin framework or a human consensus framework, or it may have amino acid sequence changes compared to the natural human framework. wild-type human immunoglobulin or human consensus framework. In some embodiments, the number of amino acid substitution changes is 10, 9, 8, 7, 6, 5, 4, 3, or 2, or 1. In some embodiments, the human VL acceptor framework is identical in sequence to the human immunoglobulin VL framework or consensus sequence. human frame area.

The hinge, hinge domain or hinge region or hinge region of an antibody refers to a heavy chain constant region domain that connects a CH1 domain to a CH2 domain and contains an upper, middle and lower portion. (Roux et al. (1998) J. Immunol. 161:4083). Depending on its amino acid sequence, the hinge provides different levels of flexibility between the antigen-binding domain and the antibody effector region, and also provides sites for the formation of intermolecular disulfide bonds between the two heavy chain constant regions. As used herein, the hinge starts at E216 and ends at G237 for all IgG isotypes (EU numbering). Id. IgG1, IgG2, IgG3, and IgG4 wild-type hinge sequences are shown in Table 1. 1.

- 9 041306

Table 1. Sequences of the hinge region

Ig type C-terminal ChI* Upper hinge Middle hinge Lower hinge IgGl VDKRV (SEQ ID NO:66) EPKSCDKTHT (SEQ ID NO:67) CPPCP (SEQ ID NO:68) APELLGG (SEQ ID NO:69) IgG2 VDKTV (SEQ ID NO:70) ERK CCVECPPCP (SEQ ID NO:71) APPVAG (SEQ ID NO:72) IgG3 (17-15-1515) VDKRV (SEQ ID NO:66) ELKTPLGDTHHT (SEQ ID NO:73) CPRCP (EPKSCDTPPPCPRCP) ₃ (SEQ ID NO:74) APELLGG (SEQ ID NO:69) IgG3 (17-15-15) VDKRV (SEQ ID NO:66) ELKTPLGDTHHT (SEQ ID NO:73) CPRCP (EPKSCDTPPPCPRCP) ₂ (SEQ ID NO:75) APELLGG (SEQ ID NO:69) IgG3 (17-15) VDKRV (SEQ ID NO:66) ELKTPLGDTHHT (SEQ ID NO:73) CPRCP (EPKSCDTPPPCPRCP)i (SEQ ID NO:76) APELLGG (SEQ ID NO:69) IgG3 (15-15-15) VDKRV (SEQ ID NO:66) EPKS (SEQ ID NO:77) CDTPPPCPRCP (EPKSCDTPPPCPRCP) ₂ (SEQ ID NO:78) APELLGG (SEQ ID NO:69) IgG3 (15) VDKRV (SEQ ID NO:66) EPKS (SEQ ID NO:77) CDTPPPPCPRCP (SEQ ID NO:79) APELLGG (SEQ ID NO:69) IgG4 VDKRV (SEQ ID NO:66) ESKYGPP (SEQ ID NO:80) CPSCP (SEQIDNO:81) APEFLGG (SEQ ID NO:82)

*C-terminal amino acid sequences of CH1 domains.

The term hinge includes wild-type hinges (such as those shown in Table 1) as well as variants thereof (eg, hinges not found in nature or modified hinges). For example, the term IgG2 hinge includes the wild-type IgG2 hinge, as shown in Table 1. 1, and variants having 1 or more mutations (e.g., substitutions, deletions, and/or additions), e.g., 1, 2, 3, 4, 5, 1-3, 1-5, 3-5, and/or no more than 5, 4, 3, 2 or 1 mutation. Exemplary variants of the IgG2 hinge include IgG2 hinges in which 1, 2, 3 or all 4 cysteines (C219, C220, C226 and C229) are replaced with another amino acid, eg serine. In certain embodiments, the IgG2 hinge region contains a C219S substitution. In some embodiments, the hinge contains sequences from at least two isotypes. For example, a hinge may comprise an upper, middle, or lower hinge of one isotype, and the remainder of the hinge of one or more other isotypes. For example, the hinge may be an IgG2/IgG1 hinge and may comprise, for example, an IgG2 upper and middle hinge and an IgG1 lower hinge. The hinge may or may not have an effector function. For example, wild-type IgG1 lower hinge provides effector function. The term CH1 domain refers to a heavy chain constant region linking a variable domain to a hinge in a heavy chain constant domain. As used herein, the CH1 domain begins at A118 and ends at V215. The term CH2 domain refers to the heavy chain constant region linking the hinge to the CH3 domain in the heavy chain constant domain. As used herein, the CH2 domain begins at P238 and ends at K340. The term CH3 domain refers to a heavy chain constant region that is C-terminal to the CH2 domain in the heavy chain constant domain. As used herein, the CH3 domain starts at G341 and ends at K447.

A bispecific or bifunctional antibody is an artificial hybrid antibody having two different binding specificities, eg two different heavy/light chain pairs, resulting in two antigen binding sites with specificities for different antigens. Bispecific antibodies can be generated by a number of methods, including hybridoma fusion or Fab' binding. See, for example, Songsivilai & Lachmann, Clin. Exp. Immunol. 79:315-321 (1990); Kostelny et al., J. Immunol. 148, 1547-1553 (1992).

As used herein, the term monoclonal antibody refers to an antibody derived from a population of substantially homogeneous antibodies, ie. individual antibodies in a population are substantially similar and bind to the same epitope(s) (e.g., antibodies exhibit the same binding specificity and affinity), except for possible variations that may occur during the production of a monoclonal antibody, which variants are usually present in small amounts. The term monoclonal refers to the nature of the antibody as derived from a substantially homogeneous population of antibodies, and does not require the production of the antibody by any particular method. The term human monoclonal antibody refers to an antibody from a population of substantially homogeneous antibodies that exhibits a single binding specificity and that has variable and optional constant regions derived from human germline immunoglobulin sequences. In one embodiment, the human monoclonal antibodies are derived from the hybridoma method. Using the hybridoma-based method, a transgenic non-human animal, such as a transgenic mouse, is exposed to

- 10 041306 antigen, and a leukocyte known as a B cell produces antibodies that bind to an antigen that is collected from a transgenic non-human animal. The isolated B cells are fused with an immortalized cell to form a hybrid cell line called a hybridoma. In one embodiment, the hybridoma has a genome containing a human heavy chain transgene and a light chain transgene fused to an immortalized cell.

Antigen binding fragments (including scFv) of such immunoglobulins are also encompassed by the term monoclonal antibody as used herein. Monoclonal antibodies are highly specific, directed against a single antigenic site. In addition, unlike conventional (polyclonal) antibody preparations, which typically include different antibodies directed against different epitopes on an antigen, each monoclonal antibody is directed against a single epitope. Monoclonal antibodies can be prepared using any of the techniques known in the art as well as the techniques described herein, such as, for example, hybridoma-based, transgenic animal, recombinant DNA techniques (see, for example, US Pat. No. 4,816,567), or using antibody phage libraries as described, for example, in US Pat. No. 7,388,088 and PCT Publication WO 00/31246). Monoclonal antibodies include chimeric antibodies, human antibodies, and humanized antibodies, and may be natural or recombinantly produced.

As used herein, the term recombinant human antibody includes all human antibodies that are made, expressed, made, or isolated by recombinant methods, such as (1) antibodies isolated from an animal (e.g., mouse) that is transgenic or transchromosomal with respect to the genes human immunoglobulins or hybridoma derived from the animal, (2) antibodies isolated from a host cell transformed to express the antibody, e.g. from a transfectoma, (3) antibodies isolated from a combinatorial library of recombinant human antibodies, and (4) antibodies, obtained, expressed, created or isolated by any other means, which will include the splicing of human immunoglobulin gene sequences with other DNA sequences. Such recombinant human antibodies contain variable and constant regions that use specific human germline immunoglobulin sequences encoded by germline genes, but include subsequent rearrangements and mutations that occur, for example, during antibody maturation. As is known in the art (see, for example, Lonberg (2005) Nature Biotech. 23(9): 1117-1125), the variable region contains an antigen-binding domain encoded by various genes that rearrange to form an antibody specific for a foreign antigen. In addition to rearrangement, the variable region can be further modified by multiple single amino acid changes (referred to as a somatic mutation or hypermutation) to increase the affinity of the antibody for the foreign antigen. The constant region will change with further antigen response (i.e., isotype switching occurs). Thus, rearranged and somatically mutated nucleic acid molecules that encode immunoglobulin light chain and heavy chain polypeptides in response to an antigen may not be identical to the original germline sequences, but will instead be substantially identical or similar (e.g., will have at least 80% identity).

As used herein, the term human antibody refers to an antibody having variable regions in which both the framework regions and the CDR regions are derived from human germline immunoglobulin sequences. In addition, if the antibody contains a constant region, the constant region is also derived from human germline immunoglobulin sequences. Anti-huICOS antibodies described herein may include amino acid residues not encoded by human germline immunoglobulin sequences (eg, mutations introduced by in vitro random or site-specific mutagenesis or by in vivo somatic mutation). However, the term human antibody is not intended to include antibodies in which CDR sequences derived from the germline of other mammalian species, such as the mouse, have been grafted onto human framework sequences. In the present description, the terms human antibodies and fully human antibodies are used interchangeably.

A humanized antibody refers to an antibody in which some, most, or all of the amino acids outside of the CDR domains of the non-human antibody are replaced with corresponding amino acids derived from human antibodies. In one embodiment of a humanized form of an antibody, some, most, or all of the amino acids outside of the CDR domains are replaced with amino acids from human antibodies, while some, most, or all of the amino acids within one or more of the CDR regions remain unchanged. Small additions, deletions, insertions, substitutions or modifications of amino acids are acceptable provided they do not interfere with the ability of the antibody to bind to a particular antigen. The humanized antibody retains antigenic specificity similar to that of the parent antibody.

A chimeric antibody refers to an antibody in which the variable regions are derived from one

- 11 041306 species, and the constant regions are derived from other species, such as an antibody in which the variable regions are derived from a mouse antibody and the constant regions are derived from a human antibody. A hybrid antibody refers to an antibody having different types of heavy and light chains, such as a mouse or hamster (parent) heavy chain and a humanized light chain, or vice versa. Chimeric or hybrid antibodies can be constructed, for example, by genetic engineering, from immunoglobulin gene segments belonging to different species.

As used herein, an isotype refers to the class of an antibody (e.g., IgG antibodies (including IgG1, IgG2, IgG3, and IgG4 antibodies), IgM, IgA (including IgA1 and IgA2), IgD, and IgE antibodies) that are encoded by heavy chain constant region genes. antibodies.

An allotype refers to naturally occurring variants within a group of defined isotypes. (See, for example, Jefferis et al. (2009) mAbs 1:1).

The phrases antibody that recognizes an antigen and antibody specific for an antigen are used interchangeably herein with the term antibody that specifically binds to an antigen.

As used herein, an isolated antibody refers to an antibody that is substantially free of other proteins and cellular materials. As used herein, effector function refers to the interaction of the Fc region of an antibody with an Fc receptor or ligand, or the biochemical event that results from such an interaction. Exemplary effector functions include C1q binding, complement-dependent cytotoxicity (CDC), Fc receptor binding, FcYR-mediated effector functions such as ADCC and antibody-dependent cell-mediated phagocytosis (ADCP), and cell surface receptor downregulation ( e.g. B-cell receptor; BCR). Such effector functions typically require the association of an Fc region with a binding domain (eg, an antibody variable domain).

An Fc receptor or FcR is a receptor that binds to the Fc region of an immunoglobulin. FcRs that bind to an IgG antibody contain receptors of the FcyR family, including allelic variants and alternatively spliced forms of these receptors. The FcyR family consists of three activating (FcyRI, FcyRIII and FcyRIV in mice; FcyRIA, FcyRIIA and FcyRIIIA in humans) and one inhibitory (FcYRIIb or equivalently FcyRIIB) receptor. Various illustrative properties of human FcyR are given in table. 2. Most innate effector cell types coexpress one or more activating FcyR and inhibitory FcYRIIb, whereas natural killer (NK) cells selectively express one activating Fc receptor (FcyRIII in mice and FcyRIIIA in humans) but do not express inhibitory FcYRIIb in mice and people. Human IgG1 binds to most human Fc receptors and is considered equivalent to mouse IgG2a in terms of the types of activating Fc receptors it binds to. Table 2 Typical properties of human FcyRs

fcy Allelic variants Affinity for human IgG Isotype preference Cell distribution FcyRI He described High (K _D ~ 10 nM) IgGl=3>4»2 Monocytes, macrophages, activated neutrophils, dendritic cells FcyRIIA H131 low medium IgGl>3>2>4 Neutrophils, monocytes, macrophages, eosinophils, dendritic cells, platelets R131 Low IgGl>3>4>2 FcyRIIIA V158 Medium IgGl=3»4>2 ΝΚ cells, monocytes, macrophages, mast cells, eosinophils, dendritic cells F158 Low IgGl=3»4>2 FcyRIIb 1232 Low IgGl=3=4>2 B cells, monocytes, macrophages, dendritic cells, mast cells T232 Low IgGl=3=4>2

As used herein, an Fc region or Fc domain or Fc refers to the C-terminal region of an antibody heavy chain that mediates binding of an immunoglobulin to host tissues or factors, including binding to Fc receptors located on various cells of the immune system (eg, effector cells), or with the first component (C1q) of the classical complement system. Thus, the Fc region contains the antibody constant region, with the exception of the first domain of the immunoglobulin constant region (eg, CH1 or CL). In the IgG, IgA, and IgD antibody isotypes, the Fc region contains CH2 and CH3 constant domains in each of the two antibody heavy chains; The Fc regions in IgM and IgE contain three heavy chain constant domains (CH domains 2-4) in each polypeptide chain. In the case of IgG, the Fc region contains the immunoglobulin domains Cy2 and Cy3, and a hinge between Cy1 and Cy2. Although the boundaries of the immunoglobulin heavy chain Fc region may differ, it has been determined that the human IgG heavy chain Fc region typically extends from an amino acid residue at position C226 or P230 (or an amino acid between

- 12 041306 for these two amino acids) to the carboxy-terminus of the heavy chain, where the numbering corresponds to the EU index, as in Kabat. (Kabat et al. (1991) Sequences of Proteins of Immunological Interest, National Institutes of Health, Bethesda, MD). The Cm domain of the human IgG Fc region extends from approximately amino acid 231 to approximately amino acid 340, while the C _n3 domain is located on the C-terminal side of the C _n2 domain in the Fc region, i.e., it extends from approximately amino acid 341 up to approximately amino acid 447 in IgG (including the C-terminal lysine). As used herein, an Fc region can be a native Fc sequence, including any allotype or Fc variant (eg, a non-naturally occurring Fc). Fc may also refer to this region alone or in the context of an Fc-containing protein polypeptide, such as a binding protein containing an Fc region, also referred to as an Fc fusion protein (eg, an antibody or an immunoadhesin).

Fc region (fragment capable of crystallization) or Fc domain'' or Fc refers to the C-terminal region of an antibody heavy chain that mediates binding of an immunoglobulin to host tissues or factors, including binding to Fc receptors located on various cells of the immune system ( for example, effector cells) or with the first component (C1q) of the classical complement system. Thus, the Fc region contains the antibody constant region, with the exception of the first domain of the immunoglobulin constant region (eg, CH1 or CL). In the IgG, IgA, and IgD antibody isotypes, the Fc region contains two identical protein fragments derived from the second (CH2) and third (CH3) constant domains of the two antibody heavy chains. In the case of IgM and IgE antibody isotypes, the Fc regions contain three heavy chain constant domains (CH domains 2-4) in each polypeptide chain. In the case of IgG, the Fc region contains the CH2 and CH3 immunoglobulin domains and a hinge between the CH1 and CH2 domains. Although the boundaries of the immunoglobulin heavy chain Fc region may differ, it is herein defined that the human IgG heavy chain Fc region typically extends from the amino acid residue at position D221 for IgG1, V222 for IgG2, L221 for IgG3, and P224 for IgG4 to the carboxy terminus of the heavy chain. , where the numbering corresponds to the EU index, as in Kabat (Kabat, et al., 1991). The CH2 domain of the human IgG Fc region extends from amino acid 237 to amino acid 340, and the CH3 domain is located on the C-terminal side of the CH2 domain in the Fc region, i.e., it extends from amino acid 341 to amino acid 447 or 446 (if C-terminal there is no lysine residue), or 445 (if there are no C-terminal glycine and lysine residues) in IgG. As used herein, an Fc region may be a native Fc sequence, including any allotype or Fc variant (eg, a non-naturally occurring Fc). Fc may also refer to this region alone or in the context of a polypeptide of an Fc-containing protein, such as a binding protein containing an Fc region, also referred to as an Fc fusion protein (eg, antibody or immunoadhesin).

The native sequence of the Fc region or the native sequence of the Fc contains an amino acid sequence that is identical to the amino acid sequence of the naturally occurring Fc region. Native sequences of the human Fc region include the native sequence of the human IgG1 Fc region (eg, SEQ ID NO: 206); native sequence of the Fc region of human IgG2; native sequence of the Fc region of human IgG3; and the native sequence of the human IgG4 Fc region, as well as natural variants thereof. Native sequence Fcs include various Fc region allotypes. (See, for example, Jefferis et al. (2009) mAbs 1:1).

The term epitope or antigenic determinant refers to a site on an antigen (eg, huICOS) to which an immunoglobulin or antibody specifically binds. Epitopes can be formed by consecutive amino acids (usually a linear epitope) or by non-consecutive amino acids adjacent due to the folding of the protein into a tertiary structure (usually a conformational epitope). Epitopes formed from consecutive amino acids are usually, but not always, retained by exposure to denaturing solvents, while epitopes formed by folding a protein into a tertiary structure are usually lost by treatment with denaturing solvents. An epitope typically comprises at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, or 22 amino acids in a unique spatial conformation .

The term epitope mapping refers to the process of identifying molecular determinants on an antigen involved in antibody-antigen recognition. Methods for determining which epitopes are bound by a given antibody are well known in the art and include, for example, immunoblotting and immunoprecipitation assays, where overlapping or contiguous peptides (eg, from ICOS) are tested for reactivity with a given antibody (eg, anti-ICOS antibody); x-ray diffraction crystallography; mutational antigen analysis, two-dimensional nuclear magnetic resonance; yeast display; and hydrogen-deuterium exchange mass spectrometry (FIDX-MS) (see, for example, Epitope Mapping Protocols in Methods in Molecular Biology, Vol. 66, G. E. Morris, Ed. (1996)).

The term binds to the same epitope, with reference to two or more antibodies, means that the antibodies bind to the same segment of amino acid residues, as defined by this

- 13 041306 in a different way. Techniques for determining whether antibodies with the same epitope on ICOS bind to the antibodies described herein include, for example, epitope mapping methods such as X-ray analyzes of crystals of antigen:antibody complexes, which provide atomic resolution of the epitope, and FIDX -MS. Other methods control antibody binding to fragments of the antigen (eg, proteolytic fragments) or mutated variations of the antigen, where loss of binding due to modification of an amino acid residue in the antigen sequence is often considered indicative of an epitope component, such as alanine-scanning mutagenesis (Cunningham & Wells (1985) Science 244:1081) or yeast display of mutant variants of the target sequence (see Example 16). In addition, computational combinatorial methods for epitope mapping can also be used. These methods rely on the ability of the antibody of interest to isolate, by affinity, individual short peptides from combinatorial phage display peptide libraries. Antibodies having the same VH and V _L or the same CDR1, 2 and 3 sequences are expected to bind to the same epitope.

Antibodies that compete with another antibody for binding to a target refer to antibodies that inhibit (partially or completely) the binding of another antibody to a target. Do two antibodies compete with each other for binding to the target, i.e. whether and to what extent one antibody can inhibit the binding of another antibody to a target can be determined using known competitive binding experiments, for example, BIACORE® Surface Plasmon Resonance (SPR) analysis. In some embodiments, the antibody competes with another antibody and inhibits the binding of the other antibody to the target by at least 50%, 60%, 70%, 80%, 90%, or 100%. The level of inhibition or competition may vary depending on which antibody is a blocking antibody (ie, an antibody that, when combined with an antigen, blocks another immunological reaction with the antigen). Competition binding assays can be performed as described, for example, in Ed Harlow and David Lane, Cold Spring Harb. Protocol. 2006; doi: 10.1101/pdb.prot4277 or Chapter 11 of Using Antibodies by Ed Harlow and David Lane, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, USA 1999. Competing antibodies bind to the same epitope, an overlapping epitope, or adjacent epitopes (eg, as evidenced by steric hindrance). Two antibodies cross-compete if the antibodies block each other in both directions by at least 50%, i.e., regardless of whether one or the other antibody contacts the antigen first in a competitive experiment.

Competitive binding assays to determine if two antibodies compete or cross-compete for binding include competition for binding to T cells expressing ICOS, for example by flow cytometry. Other methods include: surface plasmon resonance (SPR) (e.g., BIACORE®), solid phase direct or indirect radioimmunoassay (RIA), solid phase direct or indirect enzyme immunoassay (EIA), competitive sandwich assay (see Stahli et al., Methods in Enzymology 9:242 (1983)); biotin-avidin direct labeled solid phase EIA (see Kirkland et al., J. Immunol. 137:3614 (1986)); direct labeling solid phase analysis, direct labeling sandwich solid phase analysis (see Harlow and Lane, Antibodies: A Laboratory Manual, Cold Spring Harbor Press (1988)); direct labeling solid phase RIA using the I125 label (see Morel et al., Mol. Immunol. 25(1):7 (1988)); biotin-avidin direct labeled solid phase EIA (Cheung et al., Virology 176:546 (1990)); and direct labeled RIA (Moldenhauer et al., Scand. J. Immunol. 32:77 (1990)).

As used herein, the terms specific binding, selective binding, selectively binds, and specifically binds refer to the binding of an antibody to an epitope on a predetermined antigen. Typically, an antibody: (1) binds to an equilibrium dissociation constant (KD) of less than about 10 ^-7 M, such as less than about 10 ^-8 M, 10 ^-9 M, or 10 ^-10 M, or even lower at detection using, for example, Surface Plasmon Resonance (SPR) technology on a BIACORE® 2000 Surface Plasmon Resonance instrument using a predetermined antigen, such as recombinant human ICOS as analyte and antibody as ligand, or Scatchard assay of binding of antibody to antigen-positive cells, and (2) binds to the predetermined antigen with an affinity that is at least twice its binding affinity to a non-specific antigen (eg, BSA, casein) other than the predetermined antigen or a closely related antigen. Accordingly, an antibody that specifically binds to human ICOS refers to an antibody that binds to soluble or cell-bound human ICOS with a KD of 10 ^-7 M or less, such as less than about 10 ^-8 M, 10 ^-9 M, or 10 ^-10 M, or even lower. An antibody that cross-reacts with cynomolgus monkey ICOS refers to an antibody that binds to cynomolgus monkey ICOS with a KD of 10-7 M or less, such as less than about 10 ^-8 M, 10 ^-9 M, or 10 ^-10 M or even lower.

The term _kassoc or _ka as used herein refers to the association rate constant of a particular antibody-antigen interaction, while the term k _du cc or kd as

- 14 041306 used herein refers to the dissociation rate constant of a particular antibody-antigen interaction. The term Kd, as used herein, refers to the equilibrium dissociation constant, which is obtained from the ratio of kd and k _a (ie k _d /k _a ) and expressed as molar concentration (M). Antibody K _D values can be determined using methods well known in the art. Available methods for determining the K _D of an antibody include Biolayer Interferometry (BLI) analysis, e.g. using the ForteBio Octet RED device, Surface Plasmon Resonance, preferably using a biosensor system such as the BIACORE® Surface Plasmon Resonance method, or flow cytometry, and Scatchard analysis.

The term EC50 in the context of an in vitro or in vivo assay using an antibody or antigen-binding fragment thereof refers to the concentration of the antibody or antigen-binding fragment thereof that induces a response that is 50% of the maximum response, i. is in the middle between the maximum response and the baseline.

The term binds to immobilized ICOS refers to the ability of an antibody described herein to bind to an ICOS, eg expressed on a cell surface or attached to a solid support.

The term cross-react, as used herein, refers to the ability of an antibody described herein to bind to ICOS from other species. For example, an antibody described herein that binds to human ICOS may also bind ICOS from another species (eg, cynomolgus monkey ICOS). As used herein, cross-reactivity can be measured by determining specific reactivity with purified antigen in binding assays (eg, SPR, ELISA) or binding or otherwise functional interaction with cells physiologically expressing ICOS. Methods for determining cross-reactivity include standard binding assays as described herein, for example by SPR analysis using a BIACORE® 2000 SPR instrument (Biacore AB, Uppsala, Sweden) or by flow cytometry methods.

The term receptor occupancy or receptor occupancy, as used herein, refers to the amount of an agonist antibody (eg, anti-ICOS antibodies described herein) that is bound to an immunostimulatory receptor (eg, human ICOS). % receptor occupancy or % receptor occupancy can be calculated using the following formula: ([AMFI Tested] / [AMFI Total])x00. AMFI (change in mean fluorescence units) is calculated by subtracting the MFI of the isotype control antibody background staining from the MFI from the bound agonist antibody. The overall level of the receptor is determined by adding a saturating amount of agonist antibody to determine the maximum expression, and hence the MFI, of a particular immunostimulatory receptor. An alternative way to calculate total receptor expression is to use an antibody against the same immunostimulatory receptor that does not compete with the agonist antibody for which receptor occupancy is being calculated.

As used herein, the term natural refers to a substance present in nature that has not been specifically modified by man. For example, a polypeptide or polynucleotide sequence present in an organism (including viruses) that can be isolated from a natural source and that has not been intentionally modified by humans in a laboratory is natural.

The term polypeptide refers to a chain containing at least two consecutively linked amino acid residues, with no upper limit on chain length. One or more amino acid residues in a protein may contain a modification such as, but not limited to, glycosylation, phosphorylation, or a disulfide bond. The protein may contain one or more polypeptides.

The term nucleic acid molecule as used herein is intended to include a DNA molecule and an RNA molecule. The nucleic acid molecule may be single or double stranded and may be cDNA.

The term cDNA refers to a non-natural nucleic acid molecule that has been created or derived from mRNA, i.e. non-coding regions have been removed.

The term mRNA or messenger RNA is a nucleic acid intermediate that determines the amino acid sequence of a polypeptide during translation.

As used herein, the term conservative sequence modifications refers to amino acid modifications that do not significantly affect or alter the binding characteristics of an antibody that contains that amino acid sequence. Such conservative modifications include amino acid substitutions, additions and deletions. Modifications can be introduced into an antibody of the present invention by standard methods known in the art, such as site-directed mutagenesis and polymerase chain reaction (PCR) mediated mutagenesis. Conservative amino acid substitutions are

- 15 041306 kimi, in which the amino acid residue is replaced by an amino acid residue having a similar side chain. Families of amino acid residues with similar side chains have been defined in the art. These families include amino acids with basic side chains (e.g., lysine, arginine, histidine), acidic side chains (e.g., aspartic acid, glutamic acid), uncharged polar side chains (e.g., glycine, asparagine, glutamine, serine, threonine, tyrosine, cysteine, tryptophan), non-polar side chains (eg, alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine), beta-branched side chains (eg, threonine, valine, isoleucine), and aromatic side chains (eg, tyrosine, phenylalanine , tryptophan, histidine). Thus, one or more amino acid residues in the CDR regions of an antibody of the invention can be replaced with other amino acid residues from the same side chain family, and the altered antibody is analyzed for conserved function (i.e., the functions described herein above) using the functional assays described herein. In some embodiments, a precalculated non-essential amino acid residue in anti-ICOS antibodies is replaced with another amino acid residue from the family with the same side chain type. Methods for identifying nucleotide and amino acid conservative substitutions that do not preclude antigen binding are well known in the art (see, for example, Brummell et al., Biochem. 32: 1180-1187 (1993); Kobayashi et al. Protein Eng. 12( 10): 879-884 (1999); and Burks et al. Proc. Natl. Acad. Sci. USA 94:412-417 (1997)).

In relation to nucleic acids, the term significant homology indicates that, when optimally aligned and compared with corresponding nucleotide insertions or deletions, two nucleic acids or their assigned sequences are identical in at least about 80% of the nucleotides, at least about 90 to 95% or at least about 98 to 99.5% nucleotides. Alternatively, significant homology exists when the segments hybridize under selective hybridization conditions to the complement of the nucleotide chain.

With respect to polypeptides, the term significant homology indicates that two polypeptides or their assigned sequences, when optimally aligned and compared, are identical with corresponding amino acid insertions or deletions in at least about 80% of amino acids, at least about 90 to 95% or at least about 98 to 99.5% amino acids.

The percent identity between two sequences is a function of the number of identical positions shared by the sequences when the sequences are optimally aligned (i.e., % homology = number of identical positions/total number of positions x100), with the optimal alignment determined by considering the number of gaps and the length of each gap, which must be entered to optimally align the two sequences. Comparing sequences and determining percent identity between two sequences can be done using a mathematical algorithm, as described below.

The percent identity between two nucleotide sequences can be determined, for example, using the GAP program in the GCG software package, using the nwsgapdna.cmp matrix and the gap weight (gap weight) 40, 50, 60, 70 or 80, and the length weight (length weight) 1, 2, 3, 4, 5, or 6. Percent identity between two nucleotide or amino acid sequences can also be determined using the algorithm of E. Meyers and W. Miller (CABIOS, 4:11-17 (1989)), which was included in the ALIGN program (version 2.0) using the PAM120 weight residue table, a gap length penalty of 12, and a gap penalty of 4. In addition, the percent identity between two amino acid sequences can be determined using the algorithm of Needleman and Wunsch (J. Mol. Biol. (48):444-453 (1970)), which was included in the GAP program in the GCG software package, using either the Blossum 62 matrix, whether for a RAM250 matrix, and a gap weight of 16, 14, 12, 10, 8, 6, or 4, and a length weight of 1, 2, 3, 4, 5, or 6.

The nucleic acid and protein sequences described herein may additionally be used as a query sequence to perform searches on public databases, for example, to determine related sequences. Such searches can be performed using the NBLAST and XBLAST (version 2.0) programs from Altschul, et al. (1990) J. Mol. Biol. 215:403-10. BLAST nucleotide searches can be performed using the NBLAST program, score=100, wordlength=12, to obtain nucleotide sequences homologous to the nucleic acid molecules described herein. BLAST protein searches can be performed using the XBLAST program, score=50, wordlength=3, to obtain amino acid sequences homologous to the protein molecules described in this application. To obtain gapped alignments for comparative purposes, Gapped BLAST can be used as described in Altschul et al., (1997) Nucleic Acids Res. 25(17):3389-3402. When using the BLAST and Gapped BLAST programs, the default parameters of the respective programs (eg, XBLAST and NBLAST) can be used. (See, for example, National Center for Biotechnology Information

- 16 041306 (NCBI), available at https://www.ncbi.nlm.nih.gov/).

Nucleic acids may be present in whole cells, for example, in a host cell, in a cell lysate, or in partially purified or substantially pure form. The nucleic acid is isolated or appears to be substantially pure when purified from other cellular components or other contaminants, e.g., other cellular nucleic acids (e.g., other parts of the chromosome) or proteins, by standard methods including alkaline/SDS treatment, CsCl separation, column chromatography, agarose gel electrophoresis, and other methods well known in the art. (See, F. Ausubel, et al., ed. Current Protocols in Molecular Biology, Greene Publishing and Wiley Interscience, New York (1987)).

The term vector, as used herein, is intended to mean: a nucleic acid molecule capable of transporting another nucleic acid to which it is associated. One type of vector is a plasmid into which additional DNA segments can be ligated. Another type of vector is a viral vector in which additional DNA segments can be ligated into the viral genome. Some vectors are capable of autonomous replication in the host cell into which they are introduced (eg, bacterial vectors having a bacterial origin of replication and mammalian episomal vectors). Other vectors (eg, non-episomal mammalian vectors) can be integrated into the genome of a host cell when introduced into the host cell, and thus replicate along with the host genome. Moreover, some vectors are capable of directing the expression of genes to which they are operatively linked. Such vectors are referred to herein as recombinant expression vectors (or simply expression vectors). Expression vectors used in recombinant DNA methods include plasmids. As used herein, the terms plasmid and vector can be used interchangeably since plasmid is the most commonly used form of vector. However, other forms of expression vectors are also included, such as viral vectors (eg, replication-defective retroviruses, adenoviruses, and adeno-associated viruses) that perform equivalent functions.

The term host cell or the term recombinant host cell, which are used interchangeably, refers to a cell that contains a nucleic acid that is not naturally present in the cell, and may be a cell into which a recombinant expression vector has been introduced. It should be understood that such terms refer not only to the particular cell in question, but to the progeny of such a cell. Because some modifications may occur in subsequent generations due to mutation or environmental influences, such progeny may not actually be identical to the parent cell, but are still included within the scope of the term host cell as used herein.

The immune response refers to the biological response in the body against foreign agents, such as antigens, that protect the body from those agents and the diseases they cause. The immune response is mediated by the action of an immune system cell (eg, T-lymphocyte, B-lymphocyte, natural killer (NK) cell, macrophage, eosinophil, mast cell, dendritic cell, or neutrophil) and soluble macromolecules produced by any of these cells or by the liver (including antibodies, cytokines, and complement) resulting in the selective targeting, binding to, damage to, or destruction of, and/or clearance from the body of pathogens, pathogen-infected cells or tissues, cancerous or other abnormal cells, or, in cases of autoimmune disease or pathological inflammation , normal cells or tissues, including, for example, human cells or tissues. The immune response includes, for example, activation or inhibition of T cells, such as an effector T cell or Th cell, such as CD4+ or CD8+ T cells, or inhibition or depletion of a Treg cell. A T effector cell (Teff) refers to T cells (eg, CD4+ and CD8+ T cells) with cytolytic activities. T helper (Th) cells secrete cytokines and activate and direct other immune cells, but do not include regulatory T cells (Treg cells). T regulatory (Treg) cells are a subset of T cells that modulate the immune system, maintain tolerance to self antigens, and prevent autoimmune disease. Memory B cells are a subtype of B cells that form in germinal centers after a primary infection and are important for generating an accelerated and more robust antibody-mediated immune response in the event of a reinfection (also known as a secondary immune response). NK cells are a type of cytotoxic lymphocytes critical to the innate immune system. The role of NK cells is similar to that of cytotoxic T cells in the adaptive immune response of vertebrates. NK cells provide rapid responses to virus-infected cells and respond to tumor formation.

As used herein, the term T cell mediated response refers to a response mediated by T cells, eg effector T cells (eg CD8+ cells) and helper T cells (eg CD4+ cells). T cell mediated responses include, for example, T cell cytotoxicity and proliferation.

As used herein, the term cytotoxic T-lymphocyte (CTL) response

- 17 041306 refers to the immune response induced by cytotoxic T cells. CTL responses are mediated, for example, by CD8+ T cells.

An immunomodulator or immunoregulator refers to an agent, such as a component of a signaling pathway, that may be involved in modulating, regulating, or modifying an immune response. Modulation, regulation or modification of the immune response refers to any change in a cell of the immune system or in the activity of such a cell (eg, an effector T cell). Such modulation includes stimulation or suppression of the immune system, which may be manifested by an increase or decrease in the number of different types of cells, an increase or decrease in the activity of these cells, and/or any other changes that may occur within the immune system. Both inhibitory and stimulatory immunomodulators have been identified, some of which may have increased function in the tumor microenvironment. In some embodiments, the implementation of the immunomodulator is located on the surface of T cells. An immunomodulatory target or immunoregulatory target is an immunomodulator that is designed to bind and whose activity is altered by the binding of a substance, agent, fragment, compound or molecule. Immunomodulatory targets include, for example, cell surface receptors (immunomodulatory receptors) and receptor ligands (immunomodulatory ligands).

The term immunotherapy refers to the treatment of a subject suffering from or at risk of developing a disease, or suffering from a relapse of the disease, by a method, including inducing, enhancing the suppression or otherwise modifying the immune response.

Immunostimulatory therapy, or immunostimulatory therapy, refers to therapy that results in an increase (induction or enhancement) of an immune response in a subject, such as in the treatment of cancer.

Potentiating an endogenous immune response means increasing the efficacy or potency of an existing immune response in a subject. Such an increase in efficacy and potency can be achieved, for example, by overcoming mechanisms that suppress the host's endogenous immune response, or by stimulating mechanisms that enhance the host's endogenous immune response.

As used herein, the term bound refers to the association of two or more molecules. The bond may be covalent or non-covalent. The link may also be genetic (i.e., recombinantly fused). Such linkages can be achieved using a wide variety of methods known in the art, such as chemical conjugation and recombinant protein production.

As used herein, the term administration refers to the physical administration of a composition containing a therapeutic agent, such as an anti-ICOS antibody, to a subject using any of the various delivery methods and systems known to those skilled in the art. Administration includes, for example, administration to a human patient by others, such as, for example, one or more medical professionals, and self-administration by a human patient. Various routes of administration for the antibodies described herein include intravenous, intraperitoneal, intramuscular, subcutaneous, spinal, or other parenteral routes of administration, such as by injection or infusion. As used herein, the phrase parenteral administration means methods of administration other than enteral and topical administration, such as by injection, and includes, without limitation, intravenous, intraperitoneal, intramuscular, intraarterial, intrathecal, intralymphatic, intralesional, intracapsular, intraorbital, intracardiac, intradermal , transtracheal, subcutaneous, subcuticular, intraarticular, subcapsular, subarachnoid, intraspinal, epidural and intrasternal injection and infusion, as well as in vivo electroporation. Alternatively, an antibody described herein may be administered via a non-parenteral route, eg, topical, epidermal, or mucosal routes of administration, eg, intranasally, orally, vaginally, rectally, sublingually, or topically. The introduction can also be carried out, for example, once, several times and/or for one or more long periods of time.

As used herein, adjuvant or combined administration (co-administration) includes the simultaneous administration of compounds in the same or different dosage forms, or separate administration of the compounds (eg, sequential administration). Thus, the first antibody, such as an anti-ICOS antibody, and the second, third or more antibodies can be administered simultaneously as a single composition. Alternatively, the first and second (or more) antibodies can be formulated for separate administration and administered simultaneously or sequentially. Combination therapy, as used herein, means the administration of two or more therapeutic agents in a coordinated manner and includes, but is not limited to, simultaneous dosing. In particular, combination therapy encompasses both co-administration (e.g., administration of a co-drug or simultaneous administration of separate therapeutic compositions) and serial or sequential administration, provided that the administration of one therapeutic agent is carried out in the same manner as the administration of the other therapeutic agent. For example, one therapeutic

- 18 041306 agent can be entered only after the introduction of another therapeutic agent and allow him to work for a prescribed period of time. (See, for example, Kohrt et al. (2011) Blood 117:2423).

For example, an anti-ICOS antibody can be administered first, followed (eg, immediately after administration) by a second antibody, or vice versa. In one embodiment, the anti-ICOS antibody is administered prior to the administration of the second antibody. In another embodiment, the anti-ICOS antibody is administered, for example, about 30 minutes after the second antibody. Such simultaneous or sequential administration preferably results in both antibodies being simultaneously present in the treated patients.

As used herein, the terms inhibit or block are used interchangeably and encompass both partial and complete inhibition/blocking of at least about 50%, 60%, 70%, 80%, 90%, 95%, 99%, or 100%, as defined, for example, by methods described in this document.

As used herein, the term cancer refers to a broad group of diseases characterized by the uncontrolled growth of abnormal cells in the body. Unregulated cell growth or division can lead to the formation of malignant tumors or cells that invade adjacent tissues and can metastasize to distant parts of the body via the lymphatic system or bloodstream.

As used herein, the terms treat, treat, and treat refer to any type of intervention or process performed on a subject, or administering an active agent to a subject for the purpose of reversing, alleviating, alleviating, inhibiting, or slowing down or preventing the progression, development, severity, or recurrence of a symptom. , complications, conditions or biochemical parameters associated with the disease. In contrast, prophylaxis or prevention refers to administration to a subject who does not have a disease to prevent the onset of the disease. Treat, administer treatment and treatment does not cover prevention or prevention.

The term effective dose or effective dosage is defined as an amount sufficient to achieve or at least partially achieve the desired effect. A therapeutically effective amount or therapeutically effective dosage of a drug or therapeutic agent is any amount of a drug that, when used alone or in combination with another therapeutic agent, promotes disease regression as evidenced by a reduction in the severity of disease symptoms, an increase in the frequency and duration of asymptomatic periods of the disease, or prevention of impairment or disability due to disease. A prophylactically effective amount or prophylactically effective dosage of a drug is an amount of a drug that, when administered alone or in combination with another therapeutic agent to a subject at risk of developing a disease or at risk of relapse of a disease, prevents the development or recurrence of a disease. The ability of a therapeutic agent to promote regression of a disease or a prophylactic agent to prevent the development or recurrence of a disease can be assessed using a number of methods known to the practitioner in the art, for example, in human subjects during clinical trials, in animal model efficacy prediction systems in humans, or by evaluating the activity of the agent in in vitro assays.

Administration of effective amounts of an anti-ICOS antibody alone, or an anti-ICOS antibody in combination with an anti-PD-1 antibody, in combination with an anti-PD-L1 antibody, or in combination with an anti-CTLA-4 antibody, according to any one of The methods provided herein can provide at least one therapeutic effect, including, for example, reduction in tumor growth or size, reduction in the number of metastatic lesions that appear over time, complete remission, partial remission, or stable disease. For example, treatments provide a higher percentage of patients with clinical benefit rate (CBR = complete remission (CR) + partial remission (PR) + stable disease course (SD) >6 months) compared to that which was achieved without administration of anti-ICOS antibody, or compared to that achieved with administration of any of the combined antibodies, e.g., an increase in the percentage of patients with improvement in clinical parameters is about 20, 20, 30, 40, 50, 60, 70, 80 % or more.

By way of example, an anticancer agent is a drug that slows down the progression of cancer or promotes the regression of cancer in a subject. In some embodiments, a therapeutically effective amount of the drug promotes regression of the cancer to the point of eliminating the cancer. Promotes regression of cancer means that the administration of an effective amount of the drug, alone or in combination with an anticancer agent, results in a decrease in tumor growth or size, tumor necrosis, a decrease in the severity of at least one symptom of the disease, an increase in the frequency and duration of asymptomatic periods of the disease, preventing impairment or disability due to disease involvement or otherwise ameliorating the symptoms of a disease in a patient. Pharmacological efficacy

- 19 041306 (Pharmacological effectiveness), effectiveness (drug under realistic conditions) or efficacy (drug under ideal conditions) refers to the ability of a drug to promote regression of cancer in a patient. Physiological safety refers to an acceptably low level of toxicity or other undesirable physiological effects at the cellular, organ and/or organism level (adverse events) resulting from the administration of a drug.

As an example, for the treatment of tumors, a therapeutically effective amount or dosage of a drug inhibits tumor cell growth by at least about 20%, at least about 30%, at least about 40%, at least about 50%, at least about 60%, at least about 70%, at least about 80% compared to untreated subjects, or at least about 90%. In some embodiments, a therapeutically effective amount or dosage of a drug completely inhibits cell growth or tumor growth, i. inhibits cell growth or tumor growth by 100%. The ability of a compound, including an antibody, to inhibit tumor growth can be assessed using the assays described herein. Alternatively, this property of the composition can be assessed by examining the ability of the compound to inhibit cell growth; such inhibition can be measured in vitro using assays known to those skilled in the art. In some embodiments, inhibition of tumor growth may not occur immediately after treatment, but may occur only after a certain period of time or after repeated administration. In other embodiments described herein, tumor regression is observed and continues for at least about 20 days, at least about 30 days, at least about 40 days, at least about 50 days, or at least about 60 days. days or longer.

As used herein, the terms fixed dose, constant dose, and constant-fixed dose are used interchangeably and refer to the dose that is administered to a patient regardless of the patient's weight or body surface area. Thus, a fixed or constant dose is not presented as a dose in mg/kg, but rather as an absolute amount of a therapeutic agent.

As used herein, the term weight-based dose or dosing means that the dose administered to a patient is based on the weight of the patient. For example, when 3 mg/kg of anti-ICOS antibody is required for a 60 kg patient, an appropriate amount of anti-ICOS antibody (ie, 180 mg) can be calculated and applied for administration.

The term patient includes human and other mammalian subjects who are receiving therapeutic or prophylactic treatment.

The term subject refers to any human or non-human animal. For example, the methods and compositions disclosed herein can be used to treat a subject having cancer. A non-human animal includes all vertebrates, eg, mammals and non-mammals, including non-human primates, sheep, dogs, cows, chickens, amphibians, reptiles, etc. In one embodiment, the subject is a human subject.

As used herein, the term a or an before an object refers to one or more such objects, unless otherwise indicated; for example, a nucleotide sequence refers to one or more nucleotide sequences. In fact, the terms a or an, one or more, and at least one may be used interchangeably herein.

As used herein, and/or should be considered as a specific disclosure of each of the two specified features or components, with or without the other. Thus, the term and/or used in an expression such as A and/or B includes A and B, A or B, only A and only B. Similarly, the term and/or used in an expression such as A , B and/or C, covers each of the following: A, B and C; A, B or C; A or C; A or B; B or C; A and C; and A and B; B and C; only A; only in; and only S.

It should be understood that in all cases where aspects are described herein in terms of containing language, other similar aspects described in terms of consisting of and/or consisting essentially of are also provided.

Units of measurement, prefixes and symbols are indicated in the form adopted in the International System of Units (Systeme International de Unites, SI). Numeric ranges include the numbers that define the range. Unless otherwise indicated, nucleotide sequences are written from left to right in the 5' to 3' direction. Amino acid sequences are written from left to right in the direction from amino to carboxy terminus.

As used herein, the term approximately means about, about, about, or about. When a term is roughly used in conjunction with a numerical range, it modifies that range by extending its boundaries above and below the specified numerical values. As a rule, the term approximately can change the numerical value above and below the specified value with a fluctuation, for example, 10%, up or down (higher or lower).

- 20 041306

The headings provided herein are not intended to limit the various aspects of the invention that may exist from the description as a whole. Thus, the terms defined immediately below are defined in more detail based on the description as a whole. Various aspects described in this document are described in more detail in the following subsections.

I. Anti-ICOS antibodies

The present invention in some embodiments discloses antibodies, such as fully human antibodies, with desired functions or properties. Described herein are anti-human ICOS agonistic antibodies (anti-huICOS) having desirable properties for use as therapeutic agents in the treatment of diseases such as cancer. These properties include one or more of the ability to bind to human ICOS with high affinity, acceptably low immunogenicity in human subjects, the ability to bind preferentially to FcyRIIb (a specific type of IgG Fc receptor), and the absence of undesirable features in the sequence that would reduce chemical stability. antibodies. The antibodies of the invention are also useful, for example, in the diagnosis of cancer and other disorders associated with the expression and/or activity of ICOS.

Anti-ICOS antibodies, disclosed herein as an amino acid sequence, bind to specific epitopes on human ICOS as described in the examples.

Anti-huICOS antibodies with specific functional properties

Antibodies according to the invention are characterized by specific functional features or properties. For example, antibodies specifically bind to human ICOS with high affinity. In some embodiments, the antibodies specifically bind to a site on the ICOS to which ICOS-L binds. Binding to human ICOS can be assessed using one or more techniques well known in the art. For example, in some embodiments, an antibody may be tested by a flow cytometric assay, in which the antibody is interacted with a cell line expressing human ICOS, such as CHO cells transfected to express human ICOS on its cell surface. Additionally or alternatively, antibody binding, including binding kinetics (eg, K _D value) can be tested in Biacore binding assays. Yet other suitable binding assays include ELISA assays using, for example, recombinant human ICOS protein.

In one embodiment, an antibody of the invention, or an antigen-binding portion thereof, binds to an ICOS protein with a K _D value of 5x10-8 M or less, binds to an ICOS protein with a K _D value of ^2x10'8 M or less, binds to the ICOS protein with a KD value of 5x10'9 M or less, binds to an ICOS protein with a KD value of ^4x10'9 M or less, binds to an ICOS protein with a KD value of 3x10'9 M or less, binds to an ICOS protein with a value of A Kd of 2x10'9 M or less binds to an ICOS protein with a Kd value of ^1x10'9 M or less, binds to an ICOS protein with a KD value of ^5x10'10 M or less, or binds to an ICOS protein with a KD value , equal to 1x10' ¹⁰ M or less.

In another embodiment, the antibody binds to one or more SIFDPPPFKVTL (SEQ ID NO: 203) residues of human ICOS. In another embodiment, the antibody binds to an epitope that contains the amino acid residues SIFDPPPFKVTL (SEQ ID NO: 203) of human ICOS. In another embodiment, the antigen-binding portion of the antibody binds to an epitope that contains the amino acid residues SIFDPPPFKVTL (SEQ ID NO: 203) of human ICOS.

In another embodiment, the antibody binds to human human ICOS and stimulates an immune response, such as an antigen-specific T cell response. The ability of an antibody to stimulate an immune response can be tested by measuring tumor growth, for example in an in vivo tumor graft model (see, for example, examples 6, 7, 8 and 9).

In another embodiment, the antibody, or antigen-binding portion thereof, binds to a human ICOS and exhibits at least one of the following properties:

(a) binding to one or more residues within SIFDPPPFKVTL (SEQ ID NO: 203) of human ICOS;

(b) binding to the same epitope on human ICOS as antibody ICOS.33, 17C4, 9D5, 3E8, 1D7, or 2644;

(c) competing for binding to human ICOS with an ICOS.33, 17C4, 9D5, 3E8, 1D7 or 2644 antibody;

(d) a decrease in ADCC activity compared to a control IgG1 antibody;

(e) increasing the specificity of binding to the FcyRIIb receptor;

(f) blocking binding of an ICOS ligand (ICOS-L) to human ICOS;

(g) blocking the interaction of human ICOS with human ICOS-L;

(h) binding to human, cynomolgus, mouse and rat ICOS;

- 21 041306 (i) binding to activated human and cynomolgus monkey T cells;

(j) binds to human T cells with an EC50 value of approximately 0.7 nM and cynomolgus monkey T cells with an EC50 value of approximately 0.3 nM;

(k) no binding to human CD28 or human CTLA-4;

(l) activating at least one primary T lymphocyte, such as a CD4+ Teff cell, a Tfh cell, and a Treg cell;

(m) induces proliferation and production of interferon-gamma (IFN-γ) in CD25-CD4+ T cells with an EC50 value of approximately 0.01-0.16 nM in an in vitro CHOOKT3-CD32A co-culture assay;

(n) protein kinase B (pAkt) induction in an in vitro primary T cell signaling assay with an EC50 value of approximately 30 nM;

(o) induces IFN-γ production in CD25-CD4+ T cells with an EC50 value of approximately 0.002-0.4 nM in staphylococcal enterotoxin B in a co-culture assay of CD25-CD4+ T cells and B cells;

(p) induction of interleukin-10 (IL-10) production in response to staphylococcal enterotoxin B in a co-culture assay of Tfh cells and naive B cells;

(q) inducing a greater increase in proliferation of CD3-stimulated Teff cells compared to CD45RA+ Treg cells and CD45RO+ Treg cells in an in vitro assay;

(r) an increase in proliferation of Teff cells compared to CD45RA+ Treg cells (eg, when the increase in proliferation is greater in CD45RA+ Treg cells compared to CD45RO+ Treg cells);

(s) reduction of Teff suppression by Treg cells;

(t) while about 10 μg/ml of antibody does not increase the production of cytokines in a whole blood cellular assay;

(u) increasing secretion of at least one of IL-10 and IFN-g by Tfh cells in vitro; and/or (v) stimulation of ICOS-mediated signaling.

In another embodiment, the isolated antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g., human ICOS-L) with human ICOS, and induces proliferation and production of interferon -gamma (IFN-γ) in CD25-CD4+ T cells with an EC50 value of about 0.083 nM in an in vitro CHO-OKT3-CD32A co-culture assay. In another embodiment, the isolated antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g., human ICOS-L) with human ICOS and induces proliferation and production of interferon- gamma (IFN-γ) in CD25-CD4+ T cells with an EC50 value of about 0.01-0.1 nM in an in vitro CHO-OKT3-CD32A co-culture assay.

In one aspect, an isolated antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g., human ICOS-L) with human ICOS and induces the production of IFN-γ in CD25 - CD4+ T cells with an EC50 value of about 0.2 nM in staphylococcal enterotoxin B in a co-culture assay of CD25-CD4+ T cells and B cells. In another aspect, the isolated antibody is a humanized isolated antibody (or an antigen-binding portion thereof) that binds to human ICOS and blocks the binding and/or interaction of an ICOS ligand (e.g., human ICOS-L) with human ICOS and induces the production of IFN-γ in CD25 - CD4+ T cells with an EC50 value of about 0.01-0.1 nM in staphylococcal enterotoxin B in a co-culture assay of CD25-CD4+ T cells and B cells.

In some embodiments, the antibodies of the invention include humanized and fully human monoclonal antibodies. In other embodiments, the antibodies are, for example, chimeric monoclonal antibodies.

Monoclonal antibodies ICOS.33 IgGlf S267E, 17C4, 9D5, 3E8, 1D7 and 2644

In some embodiments, the antibodies of the invention are humanized and human monoclonal antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7, and 2644, which are isolated and structurally characterized as described in the following examples. The amino acid sequences of V _H IcOs.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644, and the amino acid sequences of V _L ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644 are shown in Table. 35.

Given that each of these antibodies can bind to human ICOS, the V _H and V _L sequences can be mixed and matched in mixed and matched pairs to create other anti-huICOS binding molecules of the invention. In some embodiments, when the V _H and V _L strands are mixed and matched, the V _H sequence from a particular V _H /V _L pair is replaced with a structurally similar V _H sequence. Similarly, in some embodiments, the V _L sequence in a particular V _H /V _L pair is replaced by a structurally similar sequence

- 22 041306 V _L .

Accordingly, in one aspect, this disclosure provides an isolated monoclonal antibody, or antigen-binding portion thereof, comprising:

(a) a heavy chain variable region comprising the amino acid sequence set forth in SEQ ID NO: 5, 16, 24, 32, 40, or 186; and (b) a light chain variable region comprising the amino acid sequence set forth in SEQ ID NO: 6, 17, 25, 33, 41, 48, or 189;

wherein the antibody specifically binds to human ICOS. In some embodiments, combinations of heavy and light chain variable regions include:

(a) a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 5 and a light chain variable region comprising the amino acid sequence of SEQ ID NO: 6;

(b) a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 16 and a light chain variable region comprising the amino acid sequence of SEQ ID NO: 17;

(c) a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 24 and a light chain variable region comprising the amino acid sequence of SEQ ID NO: 25;

(d) a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 32 and a light chain variable region comprising the amino acid sequence of SEQ ID NO: 33;

(e) a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 40 and a light chain variable region comprising the amino acid sequence of SEQ ID NO: 41;

(f) a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 40 and a light chain variable region comprising the amino acid sequence of SEQ ID NO: 48; or (g) a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 186 and a light chain variable region comprising the amino acid sequence of SEQ ID NO: 189.

In another aspect, this disclosure provides antibodies that comprise the heavy chain and light chain CDR1, CDR2, and CDR3 of ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7, and 2644, or combinations thereof. The amino acid sequences of CDR1 VH antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644 are shown in SEQ ID NO: 9,

18, 26, 34, 42 and 191, respectively. The amino acid sequences of CDR2 VH antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644 are shown in SEQ ID NO: 10,

19, 27, 35, 43 and 192, respectively. The amino acid sequences of CDR3 VH antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644 are shown in SEQ ID NO: 11,

20, 28, 36, 44 and 193, respectively. The amino acid sequences of CDR1 V _L antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644 are shown in SEQ ID NO: 12,

21, 29, 37, 49 and 194, respectively. The amino acid sequences of CDR2 V _L antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644 are shown in SEQ ID NO: 14,

22, 30, 38, 50 and 195, respectively. The amino acid sequences of CDR3 VL antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644 are shown in SEQ ID NO: 15,

23, 31, 39, 51 and 196, respectively. CDR regions were isolated using the Kabat system (Kabat et al., 1991) Sequences of Proteins of Immunological Interest, Fifth Edition, U.S. Department of Health and Human Services, NIH Publication No. 91-3242).

Given that each of these antibodies can bind to human ICOS, and that antigen-binding specificity is provided primarily by the CDR1, CDR2 and CDR3 regions, the CDR1, CDR2 and CDR3 V _H sequences and the CDR1, CDR2 and CDR3 V _L sequences can be mixed and mixed and matched, (i.e. CDRs from different antibodies can be mixed and matched, although each antibody must contain CDR1, CDR2 and CDR3 V _H , and CDR1, CDR2 and CDR3 V _L ) to create other anti -huICOS binding molecules according to the invention. Binding to ICOS of such mixed and matched antibodies can be tested using the binding assays described herein, including in the examples (eg, ELISA assays, Biacore®). In some embodiments, when VH CDR sequences are mixed and matched, the CDR1, CDR2, and/or CDR3 sequence from a particular VH sequence is replaced with structurally similar CDR sequence(s). Similarly, in some embodiments, when VL CDR sequences are mixed and matched, the CDR1, CDR2, and/or CDR3 sequence from a particular VL sequence is preferably replaced with structurally similar CDR sequence(s). One skilled in the art will appreciate that new V _H and V _L sequences can be created by substituting one or more of the sequences

- 23 041306 VH and/or V _L CDR regions with structurally similar sequences from the CDR sequences disclosed herein for monoclonal antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and 2644.

Accordingly, in another aspect, this disclosure provides an isolated monoclonal antibody, or antigen-binding portion thereof, comprising:

(a) a heavy chain variable region CDR1 comprising the amino acid sequence set forth in SEQ ID NOs: 9, 18, 26, 34, 42, or 191;

(b) a heavy chain variable region CDR2 comprising the amino acid sequence set forth in SEQ ID NOs: 10, 19, 27, 35, 43, or 192;

(c) a heavy chain variable region CDR3 comprising the amino acid sequence set forth in SEQ ID NOs: 11, 20, 28, 36, 44, or 193;

(d) a light chain variable region CDR1 comprising the amino acid sequence set forth in SEQ ID NOs: 12, 21, 29, 37, 49, or 194;

(e) a light chain variable region CDR2 comprising the amino acid sequence set forth in SEQ ID NOs: 14, 22, 30, 38, 50, or 195; and (f) a light chain variable region CDR3 comprising the amino acid sequence set forth in SEQ ID NOs: 15, 23, 31, 39, 51, or 196;

wherein the antibody specifically binds to human ICOS. In one embodiment, the antibody contains:

(a) a heavy chain variable region CDR1 comprising SEQ ID NO: 9;

(b) a heavy chain variable region CDR2 comprising SEQ ID NO: 10;

(c) a heavy chain variable region CDR3 comprising SEQ ID NO: 11;

(d) a light chain variable region CDR1 comprising SEQ ID NO: 12;

(e) a light chain variable region CDR2 comprising SEQ ID NO: 14; and (f) a light chain variable region CDR3 comprising SEQ ID NO: 15.

In another embodiment, the antibody contains:

(a) a heavy chain variable region CDR1 comprising SEQ ID NO: 18;

(b) a heavy chain variable region CDR2 comprising SEQ ID NO: 19;

(c) a heavy chain variable region CDR3 comprising SEQ ID NO: 20;

(d) a light chain variable region CDR1 comprising SEQ ID NO: 21;

(e) a light chain variable region CDR2 comprising SEQ ID NO: 22; and (f) a light chain variable region CDR3 comprising SEQ ID NO: 23.

In another embodiment, the antibody contains:

(a) a heavy chain variable region CDR1 comprising SEQ ID NO: 26;

(b) a heavy chain variable region CDR2 comprising SEQ ID NO: 27;

(c) a heavy chain variable region CDR3 comprising SEQ ID NO: 28;

(d) a light chain variable region CDR1 comprising SEQ ID NO: 29;

(e) a light chain variable region CDR2 comprising SEQ ID NO: 30; and (f) a light chain variable region CDR3 comprising SEQ ID NO: 31.

In another embodiment, the antibody contains:

(a) a heavy chain variable region CDR1 comprising SEQ ID NO: 34;

(b) a heavy chain variable region CDR2 comprising SEQ ID NO: 35;

(c) a heavy chain variable region CDR3 containing SEQ ID NO: 36;

(d) a light chain variable region CDR1 comprising SEQ ID NO: 37;

(e) a light chain variable region CDR2 comprising SEQ ID NO: 38; and (f) a light chain variable region CDR3 comprising SEQ ID NO: 39.

In another embodiment, the antibody contains:

(a) a heavy chain variable region CDR1 comprising SEQ ID NO: 42;

(b) a heavy chain variable region CDR2 comprising SEQ ID NO: 43;

(c) a heavy chain variable region CDR3 containing SEQ ID NO: 44;

(d) a light chain variable region CDR1 comprising SEQ ID NO: 49;

(e) a light chain variable region CDR2 comprising SEQ ID NO: 50; and (f) a light chain variable region CDR3 comprising SEQ ID NO: 51.

In another embodiment, the antibody contains:

(a) a heavy chain variable region CDR1 comprising SEQ ID NO: 191;

(b) a heavy chain variable region CDR2 comprising SEQ ID NO: 192;

(c) a heavy chain variable region CDR3 comprising SEQ ID NO: 193;

(d) a light chain variable region CDR1 comprising SEQ ID NO: 194;

(e) a light chain variable region CDR2 comprising SEQ ID NO: 195; and (f) a light chain variable region CDR3 comprising SEQ ID NO: 196.

It is well known in the art that the CDR3 domain, independently of the CDR1 and/or CDR2 domain(s), taken alone, can determine the binding specificity of an antibody for recognition.

- 24 041306 th antigen, and that multiple antibodies can be predictably generated having the same binding specificity based on a common CDR3 sequence (See, for example, Klimka et al., British J. of Cancer 83(2):252-260 (2000)). Accordingly, the present disclosure provides monoclonal antibodies comprising one or more heavy and/or light chain CDR3 domains of an antibody derived from a human or non-human animal, wherein the monoclonal antibody is capable of specifically binding to human ICOS. In some aspects, the present disclosure provides monoclonal antibodies comprising one or more heavy and/or light chain CDR3 domains of a non-human antibody, such as a mouse or rat antibody, wherein the monoclonal antibody is capable of specifically binding to human ICOS. In some embodiments, such antibodies of the invention comprising one or more non-human antibody heavy and/or light chain CDR3 domains are (a) capable of competing for binding with the corresponding parent non-human antibody; (b) retain the functional characteristics of the corresponding parental non-human antibody; (c) bind to the same epitope as the corresponding parent non-human antibody; and/or (d) have a binding affinity similar to that of the corresponding parent non-human antibody.

In other aspects, the present disclosure provides monoclonal antibodies comprising one or more heavy and/or light chain CDR3 domains of a human antibody, such as, for example, a human antibody derived from a non-human animal, wherein the human antibody is capable of specifically binding to a human ICOS. . In other aspects, the present disclosure provides monoclonal antibodies comprising one or more heavy and/or light chain CDR3 domains of a first human antibody, such as, for example, a human antibody derived from a non-human animal, wherein the first human antibody is capable of specifically binding to human ICOS, wherein the CDR3 domain from the first human antibody replaces the CDR3 domain in the human antibody lacking binding specificity for ICOS to form a second human antibody capable of specifically binding to human ICOS. In some embodiments, such inventive antibodies comprising one or more heavy and/or light chain CDR3 domains of a first human antibody are (a) capable of competing for binding with the corresponding parent non-human antibody; (b) retain the functional characteristics of the corresponding parental non-human antibody; (c) bind to the same epitope as the corresponding parent non-human antibody; and/or (d) have a binding affinity similar to that of the corresponding parental non-human antibody.

The present invention also provides anti-huICOS antibodies comprising the novel variable domain sequences disclosed herein and constant domains with modified Fc regions having increased affinity for FcyRIIb relative to their affinity for other Fc receptors. In some embodiments, such anti-huICOS agonist antibodies with enhanced FcγRIIb specificity exhibit superior efficacy in the treatment of cancer. In other embodiments, such anti-huICOS agonist antibodies with enhanced FcγRIIb specificity exhibit superior efficacy in the treatment of various disorders, such as cancer. Without intending to be bound by a mechanistic theory, such FcγRIIb-specific anti-ICOS agonist monoclonal antibodies may exhibit enhanced adjuvant effects by increasing dendritic cell maturation, thus stimulating the proliferation and activation of cytotoxic CD8+ T cells, resulting in an enhanced antitumor response. Without intending to be bound by any theory, FcR-mediated signal enhancement of anti-ICOS agonist antibodies due to increased receptor clustering or cross-linking in accordance with the present invention may be a major factor influencing therapeutic efficacy. Cross-linking of agonist antibodies against ICOS by FcR binding to the Fc portion of the antibody can increase signal intensity and thereby enhance cell activation.

The relative binding affinity of antibodies for activating (A) versus inhibitory (I) Fc receptors can be expressed as an A/I ratio and generally depends on the structure of the Fco region of an IgG antibody. See WO 2012/087928. Antibodies with increased specificity for binding to the inhibitory FcyRIIb receptor have lower A/I ratios. In some embodiments, the anti-huICOS agonist antibodies described herein have A/I ratios of less than 5, 4, 3, 2, 1, 0.5, 0.3,0,1,0,05.0, 03 or 0.01.

Examples of human IgG1 constant domains containing mutations to increase the specificity of FcyRIIb are described herein and are also provided in the sequence listing. Sequence variants are defined with reference to the human IgG1f constant region sequence set forth in SEQ ID NO: 52 and shown in FIG. 2. The nomenclature regarding the positions (numbering) of mutations in the Fc region corresponds to the EU index, as in Kabat et al., 1991 Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md., which facilitates comparison of Fc sequences at equivalent positions in

- 25 041306 antibodies with different lengths of variable domains. See also Edelman et al. (1969) Proc. Nat'l Acad. sci. (USA) 63:78; WO 2012/130831 (using the same numbering system). In table. 3 provides a summary of Fc sequence variants from which one of skill in the art will readily recognize the corresponding positions in the sequences of the antibodies disclosed herein. SE and SELF variants are described in Chu et al. (2008) Mol. Immunol. 45:3926. Variants P238D, V4, V7, V8, V9, V11 and V12 are described in Mimoto et al. (2013) Protein Engineering Design & Selection 26:589 (for example, in Table 1 in this document).

Table 3 Fc sequence variants

Designation SEQ ID: Sequence Options IgGlf 52 SE 53 S267E SELF 54 S267E L328F P238D 55 P238D V4 56 P238D P271G V4-D270E 57 P238D P271G D270E V7 58 E233D P238D P271G A330R V8 59 G237D P238D H268D P271G V9 60 G237D P238D P271G A330R V9-D270E 61 G237D P238D P271G A330R D270E vh 62 G237D P238D H268D P271G A330R

| V12 I 63 I E233D I G237D | P238D | | H268D | P271G | | A33OR | ~|

Additional Fc sequence variants with increased affinity for FcYRIIb are disclosed in Yu et al. (2013) J. Am. Chem. soc. 135:9723 (and WO 2014/184545) including V262E and V264E, for example for use in combination with S267E and L328F

Antibodies with conservative modifications

In some embodiments, an antibody of the invention comprises a heavy chain variable region comprising the sequences CDR1, CDR2, and CDR3, and a light chain variable region comprising the sequences CDR1, CDR2, and CDR3, wherein one or more of these CDR sequences comprise said amino acid sequences based on antibodies described herein (e.g., ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7, and 2644), or conservative modifications thereof, and the antibodies retain the desired functional properties of the anti-huICOS antibodies of the invention. Those skilled in the art will appreciate that certain conservative sequence modifications can be made that do not interfere with antigen binding. (See, for example, Brummell et al. (1993) Biochem 32:1180-8). Thus, this disclosure provides an isolated monoclonal antibody, or an antigen-binding portion thereof, comprising a heavy chain variable region comprising the sequences CDR1, CDR2, and CDR3, and a light chain variable region comprising the sequences CDR1, CDR2, and CDR3, wherein:

(a) the heavy chain variable region containing the sequence of CDR3 contains the amino acid sequence indicated in SEQ ID NO: 11, 20, 28, 36, 44 or 193, or conservative modifications thereof;

(b) the light chain variable region containing the sequence of CDR3 contains the amino acid sequence indicated in SEQ ID NO: 15, 23, 31, 39, 51 or 196, or conservative modifications thereof; and (c) the antibody, or antigen-binding portion thereof, specifically binds to human ICOS.

Additionally or alternatively, an antibody may have one or more of the functional properties described herein, such as high affinity binding to human ICOS and/or the ability to stimulate antigen-specific T cell responses.

In some embodiments, the heavy chain variable region containing the CDR2 sequence comprises the amino acid sequence shown in SEQ ID NOs: 10, 19, 27, 35, 43, or 192, or conservative modifications thereof; and the light chain variable region containing the CDR2 sequence contains the amino acid sequence indicated in SEQ ID NO: 14, 22, 30, 38, 50, or 195, or conservative modifications thereof. In another embodiment, the heavy chain variable region comprises a CDR1 sequence comprising the amino acid sequence set forth in SEQ ID NOs: 9, 18, 26, 34, 42, or 191, or conservative modifications thereof; and the light chain variable region containing the CDR1 sequence contains the amino acid sequence shown in SEQ ID NO: 12, 21, 29, 37, 49, or 194, or conservative modifications thereof.

In various embodiments, the antibody may be, for example, human antibodies, humanized antibodies, or chimeric antibodies.

Antibodies that bind to the same epitope as anti-huICOS antibodies

In another embodiment, this disclosure provides antibodies that bind to the same epitope on human ICOS as any of the anti-huICOS monoclonal antibodies of the invention (i.e., antibodies that have the ability to cross-compete for binding to human ICOS with any from monoclonal antibodies according to the invention). In some embodiments, monoclonal antibodies ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7, and 2644 are used as a reference antibody for cross-competition studies.

- 26 041306

Such cross-competing antibodies can be identified based on their ability to cross-compete with ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and/or 2644 in standard human ICOS binding assays. For example, standard ELISA assay methods can be used in which a recombinant human ICOS protein is immobilized on a plate, one of the antibodies is labeled with a fluorescent label, and the ability of the unlabeled antibody to compete for binding to the labeled antibody is assessed. Additionally or alternatively, the Biacore assay can be used to assess the ability of antibodies to cross-compete. The ability of the test antibody to inhibit the binding of, for example, ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and/or 2644 to human ICOS indicates that the test antibody can compete with ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and /or 2644 for binding to human ICOS and therefore binds to the same epitope on human ICOS as ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and/or 2644 the same epitope on human ICOS as ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7 and/or 2644 is a humanized or human monoclonal antibody.

As discussed further in Example 16, binding of ICOS.33 IgG1f S267E, 3E8 and 9D5 to human ICOS is mapped to ICOS residues 112-123 (SEQ ID NO: 1) or the amino acid sequence SIFDPPPFKVTL (SEQ ID NO: 203). Accordingly, in one embodiment, the invention provides an anti-huICOS antibody that binds to one or more SIFDPPPFKVTL (SEQ ID NO: 203) residues of human ICOS, for example, as determined by HDX-MS. In another embodiment, the anti-huICOS antibody binds to an epitope containing amino acid residues SIFDPPPFKVTL (SEQ ID NO: 203) of human ICOS.

Such humanized or human monoclonal antibodies may be prepared and isolated as described herein. For example, anti-huICOS antibodies that bind to the same or similar epitopes with the antibodies disclosed herein can be generated using immunization protocols such as those described herein. The resulting antibodies can be screened for high affinity binding to human ICOS. Selected antibodies can then be studied, for example, in yeast display assays, in which huICOS sequence variants are displayed on the surface of yeast cells, or by hydrogen-deuterium exchange experiments to pinpoint the epitope bound by the antibody.

Epitope determinations can be made by any method known in the art. In some embodiments, anti-huICOS antibodies are considered to bind to the same epitope as the anti-huICOS mAb described herein if they contact one or more of the same residues within at least one huICOS region; if they contact the majority of residues within at least one region of huICOS; if they are in contact with the majority of residues within each area of huICOS; if they are in contact with the majority of residues along the entire length of huICOS; if they contact in all of the same different areas of the human ICOS; if they come into contact with all residues in any one area on the human ICOS; or if they contact all the same residues in all of the same areas. Epitopic regions are clusters of residues along, but not necessarily directly adjacent to, within the primary sequence.

Methods for detecting antibodies that bind to the same epitope on huICOS as the antibodies described herein include X-ray analysis of crystals of antigen:antibody complexes, which provides atomic resolution of the epitope. Other methods control antibody binding to antigen fragments or mutated antigen variants, where loss of binding due to amino acid modification in the antigen sequence is considered to be indicative of an epitope component. Methods can also be based on the ability of the antibody of interest to affinity isolate specific short peptides (both in native three-dimensional form and in denatured form) from combinatorial peptide libraries in a phage display format or by protease digestion of a target protein. The peptides are then considered as leading to determine the epitope corresponding to the antibody used to screen the peptide library. Computational algorithms have also been developed for epitope mapping and have been shown to map conformational discontinuous epitopes.

An epitope or region containing an epitope can also be identified by screening for binding to a series of overlapping ICOS-spanning peptides. Alternatively, the method of Jespers et al. can be used. (1994) Biotechnology 12:899 as a guide for selecting antibodies having the same epitope and therefore similar properties to those of the anti-ICOS antibodies described herein. In a phage display method, first the heavy chain of an anti-ICOS antibody is paired with a repertoire of (e.g., human) light chains to select an ICOS binding antibody, and then the new light chain is paired with a repertoire of (e.g., human) heavy chains to select (e.g., , human) ICOS binding antibody having the same epitope or epitope region as the anti-huICOS antibody described herein.

- 27 041306 document. Alternatively, the antibody variants described herein can be generated by mutagenesis of cDNA sequences encoding the heavy and light chains of an antibody.

Also, alanine-scanning mutagenesis as described by Cunningham & Wells (1989) Science 244:1081, or some other form of point mutagenesis of amino acid residues in ICOS (e.g., the yeast display method presented in Example 16) can be used to determine a functional epitope. for anti-ICOS antibodies.

The epitope or epitope region (an epitope region is a region containing an epitope or an overlap with an epitope) bound by a specific antibody can also be determined by assessing the binding of the antibody to peptides containing ICOS fragments. A series of overlapping peptides spanning the sequence of an ICOS (e.g., human ICOS) can be synthesized and screened for binding, e.g., in a direct ELISA, competitive ELISA (where the peptide is evaluated for its ability to prevent binding of an antibody to an ICOS bound to the surface of a well of a plate for microtiter) or on a chip. Such peptide screening methods may not be able to detect some discontinuous functional epitopes, ie. functional epitopes that involve amino acid residues that are not sequential along the primary sequence of the ICOS polypeptide chain.

An epitope can also be identified using mass spectrometry-based protein footprinting, for example, hydrogen-deuterium exchange mass spectrometry (HDX-MS) and fast photochemical oxidation of proteins (FPOP). HDX-MS can be performed, for example, as further described in Wei et al. (2014) Drug Discovery Today 19:95, the methods of which are specifically incorporated herein by reference. FPOP can be performed as described, for example, in Hambley & Gross (2005) J. American Soc. Mass Spectrometry 16:2057, the methods of which are specifically incorporated herein by reference.

The epitope bound by anti-ICOS antibodies can also be determined using structural methods such as X-ray crystal structure determination (e.g. WO 2005/044853), molecular modeling and nuclear magnetic resonance (NMR) spectroscopy, including determination by NMR. H-D exchange rates of labile amide hydrogens in ICOS when free and complexed with the antibody of interest (Zinn-Justin et al. (1992) Biochemistry 31: 11335; Zinn-Justin et al. (1993) Biochemistry 32:6884) .

Unless otherwise indicated, and with reference to the claims, an antibody-bound epitope is an epitope determined using HDX-MS methods. Anti-huICOS antibodies derived from hamster antibodies This document describes examples of chimeric and humanized antibodies that contain CDRs and/or variable regions of heavy and/or light chains of antibodies that have been derived from hamster sequences. The chimeric or humanized antibodies described herein can be derived from the sequence of a monoclonal antibody, such as a mouse or hamster, obtained by various methods known in the art. DNA encoding immunoglobulin heavy and light chains can be obtained from the hybridoma of interest and engineered by standard molecular biology techniques to contain human immunoglobulin sequences. For example, to generate a chimeric antibody, variable regions, such as a mouse or bovine antibody, can be linked to human constant regions using methods known in the art (see, for example, US Pat. No. 4,816,567 to Cabilly et al.). To create a humanized antibody, CDR regions can be inserted into a human framework using methods known in the art (see, for example, US Pat. No. 5,225,539 to Winter and US Pat. .

Anti-ICOS antibodies that bind with high affinity

In some embodiments, the anti-huICOS antibodies of the present invention bind to huICOS with high affinity, similar to the anti-huICOS antibodies disclosed herein, making them effective therapeutic agents. In various embodiments, the anti-huICOS antibodies of the present invention bind to huICOS with a K _D of less than 10, 5, 2, 1 nM, 300, or 100 pM. In other embodiments, anti-huICOS antibodies of the present invention bind to huICOS with a K _D ranging from 2 nM to 100 pM. Standard assays for assessing the binding capacity of antibodies to huICOS include ELISA, RIA, Western blots, biolayer interferometry (BLI), and BIACORE® SPR assays (see Example 10).

Sequence variants of anti-ICOS antibodies

Anti-ICOS antibody sequence variants disclosed herein retain the desired functional properties disclosed herein. CDR regions are described using the Kabat system (Kabat, et al., 1991). In some embodiments, the present invention further provides human or humanized anti-huICOS antibodies comprising CDR sequences that are at least 70%, 75%, 80%, 85%, 90%, or 95%, 96%, 97%, 98%, or 99% identical to the CDR sequences of the antibodies described herein. The present invention also provides anti-huICOS antibodies containing the sequences

- 28 041306 heavy and/or light chain variable domains that are at least 70, 75, 80, 85, 90, 95, 96, 97, 98, or 99% sequence identical to the heavy and/or light chain variable domains antibodies disclosed herein, as well as anti-huICOS antibodies containing full length heavy and/or light chain sequences that are at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% identical to the sequences of the heavy and/or light chains of the antibodies disclosed herein.

II. Engineered and modified antibodies

VH and VL areas

Also provided are engineered and modified antibodies that can be made using an antibody having one or more of the VH and/or V _L sequences disclosed herein as a starting material for constructing a modified antibody that may have altered properties from the parent antibody. In some embodiments, an antibody described herein is engineered by modifying one or more residues in one or both of the variable regions (i.e., VH and/or V _L ), e.g., in one or more CDR regions and/or in one or more wireframe regions. Additionally or alternatively, an antibody described herein is constructed by modifying residues in the constant region(s), eg, to change the effector function(s) of the antibody.

In one embodiment, constructing the variable region includes grafting a CDR. Such grafting is particularly used to humanize non-human anti-ICOS antibodies, e.g., anti-huICOS antibodies that compete for binding with the anti-huICOS antibodies disclosed herein and/or bind to the same epitope as the selected anti- -huICOS antibodies disclosed herein. Antibodies interact with target antigens predominantly through amino acid residues that are located in the CDRs of the heavy and light chains. The CDRs are hypervariable in sequence and/or form structurally defined loops (hypervariable loops). Expression vectors can be designed to include CDR sequences from a specific reference (also called parent) antibody grafted onto framework sequences from another antibody (see, for example, Riechmann L. et al. (1998) Nature 332:323- 327; Jones P. et al. (1986) Nature 321:522-525; Queen C. et al. (1989) Proc. Natl. Acad. See U.S.A. 86:10029-10033; U.S. Patent 5,225,539 to Winter, and US Pat. In some cases, the resulting recombinant antibody has properties that are similar to the parent antibody. The engineered antibody can then be further modified to have properties different from the parent antibody. In other cases, grafting parental CDR sequences onto a framework region revokes some of the characteristics of the parent antibody such that the recombinant antibody no longer has those characteristics. One illustrative characteristic is the binding affinity for an antigen. In such cases, it may be advantageous to further modify the engineered antibody to restore the desired characteristics of the parent antibody.

Such framework sequences can be obtained from publicly available DNA databases or published references that include germline antibody gene sequences. For example, germline DNA sequences for the human heavy and light chain variable region genes can be found in the VBase human germline sequence database and also in Kabat EA, et al., 1991); Tomlinson IM, et al. (1992) The Repertoire of Human Germline V _H Sequences Reveals about Fifty Groups of V _H Segments with Different Hypervariable Loops J. Mol. Biol. 227:776-798; and Cox JPL et al. (1994) A Directory of Human Germ-line V _H Segments Reveals a Strong Bias in their Usage, Eur. J. Immunol. 24:827-836; the contents of each of these materials are expressly incorporated herein by reference.

In some embodiments, framework sequences for use in the antibodies described herein are those sequences that are structurally similar to the framework sequences used in the antibodies described herein. The CDR1, 2 and 3 VH sequences and the CDR1, 2 and 3 V _L sequences can be grafted onto framework regions that have a sequence identical to the sequence found in the germline immunoglobulin gene from which the framework sequence is derived, or the CDR sequences can be grafted onto framework regions that contain up to 20 amino acid substitutions, including conservative amino acid substitutions, compared to germline sequences. For example, it has been found in some cases to be advantageous to mutate residues in framework regions to maintain or enhance the ability of an antibody to bind to an antigen (see, for example, US Pat. Nos. 5,530,101;

The engineered antibodies described herein include antibodies in which modifications have been made to the framework residues in V _H and/or V _L , for example, to improve the properties of the antibody, such as to reduce the immunogenicity of the antibody. For example, one approach is to

- 29 041306 multiple mutation of one or more framework residues into the corresponding germline sequence. More specifically, an antibody that has been somatically mutated may contain framework residues that differ from the germline sequence from which the antibody is derived. Such residues can be identified by comparing the framework sequences of the antibody with the germline sequences from which the antibody is derived. To return the framework region sequences to their germline configuration, somatic mutations can be back-mutated into the germline sequence using, for example, site-directed mutagenesis or PCR-mediated mutagenesis. Such reverse mutated antibodies are also included in this disclosure.

Another type of framework modification involves mutating one or more residues in the framework or even one or more CDR regions to remove T cell epitopes, thereby reducing the potential immunogenicity of the antibody. This approach is also referred to as deimmunization and is described in more detail in US Patent Publication No. 20030153043 by Carr et al.

Another type of variable region modification is mutating amino acid residues in CDR regions to improve one or more binding properties (eg, affinity) of the antibody of interest. Site-directed mutagenesis or PCR-mediated mutagenesis can be performed to introduce mutation(s) and affect antibody binding or other functional property of interest. Preferably, conservative modifications are introduced. Mutations can be amino acid additions, deletions, or substitutions. In some embodiments, no more than one, two, three, four, or five residues in the CDR region are changed.

Methionine residues in the CDRs of antibodies can be oxidized, leading to potential chemical degradation and subsequent loss of antibody activity. Accordingly, anti-ICOS antibodies are also provided herein that have one or more methionine residues in the heavy and/or light chain CDRs replaced with amino acid residues that do not undergo oxidative degradation. Similarly, deamidation sites can be removed from anti-ICOS antibodies, particularly in CDRs. Also provided herein are antibodies in which potential glycosylation sites in the antigen-binding domain are removed to prevent glycosylation that may interfere with antigen binding. See, for example, US Pat. No. 5,714,350.

Antibody masking

In some embodiments, the antibodies disclosed herein are modified to limit their binding to specific cells and/or tissue. In one embodiment, such antibodies comprise a blocking peptide mask that specifically binds to the antigen-binding surface of the antibody and prevents antigen binding. In some embodiments, the mask is linked to each of the binding arms of the antibody via a protease-cleavable linker. See, for example, US Pat. No. 8,518,404 to CytomX. Antibodies with protease-cleavable linkers can be used to treat cancers in which protease levels are significantly increased in the tumor microenvironment compared to non-tumor tissues. Selective cleavage of the cleavable linker in the tumor microenvironment allows dissociation of the masking/blocking peptide, which allows the antigen to bind selectively in the tumor rather than in peripheral tissues where antigen binding can cause undesirable side effects.

In another embodiment, a bivalent binding compound (masking ligand) containing two antigen-binding domains is designed that both binds to antigen-binding surfaces of a (bivalent) antibody and interferes with antigen binding. In one embodiment, the masks of the two binding domains are linked to each other (but not to the antibody) by a cleavable linker, eg, cleavable by a peptidase. (See, for example, WO 2010/077643, Tegopharm Corp.). Masking ligands may contain or be derived from the antigen to which the antibody is intended to bind, or may be formed independently (eg, anti-idiotypic binding fragments). Such masking ligands can be used to treat cancers in which protease levels are significantly increased in the tumor microenvironment compared to non-tumor tissues. Selective cleavage of the cleavable linker in the tumor microenvironment allows the two binding domains to dissociate from each other, reducing the avidity of the antigen-binding surfaces of the antibody. The resulting dissociation of the masking ligand from the antibody allows selective binding of the antigen in the tumor rather than in peripheral tissues where antigen binding can cause undesirable side effects.

Fc and modified Fc regions

In one embodiment, the antibodies described herein may contain Fc regions selected based on the biological activities of the antibody. Salfeld (2007) Nat. Biotechnol. 25:1369. Human IgG immunoglobulins, for example, can be divided into four subclasses: IgG1, IgG2, IgG3 and IgG4. Each of these subclasses contains an Fc region that has a unique profile for binding to one or more Fcγ receptors (activating receptors FcyRI

- 30 041306 (CD64), FcyRIIA, FcyRIIC (CD32a,c); FcyRIIIA and FcyRIIIB (CD16a,b) and the inhibitory receptor FcyRIIB (CD32b)), and with the first complement component (C1q). Human IgG1 and IgG3 bind to all Fcy receptors; IgG2 binds to FcyRIIA _H131 and with lower affinity to FcyRIIA _R131 FcyRIIIA _V158 ; IgG4 binds to FcyRI, FcyRIIA, FcyRIIB, FcyRIIC and FcyRIIIA _V158 ; and the inhibitory receptor FcyRIIB has lower affinity for IgG1, IgG2 and IgG3 than all other Fcy receptors. (Bmhns et al. (2009) Blood 113:3716). Studies have shown that FcyRI does not bind to IgG2 and FcyRIIIB does not bind to IgG2 or IgG4. Id. In general, in terms of ADCC activity, human IgG1 > IgG3 >> IgG4 > IgG2. In some embodiments, an IgG1 constant domain rather than IgG2 or IgG4 may be selected, for example, for use in a therapeutic composition, since ADCC activity is desired. In other embodiments, IgG3 may be selected because activation of FcyRIIIA-expressing NK cells, monocytes, or macrophages is desired. In other embodiments, IgG4 is selected because the antibody is used to desensitize allergic patients. IgG4 can also be selected if it is desired that the antibody lack all effector functions.

The variable regions of an anti-huICOS antibody described herein can be linked (e.g., covalently linked or fused) to an Fc, e.g., an IgG1, IgG2, IgG3, or IgG4 Fc, which can be of any allotype or isoallotype, e.g., for IgG1: G1m, G1m1(a), G1m2(x), G1m3(f), G1m17(z); for IgG2: G2m, G2m23(n); for IgG3: G3m, G3m21(g1), G3m28(g5), G3m11(b0), G3m5(b1), G3m13(b3), G3m14(b4), G3m10(b5), G3ml5(s), G3m16(t), G3m6(c3), G3m24(c5), G3m26(u), G3m27(v). (See, for example, Jefferis et al. (2009) mAbs 1:1). The choice of allotype may be influenced by concerns about potential immunogenicity, for example, to minimize the formation of antibodies against the drug.

In some embodiments, the anti-ICOS antibodies of the present invention have an Fc region that binds or has increased binding to FcyRIIb, resulting in increased agonism. See, for example, WO 2012/087928; Li & Ravetch (2011) Science 333:1030; Wilson et al. (2011) Cancer Cell 19:101; White et al. (2011) J. Immunol. 187:1754. In some embodiments, the variable regions described herein can be linked to Fc variants that increase affinity for the inhibitory FcyRIIb receptor, for example, to increase apoptosis-inducing or adjuvant activity. Li & Ravetch (2012) Proc. Nat'l Acad. sci. (USA) 109:10966; patent application publication 2014/0010812. Such variants provide the antibody with immunomodulatory activities associated with FcyRIIb+ cells, including, for example, B cells and monocytes. In one embodiment, the Fc variants provide selectively increased affinity for FcyRIIb over one or more activating receptors. Such variants may also show increased FcR-mediated cross-linking resulting in increased therapeutic efficacy. Modifications to alter binding to FcyRIIb include one or more modifications, for example, at positions 234, 235, 236, 237, 239, 266, 267, 268, 325, 326, 327, 328, or 332 according to the EU index. Exemplary substitutions for increasing FcyRIIb affinity include, but are not limited to, 234D, 234E, 234F, 234W, 235D, 235F, 235R, 235Y, 236D, 236N, 237D, 237N, 239D, 239E, 266M, 267D, 268DE, 268DE, 268D 327D, 327E, 328F, 328W, 328Y and 332E. Exemplary replacements include 235Y, 236D, 239D, 266M, 267E, 268D, 268E, 328F, 328W, and 328Y. Other Fc variants to increase FcyRIIb binding include 235Y-267E, 236D-267E, 239D-268D, 239D-267E, 267E-268D, 267E-268E and 267E-328F. In particular, variants S267E, G236D, S239D, L328F and I332E, including the dual human IgG1 variant S267E-L328F, are of particular importance in specifically increasing affinity for the inhibitory FcyRIIb receptor. Chu et al. (2008) Mol. Immunol. 45:3926; patent application publication 2006/024298; WO 2012/087928. Increased specificity for FcyRIIb (as opposed to FcyRIIa _R131 ) can be obtained by adding the P238D substitution and other mutations (Mimoto et al. (2013) Protein. Eng. Des. & Selection 26:589; WO 2012/1152410) as well as V262E and V264E (Yu et al. (2013) J. Am. Chem. Soc. 135:9723 and WO 2014/184545.

He-IgG2 heavy chain constant domains with IgG2 hinge regions

In some embodiments, the anti-ICOS antibodies described herein exhibit increased agonist activity at least in part due to modifications that increase FcyRIIb binding and/or FcyRIIb specificity. An alternative approach is to design an Fc region to provide FcyR-independent enhancement of agonism. Examples of antibodies to other targets with modified IgG2 domains that provide such enhanced agonism are described in WO 2015/145360 and White et al. (2015) Cancer Cell 27:138, the disclosures of which are incorporated herein by reference in their entirety. In particular, the disulfide bonds are positioned to close the antibody into a more compact h2B conformation, resulting in increased agonism. The resulting enhanced agonism is FcyR independent. (See White et al. (2015) Cancer Cell 27:138). Such FcyR-independent agonism is preferred in the treatment of certain tumors, such as those with negligible coli.

- 31 041306 the number of cells expressing FcyR-+ (for example, a small number of NK cells or macrophages). In one embodiment, anti-ICOS antibodies are provided comprising the CDR or variable domain sequences disclosed herein linked to non-hIgG2 heavy chain constant regions (e.g., IgG1) having hIgG2 hinge regions, or variants thereof, including domain sequence variants CH1. In one embodiment, these IgG2 hinge antibodies exhibit increased agonism compared to fully constant heavy chain IgG1 antibodies, and the increased agonism is independent of FcY receptor mediated cross-linking. In some embodiments, these anti-ICOS IgG2 hinge antibodies retain substantially unchanged antigen-binding affinity. Also provided herein are methods for enhancing the FcYR-independent agonism of non-IgG2 anti-ICOS antibodies comprising replacing a non-IgG2 hinge with an IgG2 hinge. In some embodiments, the modified heavy chain constant region comprises an IgG2 isotype hinge (IgG2 hinge) and a CH1, CP2, and CH3 domain, wherein at least one of the CH1, CH2, and CH3 domains is not an IgG2 isotype. The IgG2 hinge can be a wild-type human IgG2 hinge (eg, ERKCCVECPPCPAPPVAG as set forth in SEQ ID NO: 96) or a variant thereof that also confers enhanced agonistic activity. In some embodiments, such variants of the IgG2 hinge retain the same stiffness or hardness as the wild-type IgG2 hinge. Hinge stiffness can be determined, for example, by computer simulation, electron microscopy, spectroscopy such as nuclear magnetic resonance (NMR), X-ray crystallography (B-factors), or ultracentrifugation to analytically study the sedimentation rate (AUC) to measure or compare the gyration radius of antibodies containing a hinge. In some embodiments, human IgG2 hinge variants comprise substitution(s) for one or more of the four cysteine residues (ie, C219, C220, C226, and C229), for example, with serine. In one embodiment, the IgG2 hinge variant contains a C219S mutation (eg, ERKSCVECPPCPAPPVAG as set forth in SEQ ID NO: 110). Other variants of the IgG2 hinge contain the C220S, C226S or C229S mutation, any of which can be combined with the C219S mutation).

The IgG2 loop variant may also contain non-IgG2 hinge sequence elements (chimeric hinge). In some embodiments, the stiffness of the chimeric hinge is at least similar to that of the wild-type IgG2 hinge. For example, in one embodiment, a variant IgG2 loop contains a wild-type IgG1 lower hinge (See Table 2).

In table. 4 below shows examples of human IgG2 hinge heavy chain constant region sequences that differ in the isotype origin of the CH1, CH2, and CH3 domains. As used herein, an IgG2 antibody hinge refers not only to antibodies containing hinge regions derived from IgG2, but also CH1 regions derived from IgG2 CH1. An asterisk (*) in the table. 4 means that the specified domain may be of any isotype or may be completely absent. In some embodiments, the modified heavy chain constant region contains a variant CH1 domain, for example, including the A114C and/or T173C mutations. The modified heavy chain constant region may also contain a variant CH2 domain, for example including the A330S and/or P331S mutations.

Table 4. Constructions of constant . human IgG2 hinge heavy chain regions

CH1 Hinge CH2 CH3 * IgG2 * * IgGl IgG2 * * IgG2 IgG2 * * * IgG2 IgGl * * IgG2 IgG2 * * IgG2 * IgGl * IgG2 * IgG2 IgGl IgG2 IgGl * IgGl IgG2 IgG2 * IgG2 IgG2 IgGl * IgG2 IgG2 IgG2 * IgGl IgG2 * IgGl IgGl IgG2 * IgG2

- 32 041306

IgG2 IgG2 * IgGl IgG2 IgG2 * IgG2 * IgG2 IgGl IgGl * IgG2 IgGl IgG2 * IgG2 IgG2 IgGl * IgG2 IgG2 IgG2 IgGl IgG2 IgGl IgGl IgGl IgG2 IgGl IgG2 IgGl IgG2 IgG2 IgGl IgGl IgG2 IgG2 IgG2 IgG2 IgG2 IgGl IgGl IgG2 IgG2 IgGl IgG2 IgG2 IgG2 IgG2 IgGl IgG2 IgG2 IgG2 IgG2

Examples of constant domains of antibodies containing combinations of IgG2 CH1 and hinge sequences with sequences from other isotypes and selected amino acid substitutions are shown in Table 1 below. 5.

Table 5. Examples of human IgG2 hinge heavy chain constant regions

Design SEQID NO: Description IgGlf 104 wild-type IgGlf IgGl. If 109 standard inert IgGl.If IgG2.3 105 form IgG2 A(C219S) IgG2.5 108 form IgG2B(C13IS) IgG2.3Gl-KH 107 CHI, upper hinge and lower hinge/upper CH2 IgG2.3, all other IgGlf IgG2.5Gl-KH 116 CH1, upper hinge and lower hinge/upper CH2 IgG2.5, all other IgGlf IgG2.3Gl-AY 106 CH1 and upper hinge IgG2.3, all other IgGlf IgG2.5Gl-AY 115 CH1 and upper hinge IgG2.5, all others IgGlf IgGl-G2.3Gl-KH 119 CHI IgGl, upper hinge and lower hinge/upper CH2 IgG2.3, all other IgGlf IgGl-G2.3Gl-AY 118 CHI IgGl, upper hinge IgG2.3, all other IgGlf IgG2.3Gl.lf-KH 110 CH1, lower hinge and lower hinge/upper CH2 IgG2.3, all other IgGl.If IgG2.5Gl.lf-KH 114 CH1, top hinge and bottom hinge/top CH2 IgG2.5, all other IgGl.If IgGl-deltaTHT 111 IgGl with TNT sequence removed from the hinge IgG2.3-plusTHT 112 IgG2.3 with TNT sequence (from IgGl) added to the hinge IgG2.5-plusTHT 117 IgG2.5 with TNT sequence (from IgGl) added to the hinge IgG2.3-plusGGG 113 IgG2.3 with flexible GGG sequence added to the hinge

Additional specific examples of antibody constant domains containing combinations of IgG2 CH1 and hinge sequences with sequences from other isotypes and selected amino acid substitutions are shown in Table 1. 6 below.

- 33 041306

Table 6 Additional examples of human IgG2 hinge heavy chain constant regions

Design SEQID NO: Description G2-G1-G1-G1 120 CHI domain of IgG2, with all other IgGl. Also, Cys>Ser mutant to reduce potential disulfide heterogeneity G2.5-G1-G1-G1 121 G1-G2.3-G2-G2 122 CH1 domain of IgGl with all other IgG2.3 Gl-KRGEGSSNLF 123 Swap CH1 regions in IgGl with IgG2 regions, either alone or together. Gl-KRGEGS 124 Gl-SNLF 125 IgGl-ITNDRTPR 126 Gl-SNLFPR 127 G2-RKEGSGNSFL 128 Swap CH1 regions in IgG2 with IgGl regions, either alone or together. G2-RKEGSG 129 G2-NSFL 130 IgG2-TIDNTRRP 131 G2-NSFLRP 132 G1-G1-G2-G1-AY 133 IgGl with IgG2 CH2 domain residues G1-G1-G2-G1-KH 134 G2-G2.3-G1-G2-KH 135 IgG2 with IgGl CH2 domain residues G2.5-G2.3-G1-G2-KH 136 G2-G2.3-G1-G2-AY 137 G2.5-G2.3-G1-G2-AY 138 G1-G2.3-G1-G1-KH 139 Swap hinge regions between IgGl and IgG2. G2-G1-G2-G2-AY 140 G2.5-G1-G2-G2-AY 141 G1-G2-G1-G1-AY 142 G2-G1-G2-G2-KH 143 G2.5-G1-G2-G2-KH 144 IgGl-deltaHinge 145 Hinge truncations IgG2-deltaHinge 146 IgG2.5-deltaHinge 147 IgGl-deltaG237 148 IgG2-plusG237 149 IgG2.4 150 Other IgG2.3/4 151

Anti-ICOS antibodies, including antibodies containing the CDR and/or variable domain sequences disclosed herein, may include IgG2 hinge constant domain sequences disclosed herein, for example, to enhance FcYR-independent agonist activity. Examples of such IgG2 hinge constant domains include those disclosed in Table 1. 5 (SEQ ID NOs: 104-108 and 110-119) and Table 6 (SEQ ID NOs: 120-151) as well as those disclosed in SEQ ID NOs: 101-108.

Prolongation of the half-life

In some embodiments, the anti-ICOS antibody is modified to increase its biological half-life, eg, the serum half-life of the antibody. Various approaches are known in the art. For example, the half-life of an antibody can be extended by increasing the binding affinity of the Fc region for FcRn. In one embodiment, the antibody is altered within the CH1 or CL region to contain a recycling receptor binding epitope taken from two loops of the CH2 domain in the IgG Fc region, as described in US Pat. Other exemplary Fc variants that increase FcRn binding and/or improve pharmacokinetic properties include substitutions at positions 259, 308 and 434, including, for example, 2591, 308F, 428L, 428M, 434S, 434H, 434F, 434Y and 434M. Other variants that increase Fc binding to FcRn include: 250E, 250Q, 428L, 428F, 250Q/428L (Hinton et al., 2004, J. Biol. Chem. 279(8): 6213-6216, Hinton et al. 2006 Journal of Immunology 176:346-356), 256A, 272A, 305A, 307A, 311A, 312A, 378Q, 380A, 382A, 434A (Shields et al., Journal of Biological Chemistry, 2001, 276(9):6591- 6604), 252F, 252Y, 252W, 254T, 256Q, 256E, 256D, 433R, 434F, 434Y, 252Y/254T/256E,

- 34 041306

433K/434F/436H (Dall'Acqua et al. Journal of Immunology, 2002, 169:5171-5180, Dall'Acqua et al, 2006, Journal of Biological Chemistry 281:23514-23524). See US Pat. No. 8,367,805.

Modification of certain conserved residues in an IgG Fc (I253, H310, Q311, H433, N434), such as the N434A variant (Yeung et al. (2009) J. Immunol. 182:7663), has been proposed as a way to increase FcRn affinity, thereby increasing the half-life of the antibody in the circulation. (See, for example, WO 98/023289). The combined Fc variant containing M428L and N434S has been shown to increase FcRn binding and increase serum half-life up to five times. (Zalevsky et al. (2010) Nat. Biotechnol. 28:157). The combined Fc variant containing modifications T307A, E380A and N434A also prolongs the half-life of IgG1 antibodies. (Petkova et al. (2006) Int. Immunol. 18:1759). In addition, combination Fc variants containing the M252Y-M428L, M428L-N434H, M428L-N434F, M428L-N434Y, M428L-N434A, M428L-N434M, and M428L-N434S variants have also been shown to prolong half-life. (WO 2009/086320).

In addition, the combined Fc variant containing M252Y, S254T and T256E increased the half-life by almost 4 times. (Dall'Acqua et al. (2006) J. Biol. Chem. 281:23514). An IgG1-associated modification that increased FcRn affinity but decreased pH dependence (M252Y-S254TT256E-H433K-N434F) was used to generate an IgG1 construct (MST-HN Abdeg) for use as a competitor to prevent binding of other antibodies with FcRn, which leads to an increase in the clearance of such other antibodies, which can act as endogenous IgG (for example, in the case of an autoimmune reaction) or another exogenous (therapeutic) monoclonal antibody. (Vaccaro et al. (2005) Nat. Biotechnol. 23:1283; WO 2006/130834).

Other modifications to increase FcRn binding are described in Yeung et al. (2010) J. Immunol. 182:7663-7671; 6277375; 6821505; WO 97/34631; WO2002/060919.

In some embodiments, hybrid IgG isotypes can be used to increase FcRn binding and potentially increase half-life. For example, an IgG1/IgG3 hybrid variant can be constructed by substituting positions in IgG1, in the CH2 and/or CH3 regions, with amino acids from IgG3 at positions where the two isotypes differ. Thus, a hybrid IgG antibody variant can be constructed that contains one or more substitutions, for example, 274Q, 276K, 300F, 339T, 356E, 358M, 384S, 392N, 397M, 422I, 435R, and 436F. In other embodiments described herein, an IgG1/IgG2 fusion variant can be constructed by replacing positions in IgG2, in the CH2 and/or CH3 region, with amino acids from IgG1 at positions where the two isotypes differ. Thus, a fusion variant of an IgG antibody can be constructed that contains one or more substitutions, for example one or more of the following amino acid substitutions: 233E, 234L, 235L, 236G (with reference to the glycine insertion at position 236) and 327A. See US Pat. No. 8,629,113. A fusion of IgG1/IgG2/IgG4 sequences has been generated, which is expected to increase serum half-life and improve expression. U.S. Patent No. 7,867,491 (sequence number 18 therein).

The serum half-life of the antibodies described herein can also be increased by pegylation. The antibody can be pegylated, for example, to increase the biological (eg, serum) half-life of the antibody. To pegylate an antibody, the antibody or fragment thereof is typically reacted with a polyethylene glycol (PEG), such as a reactive ester or aldehyde derivative of PEG, under conditions in which one or more PEG moieties are attached to the antibody or antibody fragment. Preferably, the pegylation is carried out by an acylation reaction or an alkylation reaction with a reactive PEG molecule (or similar reactive water-soluble polymer).

As used herein, the term polyethylene glycol includes any of the forms of PEG that have been used to derivatize other proteins, such as mono(C1-C10) alkoxy or aryloxy polyethylene glycol, or polyethylene glycol maleimide. In some embodiments, the antibody to be pegylated is an aglycosylated antibody. Methods for pegylation of proteins are known in the art and can be applied to the antibodies described herein. (See, for example, EP 0154316 Nishimura et al. and EP 0401384 Ishikawa et al.

In some cases, it may be desirable to decrease the half-life of an antibody rather than increase it. In some embodiments, the antibodies described herein include modifications to reduce their half-life. Modifications such as I253A (Hornick et al. (2000) J. Nucl. Med. 41:355) and H435A/R, 1253a or H31OA (Kim et al. (2000) Em. J. Immunol. 29:2819) in Human IgG1 Fcs can decrease FcRn binding, thereby decreasing half-life (increasing clearance) for use in situations where rapid clearance is preferred, such as in medical imaging. (See also Kenanova et al. (2005) Cancer Res. 65:622). Other means of increasing clearance include formatting the antigen-binding domains of the present invention as antibody fragments that do not bind to FcRn, such as Fab fragments. Such a modification can, for example, reduce the circulating half-life of an antibody from weeks to hours. Selective PEGylation of antibody fragments

- 35 041306 can then be used to increase the half-life of antibody fragments, if necessary. (Chapman et al. (1999) Nat. Biotechnol. 17: 780). Antibody fragments can also be fused to human serum albumin, for example in a fusion protein construct, to increase half-life. (Yeh et al. (1992) Proc. Nat'l Acad. Sci. 89:1904). Alternatively, bispecific antibodies can be constructed with a first antigen-binding domain of the present invention and a second antigen-binding domain that binds to human serum albumin (HSA). (See WO 2009/127691 and patent references therein). Alternatively, specific polypeptide sequences can be added to antibody fragments to increase half-life, such as XTEN polypeptide sequences. (Schellenberger et al. (2009) Nat. Biotechnol. 27:1186; Int'l Pat. Appl. Pub. WO 2010/091122).

Additional Fc options

In some embodiments, when using an IgG4 constant domain, it may be preferable to include an S228P substitution that mimics a hinge sequence in IgG1 and thereby stabilizes IgG4 molecules, for example, reducing Fab arm exchange between the therapeutic antibody and endogenous IgG4 in the patient being treated. (Labrijn et al. (2009) Nat. Biotechnol. 27:767; Reddy et al. (2000) J. Immunol. 164:1925).

A potential protease cleavage site in the hinge region of IgG1 constructs can be eliminated by modifications to D221G and K222S, which improves antibody stability. (WO 2014/043344).

The affinities and binding properties of the Fc variant for its ligands (Fc receptors) can be determined by various in vitro assay methods (e.g., biochemical or immunological assays) known in the art, including, but not limited to, equilibrium methods (e.g., enzyme-linked immunosorbent assay). (ELISA)), or radioimmunoassay (RIA)), or kinetic assays (e.g., BIACORE® SPR assay), as well as other methods such as indirect binding assays, competitive inhibition tests, fluorescence resonance energy transfer (FRET), gel- electrophoresis and chromatography (eg gel filtration). These and other methods may use a label on one or more components of interest and/or use a variety of detection methods, including, but not limited to, chromogenic, fluorescent, luminescent, or isotopic labels. A detailed description of binding affinities and kinetics can be found in Paul, W. E., ed., Fundamental Immunology, 4th Ed., Lippincott-Raven, Philadelphia (1999), which focuses on antibody-immunogenic interactions.

In other embodiments, glycosylation of an antibody is modified to increase or decrease effector function. For example, an aglycosylated antibody that completely lacks effector function can be generated by mutating a conserved asparagine residue at position 297 (eg, N297A), thus abolishing complement and FcyRI binding. (Bolt et al. (1993) Eur. J. Immunol. 23:403. See also Tao & Morrison (1989) J. Immunol. 143:2595 (using N297Q in IgG1 to eliminate glycosylation at position 297)).

Although aglycosylated antibodies usually lack effector functions, mutations can be introduced to restore such function. Aglycosylated antibodies, such as those generated by N297A/C/D/ or H mutations, or produced in systems (eg E. coli) that do not glycosylate proteins, may be further mutated to restore FcyR binding, such as S298G and/ or T299A/G/ or H (WO 2009/079242), or E382V and M428I (Jung et al. (2010) Proc. Nat'l Acad. Sci. (USA) 107:604).

Glycoconstruction can also be used to change the anti-inflammatory properties of the IgG construct by changing the content of α2,6-sialyl groups in carbohydrate chains attached at the Asn297 position of Fc regions, while an increase in the proportion of α2,6-sialylated forms leads to an increase in anti-inflammatory effects. (See Nimmerjahn et al. (2008) Ann. Rev. Immunol. 26:513). Conversely, reducing the proportion of antibodies having α2,6-sialylated carbohydrates may be advantageous in cases where anti-inflammatory properties are not desired. Methods for modifying the content of α2,6-sialylated carbohydrate groups in antibodies, for example by selective purification from α2,6-sialylated forms or by enzymatic modification, are described in US Patent Application Publication 2008/0206246. In other embodiments, the amino acid sequence in the Fc region can be modified to mimic the effect of α2,6 sialylation, for example by including the F241A modification. (WO 2013/095966). III.

Physical properties of antibodies

In some embodiments, the antibodies described herein contain one or more glycosylation sites in the light or heavy chain variable region. Such glycosylation sites can lead to an increase in the immunogenicity of an antibody or a change in the pharmacokinetics of an antibody due to an alteration in antigen binding (Marshall et al. (1972) Ann. Rev. Biochem. 41:673-702; Gala and Morrison (2004) J. Immunol. 172:5489 -94 Wallick et al (1988) J Exp Med 168:1099109 Spiro (2002) Glycobiology 12:43R-56R Parekh et al (1985) Nature 316:452-7 Mimura et al ( 2000) Mol Immunol 37:697-706). Glycosylation is known to occur in motifs containing sequences

- 36 041306 N-X-S/T. In some embodiments, the anti-huICOS antibody does not contain variable region glycosylation. Such antibodies can be generated by selecting antibodies that do not contain a glycosylation motif in the variable region, or by mutating residues within the glycosylation region.

In some embodiments, the antibodies described herein do not contain asparagine isomerism sites. Deamidation of asparagine can occur at NG- or D-G sequences and results in an isoaspartic acid residue that introduces a kink in the polypeptide chain and reduces its stability (known as the isoaspartic acid effect).

In some embodiments, the antibodies described herein have an isoelectric point (pI) in the pH range of 6 to 9.5. In some embodiments, the antibodies described herein have a pI in the pH range of 7 to 9.5 or 6 to 8. Antibodies having a pI within the desired pI range can be generated either by selecting antibodies with a pI in the pH range from a group of candidates, or by mutating the surface charged residues of a particular antibody.

In some embodiments, the antibodies described herein are selected and/or designed to have an initial deployment temperature (T _M1 ) greater than 60°C, greater than 65°C, or greater than 70°C. The melting point of an antibody can be measured using differential scanning calorimetry (Chen et al (2003) Pharm Res 20:1952-60; Ghirlando et al (1999) Immunol Lett. 68:47-52) or circular dichroism (Murray et al. ( 2002) J Chromatogr Sci 40:343-9).

In some embodiments, the antibodies described herein are selected and/or designed to have advantageous degradation properties, such as slow degradation in vitro and/or in vivo. Antibody degradation can be measured using capillary electrophoresis (CE) and MALDI-MS (Alexander A. J and Hughes D. E. (1995) Anal Chem. 67:3626-32). In some embodiments, the antibodies described herein are selected and/or designed to have favorable aggregation properties, e.g., antibodies that exhibit minimal in vitro and/or in vivo aggregation that may elicit an unwanted immune response and/ or altered or adverse pharmacokinetic properties. In some embodiments, the antibodies described herein show aggregation of 25% or less, 20% or less, 15% or less, 10% or less, or 5% or less compared to the aggregation of the parent antibody. Aggregation can be measured in several ways, including size exclusion chromatography (SEC), high performance liquid chromatography (HPLC), and light scattering. IV.

Nucleic acid molecules and recombinant methods

Another aspect described herein relates to nucleic acid molecules that encode the anti-huICOS antibodies described herein. The nucleic acids may be present in whole cells, eg, a host cell, in a cell lysate, or in partially purified or substantially pure form. Nucleic acid is isolated or appears to be substantially pure when purified from other cellular components or other contaminants, e.g., other cellular nucleic acids (e.g., other chromosomal DNA, e.g., chromosomal DNA that is naturally associated with isolated DNA) or proteins, by standard methods including treatment with alkali/SDS, separation in CsCl, column chromatography, restriction enzymes, agarose gel electrophoresis, and other methods well known in the art. (See, F. Ausubel, et al., ed. (1987) Current Protocols in Molecular Biology, Greene Publishing and Wiley Interscience, New York). The nucleic acid described herein may be, for example, DNA or RNA, and may or may not contain an intron sequence. In some embodiments, the nucleic acid is a cDNA molecule.

Nucleic acids described herein can be obtained using standard methods of molecular biology. With respect to antibodies expressed by hybridomas (e.g., hybridomas derived from transgenic mice carrying human immunoglobulin genes, as described further below), cDNAs encoding the light and/or heavy chain antibodies made by the hybridoma can be obtained by standard HPP amplification or cDNA cloning With respect to antibodies derived from an immunoglobulin gene library (eg using phage display methods), the nucleic acid encoding the antibody can be extracted from said library.

Once DNA fragments encoding the V _H and VL segments have been obtained, the DNA fragments can be further manipulated using standard recombinant DNA techniques, for example, to convert variable region genes into full-length antibody chain genes, Fab genes, or scFv genes. . In these manipulations, a VL or VH encoding DNA fragment is operably linked to another DNA fragment encoding a different protein, such as an antibody constant region or a flexible linker. The term operably linked, as used in this context, means that two DNA fragments are connected in such a way that the amino acid sequences

- 37 041306 encoded by these two DNA fragments remain in the reading frame.

An isolated DNA encoding a VH region can be converted into a full length heavy chain gene by operably linking the VH encoding DNA to another DNA molecule encoding the heavy chain constant regions (hinge, CH1, CH2 and/or CH3). Human heavy chain constant region gene sequences are known in the art (see, eg, Kabat, et al., 1991), and DNA fragments containing these regions can be obtained using standard PCR amplification. The heavy chain constant region can be an IgG (IgG1, IgG2, IgG3 or IgG4), IgA, IgE, IgM or IgD constant region, for example an IgG1 region. With respect to the heavy chain Fab gene, the VH-coding DNA may be operably linked to another DNA molecule encoding only the CH1 heavy chain constant region.

An isolated DNA encoding a VL region can be converted into a full length light chain gene (as well as a Fab light chain gene) by operably linking the VL encoding DNA to another DNA molecule encoding the light chain constant region, CL. Human light chain constant region gene sequences are known in the art (see, e.g., Kabat, et al., 1991) Sequences of Proteins of Immunological Interest, Fifth Edition, U.S. Department of Health and Human Services, NTH Publication No. 91-3242), and DNA fragments containing these regions can be obtained by standard PCR amplification. The light chain constant region may be a kappa or lambda constant region.

To obtain the scFv gene, V _H - and VL-encoding DNA fragments are operably linked to another fragment encoding a flexible linker, for example, encoding the amino acid sequence (Gly4-Ser) ₃ such that the VH and V _L sequences can be expressed as continuous single chain protein with V _L and V _H regions connected by a flexible linker (see, for example, Bird et al. (1988) Science 242:423-426; Huston et al. (1988) Proc. Natl. Acad. Sci. USA 85:5879-5883; McCafferty et al. (1990) Nature 348:552-554).

V. Antibody generation

Various antibodies of the present invention, for example those that bind to the same epitope as the selected anti-human ICOS antibodies disclosed herein, can be generated using various known methods such as the standard somatic cell hybridization method described by Kohler and Milstein, Nature 256: 495 (1975). Other methods for producing monoclonal antibodies can also be used, for example, viral or oncogenic transformation of B-lymphocytes, phage display method using a human antibody gene library.

An example of an animal system for producing hybridomas is the murine system. Obtaining a hybridoma in a mouse is a very well established procedure. Immunization protocols and methods for isolating immunized splenocytes for fusion are known in the art. Fusion partners (eg murine myeloma cells) and fusion procedures are also known.

The chimeric or humanized antibodies described herein can be derived from the sequence of a mouse monoclonal antibody prepared as described above. DNA encoding heavy and light chain immunoglobulins can be obtained from a mouse hybridoma of interest and engineered to contain non-mouse immunoglobulin (eg, human) sequences using standard molecular biology techniques. For example, to generate a chimeric antibody, mouse variable regions can be linked to human constant regions using methods known in the art (see, for example, US Pat. No. 4,816,567 to Cabilly et al.). To create a humanized antibody, mouse CDR regions can be inserted into a human framework using methods known in the art (see, for example, U.S. Patent 5,225,539 to Winter and U.S. Patents 5,530,101; 5,585,089; et al).

In one embodiment, the antibodies described herein are human monoclonal antibodies. Such human monoclonal antibodies directed against human ICOS can be generated using transgenic or transchromosomal mice carrying parts of the human immune system rather than the mouse system. These transgenic and transchromosomal mice include mice referred to herein as HuMAb mice and KM mice, respectively, and collectively referred to as human Ig mice.

The HuMAb® mouse (Medarex, Inc.) contains human immunoglobulin gene miniloci that encode random human k immunoglobulin heavy (μ and γ) and light chain sequences, together with targeted mutations that inactivate the endogenous μ and k chain loci (see e.g. , Lonberg, et al (1994) Nature 368(6474): 856-859). Accordingly, these mice exhibit reduced expression of mouse IgM or K, and in response to immunization, the introduced human heavy and light chain transgenes undergo class switching and somatic mutation to produce high affinity human IgGK monoclonal antibodies (Lonberg, N. et al. (1994 ), supra, reviewed by Lonberg, N. (1994) Handbook of Experimental Pharmacology 113:49-101; Lonberg, N. and Huszar, D.

- 38 041306 (1995) Intern. Rev. Immunol. 13: 65-93, and Harding F. and Lonberg, N. (1995) Ann. NY. Acad. sci. 764:536546). The production and use of HuMab mice and the genomic modifications carried by such mice are further described in Taylor, L. et al. (1992) Nucleic Acids Research 20:6287-6295; Chen, J. et al. (1993) International Immunology 5: 647-656; Tuaillon et al. (1993) Proc. Natl. Acad. sci. USA 90:3720-3724; Choi et al. (1993) Nature Genetics 4:117-123; Chen, J. et al. (1993) EMBO J. 12: 821-830; Tuaillon et al. (1994) J. Immunol. 152:2912-2920; Taylor L. et al. (1994) International Immunology 6: 579-591; and Fishwild, D. et al. (1996) Nature Biotechnology 14: 845-851, each of which is expressly incorporated herein by reference in its entirety. (See also US Pat. 03918, WO 93/12227, WO 94/25585, WO 97/13852, WO 98/24884 and WO 99/45962, all to Lonberg and Kay; and PCT Publication WO 01/14424 to Korman et al.)

In some embodiments, the antibodies described herein are grown using a mouse that carries human immunoglobulin sequences on transgenes and transchromosomes, such as a mouse that carries a human heavy chain transgene and a human light chain transchromosome. Such mice, referred to herein as KM mice, are described in detail in PCT Publication WO 02/43478, Ishida et al.

Additionally, alternative transgenic animal systems expressing human immunoglobulin genes are available in the art and can be used to generate the anti-huICOS antibodies described herein. For example, an alternative transgenic system called Xenomouse (Abgenix, Inc.) can be used; such mice are described, for example, in US patents 5939598; 6075181; 6114598; 6150584 and 6162963, Kucherlapati et al.

In addition, alternative transchromosomal animal systems expressing human immunoglobulin genes are available in the art and can be used to generate the anti-ICOS antibodies described herein. For example, mice carrying both a human heavy chain transchromosome and a human light chain transchromosome, referred to as TC mice, can be used; such mice are described in Tomizuka et al. (2000) Proc. Natl. Acad. sci. USA 97:722727. In addition, cows carrying human heavy and light chain transchromosomes have been described in the art (Kuroiwa et al. (2002) Nature Biotechnology 20:889-894) and can be used to generate the anti-huICOS antibodies described herein. .

Additional murine systems described in the art for growing human antibodies, e.g., human anti-huICOS antibodies, include (i) the VelocImmune® mouse (Regeneron Pharmaceuticals, Inc.), in which endogenous murine heavy and light chain variable regions can be replaced by homologous recombination with human heavy and light chain variable regions operably linked to endogenous mouse constant regions such that chimeric antibodies (human V/mouse C) are grown in mice and then subsequently converted into fully human antibodies using standard recombinant DNA techniques ; and (ii) a MeMo® mouse (Merus Biopharmaceuticals, Inc.), wherein the mouse contains unrearranged human heavy chain variable regions but one rearranged conventional human light chain variable region. Such mice and their use in producing antibodies are described, for example, in WO 2009/15777, US 2010/0069614, WO 2011/072204, WO 2011/097603, WO 2011/163311, WO 2011/163314, WO 2012/148873, US 2012 /0070861 and US 2012/0073004.

The human monoclonal antibodies described herein can also be generated using phage display methods to screen human immunoglobulin gene libraries. Such phage display methods for isolating human antibodies are known in the art. (See, for example, US Pat. 6555313; 6582915 and 6593081, Griffiths et al.).

The human monoclonal antibodies described herein can also be made using severe combined immunodeficiency (SCID) mice that have been injected with human immune cells such that the mice produce a human antibody upon immunization. Such mice are described, for example, in US Pat. Nos. 5,476,996 and 5,698,767 to Wilson et al.

Immunizations

To generate fully human antibodies to human ICOS, mice or transgenic or transchromosomal mice containing human immunoglobulin genes (e.g., HCo12, HCo7, or KM mice) can be immunized with a purified or enriched preparation of ICOS antigen and/or cells expressing ICOS, as described for other antigens, for example, in the work of Lonberg et al. (1994) Nature 368(6474): 856-859; Fishwild et al. (1996) Nature Biotechnology 14: 845-851 and WO 98/24884. Alternatively, mice can be immunized with DNA encoding human ICOS. Preferably, mice are 6-16 weeks old after the first infusion. For example, a purified or enriched preparation (eg, 5-50 µg) of a recombinant human antigen

- 39 041306

ICOS can be used for intraperitoneal immunization of mice. In the event that immunization with a purified or enriched ICOS antigen preparation does not produce antibodies, mice can also be immunized with ICOS expressing cells, such as a cell line, to promote immune responses.

HuMAb transgenic mice can be initially immunized intraperitoneally or subcutaneously (SC) with antigen in Ribi adjuvant, followed by IP/SC immunizations every second week (up to a total of 10 weeks) with antigen in Ribi adjuvant. The immune response can be monitored during the immunization protocol, with plasma samples obtained by retroorbital phlebotomy. Plasma can be screened by ELISA and FACS (as described below), and mice with sufficient human anti-ICOS immunoglobulin titers can be used for fusions. Mice can be subjected to intravenous antigen boosting 3 days before sacrifice and removal of the spleen and lymph nodes. Two to three fusions may be performed for each immunization. From 6 to 24 mice can be immunized for each antigen. In some embodiments, HCo7, HCo12, and KM strains are used. In addition, both HCo7 and HCo12 transgenes can be introduced together in a single mouse having two different human heavy chain transgenes (HCo7/HCo12). Obtaining hybridomas producing monoclonal antibodies to ICOS

To obtain hybridomas producing the monoclonal antibodies described herein, splenocytes and/or lymph node cells from immunized mice can be isolated and fused to an appropriate immortalized cell line, such as a mouse myeloma cell line. The resulting hybridomas can be screened for the production of antigen-specific antibodies. For example, single cell suspensions of spleen lymphocytes from immunized mice can be fused with Sp2/0 non-secreting mouse myeloma cells (ATCC, CRL 1581) with 50% PEG. Cells are plated at approximately 2x105 in a flat bottom microtiter plate followed by a two week incubation in selective media containing 10% Clone fetal serum, 18% 653 conditioned medium, 5% origen (IGEN), 4 mM L-glutamine, 1 mM sodium pyruvate, 5 mM HEPES, 0.055 mM 2-mercaptoethanol, 50 U/ml penicillin, 50 mg/ml streptomycin, 50 mg/ml gentamicin and 1X HAT (Sigma). After approximately two weeks, cells can be cultured in medium in which HAT has been replaced with HT. Individual wells can then be screened by ELISA for human monoclonal IgM and IgG antibodies. Once the hybridoma has grown extensively, the medium can be observed, typically after 10-14 days. The antibody-secreting hybridomas can be replated, screened again, and if still positive for human IgG, the monoclonal antibodies can be subcloned at least twice using limiting dilution. Stable subclones can then be cultured in vitro to produce small amounts of antibody in tissue culture growth medium for characterization.

For monoclonal antibody purification, selected hybridomas can be grown in 2 liter centrifuge flasks for monoclonal antibody purification. Supernatants can be filtered and concentrated prior to affinity chromatography using Protein A Sepharose (Pharmacia, Piscataway, NJ). The eluted IgG can be monitored by gel electrophoresis and high performance liquid chromatography to ensure purity. The buffer solution can be replaced with PBS and the concentration can be determined by OD ₂₈₀ using an extinction coefficient of 1.43. Monoclonal antibodies can be divided into aliquots and can be stored at -80°C. VI. Antibody production

Obtaining transfectomes producing monoclonal antibodies to ICOS

Antibodies of the present invention, including both specific antibodies for which sequences are provided, and other related anti-ICOS antibodies can also be produced in the transfectome of a host cell, for example, using a combination of recombinant DNA methods and gene transfection methods well known in the art. technical fields (Morrison, S. (1985) Science 229:1202).

For example, for the expression of antibodies or fragments of these antibodies, DNA encoding partial or full length light and heavy chains can be obtained by standard methods of molecular biology (for example, PCR amplification or cDNA cloning using a hybridoma that expresses the antibody of interest), and such DNA can be inserted into expression vectors such that the genes are operably linked to transcriptional and translational control sequences. As used herein, the term operably linked means that the antibody gene is ligated into the vector such that the transcriptional and translational control sequences in the vector perform their intended function of regulating transcription and translation of the antibody gene. The expression vector and expression control sequences are chosen to be compatible with the expression host cell used. The antibody light chain gene and the antibody heavy chain gene can be inserted into separate vectors, or both genes are inserted into the same expression vector. The antibody genes are inserted into the expression vector(s) by standard methods (e.g.,

- 40 041306 by ligation of complementary restriction sites on the antibody gene fragment and vector, or by ligation of blunt ends if there are no restriction sites). The light and heavy chain variable regions of the antibodies described herein can be used to generate full length antibody genes of any antibody isotype by inserting their expression vectors already encoding the heavy chain constant regions and light chain constant regions of the desired isotype such that the VH segment is operably linked to the C _H segment(s) in the vector, and the VL segment is operably linked to the CL segment in the vector. Additionally or alternatively, the recombinant expression vector may encode a signal peptide that facilitates secretion of the antibody chain from the host cell. The antibody chain gene can be cloned into a vector such that the signal peptide binds in frame to the amino terminus of the antibody chain gene. The signal peptide can be an immunoglobulin signal peptide or a heterologous signal peptide (ie, a signal peptide from a non-immunoglobulin protein).

In addition to antibody chain genes, recombinant expression vectors can carry regulatory sequences that regulate the expression of antibody chain genes in the host cell. The term regulatory sequence includes promoters, enhancers, and other expression control elements (eg, polyadenylation signals) that regulate transcription or translation of antibody chain genes. Such regulatory sequences are described, for example, in Goeddel (Gene Expression Technology. Methods in Enzymology 185, Academic Press, San Diego, CA (1990)). Those skilled in the art will appreciate that the design of the expression vector, including the choice of regulatory sequences, may depend on such factors as the choice of the host cell to be transformed, the level of expression of the desired protein, among other factors. Preferred regulatory sequences for mammalian host cell expression include viral elements that determine high levels of protein expression in mammalian cells, such as promoters and/or enhancers derived from cytomegalovirus (CMV), simian virus 40 (SV40), adenovirus (e.g., basic adenovirus (AdMLP) and polyoma late promoter.Alternatively, non-viral regulatory sequences such as the ubiquitin promoter or the β-globin promoter can be used.In addition, regulatory elements composed of sequences from different sources can be used, such as the SRa promoter system, which contains sequences from the early SV40 promoter, and the long terminal repeat of human T-cell leukemia virus type I 1 (Takebe, Y. et al. (1988) Mol. Cell. Biol. 8:466-472).

In addition to antibody chain genes and regulatory sequences, recombinant expression vectors may carry additional sequences, such as sequences that regulate replication of the vector in host cells (eg, replication origins) and selectable marker genes. The selectable marker gene facilitates the selection of host cells into which the vector has been introduced (see, for example, US Pat. For example, typically a selectable marker gene confers resistance to drugs such as G418, hygromycin, or methotrexate to the host cell into which the vector has been introduced. Examples of selectable marker genes include the dihydrofolate reductase (DHFR) gene (for use in dhfr host cells with methotrexate selection/amplification) and the peo gene (for selection in the presence of G418).

For expression of the light and heavy chains, the expression vector(s) encoding the heavy and light chains are transfected into the host cell by standard methods. The various forms of the term transfection include a wide variety of methods commonly used to introduce exogenous DNA into a prokaryotic or eukaryotic host cell, such as electroporation, calcium phosphate precipitation, DEAE-dextran transfection, and the like. Although it is theoretically possible to express the antibodies described herein in any prokaryotic or eukaryotic host cells, it is most preferred to express the antibodies in eukaryotic cells, and most preferably in mammalian host cells, since such eukaryotic cells and, in particular, cells mammals are more likely than prokaryotic cells to assemble and secrete a correctly folded and immunologically active antibody. Prokaryotic expression of antibody genes has been reported to be inefficient in obtaining high yields of active antibody (Boss, M. A. and Wood, C. R. (1985) Immunology Today 6:12-13). Antibodies of the present invention can also be obtained in glycoengineered yeast strains. (Pichia pastoris. Li et al. (2006) Nat. Biotechnol. 24:210).

Exemplary mammalian host cells for expressing the recombinant antibodies described herein include Chinese hamster ovary (CHO cells) cells (including the dhfr-CHO cells described by Urlaub and Chasin, (1980) Proc. Natl. Acad. Sci. USA 77 :4216-4220 used with a DHFR selectable marker, for example as described by R. J. Kaufman and P. A. Sharp (1982) Mol Biol 759:601-621), NSO myeloma cells, COS cells and SP2 cells. In particular for use with NSO myeloma cells, another exemplary expression system is the GS gene expression system described in WO 87/04462, WO 89/01036 and EP 338841.

- 41 041306 coding vectors encoding antibody genes into mammalian host cells, antibodies are obtained by culturing the host cells for a time sufficient to ensure expression of the antibody in the host cells or, more preferably, secretion of the antibody into the culture medium in which the cells are grown -hosts. Antibodies can be isolated from the culture medium using standard protein purification methods.

The N- and C-terminus of the polypeptide chains of an antibody of the present invention may differ from the expected sequence due to commonly observed post-translational modifications. For example, C-terminal lysine residues are often absent from antibody heavy chains. (Dick et al. (2008) Biotechnol. Bioeng. 100:1132). N-terminal glutamine residues, and to a lesser extent glutamate residues, are often converted to pyroglutamate residues on both the light and heavy chains of therapeutic antibodies. (Dick et al. (2007) Biotechnol. Bioeng. 97:544; Liu et al. (2011) JBC 28611211; Liu et al. (2011) J. Biol. Chem. 286:11211).

The amino acid sequences for various anti-huICOS agonist antibodies of the present invention are shown in the Sequence Listing, which is summarized in Table 35. For the reasons discussed above, C-terminal lysine is not included in any of the sequences in the Sequence Listing for heavy chains or heavy chain constant domains. However, in an alternative embodiment, each heavy chain for the anti-huICOS antibodies of the present invention and/or the genetic construct encoding such antibodies or their heavy or light chains includes this additional lysine residue at the C-terminus of the heavy chain(s).

VII. Analyzes

The antibodies described herein can be tested for binding to ICOS, for example, using a standard ELISA. For example, microtiter plates are coated with purified ICOS at 1-2 μg/ml in PBS and then blocked with 5% bovine serum albumin in PBS. Antibody dilutions (eg, plasma dilutions from ICOS-immunized mice) are added to each well and incubated for 1-2 hours at 37°C. The plates are washed with PBS/Tween and then incubated with a secondary reagent (e.g. for human antibodies or antibodies containing the human heavy chain constant region, polyclonal goat antibodies specific for human IgG-Fc) conjugated with horseradish peroxidase (HRP) in for 1 h at 37°C. After washing, the plates are developed with ABTS substrate (Moss Inc, product: ABTS1000) and analyzed with a spectrophotometer at OD 415-495. Sera from immunized mice are then further screened by flow cytometry for binding to a cell line expressing human ICOS, but not to a control cell line that does not express ICOS. Briefly, anti-ICOS antibody binding is assessed by incubating ICOS-expressing CHO cells with anti-ICOS antibody at a 1:20 dilution. Cells are washed and binding is determined with a PE-labeled anti-human IgG antibody. Flow cytometry analyzes were performed using FACScan flow cytometry (Becton Dickinson, San Jose, CA). Preferably, mice that develop the highest titers will be used for fusions. Similar experiments can be performed using anti-mouse detection antibodies if mouse anti-huICOS antibodies are to be detected.

An ELISA assay, for example as described above, can be used to screen for antibodies, and hence hybridomas, that produce antibodies that show positive reactivity with the ICOS immunogen. Hybridomas that produce antibodies that bind, preferably with high affinity, to ICOS can then be subcloned and further characterized. One clone from each hybridoma that retains parental cell reactivity (as determined by ELISA) can then be selected for cell bank generation as well as antibody purification.

To purify anti-ICOS antibodies, selected hybridomas can be grown in two liter spinner flasks to purify monoclonal antibodies.

Supernatants can be filtered and concentrated prior to affinity chromatography using Protein A Sepharose (Pharmacia, Piscataway, NJ). The eluted IgG can be monitored by gel electrophoresis and high performance liquid chromatography to ensure purity. The buffer solution can be replaced with PBS and the concentration can be determined by OD ₂₈₀ using an extinction coefficient of 1.43. Monoclonal antibodies can be divided into aliquots and can be stored at -80°C.

To determine if selected monoclonal anti-ICOS antibodies bind to unique epitopes, each antibody can be biotinylated using commercially available reagents (Pierce, Rockford, IL). Binding of the biotinylated mAb can be detected using a streptavidin-labeled probe. Competitive studies using unlabeled monoclonal antibodies and biotinylated monoclonal antibodies can be performed using ICOS-coated ELISA plates as described above. To determine the isotype of purified antibodies, an isotype ELISA can be performed using

- 42 041306 reagents specific for antibodies of a particular isotype. For example, to determine the isotype of a human monoclonal antibody, the wells of a microtiter plate can be coated with 1 μg/ml of anti-human immunoglobulin antibody and left at 4° C. overnight. After blocking with 1% BSA, the plates are reacted with 1 μg/ml or less of the tested monoclonal antibodies or purified isotype controls at ambient temperature for one to two hours. The wells can then be reacted with either human IgG1 or human IgM with alkaline phosphatase conjugated probes. The plates are visualized and analyzed as described above.

Flow cytometry can be used to test the binding of monoclonal antibodies to live cells expressing ICOS. Briefly, cell lines expressing membrane-bound ICOS (grown under standard growth conditions) are mixed with various concentrations of monoclonal antibodies in PBS containing 0.1% BSA and left at 4°C for 1 h. After washing, the cells are reacted with labeled phycoerythrin (PE) anti-IgG antibody under the same conditions as in the primary staining of antibodies. Samples can be analyzed with the FACScan, which uses forward scatter and side scatter properties to observe single cells to determine the binding of labeled antibodies. An alternative analysis using fluorescence microscopy (in addition to or instead of) flow cytometry analysis can be used. Cells can be stained as described above and examined by fluorescence microscopy. This method allows visualization of individual cells, but may have reduced sensitivity depending on the density of the antigen.

Anti-huICOS antibodies can be further tested for reactivity with the ICOS antigen using a Western blot. Briefly, cell extracts obtained from cells expressing ICOS can be prepared and subjected to dodecyl sulfate polyacrylamide gel electrophoresis. After electrophoresis, separated antigens are transferred to nitrocellulose membranes, blocked with 20% mouse serum, and probed with the monoclonal antibodies to be tested. IgG binding can be detected using an anti-IgG antibody-alkaline phosphatase complex and detected using BCIP/NBT substrate tablets (Sigma Chem. Co., St. Louis, MO).

Methods for analyzing the binding affinity, cross-reactivity, and binding kinetics of various anti-ICOS antibodies include standard assays known in the art, such as Bilayer Interferometry (BMI) assay and BIACORE® Surface Plasmon Resonance (SPR) analysis using the BIACORE® 2000 instrument SPR (Biacore AB, Uppsala, Sweden).

In one embodiment, the anti-huICOS antibody specifically binds to the extracellular region of the human ICOS. In one embodiment, the antibody binds to a specific domain (eg, a functional domain) within the extracellular domain of ICOS. In one embodiment, the anti-huICOS antibody specifically binds to the extracellular region of human ICOS and the extracellular region of cynomolgus monkey ICOS. In one embodiment, the anti-huICOS antibody binds to human ICOS with high affinity. VIII. Multispecific molecules

In some embodiments, the antibodies described herein may be multispecific, eg, bispecific or trispecific molecules. Multispecific antigen-binding molecules, such as multispecific antibodies, contain two or more antigen-binding sites, each of which is specific for a different epitope. Another epitope may be part of the same or different antigens. In one embodiment, one antigen binding site is specific for human ICOS and the other for another antigen. In one embodiment, an anti-huICOS antibody described herein, or antigen-binding fragments thereof, binds to another antigen-binding molecule, e.g., another peptide or protein (e.g., another antibody or antibody fragment, or ligand for a receptor) having a different binding specificity. , to form a bispecific molecule that binds to at least two different binding sites or target molecules. In one embodiment, an antibody described herein is derivatized or linked to more than one other antigen-binding molecule to form multispecific molecules that bind to more than two different binding sites and/or target molecules. Thus, provided herein are bispecific molecules comprising at least one first binding specificity for ICOS and a second binding specificity for a second target epitope. In the embodiment described herein, in which the bispecific molecule is multispecific, the molecule may further include a third binding specificity.

In one embodiment, the bispecific molecules described herein comprise, as binding specificity, at least one antibody or antibody fragment,

- 43 041306 including, for example, Fab, Fab', F(ab')2, Fv or single chain Fv. The antibody may also be a light chain or heavy chain dimer, or any minimal antibody fragment such as Fv or any single chain construct as described by Ladner et al. in US patent 4946778, the contents of which are specifically incorporated into the present description by reference.

While human monoclonal antibodies are preferred, other antibodies that can be used in the bispecific molecules described herein are murine, chimeric, and humanized monoclonal antibodies.

The bispecific molecules described herein can be prepared by conjugating the constituent binding specificities using methods known in the art. For example, each binding specificity of a bispecific molecule can be obtained separately, and then they can be conjugated to each other. When the binding specificities are proteins or peptides, a variety of binding or crosslinking agents can be used for covalent conjugation. Examples of crosslinkers include protein A, carbodiimide, N-succinimidyl-S-acetylthioacetate (SATA), 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB), ophenylenedimaleimide (oPDM), N-succinimidyl-3-(2-pyridyldithio) -propionate (SPDP) and sulfosuccinimidyl-4-(M-maleimidomethyl)cyclohaxane-1-carboxylate (sulfo-SMCC) (see e.g. Karpovsky et al. (1984) J. Exp. Med. 160:1686; Liu, MA et al (1985) Proc Natl Acad Sci USA 82:8648). Other methods include those described in Paulus (1985) Behring Ins. Mitt. no. 78, 118-132; Brennan et al. (1985) Science 229:81-83) and Glennie et al. (1987) J. Immunol. 139: 2367-2375). Preferred conjugating agents are SATA and sulfo-SMCC, both available from Pierce Chemical Co. (Rockford, IL).

When the binding specificities are antibodies, they can be conjugated through the formation of a sulfhydryl bond between the hinge regions of the C-terminus of the two heavy chains. In a particularly preferred embodiment, the hinge region is modified such that it contains an odd number of sulfhydryl residues, preferably one sulfhydryl residue, prior to conjugation.

Alternatively, both binding specificities can be encoded in the same vector and expressed and assembled in the same host cell. This method is particularly applicable when the bispecific molecule has a combination of binding specificities such as (mAbxmAb), (mAbxFab), (Fabx F(ab')2) or (ligandχFab)-fusion protein. The bispecific molecule described herein can be a single chain molecule containing one single chain antibody and a binding determinant, or a single chain bispecific molecule containing two binding determinants. Bispecific molecules may contain at least two single chain molecules. Methods for obtaining bispecific molecules are described, for example, in US patent 5260203; US Pat. No. 5,455,030; US Pat. No. 4,881,175; US Pat. No. 5,132,405; US Pat. No. 5,091,513; US Pat. No. 5,476,786; US Pat. No. 5,013,653; US Pat. No. 5,258,498; and US Patent 5,482,858.

Binding of bispecific molecules to their specific targets can be confirmed using methods known in the art, for example, enzyme-linked immunosorbent assay (ELISA), radioimmunoassay (RIA), FACS analysis, bioassay (eg, growth inhibition), or Western blot analysis. Each of these assays typically detects the presence of protein-antibody complexes of particular interest using a labeled reagent (eg, antibody) specific for the complex of interest. IX. Compositions

Further provided are compositions, eg, pharmaceutical compositions, comprising one or more anti-ICOS antibodies, or antigen-binding fragments thereof, as described herein, formulated together with a pharmaceutically acceptable carrier. Accordingly, the compositions of the present invention comprise human or humanized anti-huICOS antibodies (or antigen-binding fragments thereof) comprising CDR sequences, heavy and/or light chain variable region sequences, or full-length heavy and/or light chain sequences shown in Table 1. 35. Compositions according to the present invention also include anti-huICOS antibodies having sequences that are variants of the sequences shown in table. 35. For example, such antibodies may contain sequences that are at least 70%, 75%, 80%, 85%, 90% or 95%, 96%, 97%, 98% or 99% identical to CDR sequences, heavy and/or light chain variable region sequences, or full-length sequences of the heavy and/or light chain presented in table. 35.

Such compositions may also include a single antibody or a combination (eg, two or more different) antibodies or immunoconjugates or bispecific molecules as described herein. For example, a pharmaceutical composition described herein may contain a combination of antibodies (or immunoconjugates or bispecific antibodies) that bind to different epitopes on the target antigen or that have complementary activities.

The pharmaceutical compositions described herein may also be administered as a combination therapy, i.e. anti-ICOS antibodies in combination with other agents. For example,

- 44 041306 combination therapy may include an anti-ICOS antibody described herein in combination with at least one other anti-cancer and/or T-cell stimulating (eg, activating) agent. Examples of therapeutic agents that can be used in combination therapy are described in more detail below in the section on the use of antibodies described in this document.

In some embodiments, pharmaceutical compositions disclosed herein may include other compounds, drugs, and/or agents used in the treatment of cancer. Such compounds, drugs and/or agents may include, for example, chemotherapy drugs, small molecule drugs, or antibodies that stimulate an immune response to a given cancer. In some embodiments, the pharmaceutical composition comprises a first antibody specific for anti-huICOS and a second antibody.

In some embodiments, the first antibody and the second antibody are present in the composition in a fixed dose (ie, in a fixed ratio). In other embodiments, this fixed dose ranges between at least about 1:200 and at least about 200:1, at least about 1:150 and at least about 150:1, at least about 1:100 and at least about 100:1, at least about 1:75, and at least about 75:1, at least about 1:50 to and at least about 50:1, at least about 1:25, and at least about 25:1, at least about 1:10 and at least about 10:1, at least about 1:5 and at least about 5:1, at least about 1:4 and at least at least about 4:1, at least about 1:3 and at least about 3:1, or at least about 1:2 and at least about 2:1 mg of anti-huICOS antibody to mg of second antibody. In some embodiments, the fixed dose is at least about 1:1, about 1:2, about 1:3, about 1:4, about 1:5, about 1:6, about 1:7, about 1:8, about 1:9, about 1:10, about 1:15, about 1:20, about 1:30, about 1:40, about 1:50, about 1:60, about 1:70, about 1:80, about 1:90, about 1:100, about 1:120, about 1:140, about 1:160, about 1:180, or about 1:200 anti-huICOS antibody to the second antibody. In some embodiments, the fixed dose is at least about 2:1, about 3:1, about 4:1, about 5:1, about 6:1, about 7:1, about 8:1, about 9:1, about 10:1, about 15:1, about 20:1, about 30:1, about 40:1, about 50:1, about 60:1, about 70:1, about 80:1, about 90:1, about 100:1, about 120:1, about 140:1, about 160:1, about 180:1, or about 200:1 mg of first antibody to mg of second antibody. For example, in one embodiment, the anti-huICOS antibody and second antibody are administered as described in Example 18.

Additional antibodies include, for example, one or more of an anti-CTLA-4 antibody, an anti-PD-I antibody, an anti-PD-L1 antibody, an anti-TIGIT antibody, an anti-OX40 antibody (also known as CD134, TNFRSF4, ACT35, and /or TXGP1L), anti-LAG-3 antibodies, anti-CD73 antibodies, anti-CD137 antibodies, anti-CD27 antibodies or anti-CSF-1R antibodies.

As used herein, a pharmaceutically acceptable carrier includes any and all solvents, dispersion media, coatings, antibacterial and antifungal agents, isotonic and absorption delaying agents, and the like, which are physiologically compatible. In some embodiments, the carrier is suitable for intravenous, intramuscular, subcutaneous, parenteral, spinal, or epidermal administration (eg, by injection or infusion). In some embodiments, the carrier is suitable for intravenous administration. In other embodiments, the carrier is suitable for subcutaneous administration. In some embodiments, a composition comprising an anti-ICOS antibody is delivered subcutaneously using Halozyme's ENHANZE® drug delivery technology, which includes a recombinant human hyaluronidase (rHuPH20) enzyme that transiently degrades hyaluronan. In some embodiments, ENHANZE® drug delivery technology provides subcutaneous administration of compositions that are faster than intravenous administration. In other embodiments, depending on the route of administration, the active compound, ie, antibody, immunoconjugate, or bispecific molecule, may be coated with a material to protect the compound from acids and other environmental conditions that may inactivate the compound.

The pharmaceutical compounds described herein may include one or more pharmaceutically acceptable salts. A pharmaceutically acceptable salt refers to a salt that retains the desired biological activity of the parent compound and does not cause any undesirable toxicological effects (see, for example, Berge, S. M., et al. (1977) J. Pharm. Sci. 66:1-19). Examples of such salts include acid addition salts and base addition salts. Acid addition salts include salts formed from non-toxic inorganic acids such as

- 45 041306 as hydrochloric acid, nitric, phosphoric, sulfuric, hydrobromic, hydroiodic, phosphorous, etc., as well as from non-toxic organic acids such as aliphatic mono- and dicarboxylic acids, phenyl-substituted alkanoic acids, hydroxyalkanoic acids, aromatic acids, aliphatic and aromatic sulfonic acids, and the like. Base addition salts include salts derived from alkaline earth metals such as sodium, potassium, magnesium, calcium, etc., as well as from non-toxic organic amines such as CD'-dibenzylethylenediamine, N-methylglucamine, chloroprocaine, choline, diethanolamine, ethylenediamine , procaine, etc.

The pharmaceutical composition described herein may also include a pharmaceutically acceptable antioxidant. Examples of pharmaceutically acceptable antioxidants include: (1) water-soluble antioxidants such as ascorbic acid, cysteine hydrochloride, sodium bisulfate, sodium metabisulfite, sodium sulfite, and the like; (2) oil-soluble antioxidants such as ascorbyl palmitate, butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), lecithin, propyl gallate, alpha-tocopherol, and the like; and (3) metal chelating agents such as citric acid, ethylenediaminetetraacetic acid (EDTA), sorbitol, tartaric acid, phosphoric acid, and the like.

Examples of suitable aqueous and non-aqueous carriers that may be used in the pharmaceutical compositions described herein include water, ethanol, polyols (such as glycerol, propylene glycol, polyethylene glycol, and the like) and suitable mixtures thereof, vegetable oils such as olive oil, and injectable organic esters such as ethyl oleate. Proper fluidity can be maintained, for example, by using coating materials such as lecithin, by maintaining the required particle size in the case of dispersions and the use of surfactants.

These compositions may also contain adjuvants such as preservatives, wetting agents, emulsifying agents and dispersing agents. Prevention of the presence of microorganisms can be ensured both by sterilization procedures, supra, and by the inclusion of various antibacterial and antifungal agents, for example, paraben, chlorobutanol, phenol sorbic acid, and the like. It may also be desirable to include isotonic agents such as sugars, sodium chloride, and the like in the compositions. In addition, prolonged absorption of the injectable pharmaceutical form can be accomplished by the inclusion of absorption delaying agents such as aluminum monostearate and gelatin.

Pharmaceutically acceptable carriers include sterile aqueous solutions or dispersions and sterile powders for immediate preparation of sterile injectable solutions or dispersions. Exemplary pharmaceutically acceptable carriers herein further include interstitial drug dispersing agents such as soluble neutrally active hyaluronidase glycoproteins (sHASEGP), e.g. human soluble hyaluronidase PH-20 glycoproteins such as rHuPH20 (HYLENEX™, Baxter International, Inc. .). Some exemplary sHASEGP and uses, including rHuPH20, are described in US Patent Publications 2005/0260186 and 2006/0104968. In one aspect, sHASEGP is combined with one or more additional glycosaminoglycans, such as chondroitinases.

The use of such media and agents for pharmaceutically active substances is known in the art. Unless any conventional medium or agent is incompatible with the active compound, its use in the pharmaceutical compositions described herein is contemplated. Additional active compounds may also be included in the compositions.

Therapeutic compositions generally must be sterile and stable under the conditions of preparation and storage. The composition may be formulated as a solution, microemulsion, liposome, or other ordered structure suitable for high drug concentration. The carrier may be a solvent or dispersion medium containing, for example, water, ethanol, polyol (eg, glycerol, propylene glycol and liquid polyethylene glycol, etc.) and suitable mixtures thereof. Proper fluidity can be maintained, for example, using a coating such as lecithin, maintaining the desired particle size in the case of a dispersion, and using surfactants. In many cases, it is preferred to include isotonic agents, eg sugars, polyalcohols such as mannitol, sorbitol or sodium chloride, in the compositions. Prolonged absorption of injectable compositions can be accomplished by including in the composition an agent that delays absorption, such as monostearate salts and gelatin.

Sterile injectable solutions can be prepared by incorporating the active compound in the required amount in a suitable solvent with one or a combination of the ingredients listed above, if necessary, followed by microfiltration sterilization. Typically, dispersions are prepared by incorporating the active compound into a sterile vehicle which contains the dispersion medium as a base and the required other ingredients from those described above. In the case of sterile powders for the preparation of sterile injectable solutions, the preferred methods of preparation are vacuum drying and freeze-drying (lyophilization).

- 46 041306 tsiya), which give a powder of the active ingredient with any additional desired ingredient from its pre-sterile filtered solution.

The amount of active ingredient which may be combined with the carrier material to prepare a single dosage form will vary depending on the subject to be treated and the particular route of administration. The amount of active ingredient that can be combined with the carrier material to prepare a single dosage form is usually that amount of the composition that produces a therapeutic effect. On a one hundred percent basis, such an amount may range from about 0.1% to about 99% of the active ingredient, for example, from about 0.1% to about 70%, for example, from about 1% to about 30% of the active ingredient in combination with pharmaceutically acceptable carrier.

In some embodiments, the composition comprises an anti-ICOS antibody, such as the ICOS.33 IgG1f S267E antibody, at a concentration of 10 mg/mL. The composition is a sterile, non-pyrogenic, disposable, preservative-free, isotonic aqueous solution for intravenous administration. The composition may be administered neat or further diluted with 0.9% sodium chloride solution to desired protein concentrations prior to infusion. In some embodiments, the anti-ICOS antibody comprises the following excipients: L-histine, L-histidine hydrochloride monohydrate, sucrose, pentetic acid (also known as diethylenetriaminepentaacetic acid), polysorbate 80, and water for injection.

Dosing regimens are selected in such a way as to provide the optimal desired response (eg, therapeutic response). For example, a single bolus may be administered, several divided doses over time, or the dose may be proportionally reduced or increased depending on the urgency of the therapeutic situation. Parenteral compositions are particularly preferably formulated in unit dosage forms for ease of administration and dose uniformity. As used herein, unit dosage form refers to physically discrete units suitable as a single dose for subjects to be treated; each dosage unit contains a predetermined amount of active compound, calculated to obtain the desired therapeutic effect, together with the required pharmaceutical carrier. The specification for the unit dosage forms described herein is dictated by (a) and dependent on the unique characteristics of the active compound and the specific therapeutic effect to be achieved, and (b) the limitations inherent in the art of formulating such an active compound with respect to treatment of sensitivity in individuals.

For administration of the antibody, the dose may be in the range of about 0.0001 to 100 mg/kg, and more typically 0.01 to 5 mg/kg of the host's body weight. For example, doses may be 0.3 mg/kg body weight, 1 mg/kg body weight, 3 mg/kg body weight, 5 mg/kg body weight, or 10 mg/kg body weight, or in the range of 1-10 mg/day. kg. Alternatively, administration of the antibody is a fixed dose that can vary from 2 to 800 mg, such as a dose of 25, 80, 200, or 400 mg. An exemplary treatment regimen includes administration once a week, once every two weeks, once every three weeks, once every four weeks, once a month, once every two months, once every three months, once every four months , once every five months or once every six months. In some embodiments, the treatment regimen includes an initial dose followed by a maintenance dose in varying dose amounts at intervals of dosing.

In some embodiments, two or more monoclonal antibodies with different binding specificities are administered simultaneously, in which case the dose of each antibody administered is within the ranges mentioned. In some embodiments, the therapeutic antibody is administered on multiple occasions. The intervals between individual doses can be, for example, weekly, once every three weeks, once every four weeks, monthly, every three months or once a year. The intervals may also be irregular, depending on the measurement of blood levels of the antibody to the target antigen in the patient. In some embodiments, the dose is adjusted to achieve a plasma antibody concentration of about 1-1000 µg/ml, and in some methods about 25-300 µg/ml.

In some embodiments, the antibody may be administered as a sustained release formulation. Administration via sustained release formulations may require less frequent administration. The dose and frequency of administration will vary depending on the half-life of the antibody in the patient. The dose and frequency of administration may vary depending on whether the treatment is prophylactic or therapeutic. In prophylactic applications, a relatively low dose is administered at relatively infrequent intervals over a long period of time. Some patients continue to receive treatment for the rest of their lives. In some embodiments, a relatively high dose is administered for therapeutic treatment at relatively short intervals. In some embodiments, a relatively high dose is administered until the progression of the disease is reduced or stopped, for example, until the patient shows partial or complete relief of the symptoms of the disease. In some embodiments, the implementation of prophylactic treatment is carried out in a patient after therapeutic treatment.

- 47 041306

Actual dosage levels of active ingredients in the pharmaceutical compositions described herein may vary so as to obtain an amount of the active ingredient that is effective to achieve the desired therapeutic response for a particular patient, composition and route of administration, without causing toxicity to that patient. The selected dose level will depend on various pharmacokinetic factors, including the activity of the specific compositions used herein or their esters, salts or amides, the route of administration, the time of administration, the rate of elimination of the particular compound used, the duration of treatment, other drugs, compounds and / or materials used in combination with the specific compositions employed, age, sex, weight and disease, general health, and previous medical history of the patient being treated, as well as similar factors well known in the medical field.

A therapeutically effective dose of the anti-ICOS antibody described herein preferably results in a reduction in the severity of the symptoms of the disease, an increase in the frequency and duration of asymptomatic periods of the disease, or the prevention of aggravation or disability due to the disease. In the context of cancer, a therapeutically effective dose preferably prevents further deterioration of the physical symptoms associated with the cancer. Symptoms of cancer are well known in the art and include, for example, unusual features of the mole, change in the appearance of the mole, including asymmetry, borders, color and/or diameter, newly pigmented areas of the skin, abnormal mole, darkened area under the nail, breast lump, changes nipples, breast cyst, chest pain, death, weight loss, weakness, excessive fatigue, difficulty eating, loss of appetite, chronic cough, worsening dyspnea, coughing up blood, blood in the urine, blood in the stool, nausea, vomiting, metastases in liver, lung metastases, bone metastases, abdominal fullness, bloating, fluid in the abdomen, vaginal bleeding, constipation, bloating, colonic perforation, acute peritonitis (infection, fever, pain), pain, blood in vomit, profuse sweating, fever, high blood pressure, anemia, diarrhea, jaundice, dizziness, chills, muscle spasms, colon metastases, lung metastases, bladder metastases, m liver metastases, bone metastases, kidney metastases and pancreatic metastases, difficulty in swallowing, and the like. Therapeutic efficacy may be observed immediately after the first administration of the anti-huICOS agonistic monoclonal antibody of the present invention, or it may be observed only after a certain period of time and/or a series of doses. This delay in efficacy is observed only after several months of treatment, for example up to 6, 9 or 12 months.

A therapeutically effective dose may prevent or delay the onset of cancer, which may be desirable when early or early signs of the disease are present. Accordingly, any clinical or biochemical assay that controls for any of the above indications may be used to determine whether a particular treatment is a therapeutically effective dose for treating cancer. One of skill in the art will be able to determine such amounts based on factors such as the size of the subject, the severity of the subject's symptoms, and the particular composition or route of administration.

The composition described herein can be administered via one or more routes of administration using one or more of the various routes known in the art. As will be appreciated by one of skill in the art, the route and/or schedule of administration will vary depending on the desired results. Examples of the routes of administration of antibodies described herein include intravenous, intramuscular, intradermal, intraperitoneal, subcutaneous, spinal, or other parenteral routes of administration, such as by injection or infusion.

Alternatively, the antibody described herein may be administered by a non-parenteral route, such as a topical, epidermal, or transmucosal route, for example, intranasally, orally, vaginally, rectally, sublingually, or topically.

Active compounds can be formulated with carriers that will protect the compound against rapid release, such as a controlled release formulation, including implants, transdermal patches, and microencapsulated delivery systems. Biodegradable, biocompatible polymers such as ethylene vinyl acetate, polyanhydrides, polyglycolic acid, collagen, polyorthoesters, and polylactic acid can be used. Many methods for making such formulations are patented or commonly known to those skilled in the art. See, for example, Sustained and Controlled Release Drug Delivery Systems, J.R. Robinson, ed., Marcel Dekker, Inc., New York, 1978.

Therapeutic compositions can be administered using medical devices known in the art. For example, in a preferred embodiment, the therapeutic composition described herein can be administered using a needleless hypodermic injection device such as those disclosed in US Pat. Nos. 5,399,163; 5383851; 5312335; 5064413; 4941880; 4790824; or 4596556. Examples of well-known implants and modules for

- 48 041306 variations with anti-huICOS antibodies described herein include US Pat. No. 4,487,603, which describes an implantable microinfusion pump for drug administration at a controlled rate; US Pat. No. 4,486,194, which discloses a therapeutic device for administering a drug through the skin; US Pat. No. 4,447,233, which describes a drug infusion pump for drug delivery at a precise infusion rate; US Pat. No. 4,447,224, which discloses an implantable AC infusion device for continuous drug delivery; US Pat. No. 4,439,196, which describes an osmotic drug delivery system having multi-chamber compartments; and US Pat. No. 4,475,196, which describes an osmotic drug delivery system. These patents are incorporated herein by reference. Many other similar implants, delivery systems and modules are known to those skilled in the art.

In some embodiments, the anti-huICOS antibodies described herein can be formulated to ensure proper distribution in vivo. For example, the blood-brain barrier (BBB) excludes many highly hydrophilic compounds. To ensure that the therapeutic compounds described herein cross the BBB (if desired), they can be formulated in, for example, liposomes. For methods of preparing liposomes, see, for example, US Pat. Nos. 4,522,811; 5374548; and 5,399,331. Liposomes may contain one or more fragments that are selectively transported to specific cells or organs and thus enhance targeted drug delivery (see, for example, V.V. Ranade (1989) J. Clin. Pharmacol. 29:685). Exemplary targeting moieties include folate or biotin (see, for example, US Pat. No. 5,416,016 to Low et al.); mannosides (Umezawa et al., (1988) Biochem. Biophys. Res. Commun. 153:1038); antibodies (P. G. Bloeman et al. (1995) FEBS Lett. 357:140; M. Owais et al. (1995) Antimicrob. Agents Chemother. 39:180); surfactant protein A receptor (Briscoe et al. (1995) Am. J. Physiol. 1233:134); pl20 (Schreier et al. (1994) J. Biol. Chem. 269:9090); see also K. Keinanen; M.L. Laukkanen (1994) FEBS Lett. 346:123; J.J. killion; I.J. Fidler (1994; Immunomethods 4:273.

Also included within the scope of the present invention are kits containing compositions of the antibodies described herein (eg, human antibodies, bispecific or multispecific molecules, or immunoconjugates) and instructions for use. In addition, the kit contains at least one additional reagent, or one or more additional human antibodies described herein. Kits may include a label indicating the intended use of the contents of the kit. The term label includes any written or recorded material supplied on or with a kit, or otherwise attached to a kit.

X. Methods of application

The antibodies, antibody compositions, and methods described herein have numerous in vitro and in vivo uses, including, for example, enhancing an immune response by stimulating ICOS signaling. In one embodiment, the antibodies described herein are monoclonal human or humanized antibodies. In one embodiment, the anti-huICOS antibodies described herein (e.g., ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7, and 2644) can be administered to cells in culture, in vitro or ex vivo, or to human subjects. to enhance immunity in various diseases. In a specific embodiment, the anti-huICOS antibodies are agonist antibodies, ie anti-huICOS agonist antibodies. Provided herein are methods for modifying an immune response in a subject, comprising administering to the subject an antibody as described herein, or an antigen-binding fragment thereof, such that the subject's immune response is enhanced, stimulated, or upregulated. In one embodiment, administration of an anti-huICOS antibody (ie, an anti-huICOS agonist antibody) according to the methods described herein enhances co-stimulation of T cell responses. In one embodiment, administration of an anti-huICOS antibody according to the methods described herein stimulates, enhances, or upregulates antigen-specific T cell responses to a tumor. The tumor may be a solid tumor or a liquid tumor, such as a hematological malignancy. In some embodiments, the tumor is an immunogenic tumor. In some embodiments, the implementation of the tumor is non-immunogenic. In some embodiments, the tumor is PD-L1 positive. In some embodiments, the tumor is PD-L1 negative. The subject may also be a virus-carrying subject and the immune response is enhanced against the virus. In one embodiment, administration of an anti-huICOS antibody according to the methods described herein stimulates, enhances, or upregulates CD4+ and CD8+ T cell responses. T cells can be Teff cells, eg CD4+ Teff cells, CD8+ Teff cells, T helper (T _h ) cells and T cytotoxic (T _c ) cells.

In one embodiment, the methods result in an enhancement of the immune response in a human subject, such enhancement having the desired effect. In one embodiment, the human subject is a human patient having a disorder that can be treated by enhancing an immune response, eg, a T-cell mediated immune response. IN

- 49 041306 in a specific embodiment, the human patient has cancer. In one embodiment, the anti-huICOS antibodies described herein may be administered together with the antigen of interest, or the antigen may already be present in the subject to be treated, such as a tumor-bearing or virus-bearing subject. When antibodies to ICOS are administered together with another agent, they can be administered separately or simultaneously.

In addition, methods are provided for inhibiting the growth of a tumor cell in a subject, comprising administering to the subject an anti-huICOS antibody described herein such that growth of the tumor cell is inhibited in the subject. Also provided are methods for treating a chronic viral infection in a subject, comprising administering to the subject an anti-huICOS antibody described herein such that the chronic viral infection is cured in the subject.

In some embodiments, the anti-huICOS agonist antibody is administered to a subject, such as a human patient, as an add-on therapy, adjuvant therapy, or neoadjuvant therapy. In some embodiments, treatment of subjects with cancer with an anti-huICOS antibody may result in a long-term sustained response compared to current standard therapy; long-term survival of at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 years or more, relapse-free survival of at least 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 years or more. In some embodiments, treatment of a subject having cancer with an anti-huICOS antibody prevents cancer recurrence or delays cancer recurrence, for example, by 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 years or more. Anti-ICOS antibody treatment can be used as first, second or subsequent line therapy.

The data and other methods described herein are discussed in more detail below.

Cancer

Provided herein are methods of treating a subject having cancer, comprising administering to the subject an anti-huICOS antibody described herein such that the subject is treated, for example, such that the growth of cancerous tumors is inhibited or reduced and/or the tumors regress. An anti-huICOS antibody can be used alone to inhibit the growth of cancerous tumors. Alternatively, the anti-huICOS antibody may be used in combination with another agent, such as other immunogenic agents, conventional cancer treatments, or other antibodies, as described below. A combination with a PD-1 inhibitor such as an anti-PD-1 or anti-PD-L1 antibody is also provided. A combination with a CTLA-4 inhibitor such as an anti-CTLA4 antibody is also provided. A combination with a PD-1 inhibitor and a CTLA-4 inhibitor is also provided.

In one aspect, provided herein are methods of treating cancer in a subject, comprising administering to the subject a therapeutically effective amount of an anti-huICOS antibody described herein, for example, a humanized form of a hamster anti-ICOS antibody, or an antigen-binding fragment thereof. In one embodiment, the anti-huICOS antibody can be a chimeric antibody, a human antibody, or a humanized anti-huICOS antibody, such as any of the humanized anti-huICOS antibodies described herein. In one embodiment, the cancer treatment methods described herein include administering a humanized anti-huICOS antibody that contacts human ICOS at one or more amino acid residues of SEQ ID NO: 203 of the human ICOS protein. In another embodiment, the method comprises administering the ICOS.33 IgG1f S267E antibody. In another embodiment, the method comprises administering a composition comprising the ICOS.33 IgG1f S267E antibody.

Examples of cancer include, but are not limited to, squamous cell carcinoma, small cell lung cancer (SCLC), non-small cell lung cancer, squamous non-small cell lung cancer (NSCLC), non-NSCLC, glioma, gastrointestinal cancer, kidney cancer (e.g., clear cell carcinoma), ovarian cancer, liver cancer, colorectal cancer, endometrial cancer, kidney cancer (eg, renal carcinoma (RCC)), prostate cancer (eg, hormone resistant prostate adenocarcinoma), thyroid cancer, neuroblastoma, pancreatic cancer, glioblastoma (glioblastoma multiforme) ), cervical cancer, stomach cancer, bladder cancer, hepatoma, breast cancer, colon cancer, head and neck cancer (or carcinoma), stomach cancer, germ cell tumor, pediatric sarcoma, sinonasal natural killer, melanoma (eg, metastatic malignant melanoma such as cutaneous or intraocular malignant melanoma), bone cancer, skin cancer, uterine cancer, anal cancer, ovarian cancer check, fallopian tube cancer, endometrial carcinoma, cervical carcinoma, vaginal carcinoma, vulvar carcinoma, esophageal cancer, small intestine cancer, endocrine system cancer, parathyroid cancer, adrenal cancer, soft tissue sarcoma, bladder cancer, penile cancer, solid childhood tumors, ureteral cancer, renal pelvis cancer, central nervous system (CNS) neoplasm, primary CNS lymphoma, tumor angiogenesis, spinal cord tumor, brain cancer including brain glioma, pituitary adenoma, Kaposi's sarcoma, epidermal cancer, squamous cell carcinoma , T-cell lymphoma, environmentally induced malignancies, including asbestos-induced, virus-associated malignancies (e.g., human papillomavirus (HPV)-associated tumors), and hematologic malignancies derived from any of the two major blood cell lines, That

- 50 041306 have a myeloid cell line (which produces granulocytes, erythrocytes, platelets, macrophages and mast cells) or a lymphoid cell line (which produces B, T, NK and plasma cells), such as all types of leukemias, lymphomas and myelomas, such as acute , chronic, lymphocytic and/or myelogenous leukemias such as acute leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL) and chronic myelogenous leukemia (CML), undifferentiated AML (M0), myeloid leukemia (M1) , myeloblastic leukemia (M2; with cell maturation), promyelocytic leukemia (M3 or M3 variant [M3V]), myelomonocytic leukemia (M4 or M4 variant with eosinophilia [M4E]), monocytic leukemia (M5), erythroleukemia (M6), megakaryoblastic leukemia (M7), isolated granulocytic sarcoma and chlorleukemia; lymphomas such as Hodgkin's lymphoma (HL), non-Hodgkin's lymphoma (NHL), B-cell lymphomas, T-cell lymphomas, lymphoplasmacytoid lymphoma, monocytoid B-cell lymphoma, mucosal associated lymphoid tissue (MALT) lymphoma, anaplastic (eg, Ki 1+) large cell lymphoma, adult T-cell lymphoma/leukemia, mantle cell lymphoma, angioimmunoblastic T-cell lymphoma, angiocentric lymphoma, intestinal T-cell lymphoma, primary mediastinal B-cell lymphoma, T-lymphoblastic cell lymphoma precursors, T-cell-lymphoblastic lymphoma; and lymphoblastic leukemia (T-Lbly/T-ALL), peripheral T-cell lymphoma, lymphoblastic lymphoma, lymphoproliferative disorders after transplantation, true histiocytic lymphoma, primary central nervous system lymphoma, primary effusion lymphoma, lymphoblastic lymphoma (LBL), hematopoietic tumors of the lymphoid origin, acute lymphoblastic leukemia, diffuse large B-cell lymphoma, Burkitt's lymphoma, follicular lymphoma, diffuse histiocytic lymphoma (DHL), immunoblastic large cell lymphoma, progenitor B-cell lymphoblastic lymphoma, cutaneous T-cell lymphoma (CTLC) (also called fungoid mycosis or Cesari syndrome), and lymphoplasmacytoid lymphoma (LPL) with Waldenström's macroglobulinemia; myelomas such as IgG myeloma, light chain myeloma, non-secreting myeloma, smoldering myeloma (also called indolent myeloma), single plasmacytoma and multiple myelomas, chronic lymphocytic leukemia (CLL), hairy cell lymphoma; hematopoietic tumors of myeloid origin, tumors of mesenchymal origin, including fibrosarcoma and rhabdomyosarcoma; seminoma, teratocarcinoma, tumors of the central and peripheral nervous system, including astrocytoma, schwannomas; tumors of mesenchymal origin, including fibrosarcoma, rhabdomyosarcoma and osteosarcoma; and other tumors, including melanoma, xeroderma pigmentosa, keratoacanthoma, seminoma, follicular thyroid cancer, and teratocarcinoma, hematopoietic tumors of lymphoid origin, such as T-cell and B-cell tumors, including, but not limited to, T-cell disorders such as T-prolymphocytic leukemia (T-PLL), including small cell and medulla cell type; large granular lymphocytic leukemia (LGL), preferably T-cell type; a/d T-NHL hepatosplenic lymphoma; peripheral/post-thymus T-cell lymphoma (pleomorphic and immunoblastic subtypes); angiocentric (nasal) T-cell lymphoma; head or neck cancer, kidney cancer, rectal cancer, thyroid cancer; acute myeloid lymphoma, as well as any combination of these types of cancer. In one embodiment, the methods described herein can also be used to treat metastatic cancers, refractory cancers (e.g., cancer refractory to previous immunotherapy, e.g., with a CTLA-4 and/or PD-1 blocking antibody), and recurrent malignancies.

In one embodiment, the anti-huICOS antibody can be administered as monotherapy. In one embodiment, an anti-huICOS agonist antibody is administered as the sole immunostimulatory agent. In one embodiment, an anti-human ICOS agonist antibody is administered to a patient with another agent. In one embodiment, the anti-huICOS antibody is administered with an immunogenic agent. In one embodiment, an anti-human ICOS agonist antibody is administered in combination with a cancer vaccine. In some embodiments, the cancer vaccine comprises cancer cells, purified tumor antigens (including recombinant proteins, peptides, and carbohydrate molecules), cells, and cells transfected with genes encoding immunostimulatory cytokines (He et al. (2004) J. Immunol. 173:4919- 28). In some embodiments, the cancer vaccine is a cancer peptide vaccine, which in some embodiments is a personalized peptide vaccine. In some embodiments, the cancer peptide vaccine is a multivalent long peptide, multipeptide, peptide cocktail, fusion peptide, or peptide-activated dendritic cell vaccine (see, for example, Yamada et al., Cancer Sci, 104:14-21, 2013 ). In some embodiments, an anti-human ICOS agonist antibody may be administered in combination with an adjuvant. Non-limiting examples of cancer vaccines that can be used include melanoma antigen peptides such as gp100 peptides, MAGE, Trp-2, MART1 and/or tyrosinase antigens, or tumor cells transfected to express the cytokine GM-CSF. Many experimental tumor vaccination strategies have been developed (see Rosenberg, S., 2000, Development of Cancer Vaccines, ASCO Educational Book Spring:

- 51 041306

60-62; Logothetis, C., 2000, ASCO Educational Book Spring: 300-302; Khayat, D. 2000, ASCO Educational Book Spring: 414-428; Foon, K. 2000, ASCO Educational Book Spring: 730-738; see also Restifo, N. and Sznol, M., Cancer Vaccines, Ch. 61, pp. 3023-3043 in DeVita et al. (eds.), 1997, Cancer: Principles and Practice of Oncology, Fifth Edition). In one such strategy, a vaccine is prepared using autologous or allogeneic tumor cells. These cell vaccines have been shown to be most effective when tumor cells are transduced to express GM-CSF. GM-CSF has been shown to be a strong activator of antigen presentation for antitumor vaccination. Dranoff et al. (1993) Proc. Natl. Acad. sci. U.S.A. 90:3539-43.

Other cancer vaccines may include proteins from viruses implicated in human cancers such as human papillomaviruses (HPV), hepatitis viruses (HBV and HCV), and Kaposi's sarcoma associated with herpes virus (KHSV). Another form of tumor-specific antigen that can be used in combination with ICOS inhibition is purified heat shock proteins (HSPs) isolated from the tissue of the tumor itself. These heat shock proteins contain fragments of proteins from tumor cells and such HSPs are highly effective in delivering to antigen presenting cells to induce antitumor immunity (Suot & Srivastava (1995) Science 269:1585-1588; Tamura et al. (1997) Science 278: 117-120).

Dendritic cells are strong antigen-presenting cells that can be used to generate antigen-specific responses. Dendritic cells can be produced ex vivo and loaded with various protein and peptide antigens as well as tumor cell extracts (Nestle et al. (1998) Nature Medicine 4: 328-332). Dendritic cells can also be genetically transduced to express these tumor antigens. Dendritic cells have also been fused directly to tumor cells for immunization purposes (Kugler et al. (2000) Nature Medicine 6:332-336). As a vaccination method, immunization with dendritic cells can be effectively combined with ICOS agonism to activate (release) stronger antitumor responses.

In some embodiments, an anti-human ICOS agonist antibody is administered in combination with a standard of care, such as surgery, radiation, and/or chemotherapy. In some embodiments, the anti-ICOS antibody may be administered in combination with a chemotherapeutic agent. In some embodiments, the anti-ICOS antibody is administered in combination with one or more of carboplatin, cisplatin, paclitaxel, nab-paclitaxel, gemcitabine, or FOLFOX. In some embodiments, an anti-human ICOS agonist antibody may be administered in combination with carboplatin or nab-paclitaxel. In some embodiments, an anti-human ICOS agonist antibody may be administered in combination with carboplatin and paclitaxel. In some embodiments, an anti-human ICOS agonist antibody may be administered in combination with cisplatin and pemetrexed. In some embodiments, an anti-human ICOS agonist antibody may be administered in combination with cisplatin and gemcitabine. In some embodiments, an anti-human ICOS agonist antibody may be administered in combination with FOLFOX. In some embodiments, an anti-human ICOS agonist antibody may be administered in combination with FOLFIRI. In one embodiment, an anti-huICOS antibody is administered in combination with decarbosine to treat melanoma. In some embodiments, cisplatin is administered intravenously at a dose of 100 mg/ml once every four weeks. In some embodiments, an anti-human ICOS agonist antibody may be administered in combination with doxorubicin (adriamycin), cisplatin, bleomycin sulfate, carmustine, chlorambucil, dacarbazine, and/or hydroxyurea cyclophosphamide. In some embodiments, adriamycin is administered intravenously at a dose of 60-75 mg/ml once every 21 days. In one embodiment, an anti-huICOS antibody is administered to a human patient who is resistant to at least one drug treatment, wherein administration of the anti-huICOS antibody reduces, attenuates, or abolishes resistance to at least one drug.

The combination therapies noted above encompass combined administration (when two or more therapeutic agents are included in the same formulation or separate formulations) and separate administration, in which case administration of an antibody of the invention may occur before, simultaneously and/or after administration of an additional therapeutic agent and/or adjuvant. The antibodies of the invention can also be used in combination with radiation therapy.

Another example of such a combination is an anti-huICOS antibody in combination with interleukin 2 (IL-2). In some embodiments, the combination of anti-huICOS antibody and IL-2 is for the treatment of various cancers, including the treatment of renal cell carcinoma and melanoma. In some embodiments, the anti-huICOS antibodies discussed herein are combined with an IL-2 pathway agonist to treat various cancers. The combination includes various IL-2 pathway agonists such as those described in WO 2012/065086 (Nektar Therapeutics) and WO 2015/125159 (Nektar Therapeutics), the entire contents of which are incorporated by reference. WO 2006/138572 (Nektar Therapeutics) provides conjugates having a degradable bond and polymeric reagents useful

- 52 041306 to obtain such conjugates, as well as methods for obtaining polymeric reagents and conjugates, and are included in this description by reference in its entirety.

In some embodiments, a combination of an anti-huICOS antibody as described herein, such as ICOS.33 IgG1 S267E, and an IL-2 pathway agonist, such as NKTR-214, is administered to patients for the treatment of cancer. As described in more detail below, NKTR-214 is prepared by conjugation of an average of approximately six polyethylene glycol (PEG) FMOC (fluorenylmethyloxycarbonyl chloride) reagents having the following structure (mPEG2-C2-fomc-20K-N-hydroxysuccinimidate derivative, 20 kDa, ( mPEG2-C2-fmoc-20K-NHS):

with a protein having the following sequence of 132 amino acids:

PTSSSTKKTQLQLEHLLLDLQMILNGINNYKNPKLTRMLTFKFYMPKKATELKHLQCLEE 60

ELKPLEEVLNLAQSKNFHLRPRDLISNINVIVLELKGSETTFMCEYADETATIVEFLNRW 120 ITFSQSIISTLT 132 (SEQ ID NO: 219).

WO 2012/065086 provides conjugates of an IL-2 fragment and one or more non-peptide water-soluble polymers, including polyethylene glycol or a derivative thereof. In particular, Example 2 (paragraphs 202-204) of WO 2012/065086 describes the PEGylation of rIL-2 with mPEG2-C2-fmoc-20KNHS to give the mPEG2-C2-fmoc-20K-NHS structure above. Example 1 (paragraphs 63-66) of WO 2015/125159 describes an extended approach to pegylation of IL-2 with mPEG2-C2fmoc-20K-NHS resulting in RSLAIL-2 (NKTR-214). NKTR-214 is a cytokine that is designed to target CD122 (also known as the interleukin-2 receptor subunit, IL-2RP), a protein found on certain cells of the immune system (for example, CD8+ T cells and NK cells), to multiply these cells to stimulate their antitumor effects.

In some embodiments, the anti-huICOS antibody may be administered in combination with an angiogenesis inhibitor.

Other combination therapies that may result in synergy with ICOS agonism through cell death are radiation, surgery, and endocrine shutdown.

In some embodiments, anti-huICOS antibodies described herein can be administered in combination with bispecific antibodies. Bispecific antibodies can be used to target two separate antigens. In one embodiment, anti-huICOS antibodies are used in combination with bispecific antibodies that target Fca or Fcγ receptor expressing effector cells to tumor cells (see, for example, US Pat. Nos. 5,922,845 and 5,837,243). For example, anti-Fc receptor/anti-tumor antigen bispecific antibodies (eg, Her-2/neu) have been used to target macrophages to tumor sites. In one embodiment, the T cell arm of these responses is increased by ICOS agonism with an anti-huICOS antibody. Alternatively, the antigen can be delivered directly to the DC using bispecific antibodies that bind to the tumor antigen and a dendritic cell-specific cell surface marker. In some embodiments, anti-huICOS antibodies are used in combination with antibodies that reduce or inactivate immunosuppressive proteins expressed by the tumor, such as anti-TGF-P antibodies, anti-IL-10 antibodies, and anti-Fas ligand antibodies.

Chronic viral infections

In another aspect, the invention described herein provides a method of treating an infectious disease in a subject, comprising administering to the subject an anti-huICOS antibody, or antigen-binding fragment thereof, such that the subject is treated for the infectious disease.

Similar to its use in tumors, as discussed above, antibody-mediated ICOS agonism can be used alone or, as an adjuvant, in combination with vaccines to stimulate an immune response to pathogens, toxins, and self-antigens. Examples of pathogens for which this therapeutic approach may be particularly useful include pathogens for which there is currently no effective vaccine, or pathogens for which conventional vaccines are not fully effective. Pathogens include, but are not limited to, HIV, hepatitis viruses (A, B, and C), influenza viruses, herpes viruses, giardiasis, malaria, leishmaniasis, Staphylococcus aureus, Pseudomonas aeruginosa. ICOS agonism is particularly applicable to established infections caused by such agents, such as HIV, in which the course of the infection is accompanied by a change in antigens. These new epitopes are recognized as foreign at the time of administration of the anti-human ICOS antibody, thereby eliciting a strong T cell response.

- 53 041306

Some examples of pathogenic viruses that cause infections that can be treated with the methods described herein include HIV, hepatitis virus (A, B, or C), herpes virus (e.g., VZV, HSV-1, HAV-6, HSV-II and CMV, Epstein-Barr virus), adenovirus, influenza virus, flaviviruses, echovirus, rhinovirus, coxsackievirus, coronavirus, respiratory syncytial virus, mumps virus, rotavirus, measles virus, rubella virus, parvovirus, vaccinia virus, HTLV virus , dengue virus, human papillomavirus, molluscum contagiosum virus, poliovirus, rabies virus, JC virus, and arbovirus encephalitis virus.

Some examples of pathogenic bacteria causing infections that can be treated with the methods described herein include Chlamydia, Rickettsial, Mycobacteria, Staphylococcus, Streptococcus, Pneumococcus, Meningococcus and Gonococcus, Klebsiella, Proteus, Serratia, Pseudomonas, Legionella, Diphtheria , Salmonella, bacilli, cholera, tetanus, botulism, anthrax, plague, leptospirosis, and the bacteria that cause Lyme disease.

Some examples of pathogenic fungi that cause infections that can be treated with the methods described herein include Candida (albicans, krusei, glabrata, tropicalis, etc.), Cryptococcus neoformans, Aspergillus (fumigatus, niger, etc.) , genus Mucorales (mucor, absidia, rhizopus), Sporothrix schenkii, Blastomyces dermatitidis, Paracoccidioides brasiliensis, Coccidioides immitis and Histoplasma capsulatum

Some examples of pathogenic parasites that cause infections that can be treated with the methods described herein include Entamoeba histolytica, Balantidium coli, Naegleriafowleri, Acanthamoeba sp., Giardia lambia, Cryptosporidium sp., Pneumocystis carinii, Plasmodium vivax, Babesia microti, Trypanosoma brucei, Trypanosoma cruzi, Leishmania donovani, Toxoplasma gondii, Nippostrongylus brasiliensis

The methods of administering anti-huICOS antibodies described herein can be combined with other forms of immunotherapy such as cytokine (eg, interferons, GM-CSF, G-CSF, IL-2) therapy or bispecific antibody therapy.

Combination Therapies

In one aspect, provided herein are methods of combination therapy, e.g., for the treatment of cancer, in which an anti-huICOS antibody (e.g., an anti-huICOS agonist antibody) is co-administered with one or more additional agents, e.g., antibodies, that are effective in stimulating immune responses, thereby further improving, stimulating or activating immune responses in the subject. Provided herein are methods for treating or slowing the progression of cancer in a subject, comprising administering to the subject an anti-huICOS antibody (e.g., ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7, and 2644) in combination with another anti-cancer agent or antitumor therapy. In some embodiments, the anti-huICOS antibody may be administered in combination with chemotherapy or a chemotherapeutic agent, or in combination with radiation therapy or a radiotherapeutic agent, as described above. In some embodiments, the anti-huICOS antibody may be administered in combination. In some embodiments, the anti-huICOS antibody may be administered in combination with a targeted therapy or a targeted therapeutic agent. In some embodiments, the anti-huICOS antibody may be administered in combination with immunotherapy or an immunotherapeutic agent, such as a monoclonal antibody.

In some embodiments, an anti-huICOS antibody described herein can be combined with (i) another costimulatory receptor agonist and/or (ii) an inhibitory signal antagonist on T cells. In some embodiments, a combination therapy comprising an anti-huICOS antibody and an agonist and/or antagonist results in an enhanced antigen-specific T cell response in the subject. In some embodiments, the anti-ICOS antibodies described herein can be administered in combination with an agent that targets co-stimulatory and co-inhibitory molecules that are members of the immunoglobulin superfamily (IgSF) to increase the immune response. In some embodiments, the anti-ICOS antibodies (e.g., ICOS.33 IgG1f S267E, 17C4, 9D5, 3E8, 1D7, and 2644) described herein may be administered in combination with an agent that targets the ligand of a co-stimulatory or co-inhibitory molecule. The family of membrane-bound ligands that bind to co-stimulatory or co-inhibitory receptors is the B7 family, which includes B7-1, B7-2, B7-H1 (PD-L1), B7-DC (PD-L2), B7-H2 (ICOS -L), B7-H3, B7-H4, B7-H5 (VISTA) and B7H6. Another family of membrane-bound ligands that bind to co-stimulatory or co-inhibitory receptors is the TNF family of molecules that bind to related members of the TNF receptor family, which include CD40, CD40L, OX-40, OX-40L, CD70, CD27L, CD30, CD30L, 4-1BBL, CD137/4-1BB, TRAIL/Apo2-L, TRAILR1/DR4, TRAILR2/DR5, TRAILR3, TRAILR4, OPG, RANK, RANKL, TWEAKR/Fn14, TWEAK, BAFFR, EDAR, XEDAR, TACI, APRIL, BCMA, LTeR, LIGHT, DcR3, HVEM, VEGI/TL1A, TRAMP/DR3, EDAR, EDA1, XEDAR, EDA2, TNFR1,

- 54 041306 lymphotoxin α/TNFe, TNFR2, TNFa, LTeR, lymphotoxin α1β2, FAS, FASL, RELT, DR6, TROY, NGFR.

In another aspect, anti-huICOS antibodies can be used in combination with cytokine antagonists that inhibit T cell activation (e.g., IL-6, IL-10, TGF-β, VEGF; or other immunosuppressive cytokines, or cytokines that stimulate T cell activation). -cells), to stimulate an immune response, for example in the treatment of proliferative diseases such as cancer.

In one aspect, T cell responses are stimulated by a combination of an anti-huICOS antibody described herein and one or more of (i) a protein antagonist that inhibits T cell activation (e.g., immune checkpoint inhibitors), such as CTLA-4 , PD-1, PD-L1, PD-L2, LAG-3, TIM-3, Galectin 9, SEACAM-1, BTLA, CD69, Galectin-1, TIGIT, CD113, GPR56, VISTA, 2B4, CD48, GARP, PD1H, LAIR1, TIM-1 and TIM-4, and (ii) a protein agonist that stimulates T cell activation such as B7-1, B7-2, CD28, 4-1BB (CD137), 4-1BBL, CD40 , ICOS-L, OX40, OX40L, GITR, GITRL, CD70, CD27, DR3 and CD28H.

Examples of agents that modulate one of the above proteins and can be combined with anti-huICOS agonist antibodies, such as the antibodies described herein, for the treatment of cancer include: YERVOY®/ipilimumab or tremelimumab (to CTLA-4), galiximab (to B7.1), BMS-936558 (to PD-1), pidilizumab/CT-011 (to PD-1), KEYTRUDA®/pembrolizumab/MK3475 (to PD-1), AMP224 (to B7-DC/PD -L2), BMS-936559 (to B7-H1), MPDL3280A (to B7-H1), CP-870893 or dacetuzumab/SGN-40 (CD40 - Kirkwood et al. (2012) C.A. Cancer J. Clin. 62 :309; Vanderheide & Glennie (2013) Clin. Cancer Res. 19:1035), AMG557 (to B7H2), MGA271 (to B7H3 - WO 11/109400), IMP321 (to LAG3), urelumab/BMS-663513 and PF- 05082566 (to CD137/4-1BB), varlilumab/CDX-1127 (to CD27), MEDI-6383 and MEDI-6469 (to OX40), RG-7888 (to OX40L - WO 06/029879), atcicept (to TACI) , muromonab-CD3 (to CD3), ipilimumab (to CTLA-4). Accordingly, in one embodiment, an anti-huICOS antibody (such as ICOS.33 IgG1f S267E) is combined with an anti-PD-1 antibody (such as nivolumab) and/or an anti-CTLA-4 antibody (such as ipilimumab).

Other molecules that can be combined with anti-huICOS agonist antibodies for cancer treatment include inhibitory receptor antagonists on NK cells or activating receptor agonists on NK cells. For example, anti-huICOS agonist antibodies can be combined with KIR antagonists (eg, lirilumab).

Still other agents for combination therapies include agents that inhibit or deplete macrophages or monocytes, including but not limited to CSF-1R antagonists such as CSF-1R antagonist antibodies, including RG7155 (WO 11/70024, WO 11/107553, WO 11/131407, WO 13/87699, WO 13/119716, WO 13/132044) or FPA-008 (WO 11/140249; WO 13/169264; WO 14/036357).

In some embodiments, the anti-huICOS agonist antibodies described herein are used together with one or more of agonist agents that ligate positive co-stimulatory receptors, blocking agents that attenuate signal transduction through inhibitory receptors, and one or more agents that systemically increase the frequency of antitumor T cells, agents that traverse various immunosuppressive pathways in the tumor microenvironment (eg, block an inhibitory receptor interaction (eg, PD-L1/PD-1 interactions), deplete or inhibit Tregs (eg, using monoclonal anti-CD25 antibodies (eg, daclizumab) or by ex vivo depletion of anti-CD25 antibody coated beads), inhibit metabolic enzymes such as IDO, or reverse/prevent T cell anergy or depletion) and agents that induce innate immune activation and /or inflammation in areas of the tumor.

Provided herein are methods of stimulating an immune response in a subject, comprising administering to the subject an ICOS agonist, e.g., an antibody, and one or more additional immunostimulatory antibodies, such as a PD-1 antagonist, e.g., an antagonist antibody, a PD-L1 antagonist, e.g. , a CTLA-4 antagonist, such as an antagonist antibody, and/or a LAG3 antagonist, such as an antagonist antibody, such that an immune response is stimulated in the subject, such as to inhibit tumor growth or to stimulate an antiviral response. In one embodiment, an anti-huICOS agonist antibody and an anti-PD-1 antagonist antibody are administered to the subject. In one embodiment, an anti-huICOS agonist antibody and an anti-PD-L1 antagonist antibody are administered to the subject. In one embodiment, an anti-huICOS agonist antibody and an anti-CTLA4 antagonist antibody are administered to the subject. In one embodiment, at least one additional immunostimulatory antibody (e.g., anti-PD-1 antagonist antibody, anti-PD-L1 antagonist antibody, anti-CTLA-4 antagonist antibody, and/or anti-LAGS antagonist antibody) is a human antibody. . Alternatively, at least one additional immunostimulatory antibody may be, for example, a chimeric or humanized antibody (e.g., derived from mouse or hamster anti-PD-1, anti-PD-L1, anti-CTLA-4, and/or anti-LAG3 antibodies).

This document provides methods for the treatment of hyperproliferative disease (on

- 55 041306 example, cancer) comprising administering to the subject an anti-huICOS agonist antibody and an anti-PD-l antagonist antibody. In some embodiments, the cancer is non-small cell lung cancer (NSCLC) or colorectal cancer (CRC). In some embodiments, the cancer is characterized by tumors (i) overexpressing CD32A/CD32B (FcyRIIa/Fcy) and/or (ii-a) overexpressing ICOS, or (ii-b) underexpressing ICOS-L, e.g. cytometry or immunohistochemistry (IHC). Tumor types with moderate to high expression of ICOS RNA include head and neck, lung, cervix, kidney, pancreas, breast, and colorectal cancers, suggesting that these cancer types may also show increased expression of the ICOS protein. In some embodiments, the agonist is administered at a subtherapeutic dose, the anti-PD-1 antibody is administered at a subtherapeutic dose, or both are administered at a subtherapeutic dose. Also provided herein are methods for modifying an adverse event associated with treatment of a hyperproliferative disease with an immunostimulatory agent. In one embodiment, the method includes administering to the subject an anti-huICOS agonist antibody and a subtherapeutic dose of an anti-PD-1 antibody. In some embodiments, the subject is a human. In some embodiments, the anti-PD-1 antibody is a human monoclonal antibody and the anti-huICOS agonist antibody is a humanized monoclonal antibody, such as an antibody containing the CDRs or variable regions of the antibodies described herein.

Anti-PD-I antibodies that are known in the art can be used in the methods described herein. Various human monoclonal antibodies that specifically bind to PD-1 with high affinity are described in US Pat. No. 8,008,449. Human anti-PD-1 antibodies disclosed in US Pat. No. 8,008,449 have been shown to exhibit one or more of the following characteristics: (a) bind to human PD-1 with a KD value of 1x10'7 M or less as measured by surface plasmon resonance analysis using the Biacore biosensor system; (b) do not significantly bind to human CD28, CTLA-4 or ICOS; (c) increase T cell proliferation in a mixed culture lymphocyte response (MLR) assay; (d) increase interferon-y production in the MLR assay; (e) increase IL-2 secretion in the MLR assay; (f) bind to human PD-1 and cynomolgus monkey PD-1; (g) inhibit the binding of PD-L1 and/or PD-L2 to PD-1; (h) stimulate antigen-specific memory responses; (i) stimulate antibody responses; and (j) inhibit the growth of tumor cells in vivo. Anti-PD-1 antibodies useful in the present invention include monoclonal antibodies that specifically bind to human PD-1 and exhibit at least one, in some embodiments, at least five of the preceding characteristics.

Other monoclonal anti-PD-l antibodies are described, for example, in US Pat. , Wo 2015/112800, Wo 2014/206107, Wo 2015/35606, WO 2015/085847, Wo 2014/179664, Wo 2017/020291, Wo 2017/020858, Wo 2016/197367, Wo 2017/024515, Wo 2017/025051 , Wo 2017/123557, Wo 2016/106159, Wo 2014/194302, Wo 2017/040790, Wo 2017/133540, Wo 2017/132827, Wo 2017/024465, Wo 2017/025016, Wo 2017/106061, Wo 2017/19846 , WO 2017/024465, WO 2017/025016, WO 2017/132825 and WO 2017/133540, each of which is incorporated herein by reference in its entirety.

In some embodiments, the anti-PD-1 antibody is nivolumab (also known as OPDIVO®, 5C4, BMS-936558, MDX-1106, and ONO-4538), pembrolizumab (Merck; also known as KEYTRUDA®, lambrolizumab, and MK-3475 ; see WO 2008/156712), PDR001 (Novartis; see WO 2015/112900), MEDI-0680 (AstraZeneca; also known as AMP-514; see WO 2012/145493), cemiplimab (Regeneron; also known as REGN -2810; see WO 2015/112800), JS001 (TAIZHOU JUNSHI PHARMA; see Si-Yang Liu et al., J. Hematol. Oncol. 10:136 (2017)), BGB-A317 (Beigene; see WO 2015/35606 and US 2015/0079109), INCSHR1210 (Jiangsu Hengrui Medicine; also known as SHR-1210; see WO 2015/085847; Si-Yang Liu et al, J. Hematol. Oncol. 10:136 (2017)) , TSR-042 (Tesaro Biopharmaceutical; also known as ANB011; see WO2014/179664), GLS-010 (Wuxi/Harbin Gloria Pharmaceuticals; also known as WBP3055; see Si-Yang Liu et al., J. Hematol. Oncol 10:136 (2017)), AM-0001 (Armo), STI-1110 (Sorrento Therapeutics; see WO 2014/194302), AGEN2034 (Agenus; see WO 2017/040790), MGA012 (Macrogenics, see WO 2017/19846) or IBI308 (Innovent; see WO 2017/024465, WO 2017/025016, WO 2017/132825 and WO 2017/133540).

In one embodiment, the anti-PD-1 antibody is nivolumab. Nivolumab is an immune checkpoint inhibitor, which is a fully human IgG4 (S228P) anti-PD-1 antibody that selectively prevents interaction with PD1 ligands (PD-L1 and PD-L2), thereby blocking the downregulation of antitumor functions of T cells (patent). US 8008449; Wang et al., 2014 Cancer Immunol Res. 2(9):846-56).

In another embodiment, the anti-PD-1 antibody is pembrolizumab. Pembrolizumab is a humanized IgG4 (S228P) monoclonal antibody directed to

- 56 041306 for the human cell surface receptor PD-1 (programmed death-1 or programmed cell death-1). Pembrolizumab is described, for example, in US Pat. Nos. 8,354,509 and 8,900,587.

Anti-PD-1 antibodies used in the disclosed methods also include isolated antibodies that specifically bind to human PD-1 and cross-compete for binding to human PD-1 with any anti-PD-1 antibody disclosed herein, for example, nivolumab (see, for example, US Pat. Nos. 8,008,449 and 8,779,105; WO 2013/173223). In some embodiments, an anti-PD-1 antibody binds to the same epitope as any of the anti-PD-1 antibodies described herein, such as nivolumab. The ability of antibodies to cross-compete for binding to an antigen indicates that these monoclonal antibodies bind to the same region of an antigen epitope and sterically make it difficult for other cross-competing antibodies to bind to that particular epitope region. These cross-competing antibodies are expected to have functional properties very similar to those of a reference antibody such as nivolumab due to their binding to the same epitope region of PD-1. Cross-competing antibodies can be easily identified based on their ability to cross-compete with nivolumab in standard PD-1 binding assays such as Biacore assay, ELISA assays or flow cytometry (see, for example, WO 2013/173223).

In some embodiments, antibodies that cross-compete for binding to human PD-1 or bind to the same epitope region of a human anti-PD-1 antibody as nivolumab are monoclonal antibodies. For administration to human subjects, these cross-competing antibodies are chimeric antibodies, engineered antibodies, or humanized antibodies, or human antibodies. Such chimeric, engineered, humanized, or human monoclonal antibodies can be made and isolated by methods well known in the art.

Anti-PD-I antibodies used in the methods of the disclosed invention also include antigen-binding portions of the above antibodies. It has been clearly demonstrated that the antigen-binding function of an antibody can be performed by fragments of a full length antibody.

Anti-PD-I antibodies suitable for use in the disclosed methods or compositions are antibodies that bind to PD-1 with high specificity and affinity, block PD-L1 and/or PD-L2 binding, and inhibit the immunosuppressive effect of the PD signaling pathway. -1. In any of the compositions or methods disclosed herein, an anti-PD-1 antibody comprises an antigen-binding portion or fragment that binds to the PD-1 receptor and exhibits functional properties similar to those of whole antibodies in inhibiting ligand binding and upregulating immune systems. In some embodiments, the anti-PD-1 antibody, or antigen-binding portion thereof, cross-competes with nivolumab for binding to human PD-1.

Provided herein are methods for treating a hyperproliferative disease (eg, cancer) comprising administering an anti-huICOS agonist antibody and an anti-PD-L1 antagonist antibody to a subject. In some embodiments, the anti-huICOS agonist antibody is administered at a subtherapeutic dose, the anti-PD-L1 antibody is administered at a subtherapeutic dose, or both are administered at a subtherapeutic dose. Provided herein are methods for altering an adverse event associated with treatment of a hyperproliferative disease with an immunostimulatory agent, comprising administering an anti-huICOS agonist antibody and a subtherapeutic dose of an anti-PDL1 antibody to a subject. In certain embodiments, the subject is a human. In some embodiments, the anti-PD-L1 antibody is a human monoclonal antibody and the anti-huICOS agonist antibody is a humanized monoclonal antibody, such as an antibody containing the CDRs or variable regions of the antibodies disclosed herein.

Anti-PD-LI antibodies that are known in the art can be used in the methods of the present invention. Examples of anti-PD-L1 antibodies useful in the methods of the present invention include those disclosed in US Pat. No. 9,580,507. Human monoclonal anti-PD-L1 antibodies disclosed in US Pat. : (a) bind to human PD-L1 with a KD value of 1x10'7 M or less as measured by SPR using the Biacore biosensor system; (b) increase T cell proliferation in the Mixed Lymphocyte Reaction (MLR) assay; (c) increase interferon-y production in the MLR assay; (d) increase IL-2 secretion in the MLR assay; (e) stimulate antibody responses; and (f) reverse the effect of regulatory T cells on effector and/or dendritic T cells. Anti-PD-L1 antibodies useful in the present invention include monoclonal antibodies that specifically bind to human PD-L1 and exhibit at least one, in some embodiments, at least five of the preceding characteristics.

In some embodiments, the anti-PD-L1 antibody is BMS-936559 (also known as 12A4, MDX-1105; see, for example, US Pat. as TECENTRIQ®, MPDL3280A, RG7446, see US 8217149, see also Herbst et al (2013) J Clin Oncol 31(suppl): 3000), durvalumab (AstraZeneca; also known as IMFINZI™, MEDI-4736; see WO 2011/066389), avelumab (Pfizer; also known as BAVENCIO®, MSB0010718C; see WO 2013/079174), STI-1014 (Sorrento; see WO 2013/181634), CX-072 (Cytomx; see WO 2016/149201), KN035 (3D Med/Alphamab; see Zhang et al., Cell Discov. 7:3 (March 2017), LY3300054 (Eli Lilly Co.; see e.g. WO 2017/034916) or CK -301 (Checkpoint Therapeutics; see Gorelik et al., AACR: Abstract 4606 (Apr 2016)).

In some embodiments, the anti-PD-L1 antibody is atezolizumab (TECENTRIQ®). Atezolizumab is a fully humanized anti-PD-L1 IgG1 monoclonal antibody.

In some embodiments, the anti-PD-L1 antibody is durvalumab (IMFINZI™). Durvalumab is a human anti-PD-L1 IgG1-kappa monoclonal antibody.

In some embodiments, the anti-PD-L1 antibody is avelumab (BAVENCIO®). Avelumab is a human anti-PDL1 IgG1-lambda monoclonal antibody.

In other embodiments, the αntu-PD-L1 monoclonal antibody is 28-8, 28-1, 28-12, 29-8, 5H1, or any combination thereof.

Anti-PD-LI antibodies used in the disclosed methods also include isolated antibodies that specifically bind to human PD-L1 and cross-compete for binding to human PD-L1 with any anti-PD-L1 antibody disclosed herein, e.g. , atezolizumab, durvalumab and/or avelumab. In some embodiments, an anti-PD-L1 antibody binds to the same epitope as any of the anti-PD-L1 antibodies described herein, such as atezolizumab, durvalumab, and/or avelumab. The ability of antibodies to cross-compete for binding to an antigen indicates that these antibodies bind to the same epitope region of the antigen and sterically make it difficult for other cross-competing antibodies to bind to that particular epitope region. These cross-competing antibodies are expected to have functional properties very similar to those of a reference antibody such as atezolizumab and/or avelumab due to their binding to the same epitope region of PD-L1. Cross-competing antibodies can be easily identified based on their ability to cross-compete with atezolizumab and/or avelumab in standard PD-L1 binding assays such as Biacore assay, ELISA assays, or flow cytometry (see e.g. WO 2013/173223).

In some embodiments, antibodies that cross-compete for binding to human PD-L1 or bind to the same epitope region of a human anti-PD-L1 antibody as atezolizumab, durvalumab, and/or avelumab are monoclonal antibodies. For administration to human subjects, these cross-competing antibodies are chimeric antibodies, engineered antibodies, or humanized antibodies, or human antibodies. Such chimeric, engineered, humanized, or human monoclonal antibodies can be made and isolated by methods well known in the art.

Anti-PD-LI antibodies used in the methods of the disclosed invention also include the antigen-binding portions of the above antibodies. It has been clearly demonstrated that the antigen-binding function of an antibody can be performed by fragments of a full length antibody.

Anti-PD-LI antibodies suitable for use in the disclosed methods or compositions are antibodies that bind to PD-L1 with high specificity and affinity, block PD-1 binding, and inhibit the immunosuppressive effect of the PD-1 signaling pathway. In any of the compositions or in any of the methods disclosed herein, an anti-PD-L1 antibody comprises an antigen-binding portion or fragment that binds to PD-L1 and exhibits functional properties similar to those of whole antibodies in inhibiting receptor binding and upregulating immune system. In some embodiments, the anti-PD-L1 antibody, or antigen-binding portion thereof, cross-competes with atezolizumab, durvalumab, and/or avelumab for binding to human PD-L1.

In one embodiment, an anti-huICOS agonist antibody of the present invention is combined with a PD-1/PD-L1 signaling antagonist, such as a PD-1 antagonist (e.g., nivolumab, also known as MDX1106, as described in WO 06/121168) or an antagonist PD-L1, in combination with a third immunotherapeutic agent (eg, an anti-ICOS antibody such as ICOS.33 IgG1f S267E in combination with nivolumab and ipilimumab). In one embodiment, the third immunotherapeutic agent is an anti-CTLA-4 antagonist antibody. In some embodiments, the anti-CTLA-4 antibody is YERVOY® (ipilimumab or the 10D1 antibody described in PCT Publication WO 01/14424) or tremelimumab (formerly ticilimumab, CP-675,206). In one embodiment, the third immunotherapeutic agent is a GITR antagonist or OX-40 antagonist, such as anti-GITR or anti-OX40 antibodies disclosed in

- 58 041306 of this document. In one embodiment, the third immunotherapeutic agent is a GITR agonist, such as a GITR agonist antibody. Suitable anti-GITR antibodies include, for example, BMS-986153, BMS-986156, TRX-518 (WO 06/105021, WO 09/009116) and MK-4166 (WO11/028683). In one embodiment, the third immunotherapeutic agent is an IDO antagonist. Suitable Ido antagonists include, for example, INCB-024360 (WO 2006/122150, WO 07/75598, WO 08/36653, WO 08/36642), indoxymod or NLG-919 (WO 09/73620, WO 09/1156652, WO 11 /56652, WO 12/142237).

Provided herein are methods for treating a hyperproliferative disease (eg, cancer) comprising administering to a subject an anti-huICOS agonist antibody described herein and an anti-CTLA-4 antagonist antibody. In some embodiments, the anti-huICOS agonist antibody is administered at a subtherapeutic dose, the anti-CTLA-4 antibody is administered at a subtherapeutic dose, or both are administered at a subtherapeutic dose. Provided herein are methods for modifying an adverse event associated with treatment of a hyperproliferative disease with an immunostimulatory agent, comprising administering an anti-huICOS agonist antibody and a subtherapeutic dose of an anti-CTLA-4 antibody to a subject. In certain embodiments, the subject is a human.

Ahtu-CTLA-4 antibodies, which are known in the art, can be used in the methods of the present invention. The Ahtu-CTLA-4 antibodies of the present invention bind to human CTLA-4 in such a way that the interaction of CTLA-4 with the human B7 receptor is disrupted. Since the interaction of CTLA-4 with B7 transforms the signal, leading to the inactivation of T cells bearing the CTLA-4 receptor, disruption of the interaction effectively induces, enhances or prolongs the activation of such T cells, thereby inducing, enhancing or prolonging the immune response.

Human monoclonal antibodies that specifically bind to CTLA-4 with high affinity are disclosed in US Pat. No. 6,984,202. Other monoclonal antibodies against CTLA-4 are described, for example, in US Pat. /122444, WO 2007/113648, WO 2016/196237 and WO 2000/037504, each of which is incorporated herein by reference in its entirety. Human monoclonal anti-CTLA-4 antibodies disclosed in US Pat. No. 6,984,202 have been shown to exhibit one or more of the following characteristics: (a) bind specifically to human CTLA-4 with a binding affinity expressed as an equilibrium association constant (Ka) , constituting at least about 10 ⁷ M ^-1 or about 10 ⁹ M ^-1 , or from about 10 ¹⁰ M ^-1 to 10 ¹¹ M ^-1 or higher, according to analysis on Biacore; (b) the association kinetic constant (k _a ) is at least about 10 ³ , about 10 ⁴ , or about 105 m ^-1 s ^-1 ; (c) the kinetic dissociation constant (kd) is at least about 10 ³ , about 10 ⁴ or about 105 m ^-1 s ^-1 and (d) inhibits CTLA-4 binding to B7-1 (CD80) and B7-2 ( CD86). Anti-CTLA-4 antibodies useful for the present invention include monoclonal antibodies that specifically bind to human CTLA-4 and exhibit at least one, at least two, or at least three of the preceding characteristics.

In some embodiments, the anti-CTLA-4 antibody is ipilimumab (also known as YERVOY®, MDX-010, 10D1; see US Pat. No. 6,984,202), MK-1308 (Merck), AGEN-1884 (Agenus Inc.; see WO 2016/196237) or tremelimumab (AstraZeneca; also known as ticilimumab, CP-675,206; see WO 2000/037504 and Ribas, Update Cancer Ther. 2(3): 133-39 (2007)). In specific embodiments, the anti-CTLA-4 antibody is ipilimumab.

In specific embodiments, the anti-CTLA-4 antibody is ipilimumab for use in the methods disclosed herein. Ipilimumab is a fully human IgG1 monoclonal antibody that blocks the binding of CTLA-4 to its B7 ligands, thereby stimulating T cell activation and improving overall survival (OS) in patients with advanced melanoma.

In specific embodiments, the anti-CTLA-4 antibody is tremelimumab.

In specific embodiments, the anti-CTLA-4 antibody is MK-1308.

In specific embodiments, the anti-CTLA-4 antibody is AGEN-1884.

Anti-CTLA-4 antibodies used in the disclosed methods also include isolated antibodies that specifically bind to human CTLA-4 and cross-compete for binding to human CTLA-4 with any anti-CTLA-4 antibody disclosed herein, e.g. , ipilimumab and/or tremelimumab. In some embodiments, the anti-CTLA-4 antibody binds to the same epitope as any of the anti-CTLA-4 antibodies described herein, eg, ipilimumab and/or tremelimumab. The ability of antibodies to cross-compete for binding to an antigen indicates that these antibodies bind to the same epitope region of the antigen and sterically make it difficult for other cross-competing antibodies to bind to that particular epitope region. These cross-competing antibodies are expected to have functional properties very similar to those of a reference antibody such as ipilimumab and/or tremelimumab due to their binding to the same CTLA-4 epitope region. Cross-competing antibodies can be easily identified based on their ability to

- 59 041306 cross-compete with ipilimumab and/or tremelimumab in standard CTLA-4 binding assays such as Biacore assay, ELISA assays or flow cytometry (see, for example, WO 2013/173223).

In some embodiments, antibodies that cross-compete for binding to human CTLA-4 or bind to the same epitope region of a human anti-CTLA-4 antibody as ipilimumab and/or tremelimumab are monoclonal antibodies. For administration to human subjects, these cross-competing antibodies are chimeric antibodies, engineered antibodies, or humanized antibodies, or human antibodies. Such chimeric, engineered, humanized, or human monoclonal antibodies can be made and isolated by methods well known in the art.

Ahtu-CTLA-4 antibodies useful in the methods of the disclosed invention also include antigen-binding portions of the above antibodies. It has been clearly demonstrated that the antigen-binding function of an antibody can be performed by fragments of a full length antibody.

Ahtu-CTLA-4 antibodies suitable for use in the disclosed methods or compositions are antibodies that bind to CTLA-4 with high specificity and affinity, block CTLA-4 activity, and disrupt the interaction of CTLA-4 with the human B7 receptor. In any of the compositions or in any of the methods disclosed herein, an anti-CTLA-4 antibody comprises an antigen-binding portion or fragment that binds to CTLA-4 and exhibits functional properties similar to those of whole antibodies in inhibiting the interaction of CTLA-4 with human B7 receptor and upregulation of the immune system. In some embodiments, the anti-CTLA-4 antibody, or antigen-binding portion thereof, cross-competes with ipilimumab and/or tremelimumab for binding to human CTLA-4.

In one embodiment, an anti-huICOS agonist antibody of the present invention is combined with an anti-CTLA-4 antibody in combination with a third immunotherapeutic agent. In one embodiment, the third immunotherapeutic agent is a GITR antagonist or OX-40 antagonist, such as anti-GITR or anti-OX40 antibodies disclosed herein. In one embodiment, the third immunotherapeutic agent is a GITR agonist, such as an anti-GITR agonist antibody. Suitable anti-GITR antibodies include, for example, BMS-986153, BMS-986156, TRX-518 (WO 06/105021, WO 09/009116) and MK-4166 (WO 11/028683). In one embodiment, the third immunotherapeutic agent is an IDO antagonist. Suitable IDO antagonists include, for example, INCB-024360 (WO 2006/122150, WO 07/75598, WO 08/36653, WO 08/36642), indoxymod or NLG-919 (WO 09/73620, WO 09/1156652, WO 11 /56652, WO 12/142237).

Provided herein are methods for treating a hyperproliferative disease (eg, cancer) comprising administering an anti-huICOS agonist antibody and an anti-LAG-3 antibody to a subject. In other embodiments, the anti-huICOS agonist antibody is administered at a subtherapeutic dose, the anti-LAG-3 antibody is administered at a subtherapeutic dose, or both are administered at a subtherapeutic dose. Provided herein are methods for altering an adverse event associated with treatment of a hyperproliferative disease with an immunostimulatory agent, comprising administering an anti-huICOS agonist antibody and a subtherapeutic dose of an anti-LAG-3 antibody to a subject. In some embodiments, the subject is a human. In some embodiments, the anti-LAG-3 antibody is a human sequence monoclonal antibody and the anti-huICOS agonist antibody is a humanized monoclonal antibody, such as an antibody containing the CDRs or variable regions of the antibodies disclosed herein. Examples of anti-LAG3 antibodies include antibodies containing the CDRs or variable regions of antibodies 25F7, 26H10, 25E3, 8B7, 11F2, or 17E5, which are described in US Patent Publications US 2011/0150892 and WO 2014/008218. In one embodiment, the anti-LAG-3 antibody is BMS-986016. Other anti-LAG-3 antibodies that can be used include IMP731 described in US 2011/007023 or IMP-321. Ahtu-LAG-3 antibodies that compete with or bind to the same epitope as any of these antibodies can also be used in combination treatments.

In some embodiments, the anti-LAG-3 antibody binds to human LAG-3 with a K _D of 5x10'8 M or less, binds to human LAG-3 with a K _D of 1x10'8 M or less, binds to human LAG-3 with a K _D 5x10'9 M or less, or binds to human LAG-3 with a K _D ranging from 1x10'8 M to 1x10' ¹⁰ M or less.

Administration of agonist anti-huICOS antibodies described herein and antagonists, e.g. antagonist antibodies, to one or more second target antigens such as LAG-3 and/or CTLA-4 and/or PD-1 and/or PD-L1 may enhance the immune response to cancer cells in a patient. Types of malignant tumors, the growth of which can be inhibited using antibodies according to the present disclosure, include types of cancer that usually respond to immunotherapy. Examples of cancers to be treated with the combination therapy described in this

- 60 041306 herein include, but are not limited to, those listed above in the discussion of anti-huICOS agonist antibody monotherapy.

In some embodiments, the combination of therapeutic antibodies discussed herein can be administered simultaneously as a single composition in a pharmaceutically acceptable carrier, or simultaneously as separate compositions in which each antibody is in a pharmaceutically acceptable carrier. In another embodiment, the combination of therapeutic antibodies can be administered sequentially. For example, an anti-CTLA-4 antibody and an anti-huICOS agonist antibody can be administered sequentially, for example, an anti-CTLA-4 antibody is administered first and an anti-huICOS agonist antibody is administered second, or an anti-huICOS agonist antibody is administered first and an anti-CTLA-4 antibody enter second. Additionally or alternatively, the anti-PD-1 antibody and the anti-huICOS agonist antibody can be administered sequentially, for example, the anti-PD-1 antibody is administered first and the anti-huICOS agonist antibody is administered second, or the anti-huICOS agonist antibody is administered first and the anti- The PD-1 antibody is administered second. Additionally or alternatively, the anti-PD-L1 antibody and the anti-huICOS agonist antibody can be administered sequentially, for example, the anti-PD-L1 antibody is administered first and the anti-huICOS agonist antibody is administered second, or the anti-huICOS agonist antibody is administered first and the anti-PD- L1 antibody is administered second. Additionally or alternatively, the anti-LAG-3 antibody and the anti-huICOS agonist antibody can be administered sequentially, for example, the anti-LAG-3 antibody is administered first and the anti-huICOS agonist antibody is administered second, or the anti-huICOS agonist antibody is administered first and the anti- LAG-3 antibody is administered second.

In addition, if more than one dose of combination therapy is administered consecutively, the order of sequential administration may be reversed or the same administration order may be maintained at each administration time point, consecutive administrations may be combined with co-administrations, or any combination thereof. For example, the first administration of a combination of an anti-CTLA-4 antibody and an anti-huICOS agonist antibody may be co-administration, a second administration may be sequential with an anti-CTLA-4 antibody first and an anti-huICOS agonist antibody second, and a third administration may be sequential with an anti-huICOS agonist antibody. huICOS antibody first and anti-CTLA-4 antibody second, and so on. Additionally or alternatively, the first administration of the combination of anti-PD-1 antibody and anti-huICOS agonist antibody may be co-administered, the second administration may be sequential with the anti-PD-1 antibody first and the anti-huICOS agonist antibody second, and the third administration may be sequential with the anti-huICOS agonist. antibody first and anti-PD-1 antibody second, and so on. Additionally or alternatively, the first administration of the combination of the anti-PD-L1 antibody and the anti-huICOS agonist antibody may be concurrent, the second administration may be sequential with the anti-PD-L1 antibody and the second anti-HuICOS agonist antibody, and the third administration may be sequential with an anti-huICOS agonist antibody first and an anti-PD-Ll antibody second; and so on. Additionally or alternatively, the first administration of the combination of the anti-LAG-3 antibody and the anti-huICOS agonist antibody may be concurrent, the second administration may be sequential with the anti-LAG-3 antibody first and the anti-huICOS agonist antibody second, and the third administration may be sequential with anti-huICOS agonist antibody first and anti-LAG-3 antibody second, etc. Another representative dosing regimen may include a first administration that is sequential with an anti-huICOS agonist antibody first and an anti-CTLA-4 antibody second (and/or an anti-PD-1 antibody and/or an anti-PD-L1 antibody, and/or anti-LAG-3 antibody) second, and the following administrations may be simultaneous.

In one embodiment, an anti-huICOS agonist antibody as the sole immunotherapeutic agent, or a combination of an anti-huICOS agonist antibody and one or more additional immunotherapeutic antibodies (e.g., anti-CTLA-4 and/or anti-PD-1 and/or anti- PD-L1 and/or anti-LAG-3 antibody) can be further combined with an immunogenic agent such as cancer cells, purified tumor antigens (including recombinant proteins, peptides and carbohydrate molecules), cells and cells transfected with genes encoding immunostimulatory cytokines ( He et al (2004) J Immunol 173:4919-28). Non-limiting examples of cancer vaccines that can be used include melanoma antigen peptides such as gp100 peptides, MAGE, Trp-2, MART1 and/or tyrosinase antigens, or tumor cells transfected to express the cytokine GM-CSF (discussed further below). An ICOS agonist and one or more additional antibodies (eg blockade of CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG-3) can also be further combined with standard cancer therapies. For example, an ICOS agonist and one or more additional antibodies (eg, blockade of CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG-3) can be combined with chemotherapy regimens. In one embodiment, an anti-huICOS agonist antibody is administered to a patient with an anti-CTLA-4 antibody and/or an anti-PD-1 antibody and/or an anti-PD-L1 antibody and/or an anti-LAG-3 antibody in combination with decarbazine for the treatment of melanoma. In one embodiment, an anti-huICOS agonist antibody is administered to a patient with an anti-CTLA-4 antibody and/or an anti-PD-1 antibody and/or an anti-PD-L1 antibody,

- 61 041306 and/or aHTu-LAG-3 antibody in combination with interleukin-2 (IL-2) for the treatment of cancer, including melanoma. Without wishing to be bound by any theory, the combined use of ICOS agonism and CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG-3 antagonism with chemotherapy may function synergistically since the cytotoxic effects of most chemotherapeutic compounds may lead to increased levels of tumor antigen in the way of antigen presentation. Other combination therapies that may result in synergism with or without combined ICOS agonism with or without CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG-3 antagonism via cytotoxicity include radiation, surgical interference or shutdown of endocrine function. In another embodiment, angiogenesis inhibitors can be combined with an anti-huICOS antibody and CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG-3 antagonism.

In one embodiment, an anti-huICOS antibody as the sole immunotherapeutic agent or a combination of an anti-huICOS antibody and antibodies blocking CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG-3 can also be used. in combination with bispecific antibodies that target Fca or Fcy receptor-expressing effector cells to tumor cells. See, for example, US Pat. Nos. 5,922,845 and 5,837,243. Bispecific antibodies can be used to target two separate antigens. The T cell arm of these responses will be increased using a combination of ICOS agonism and blockade of CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG3.

In one embodiment, an anti-ICOS antibody as the sole immunotherapeutic agent, or a combination of an anti-ICOS antibody and an additional immunostimulatory agent, e.g., an anti-CTLA-4 antibody and/or an anti-PD-1 antibody, and/or an anti-PD-L1 antibodies and/or antiLAG-3 antibodies can be used in combination with an anticancer agent such as RITUXAN® (rituximab), HERCEPTIN® (trastuzumab), BEXXAR® (tositumomab), ZEVALIN® (ibritumomab), CAMPATH® (alemtuzumab), LYMPHOCIDE ® (eprtuzumab), AVASTIN® (bevacizumab), and TARCEVA® (erlotinib). By way of example and without wishing to be bound by any theory, treatment with an anti-cancer antibody or a toxin-conjugated anti-cancer antibody can result in the death of cancer cells (e.g., tumor cells), which can potentiate an immune response mediated by an immunostimulatory agent, such as an anti-ICOS antibody. , an anti-TIGIT antibody, an anti-CTLA-4 antibody, an anti-PD-1 antibody, an anti-PD-L1 antibody, or an anti-LAG-3 antibody. In one embodiment, treatment of a hyperproliferative disease (e.g., cancer) may include an anti-cancer agent, e.g., an antibody, in combination with an anti-huICOS agonist antibody and optionally an additional immunostimulatory agent, e.g., an anti-CTLA-4 antibody and/or an anti -PD-1 antibody and/or anti-PD-L1 antibody and/or anti-LAG-3 antibody, simultaneously or sequentially, or any combination thereof, which can potentiate antitumor immune responses in the host.

Provided herein are methods of reducing, attenuating, or eliminating an adverse event associated with treatment of a hyperproliferative disease (e.g., cancer) with an immunostimulatory agent, comprising administering to a subject an anti-huICOS agonist antibody with antu-CTLA-4 and/or anti-PD-1, and /or anti-PD-L1 and/or anti-LAG-3 antibody or without it. In one embodiment, the method provides for reducing the incidence of immunostimulatory therapeutic antibody-induced colitis or diarrhea by administering a non-absorbable steroid to the patient. As used herein, a non-absorbable steroid is a glucocorticoid that exhibits extensive first pass metabolism such that after hepatic metabolism, the bioavailability of the steroid is low, i. less than about 20%. In one embodiment described herein, the non-absorbable steroid is budesonide. Budesonide is a locally acting glucocorticosteroid that is extensively metabolized primarily by the liver after oral administration. ENTOCORT EC® (AstraZeneca) is a pH and time dependent oral formulation of budesonide designed to optimize drug delivery to the ileum and colon. ENTOCORT EC® is approved in the US for the treatment of mild to moderate Crohn's disease involving the ileum and/or ascending colon. The usual oral dose of ENTOCORT EC® for the treatment of Crohn's disease is 6 to 9 mg/day. ENTOCORT EC® is released in the intestine before absorption and is stored in the intestinal mucosa. After passing through the target tissue of the intestinal mucosa, ENTOCORT EC® is extensively metabolized by the cytochrome P450 system in the liver to metabolites with negligible glucocorticoid activity. Thus, bioavailability is low (about 10%). The low bioavailability of budesonide results in an improved therapeutic index compared to other glucocorticoids with less extensive first pass metabolism. Budesonide results in fewer side effects, including less hypothalamic-pituitary suppression, than systemically acting corticosteroids. However, prolonged administration of ENTOCORT EC® may result in systemic effects of glucocorticoids, such as suppression of hypercortisolism and adrenal suppression. See PDR ^58th ed. 2004; 608-62 041306

610.

In one embodiment, an anti-ICOS antibody with or without a CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG-3 antagonist (i.e., immunostimulatory therapeutic antibodies against ICOS and optionally anti -CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 and/or anti-LAG-3 antibodies) in combination with a non-absorbable steroid can be further combined with salicylate. Salicylates include 5-ASA agents such as, for example, sulfasalazine (AZULFIDINE®, Pharmacia &UpJohn); olsalazine (DIPENTUM®, Pharmacia &UpJohn); balsalazid (COLAZAL®, Salix Pharmaceuticals, Inc.); and mesalamine (ASACOL®, Procter & Gamble Pharmaceuticals; PENTASA®, Shire US; CANASA®, Axcan Scandipharm, Inc.; ROWASA®, Solvay).

According to the methods described herein, salicylate administered in combination with an anti-huICOS antibody, with or without antibodies against CTLA-4 and/or PD-1 and/or PD-L1 and/or LAG-3, and a non-absorbable steroid may include any overlapping or sequential administration of a salicylate and a non-absorbable steroid to reduce the incidence of immunostimulatory antibody-induced colitis. Thus, for example, methods for reducing the incidence of colitis induced by immunostimulatory antibodies described herein include administering salicylate and a non-absorbable steroid simultaneously or sequentially (e.g., salicylate is administered 6 hours after administration of a non-absorbable steroid), or any combination thereof. In addition, the salicylate and the non-absorbable steroid may be administered by the same route (eg, both are administered orally) or by different routes (eg, the salicylate is administered orally and the non-absorbable steroid is administered rectally), which may differ from the routes used to administer anti-huICOS. and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 and/or antiLAG-3 antibodies.

The anti-huICOS agonist antibodies and antibody combination therapies described herein can also be used in combination with other well known therapies, which are selected for their particular applicability to the indication being treated (eg, cancer). Combinations of agonistic anti-huICOS antibodies described herein can be used sequentially with known pharmaceutically acceptable agent(s).

In one embodiment, the anti-huICOS agonist antibodies and combination antibody therapies described herein can be used in combination (e.g., simultaneously or separately) with additional treatments such as radiation, chemotherapy (e.g., using camptothecin (CPT-11 ), 5-fluorouracil (5-FU), cisplatin, doxorubicin, irinotecan, paclitaxel, gemcitabine, cisplatin, paclitaxel, carboplatin/paclitaxel (taxol), doxorubicin, 5-fu or camptothecin + apo21/TRAIL (6X combo)), one or multiple proteasome inhibitors (eg, bortezomib or MG132), one or more Bcl-2 inhibitors (eg, BH3I-2' antagonists (bcl-xl inhibitor), indolamine dioxygenase-1 (IDO1) inhibitors (eg, INCB24360), AT -101 (R-(-)-gossypol derivative), ABT-263 (small molecule), GX15-070 (obatoclax) or MCL-1 (myeloid leukemia cell differentiation protein-1)), iAP antagonists (inhibitor of apoptosis proteins) ( e.g. smac 7, smac4, small molecular weight smac mimetic, synthetic smac peptides (see Fulda et al., Nat Med 2002;8:808-15), ISIS23722 (LY2181308) or AEG-35156 (GEM-640)), HDAC inhibitors, anti-CD20 antibodies (eg rituximab), angiogenesis inhibitors (eg , bevacizumab), antiangiogenic agents targeting VEGF and VEGFR (e.g. AVASTIN®), synthetic triterpenoids (see Hyer et al., Cancer Research 2005;65:4799-808), modulators of c-FLIP (cellular FLICE inhibitory protein) (e.g. natural and synthetic PPARy ligands (peroxisome proliferator-activated receptor γ), 5809354 or 5569100), kinase inhibitors (e.g. sorafenib), trastuzumab, cetuximab, temsirolimus, mTOR inhibitors such as rapamycin and temsirolimus, bortezomib, JAK2 inhibitors, HSP90, PI3K-AKT inhibitors, lenalidomide, GSK3P inhibitors, IAP inhibitors and/or genotoxic drugs.

The anti-huICOS agonist antibodies and combination antibody therapies described herein may additionally be used in combination with one or more antiproliferative cytotoxic agents. Classes of compounds that can be used as antiproliferative cytotoxic agents include, but are not limited to, the following:

Alkylating agents (including but not limited to nitrogen mustard, ethyleneimine derivatives, alkylsulfonates, nitrosoureas and triazenes): uracil mustard, chlormethine, cyclophosphamide (CYTOXAN™) phosphamide, melphalan, chlorambucil, pipobroman, triethylenemelamine, triethylenethiophosphoramine, busulfan, carmustine, lomustine, streptozocin, dacarbazine and temozolomide.

Antimetabolites (including, without limitation, folic acid antagonists, pyrimidine analogs, purine analogs, and adenosine deaminase inhibitors): methotrexate, 5-fluorouracil, floxuridine, cytarabine, 6-mercaptopurine, 6-thioguanine, fludarabine phosphate, pentostatin, and gemcitabine.

- 63 041306

Suitable anti-proliferative agents for combination with anti-huICOS agonist antibodies include, but are not limited to, taxanes, paclitaxel (paclitaxel is commercially available as TAXOL™), docetaxel, discodermolide (DDM), dictiostatin (DCT), peloruside A, epothilones, epothilone A, epothilone B , epothilone C, epothilone D, epothilone E, epothilone F, furanoepothilone D, deoxyepothilone B1, [17]-dehydrodeoxyepothilone B, [18]dehydrodeoxyepothilones B, C12,13-cyclopropylepothilone A, C6C8 bridged epothilone A, trans-9, 10-dehydroepothilone D, cis-9,10dehydroepothilone D, 16-desmethylepothilone B, epothilone B10, disdermomolide, patupilone (EPO-906), KOS862, KOS-1584, ZK-EPO, ABJ-789, XAA296A (discodermolide), TZT- 1027 (soblidotin), ILX-651 (tasidotin hydrochloride), halichondrin B, eribulin mesilate (E-7389), chemiasterlin (HTI-286), E-7974, cyrptophycins, LY-355703, maytansinoid immunoconjugates (DM-1), MKC -1, ABT-751, T1-38067, T900607, SB-715992 (ispinesib), SB-743921, MK-0731, STA-5312, eleutherobin, 17 -beta-acetoxy-2ethoxy-6-oxo-B-homo-estra-1,3,5(10)-trien-3-ol, cyclostreptin, isolaulimalide, laulimalide, 4-epi-7dehydroxy-14,16-didemethyl- (+)-discodermolides and cryptothilon 1, in addition to other microtubule stabilizing agents known in the art.

In some embodiments, it may be desirable to quiescent aberrantly proliferating cells in conjunction with or prior to treatment with the anti-huICOS agonist antibodies described herein, for example, by administering hormones and steroids (including synthetic analogs) to the patient, such as 17a-ethinyl estradiol, diethylstilbestrol, testosterone, prednisone, fluoxymesterone, dromostanolone propionate, testolactone, megestrol acetate, methylprednisolone, methyltestosterone, prednisolone, triamcinolone, chlortrianisene, hydroxyprogesterone, aminoglutethimide, estramustine, medroxyprogesterone acetate, leuprolide, flutamide, toremifene, ZOLADEX™. When using the methods or compositions described herein, other agents used to modulate tumor growth or metastases in the clinical setting, such as antimimetics, can also be administered as desired.

Methods for safely and effectively administering chemotherapeutic agents are known to those skilled in the art. In addition, their administration is described in the standard literature. For example, the administration of many chemotherapeutic agents is described in the Physicians' Desk Reference (PDR), eg 1996 edition (Medical Economics Company, Montvale, N.J. 07645-1742, USA); the disclosure of which is incorporated herein by reference.

The chemotherapeutic agent(s) and/or radiation therapy may be administered in accordance with therapeutic protocols known in the art. Those skilled in the art will appreciate that the administration of the chemotherapeutic agent(s) and/or the use of radiation therapy may vary depending on the disease being treated and the known effects of the chemotherapeutic agent(s) and/or radiation therapy on that disease. In addition, according to the knowledge of the skilled physician, therapeutic protocols (eg, number of dosages and time of administration) may vary depending on the observed effects of the administered therapeutic agents on the patient and taking into account the observed responses of the disease to the administered therapeutic agents.

results

Tumor response was determined, for example, using the modified Response Evaluation Criteria in Solid Tumors (RECIST) established by the NCI.

With regard to targeted tumor lesions, responses to therapy may include the following:

- 64 041306

Complete Response (CR) (RECIST VI . 1) Disappearance of all targeted tumor lesions. Any pathological lymph nodes (whether targeted or non-targeted) should show short-axis reduction to < 10 mm. Partial Response (PR) (RECIST VI . 1) Reducing the sum of target tumor lesion diameters by at least 30%, using the baseline total diameters as reference. Progressive Disease (PD) (RECIST VI.1) Increasing the sum of the target tumor lesion diameters by at least 20%, using the smallest sum obtained during the study as a reference (includes the original sum if it is the smallest during the study). In addition to a relative increase of 20%, the sum must also show an absolute increase of at least 5 mm. (Note: the appearance of one or more new lesions is also considered progression). Stable Disease (SD) (RECIST VI.1) There is neither a sufficient decrease in size to qualify for PR, nor a sufficient increase to qualify for PD, taking as a benchmark the smallest total diameters over

- 65 041306

research. Immune-related Complete Response (irCR) (irRECIST) Disappearance of all targeted tumor lesions. Any pathological lymph nodes (whether targeted or non-targeted) should show short-axis reduction to < 10 mm. Immune-related Partial Response (irPR) (irRECIST) Reduction in the sum of target tumor lesion diameters and all newly measured tumor lesions (ie, percentage change in tumor mass) by at least 30%, using the original total diameters as a reference. Note: The appearance of new measurable tumor lesions is decomposed into a total tumor mass, but does not automatically qualify as disease progression until the sum of the diameters increases by >20% from the lowest value. Immunerelated Progressive Disease (irPD) (irRECIST) At least 20% increase in tumor mass (i.e. the sum of the target tumor lesion diameters plus any new measurable tumor lesions) using the smallest sum over the course of the study as the benchmark (this includes the original sum if it is the smallest during the study). In addition to a relative increase of 20%, the sum must also show an absolute increase of at least 5 mm. Tumor scores using immune response-related criteria for disease progression include the contribution of newly measured tumor lesions. Each net percentage change in tumor mass per score takes into account the size and growth kinetics of both old and new tumor lesions as they appear. Immune-relat Stable Disease (irSD) There is neither a sufficient decrease in size to qualify for irPR, nor a sufficient increase to (irRECIST) qualify for irPD by taking the smallest total diameters over the course of the study as a benchmark.

For off-target tumor lesions, responses to therapy may include the following:

- 66 041306

Complete Response (CR) (RECISTV1.1) Disappearance of all non-target tumor lesions. All lymph nodes should be non-pathological in size (<10 mm on the short axis). He-CR/He-PD (RECIST V1.1) Persistence of one or more non-target tumor lesions(s). Disease Progression (PD) (RECIST V1.1) Unequivocal progression of existing non-target tumor lesions. The appearance of one or more new lesions is also considered progression. Immune response-associated complete response (irCR) (irRECIST) Disappearance of all non-target tumor lesions. All lymph nodes should be non-pathological in size (< 10 mm on the short axis). Immune Response Related Disease Progression (irPD) (irRECIST) An increase in the number or size of the non-target tumor lesion(s) does not represent disease progression unless/until the tumor mass increases by 20% (i.e., the sum of the diameters of the smallest target tumor lesions and any newly measured tumor lesions increases by required amount). Non-target tumor lesions do not count towards Stable Disease and Partial Response.

Patients treated in accordance with the methods disclosed herein will advantageously experience an improvement in at least one manifestation of the cancer. In one embodiment, improvement is measured by a reduction in the number and/or size of measurable tumor lesions. In another embodiment, lesions can be measured on chest x-rays or CT or MRI films. In another embodiment, cytology or histology may be used to assess response to therapy.

In one embodiment, the patient being treated exhibits a complete response (CR), partial response (PR), stable disease (SD), immune response associated complete disease (irCR), immune response associated partial response (irPR), or immune response associated stable disease (irSD). In another embodiment, the patient being treated experiences a reduction in tumor size and/or a decrease in growth rate, i. suppression of tumor growth. In another embodiment, the proliferation of unwanted cells is reduced or suppressed. In another embodiment, one or more of the following may occur: the number of cancer cells may be reduced; tumor size can be reduced; infiltration of cancer cells into peripheral organs can be inhibited, inhibited, slowed down or stopped; tumor metastasis can be slowed down or inhibited; tumor growth can be inhibited; tumor recurrence can be prevented or delayed; one or more of the symptoms associated with the cancer may be relieved to some extent.

In other embodiments, administering effective amounts of an anti-ICOS antibody (or combinations of an anti-ICOS antibody and at least one additional antibody, e.g., an anti-PD-1 antibody or an anti-CTLA-4 antibody) according to any of the methods provided herein. document leads to a reduction in tumor size, a reduction in the number of metastatic lesions that appear over time, complete remission, partial remission, or stabilization of the disease. In other embodiments, the treatments provide a higher Clinical benefit rate (CBR = CR+ PR+ SD>6 months) compared to the clinical benefit rate achieved with the anti-ICOS antibody alone (or any of the above). alone combined antibodies). In other embodiments, the improvement in clinical efficacy rate is about 20%, 30%, 40%, 50%, 60%, 70%, 80% or more compared to the anti-ICOS antibody (or any of the combined antibodies) taken alone.

Vaccine adjuvants

Anti-huICOS antibodies described herein can be used to enhance antigen-specific immune responses by co-administering an anti-huICOS antibody with an antigen of interest, such as a vaccine. Accordingly, provided herein are methods for enhancing an immune response to an antigen in a subject, comprising administering to the subject: (i) an antigen; And

- 67 041306 (ii) an anti-huICOS antibody or antigen-binding fragment thereof, such that the subject's immune response to the antigen is enhanced. The antigen may be, for example, a tumor antigen, a viral antigen, a bacterial antigen, or an antigen from a pathogen. Non-limiting examples of such antigens include those discussed in the sections above, such as tumor antigens (or tumor vaccines) discussed above, or antigens from viruses, bacteria or other pathogens described above.

Detection and diagnostics

In another aspect, provided herein are methods for detecting the presence of a human ICOS antigen in a sample, or for measuring the amount of human ICOS antigen, comprising contacting the sample and a control sample with an anti-ICOS antibody, e.g., an anti-human ICOS monoclonal antibody or an antigen-binding fragment thereof, that specifically binds to human ICOS, under conditions that allow the formation of a complex between the antibody or its fragment and human ICOS. Complex formation is then detected, with a difference in complex formation between the sample compared to the control sample indicating the presence of the human ICOS antigen in the sample. In addition, the anti-ICOS antibodies described herein can be used to purify human ICOS by immunoaffinity purification.

The present disclosure is further illustrated by the following examples, which should not be construed as limiting. The contents of all figures and all references, Genbank sequences, patents and published patent applications cited in this application are expressly incorporated herein by reference.

Examples

The following are non-limiting examples of antibodies, compositions, and methods of the invention. It is understood that various other embodiments may be practiced in accordance with the general description provided herein.

Example 1 Production of fully human anti-huICOS antibodies

Disclosed herein are fully human anti-huICOS monoclonal antibodies and fully human antibodies that bind to the same epitope and/or cross-block the binding of fully human anti-ICOS antibodies. Such antibodies can be generated using transgenic mice that express human antibody genes, as described in the following example.

A. Hybridoma technology using HuMab® mouse and/or Kunming (KM)® mouse Generation of anti-ICOS antibodies

Human monoclonal anti-ICOS antibodies were generated by immunizing the HC2/KCo7 strain of HuMAb® transgenic mice (HuMAb® is a trademark of Medarex, Inc., Princeton, New Jersey) and KM mice (the KM® mouse strain contains the SC20 transchromosome as described in WO 02/43478) 1) soluble human ICOS antigen and 2) a Hek293T cell line that was transfected with the human ICOS gene that expresses human ICOS, a Chinese Hamster Ovary (CHO) cell line that expresses ICOS, and a 300-19 cell line that expresses ICOS. HC2/KCo7 HuMAb and KM mice were prepared as described in US Pat.

Antigen and immunization

The antigens were a soluble fusion protein containing an ICOS extracellular domain fused to the Fc domain of an antibody (recombinant chimeric human ICOS protein - mouse Fc), Hek293T cells, CHO cells, or 300-19 cells that had been transfected to surface express human ICOS. Antigens were mixed in an RTBI monophosphoryl lipid A (MPL) plus TDM (Sigma) adjuvant system for immunization. Mice described above were immunized with soluble ICOS protein in 15-25 µg of soluble recombinant ICOS antigen in PBS, or 1x107 CHO cells, Hek293T cells, or 300-19 cells transfected to surface express human ICOS in PBS mixed 1:1 with adjuvant. Mice were injected with 200 μl of the prepared antigens in the abdominal cavity or subcutaneously or in the ball of the foot every two to fourteen days. Mice were injected with 100-200 μl of recombinant mouse IL21 after immunization with ICOS antigens. Mice that developed anti-ICOS antibody titers were injected intravenously and/or injected into the ball of the foot with 10-20 μg of soluble recombinant ICOS antigen, or 5x106 CHO cells or 300-19 cells transfected to surface expression of human ICOS, or plus intraperitoneal injection of 15 μg of recombinant mouse IL21 protein in 100 μl of PBS three to two days before the fusion. Mouse lymph nodes and/or spleens were harvested and the isolated lymph node and/or splenocyte cells were used to prepare a hybridoma.

Selection of HuMab® or KM® mice that produced anti-ICOS antibodies

To select HuMab® or KM® mice that produced ICOS-binding antibodies, the sera of the immunized mice were tested using an enzyme-linked immunosorbent assay.

- 68 041306 (ELISA). Briefly, microtiter plates were coated with purified recombinant human ICOS-mouse Fc at a concentration of 1-2 μg/ml in PBS; 50 μl/well was incubated at 4° C. overnight, then blocked with 200 μl/well 5% chicken serum in PBS/Tween (0.05%). Plasma dilutions from ICOS immunized mice were added to each well and incubated for one hour at ambient temperature. The plates were washed with PBS/Tween and then incubated with goat anti-human Fc IgG polyclonal horseradish peroxidase (HRP) conjugated for one hour at room temperature. After washing, the plates were developed with ABTS substrate (Moss Inc., product: ABTS-1000) and an optical density index (OD) of 415-495 was determined on a spectrophotometer. The sera from the immunized mice were then further screened by flow cytometry for binding to a cell line that expressed human ICOS, but not a control cell line that did not express ICOS. Briefly, anti-ICOS antibody binding was assessed by incubating ICOS-expressing CHO cells or 300-19 cells with anti-ICOS antibody at a 1:20 dilution. Cells were washed and binding was detected with phycoerythrin (PE) labeled anti-human IgG. Flow cytometry analyzes were performed using a FACScan™ flow cytometer (Becton Dickinson, San Jose, CA). Mice that developed the highest titers of anti-ICOS antibodies were used for fusions. Mergers were performed as described below. Hybridoma supernatants were tested for activity against ICOS by ELISA and flow cytometry using fluorescence activated cell sorting (FACS).

Obtaining a hybrid

Mouse splenocytes and/or lymphocytes isolated from HuMab® and/or KM® mice were fused to a mouse myeloma cell line using electric field-based electrofusion using a Cyto Pulse Large Chamber Cell Fusion Electroporator (Cyto Pulse Sciences, Inc. Glen Burney, MD). Briefly, single cell suspensions of spleen lymphocytes from immunized mice were fused with an equal number of non-secreting Sp2/0 mouse myeloma cells (ATCC, CRL 1581 cell lines). Cells were seeded at a density of approximately 2x10 ⁴ cells/well in flat-bottomed microtiter plates and then incubated for approximately two weeks in selective media containing 10% fetal bovine serum, 10% P388D1 conditioned media (ATCC, CRL TIB-63), 3- 5% origen (IGEN) in DMEM (Mediatech, CRL 10013, high glucose, L-glutamine and sodium pyruvate) plus 5 mM HEPES, 0.055 mM 2-mercaptoethanol, 50 mg/ml gentamicin, and 1x hypoxanthine-aminopterin-thymidine medium (HAT) (Sigma, CRL P-7185). After one to two weeks, the cells were cultured in medium in which the HAT medium was replaced with a medium containing hypoxanthine and thymidine (HT). 10-14 days after cell seeding, supernatants from individual wells were screened first for anti-human gamma and kappa antibodies. Supernatants that were positive for anti-human gamma and kappa antibodies were then sequentially screened by ELISA and FACS for anti-ICOS human IgG monoclonal antibodies. Antibody-secreting hybridomas were transferred to 24-well plates, screened again and, if still positive for anti-ICOS human monoclonal antibodies, subcloned at least twice by serial dilutions. Stable subclones were then cultured in vitro to generate minute amounts of antibody in tissue culture medium for further characterization. The resulting human monoclonal antibodies were then purified by protein A column chromatography. Isolated antibodies of particular interest were designated 17C4, 9D5, 3E8, 1B7-a, and 1D7-b as described in Table 1. 7 below.

Table 7. Isolated antibodies

Name of the antibody SEQ ID NO CDR 1, 2 and 3 heavy chain SEQ ID NO CDR 1, 2 and 3 light chain SEQ ID NO of heavy chain variable domain SEQ ID NO of light chain variable domain 17С4 18, 19 and 20 21, 22 and 23 16 17 9D5 26, 27 and 28 29, 30 and 31 24 25 3E8 34, 35 and 36 37, 38 and 39 32 33 Yu7-a 42, 43 and 44 45, 46 and 47 40 41 lD7-b 42, 43 and 44 49, 50 and 51 40 48

B. PROfusion® mRNA Display System

Mice KM #333819 and #333821 were immunized with CHO cells overexpressing human ICOS and then the spleen and lymph nodes were harvested. Total RNA was extracted from spleen and lymph node cells and reverse transcribed using primers specific for antibody constant regions. The antibody cDNA was used to create a single chain variable fragment (scFv) library that was expressed in an mRNA display where each scFv protein was fused to its coding mRNA via a puromycin bond. The library was selected from 10 nM recombinant human ICOS-Fc, any bound molecules were recovered using Protein G magnetic bead capture and amplified by polymerase chain

- 69 041306 reactions (PCR) to move to the next cycle. A total of six cycles were performed, after which a significant signal binding to ICOS was observed using quantitative PCR (qPCR). The final population was sequenced and unique variable regions cloned into IgG expression vectors. IgG proteins were expressed using transient transfection of Hek293T cells to generate material for binding and functional assays. Was selected antibody IgG-2644, as described in table. 8 below.

Table 8. Antibody IgG-2644

Name of the antibody SEQ ID NO CDR 1, 2 and 3 heavy chain SEQID NOCDR 1,2uF light chain SEQ ID NO of heavy chain variable domain SEQ ID NO of light chain variable domain SEQ ID NO of heavy chain domain SEQID NO of the light chain domain 2644 191, 192 and 193 194, 195 and 196 186 189 185 188

Example 2 Generation of Humanized Anti-ICOS Antibodies

Sequencing of Hamster Anti-ICOS Antibody C398.4A A hamster anti-rat ICOS monoclonal antibody, C398.4A monoclonal antibody (anti-H4/ICOS), herein referred to as parental hamster antibody or C398.4A antibody, was obtained from BioLegend®. The C398.4A antibody was sequenced using mass spectrometry. Specifically, C398.4A was denatured in 5.3 M guanidine-HCl, reduced with dithiothreitol (40 mM) and alkylated with iodoacetamide (80 mM). After desalting on a Zeba desalting column with a 6 kDa MW cutoff, the antibody was enzymatically digested with trypsin, chymotrypsin, pepsin, Lys-C, AspN or GluC and analyzed by mass spectrometry. Peptide mapping and MS/MS were used to identify the resulting peptides and to confirm the amino acid sequence. Masses of intact heavy and light chains were obtained by cleavage of the glycan with PNGaseF, reduction of the antibody with dithiolthreitol, and alkylation with iodoacetic acid. The resulting antibody chains were analyzed by LC-MS.

The obtained data on peptide fragmentation were compared with a custom protein database consisting of three light and heavy chain sequences of antibodies for Cricetulus migratorius present in GenBank, as well as with antibody sequences determined in-house by RNA sequencing of monoclonal antibodies obtained from Armenian hamsters. A database search identified the GenBank sequence gene locus CMU17870 (Accession U17870) as similar to the C398.4A light chain. Amino acid substitutions in CDR3 and the framework region were observed in the C398.4A sequence when compared to the CMU17870 light chain sequence. A database search identified the GenBank sequence gene locus CMU17166 (Accession U17166) as similar to the C398.4A heavy chain variable region. The J region corresponded to a self-identified hamster HA-VH-7 sequence. The heavy chain constant region corresponded to the same isotype as the HL4E10 antibody (Accession HM369133). The D region was determined to be novel and was identified by de novo sequencing of peptide fragmentation data. Amino acid substitutions in CDR1, CDR2, CDR3 and the variable framework region were observed when compared to the heavy chain sequences CMU17166 and HA-VH-7.

Generation and evaluation of the chimeric ICOS.4 antibody based on the C398.4A antibody

The protein sequence of the C398.4A antibody was translated back into a cDNA sequence. The isoleucine/leucine (I/L) residue at position 96 in the D region (CDRH3) was expressed with either isoleucine or leucine at that position. The variable regions were cloned into expression vectors containing the human IgG1f signal sequence and constant regions and transfected into CHO-S cells to express the chimeric human antibody, ICOS.4. The chimeric antibody was purified using 2 L of supernatant, each using a 250 ml Protein A column on an AKTA Avant system, and screened for activity in a CHO-OKT3-CD32a/CD25-CD4+ T cell assay. The CHO-OKT3-CD32a/CD25-CD4+ T cell assay was a co-culture of irradiated (stunted) CHO cells transfected with low single-stranded CD3 (clone OKT3) and higher CD32A (for antibody cross-linking) with CD25-depleted CD4+ T -cells at a ratio of CHO:T cells equal to 1:4. The CHO cell line was grown in mixing flasks and irradiated on the day of the assay. T cells were harvested from fresh buffy coat (Stanford Blood Bank) using the RosetteSep® CD4+ T cell isolation kit (catalog number 15062) followed by CD25+ cell depletion using Miltenyi® CD25 microbeads (catalog number 130-092- 983) following the AutoMACS® depletion instructions provided in the kit.

Anti-ICOS antibody or isotype control was titrated from 5 μg/ml by 5-fold serial dilutions, with each dilution performed in triplicate. Cultures were plated in flat-bottomed TC-treated 96-well Costar® plates at a density of 5x104 T cells and 1.25x104 cells.

- 70 041306

CHO in 200 µl of complete medium (RPMI-1640 (Corning®, part number 10-040-CM) + 10% fetal bovine serum (FBS) (Gibco®, part number 25140) + 1x Pen Strep (Corning, part number 30002-CL) + 10 mM HEPES (Corning, cat. no. 25-000-CL) + 1 mM sodium pyruvate (Corning, cat. no. 25-000Cl) + 1xMEM (Corning, cat. no. 25030-CL)) per well and incubated for three days at 37°C and 5% CO ₂ .

Culture supernatants (50 µl/well) were collected on Day 3 for analysis of interferon-gamma concentrations using Homogeneous Time-Resolved Fluorescence (HTRF) (Cisbio®), read using a Rubystar® microplate reader, and calculate concentrations from a standard curve using Softmax software Pro®. The ICOS.4 antibody was tested in a functional T cell assay using CHO-OKT3-CD32 and CD4+CD25-T cells with the antibody titrated to compare relative levels of dose-dependent costimulation as measured by interferon-gamma secretion. ICOS.4 showed an EC50 value of 0.018 µg/ml.

Isotype selection

The choice of isotype for immuno-oncological therapeutic antibodies, and in particular for agonist targets, is influenced by two different considerations after FcR binding. As detailed by Ravetch and colleagues (Li and Ravetch, Science 2011;333:1030-4; Often et al., J Immunol. 2008;181:6829-36), binding of antibodies to activating receptors can lead to antibody-dependent cell-mediated cytotoxicity (ADCC ) or antibody-dependent cellular phagocytosis (ADCP) of cells expressing the target. On the other hand, binding of antibodies preferentially to an inhibitory FcR may mediate multivalent cross-linking of receptor and agonist signals. Since ICOS can be highly expressed on CD8+ and CD4+ Teff in the tumor microenvironment, the use of an isotype that can mediate ADCC or ADCP was considered a less attractive option. The in vitro ADCC activity of anti-ICOS antibodies also suggests that anti-ICOS antibodies were highly competent at mediating ADCC and supports the idea that ADCC-inducing isotypes should be avoided. Antibodies that increase the affinity of human IgG1 for CD32B, on the contrary, have been considered as alternative isotypes. The isotypes considered were the S267E IgG1 mutation, the SELF mutations, and the V12 mutations of human IgG1, as shown in Table 1. 3 above. All of these mutations increase affinity for CD32B and to varying degrees for CD32A while decreasing affinity for CD16 (as shown in Table 9). This decrease was predicted to decrease ADCC activity, as FcR likely mediates depletion of T cells in the tumor.

Table 9 Comparison of binding properties of wild-type and human IgG1 variant S267E (μM Kd)

Protein IgGlf IgGlf-S267E CD16-V 97 950 CD16-F 200 >5000 CD32A-H131 530 650 CD32A-R131 960 31 CD32B 3400 87 CD64 0.2 0.2 Clq + ++

In vitro activity in the SEB assay using CD4+ T cells and B cells showed higher activity of the S267E IgG1f antibody compared to human IgG1 and the other isotypes described above. Based on these functional experiments, IgG1f S267E was selected as the lead antibody. One difficulty in selecting IgG1f S267E was that this isotype binds complement C1q with a higher affinity than human IgG1, which poses a possible increased risk of complement-dependent cytolysis (CDC). Surprisingly, IgG1f S267E did not have increased CDC activity compared to human IgG1 in in vitro tests. Thus, the S267E mutation did not lead to an increased risk of CDC.

Humanization of the ICOS antibody. 4

The ICOS.4 antibody was humanized by grafting hamster CDRs onto human germline genes (FIG. 3), VH3-15 was selected for the heavy chain, and VKI 018 was selected for the light chain based on framework sequence homology. The human FW4 germline, JK3, was also selected for the light chain based on sequence homology. The human FW4 germline, JH4, was selected for the heavy chain based on sequence similarity and did not contain residues that could pose a potential liability risk. A panel of 26 antibodies was evaluated in the CHO-OKT3-CD32A/CD4 + CD25- T cell assay, with the antibody range starting at 0.2 μg/mL and titrated by four-fold dilutions to identify humanized sequences that retain binding similar to parental hamster antibody (C398.4A, i.e. parental hamster antibody having the heavy and light chain region sequences set forth in SEQ ID NOs: 3 and 4, respectively).

One amino acid residue substitution (T94A) has been identified to restore binding of the humanized CDR-grafted antibody and is located at the junction of FR3 and CDRH3. Except

In addition, three chimeric antibodies with disease-causing mutations in the sequence D56, G57 were also evaluated to see if a potential isomerization site in VL could be removed without affecting activity. The D56E residue substitution was chosen to eliminate a potential isomerization site (D56, G57) in the light chain and inserted into the humanized sequence. Humanized antibodies were screened with the IgG1f isotype, however ICOS.33 IgGlf S267E was re-expressed using the IgG1f S267E isotype. The description of the obtained antibodies is given below in table. 10.

Table 10. Summary of obtained antibodies

Name of the antibody Description 1 C398.4A Parent hamster antibody 2 ICOS.lmGl Mouse IgGl anti-mouse ICOS antibody derived from rat 17G9 (does not bind to human ICOS) ICOS.4 C398.4A Variable Region Chimeric Antibody Produced in Four Different Variants (listed below) 4a ICOS.4 mgl Mouse variant of IgGl ICOS.4 4b ICOS.4mIgG2a Mouse variant of IgG2a ICOS.4 4s ICOS.4hgl Human version of IgGl ICOS.4 4d ICOS.4 hgl SE Human version of IgGl ICOS.4 with mutation S267E 5 ICOS.33 Humanized (IgGl isotype) ICOS.4 with parental CDRs grafted onto a human framework and mutations T94A and D56E 6 ICOS.33 IgGlf S267E ICOS.33 with S267E replacement 7 ICOS.34 Gif Humanized (IgGl isotype) ICOS.4 with parental CDRs grafted onto a human framework (also referred to as "C398.4A-03") 8 ICOS.35 Gif ICOS.34 plus T94A mutation

A set of four humanized antibodies based on C398.4A was tested in a CHO-OKT3-CD32a/CD4+CD25- T cell functional assay compared to the parent hamster chimeric antibody as described in Example 3 below. ICOS.33 IgGlf S267E was chosen for further characterization and development. The heavy and light chain variable region sequences for ICOS.33 IgGlf S267E are shown in SEQ ID NOs: 5 and 6, respectively, and in FIG. 4.

Example 3 Antibody Selection

Analysis of CHO-scFv-CD3-CD32A/CD25-CD4+ T cells

The initial functional assay for screening was performed using CHO cells expressing anti-CD3 (OKT3) and human CD32A single chain variable fragment (scFv) to stimulate primary human T cells. This assay involved co-cultivation of irradiated (stunted) CHO cells transfected with low levels of CD3 single-chain variable fragment (clone OKT3) and higher levels of CD32A (for antibody cross-linking) with CD25-depleted CD4+ T cells at a ratio of CHO: T cells, equal to 1:4. The CHO cell line was grown in mixing flasks and irradiated on the day of the assay. T cells were harvested from fresh buffy coat (Stanford Blood Bank) using the RosetteSep® CD4+ T Cell Isolation Kit. CD25+ was depleted using Miltenyi CD25 microbeads following the kit's instructions for AutoMACS depletion.

An anti-ICOS antibody or an isotype control (i.e., an antibody of the same isotype as an anti-ICOS antibody, but which does not bind any naturally occurring human protein, e.g., antibodies against keyhole limpet hemocyanin (KLH), diphtheria toxin, etc.) was titrated from 5 μg/ml by 5-fold serial dilutions, with each dilution being made in triplicate using T cells from two donors. Cultures were plated in TC-treated 96-well flat-bottomed plates (Costar) with 5 x 10 ⁴ T cells and 1.25 x 10 4 CHO cells in 200 µl of complete medium per well and incubated for three days at 37°C and 5%. _CO2 .

Culture supernatants (50 μl/well) were collected on Day 3 for analysis of interferon-gamma (IFN-γ) concentrations using homogeneous time-resolved fluorescence (HTRF) analysis (Cisbio®). Concentrations were determined using a Rubystar microplate reader and calculated from a standard curve using Softmax Pro software. Then, 0.5 µCi of tritium-labeled thymidine was added to each well of the plate for eight hours and frozen. Cells were harvested into filter plates (Perkin Elmer) for tritiated thymidine incorporation assay to assess proliferation.

FcR CD32A cross-linked antibodies regardless of the Fc subtype of the antibody.

- 72 041306

This cross-linking provides costimulation of T cells through ICOS agonism, resulting in enhanced proliferation and cytokine release compared to cells treated with isotype control. This activity was observed on CD8+, CD4+ and CD25-CD4+ T cells. Because of the excellent signal-to-noise ratio in the CD25-CD4+ depleted T cell assay, these cells were used to screen for hybridomas. The most effective antibodies were selected for subcloning, purification and further characterization. Parental hamster monoclonal antibody was included in the antibody panel analysis. As described above, activity in the CHO-CD3CD32 assay was used to select a master panel of antibodies that were re-expressed as human IgG1 antibodies or other modified versions of human IgG1. ICOS.33 IgGlf S267E showed a dose-dependent induction of IFN-γ secretion and proliferation in the CHO-scFv-CD3CD32A/CD25-CD4+ T cell assay, as shown in FIG. 5. The mean EC50 of this effect was 0.083 nM (±0.067, n=6) for proliferation and 0.083 nM (±0.074, n=6) for IFN-γ induction. Proliferation induction ranged from 2-fold to 5-fold at the highest three concentrations tested in a total of 6 donors, while IFN-γ induction ranged from 2-fold to 9-fold in the same experiments compared to controls. Previous experiments using CHO-scFv-CD3 (without CD32A) confirmed that cross-linking is necessary for the agonistic activity of all tested anti-ICOS antibodies.

Analysis of CD25-CD4+ T cells and B cells-SEB

Further characterization of the functional activity of anti-ICOS antibodies was performed using staphylococcal enterotoxin B (SEB) as a T-cell receptor (TCR) stimulus and adding anti-ICOS antibodies for costimulation testing. When human peripheral blood cells (PBMCs) were used in the assay, anti-ICOS antibodies showed no functional activity. However, when CD4+ T cells (CD25 depleted or total CD4+ T cells) were used together with purified B cells, anti-ICOS antibodies showed increased secretion of interferon gamma (IFN-γ) compared to control antibodies.

This assay included co-culture of autologous CD25-CD4+ T cells and B cells. SEB was added to a final fixed concentration of 85 ng/mL to provide submaximal stimulation, and ICOS antibody was titrated to demonstrate a dose-dependent costimulation effect. The aim of this assay was to assess the ability of ICOS antibodies to enhance T cell activation in the context of the primary activating signal (SEB+ B cells) as evidenced by levels of IFN-γ induction. It is useful to induce higher levels of IFN-γ because it is an indicator of T-cell activation reflecting the activity of various antibodies exhibiting ICOS receptor agonism, and IFN-γ is a known mediator of anti-tumor immunity.

T cells were isolated by positive selection from two fresh buffy coats followed by particle separation to generate intact CD4+ T cells (Invitrogen). CD25+ cells were then depleted from CD4+ T cells using CD25 microbeads (Miltenyi) following the instructions in the depletion kit using AutoMACS. The negative fractions from the CD4 secretions were then used to isolate autologous B cells using Miltenyi CD20 beads following the instructions in the AutoMACS positive selection kit.

T cells were seeded in flat-bottomed TC-treated 96-well culture plates at a density of 5 x 10 ⁴ cells/well with autologous B cells at 3-5 x 10 ⁴ cells/well (depending on yield from each donor) with SEB included for final concentration 85 ng/ml. Anti-ICOS antibody or isotype control was titrated from 5 μg/ml by 5-fold serial dilutions for a total of seven points, each of which was tested in triplicate. The assay was performed in complete medium with a final volume of 200 μl/well. The plates were incubated for 3 days at 37°C and 5% CO ₂ .

Culture supernatants (50 μl/well) were collected on Day 3 for analysis of IFN-γ concentrations using HTRF assay (Cisbio). Concentrations were determined using a Rubystar microplate reader and calculated from a standard curve using Softmax Pro software.

In co-cultivation experiments with SEB, IFN-γ production was compared for anti-ICOS antibodies ICOS.33 IgG1f, ICOS.33 IgGlf S267E, 9D5 IgGlf, 9D5 IgGlf S267E, 2644 IgGlf S267E and control Iglf KLH control antibody (Fig. 6). The humanized ICOS.33 lead antibody with the S267E mutation (ICOS.33 IgG 1f S267E, denoted by the solid triangle pointing down in Fig. 6) induced higher levels of IFN-γ than the same wild-type Fc antibody (ICOS.33 IgGlf) . The other reference antibodies tested, ie 95D IgGlf, 9D5 IgGlf S267E and 2644 IgGl, also showed lower activity than ICOS.33 IgGlf S267E. Control IgLf KLH showed no activity. The ICOS.33 IgGlf S267E antibody increased IFN-γ production by 2.3 times compared to the control antibody in a dose dependent manner in co-cultures of CD25-CD4+ T cells and B cells stimulated with a suboptimal dose of SEB. A total of 20 donors were tested using this assay, all of which showed the highest agonist activity for

- 73 041306

ICOS.33 IgG1f S267E with an EC50 value of 0.020 nM (±0.018).

The activity of antibodies to ICOS on follicular T-helper cells (Tfh) was tested in this way. Compared to the control antibody 1D12, increased secretion of IL-10 was observed after the addition of anti-ICOS antibodies 9D5 and ICOS.4. Tfh cells were sorted from PBMCs after enrichment by CD4 selection (Invitrogen kit) by staining CD4-enriched cells for CD4, CD14, CXCR5, CD45RA and CD123, and sorting for Tfh cells (CD4 + CXCR5 + CD45RA-CD123-CD14-) with using Aria II FACS. Naive B cells were isolated from the CD4 negative fraction using the Miltenyi kit. Tfh and naïve B cells were co-cultured in flat-bottom TC 96-well plates at a density of 5 x ^{10 -4} cells/well and stimulated with SEB for 2 days, then IL-10 and IFNy secretion was measured by ELISA (BD), and it was shown that it is enhanced by antibodies to ICOS. This enhanced cytokine secretion does not require an exogenous crosslinker and can be enhanced by incorporating the S267E mutation into human IgG1 (FIGS. 7A and 7B).

COS.33 IgG1f S267E was chosen for further development because of its ability to stimulate IFN-γ production in the FcR CHO assay and induce cell proliferation (Figure 5), and because of its higher functional activity in the SEB assay compared to others. tested with anti-ICOS antibodies (FIG. 6).

Example 4 Reversal of Regulatory T Cell Suppression with ICOS.33 IgG1f S267E Antibody

The aim of this study is to determine the effect of ICOS.33 IgG1f S267E on effector T cell proliferation (Teff) and Treg-mediated suppression.

U-bottom plates were coated for 3 hours at 37°C with anti-CD3 (3 μg/ml) in combination with ICOS.33 IgG1f S267E (10 μg/ml) or anti-KLH, an isotype control that does not bind protein ICOS (10 µg/ml), in PBS. CD4+ T cells were isolated from whole fresh buffy coats using Rosette Sep CD4+ T Cell Enrichment Mix in combination with FicollPaque separation following RosetteSep manufacturer's instructions. CD4+ enriched T cells were stained with fluorophore-conjugated monoclonal antibodies directed against CD4, CD25, CD127, CD45RA and CD45RO in FACS sorting buffer. CD4+ T cells were then sorted into Teff (CD4+CD251oCD127hi), RA+Treg (CD4+CD25hiCD127lo/CD45RA+/CD45RO-) and RO+Treg (CD4+CD25hiCD1271o/CD45RA-/CD45RO+) cell populations using a FACSAria II cell sorter. Sorted Tregs were labeled with CellTrace™ Violet (CTV) proliferation dye according to the manufacturer's instructions at a concentration of 5 μM. Sorted Teffs were labeled with CellTrace CFSE™ Proliferation Evaluation Dye (CFSE) according to the manufacturer's instructions, except that it was used at a higher dilution of 1.25 μM to reduce the cytotoxic effects observed in previous experiments.

Fifty thousand sorted and CFSE-labeled Teff in 100 µl of complete medium was added to each well of a 96-well plate coated with anti-CD3 and ICOS.33 IgG1f S267E or isotype control. They were prepared with or without anti-CD28 added at 2 μg/ml (for a final concentration of 1 μg/ml). Titration amounts of sorted and CTV-labeled Tregs in 100 µl of complete medium were then added to each well, starting at 5x10 ⁴ Tregs (1:1 Tregs to Teff) and decreasing 2-fold in subsequent wells (1:2, 1:4, etc.). .d.).

Cultures were incubated for six days at 37°C, after which the cells were stained with Fixable Viability Dye Ghost Red-780 to exclude dead cells. Flow cytometry data were collected using a BD FACSCanto II flow cytometer. Percent proliferation Teff was determined using FACSDiva flow cytometric analysis software. Percentage Teff proliferation was determined by gating Teff that had diluted their dye to assess CellTrace CFSE proliferation after at least one round of division.

This example showed that ICOS.33 IgG1f S267E altered Treg-mediated suppression and increased Teff proliferation as shown in FIG. 8A and 8B. The values shown in the legend in FIG. 8A and 8B are Teff:Teg ratios. Accordingly, a value of 1 means 1 Teff to 1 Treg, a value of 2 means 2 Teff to 1 Treg, and so on, essentially titrating down the Treg. At a 1:1 ratio, ICOS.33 IgG1f S267E showed an approximately 4-fold increase in proliferation and a clear change in RA+ Treg-mediated suppression. Also, there was a 7-fold increase in proliferation and a clear change in RO+ Treg-mediated suppression as measured by the difference in percent dividing Teff between isotype control and anti-ICOS antibody. As Tregs were titrated, there was a proportional decrease in observed suppression, and in the absence of Tregs, there was an approximately 1.5-fold increase in the percentage of separated Teff compared to the isotype control. The effect in the presence of Treg may result from a decrease in Teff's susceptibility to Treg suppression or a decrease in Treg suppression capacity. Advantageously, ICOS.33 IgG1f S267E altered Treg-mediated suppression and increased Teff proliferation, as ICOS.33 IgG1f S267E was shown to stimulate an immune response.

Example 5 In vitro Fc effector function of ICOS.33 IgG1f S267E

The aim of this study was to evaluate antibody-dependent cellular cytotoxicity

- 74 041306 (ADCC) and activities of ICOS.33 IgG1f S267E in relation to the binding of complement factor C1q.

Target cell labeling with calcein AM

CD4+ T cells from Donor 2 (Stanford Blood ID WO 70516511239) were isolated, activated and labeled with Calcein AM. Briefly, peripheral blood mononuclear cells (PBMC) were purified from heparinized buffy coat by density gradient centrifugation and washed with phosphate buffered saline (PBS) supplemented with 2% FBS (HyClone). CD4+ T cells were isolated by negative selection using a magnetic bead separation kit (StemCell Technologies) and a RoboSep automatic cell separator (StemCell Technologies). After isolation of CD4+ T cells, CD25+ Tregs were depleted using a magnetic bead separation kit (Miltenyi Biotec). Purified CD4+ T cells were resuspended at 2.5x106 cells/ml in R10 medium and activated with a T cell activation/expansion kit (Miltenyi Biotec) one bead per two cells for three days at 37°C. On day 3, cells were counted, pelletized and resuspended at 1 x 106 cells/ml in PBS in a 15 ml conical tube. Calcein AM reagent was prepared by adding 20 µl of ultrapure DMSO to a reagent tube containing 50 µg of lyophilized reagent. A volume equal to 2 μl of reduced calcein AM was added to the suspended cells for every 1 ml of volume. Cells were vortexed and placed in an incubator at 37°C for 30 min. After the incubation period, the labeled target cells were washed three times with the ADCC assay medium and adjusted to 105 cells/ml in the assay medium.

Antibody-Dependent Cellular Cytotoxicity (ADCC) Assay Using Activated CD4+ T Cells as Targets

Primary human NK effector cells were purified from fresh PBMCs from two different donors (BDC Donors 9 and 12) and stimulated with IL-2. Briefly, PBMCs were purified from heparinized whole blood samples by density gradient centrifugation and washed with PBS supplemented with 2% FBS (HyClone). NK cells were isolated from PBMCs by negative selection using a magnetic bead separation kit (Miltenyi Biotech) and an autoMACs Separator (Miltenyi Biotech). Purified NK cells were resuspended at 1x106 cells/ml in MyeloCult medium supplemented with 500 IU/ml IL-2 and incubated overnight at 37°C.

The next day, activated NK effector cells were washed twice in assay medium and adjusted to 4.33-5x105 cells/ml in assay medium. Labeled target cells (50 μl/well) were added to a 96-well U-bottomed plate containing 50 μl/well of test or control antibody. Activated effector NK cells (100 µl/well) were then added to obtain a final ratio of effector cells to target cells (E:T) of 10:1 and a final antibody concentration in the range of 0.0002 µg/ml to 1 µg/ ml. The plate was then placed in a humidified incubator at 37°C for 2 h. The supernatant (50 μl/well) was transferred to an optical 96-well black plate and fluorescence intensity was read on an EnVision plate reader set to 485 excitation and 535 emission filters.

Target cells incubated with effector cells in the absence of antibody provided control for a background of antibody-independent lysis (spontaneous lysis), while target cells lysed with 20 μl or 100 μl/well of Delfia Lysis buffer represented the maximum release in the assay.

The percentage of antibody-dependent cell lysis was calculated based on the mean fluorescence intensity (MFI) using the following formula:

(tested MFI - average background value) / (average maximum value - average background value) x 100.

Target cell lysis percentages were plotted for each antibody using Prism v5.01 software from GraphPad Inc.

results

ADCC mediated by primary NK cells targeting activated CD4+ T cells

Anti-ICOS antibody ICOS.33 IgG1f S267E was tested for its ability to induce ADCC of ICOS-expressing CD4+ T cells as targets and compared to ADCC induced by ICOS.33 IgG1. Two experiments were performed using target cells and pairs of NK cell donors. In each case, ICOS.33 IgG1f S267E with modified IgG1 isotype induced less ADCC of activated CD4+ T cells than ICOS.33 IgG1. The data of these experiments are summarized in Table. 11 and in FIG. 9A and 9B.

- 75 041306

Table 11. Comparison of ADCC mediated by anti-ICOS IgG1 and modified IgG1 isotypes

Donor CD4+ target cells Effector cell donor Antibody Concentration (µg/ml) Percentage of target cell lysis Hole 1 Hole 2 Hole 3 2 12 ICOS.33 1.0000 31 16 42 IgGlf S267E 0.0625 20 23 26 0.0039 6 1 7 ICOS.33 IgGl 1.0000 57 56 55 0.0625 52 47 45 0.0039 42 38 29 isotypic 1.0000 16 13 18 control 2 9 ICOS.33 1.0000 41 -AND 16 IgGlf S267E 0.2500 8 22 1 0.0625 1 10 -5 0.0156 21 -3 -4 0.0039 17 -1 1 0.0010 7 9 -6 0.0002 8 20 -10 ICOS.33 IgGl 1.0000 36 43 62 0.2500 48 29 51 0.0625 61 47 51 0.0156 59 34 45 0.0039 41 -3 59 0.0010 46 20 27 0.0002 25 -6 9 isotypic 1.0000 -3 2 -8 control

C1q binding assay

Binding of ICOS.33 IgG1f S267E to human C1q was examined by ELISA. All antibodies were loaded onto a high binding immunoassay plate at 10 μg/ml in PBS at 50 μl per well. A non-specific binding control was included with wells coated with PBS only. The tablet was incubated overnight at 4°C. The next day, the plate and all reagents were equilibrated to room temperature; all subsequent steps were carried out at room temperature. Unoccupied protein binding sites were blocked with SmartBlock® at 200 µl per well for 30 minutes. The plate was washed 3 times with wash solution (PBS + 0.05% Tween-20) at 200 µl/well. Increasing doses of human C1q (48.00 to 0.76 μM) in ELISA assay buffer were added at 50 μl/well. The plate was incubated for 2 hours and washed three times with wash solution. Binding of human C1q to immobilized antibodies was detected with a biotinylated mouse anti-C1q mAb diluted 1:1000 in ELISA assay buffer and incubated for 1 h. conjugate, at 50 µl/well and incubated for 30 minutes. The final washing step was completed and the plate was developed with TMB substrate at 50 µl/well for 5 minutes. Absorbance was read at 650 nm on a SpectraMax 340PC384 Microplate Reader (Molecular Device). Data were plotted using Prism, Version 5.01.

results

ICOS.33 IgG1f S267E binds to human complement component C1q

Anti-ICOS antibody ICOS.33 IgG1f S267E was tested for its ability to bind to the human complement component C1q compared to ICOS.33 IgG1 in an ELISA assay. ICOS.33 IgG1f S267E was found to bind to human C1q with higher affinity than ICOS.33 IgG1. The data is shown summarized in FIG. 10 and in table. 12.

- 76 041306

Table 12. ICOS.33 IgGlf S267E binds to human C1q

OD 650 nm Clq, (µg/ml) ICOS33 IgGl ICOS.33 IgGlf S267E background value 20.000 0.8363 0.8463 0.8704 0.9977 1.0367 1.0318 0.1 168 10.000 0.7742 0.7153 0.7053 0.9207 1.0096 0.8919 n/a 5.000 0.5254 0.4921 0.5139 0.7107 0.7528 0.7405 n/a 2.500 0.3531 0 3600 0 3591 0.5393 0.5667 0.5590 n/a 1.250 0.2561 0.2298 0.2309 0.4255 0.4460 0.4259 n/a 0.625 0.1724 0.1616 0.1630 0.3306 0.3415 0.3302 n/a 0.313 0.1249 0.1260 0.1230 0.2843 0.2784 0.2815 n/a

Conclusion

COS.33 IgG1f S267E with modified IgG1 induced less ADCC-mediated killing of ICOS-expressing CD4+ T cells, but associated with higher affinity for human C1q than anti-ICOS antibody with wild-type IgG1.

Example 6 Antitumor activity in vivo

Antitumor activity of anti-mouse ICOS alone or in combination with other agents

Variations in the isotype of antibodies specific for T-cell surface receptors (both co-stimulatory and co-inhibitory) can alter antitumor activity. Mouse Fc variants of the 17G9 and ICOS.4 isotypes were generated and expressed as mouse IgG2a isotypes. Both showed superior antitumor activity compared to murine IgG1 variants, as described below. Although, without being bound by any theory, this was probably due to the depletion of regulatory T cells (Treg) at the tumor site, as well as the expansion of effector T cells (Teff) due to antibody-mediated ICOS agonism. Mouse studies using 17G9 Ab have also shown downregulation of the ICOS receptor on T cell populations in the spleen and tumor. It was observed that ICOS expression was lower in mice treated with Ab isotypes that bind to FcR (mIgG1 and mIgG2a), while receptor levels remained unchanged in mice treated with non-FcyR binding Ab (mIgG1 D265A, also referred to as ICOS.1 D265A). The dependence on interaction with FcyR suggests that cross-linking is required for this down-regulation. Importantly, antitumor activity was demonstrated even when the receptor was downregulated.

Both murine IgG1 variants of both 17G9 (ICOS.1) and parental hamster antibody (ICOS.4) were expected to have agonistic activity due to their ability to bind to FcRII (inhibitory receptor). As shown in Table. 13, both mouse IgG1 variants showed detectable but lower levels of antitumor activity compared to the IgG2a isotype, and less reduction in Treg in the tumor. In contrast, an anti-ICOS antibody that does not bind to FcR (17G9-IgG1-D265A) showed no antitumor activity. Although a debilitating isotype such as murine IgG2a has shown higher antitumor activity than agonistic mIgG1, the high expression of ICOS on Teff makes this mechanism of action less favorable when considered in combination with other treatments such as anti-CTLA treatment. -4 antibody, which is expected to deplete Treg more selectively. Consistent with in vitro findings of better agonist activity in isotypes with the S267E mutation, these isotypes also showed slightly higher or equivalent antitumor activity compared to human IgG1 in human FcR transgenic mice.

Table 13 Summary of efficacy studies using anti-mouse ICOS alone or in combination with other agents

Tumor mAbs Antitumor activity Toyota gut ST26 Anti-ICOS parent-nlIgG2a or mlgGl (ICOS.4) Anti-PD-l 4H2-mIgGl-D265A Ahth-CTLA-4 9D9-mIgG2b Anti-ICOS IgG2a monotherapy 69% TGI, 1/10 TF and ICOS IgGl 15% TGI, 0/10 TF Monotherapy 5% TGI, 0/10 TF in combination with anti-ICOS IgGl 83% TGI, 3/10 TF Monotherapy 15% TGI, 0/9 TF in combination with anti-ICOS IgGl 48% TGI, 0/9 TF Toyota Gut MS38 Anti-ICOS parent-InIgG2a or mlgGl or anti-ICOS 17G9 rat IgG2b Anti-PD-l 4H2 mIgGl-D265A Monotherapy with ICOS rat IgG2b 25% TGI 0/9 TF and anti-ICOS IgGl 6% TGI, 0/10 TF Monotherapy with 73% TGI, 1/9 TF combined with anti-ICOS rat IgG2b 92% TGI, 5/9 TF Thymoma EG7 Anti-ICOS parent-InIgG2a or mlgGl Monotherapy ICOS IgG2a 69% TGI, 2/8 TF and anti-ICOS IgGl 11% TGI, 1/8 TF Sarcoma 1956 Anti-ICOS parent-mIgG2a or mlgGl Monotherapy anti-ICOS IgG2a 94% TGI, 0/8 TF and anti-ICOS IgGl 50% TGI, 2/10 TF Fibrosarcoma a SA1N Anti-ICOS parent-InIgG2a or mlgGl Monotherapy anti-ICOS IgG2a 84% TGI, 6/10 TF and anti-ICOS IgGl 55% TGI, 5/10 TF

Antitumor Activity of Fc IgG1 ICOS Variants

To determine if the human IgG1 S267E antibody behaves similarly to mouse IgG1 antibodies, but with higher activity for FcR binding to CD32 and agonist recruitment

- 77 041306 cal receptor, conducted additional experiments on tumor models. In particular, to evaluate anti-human ICOS isotype variants in mice transgenic for the human Fc receptor (FcR), the following antibodies were constructed:

(a) anti-ICOS hIgG1 - anti-mouse ICOS monoclonal antibody, chimeric hamster/mouse anti-mouse ICOS antibody, IgG1 isotype (ICOS.4 hg1);

(b) anti-ICOS hlgG1 SE, an anti-mouse ICOS monoclonal antibody, a chimeric hamster/mouse anti-mouse ICOS antibody, isotype IgG1 SE (ICOS.4 hg1 SE) which has a mutation allowing it to bind to CD32R and CD32B better than unmodified version; and (c) IgG1 isotype control - fully human IgG1 isotype control (DT1D12 hgl).

MC38 mouse colon carcinoma cells were implanted subcutaneously in the right flanks of mice. Mice were divided into three treatment groups and dosed with 60 µg of (1) anti-ICOS IgG1 or (2) anti-ICOS IgG1 SE, or (3) IgG1 isotype control antibody (i.e., an antibody of the same isotype as the anti-ICOS antibody, but which does not bind to any of the natural mouse proteins, eg, anti-KLH antibodies, diphtheria toxin, etc.) on Days 7, 10 and 14 after implantation. Body weight and tumor size were measured twice a week until the end of the study on Day 52. If the tumors reached >2000 mm ³ or appeared ulcerated, the animal was euthanized. An increase in antitumor activity was observed when treated with anti-ICOS IgG1 SE mAb at a dose of 60 μg per mouse; the mean tumor growth inhibition (TGI) was 76% compared to 63% for anti-ICOS IgG1 without SE modification, as shown in Table 1. 14 and in FIG. 11A-C. Significant changes in body weight associated with treatment, or any clear signs of clinical toxicity were observed.

results

In a human FcR transgenic mouse model, administration of the monoclonal antibodies ICOS IgG1 SE and ICOS IgG1 resulted in a mean tumor growth inhibition (TGI) of 76% and 63%, respectively (as shown in Table 14). Five complete regressions were observed in each group at the tested dose level (60 μg/mouse) (Table 14 and FIGS. 11A-C). No physical signs of toxicity or weight loss were observed.

Table 14. Antitumor activity of Fc variants of IgG1 ICOS

Treatment (µg/mouse) Average % TGI on Day 30 Complete regressions ^a Isotype control IgGl, 60 mcg N/A 0/9 Anti-ICOS.4 hgl ^b , 60 mcg 63 5/9 Anti-ICOS. 4 hgl ^with SE, 60 mcg 76 5/9

a Complete regression = mouse with tumors <20 mm ³ for at least 3 measurements on the last day of the study.

^b IC OS.4 with human IgG1 ^c IC OS.4 with human IgG1 and S267E mutation

Conclusions

In a study in the Ravech syngeneic tumor model (summarized in Table 15), both anti-ICOS monotherapies stimulated modest antitumor activity, with anti-ICOS IgG1 SE showing slightly higher efficacy at Day 30 (76% versus 63% mean TGI) (Table 14). There were no significant changes in body weight associated with treatment or overt signs of clinical toxicity. Overall, anti-ICOS monotherapy stimulated antitumor activity, with anti-ICOS IgG1 SE showing slightly higher efficacy at Day 30 (76% versus 63% mean TGI). Both treatments resulted in five mice rejecting their tumors. There were no significant changes in body weight associated with treatment or overt signs of clinical toxicity.

Type Scheme of administration / Method of administration / Range Animals research/ Study duration doses henna Views/Line Carrier/ (µg/mouse group Dosage form ) (M/F) Anticancer Antibodies were administered IP after 60 9 on th activity implantation on Days 7, 10 and 14 µg/mouse group. tumor model Isotype control of IgGl groups with MS38/ human mixed transgenic C57/B6 mice with Anti-ICOS 4 hgl m floor Human FcyR Anti-ICOS 4 hgl SE

Pharmacological research

Example 7 Tumor Model Sa1N

The Sa1N mouse model of fibrosarcoma was used to evaluate the antitumor activity of chimeric anti-ICOS monoclonal antibodies. ICOS.4 mIgG1 is a good surrogate for

- 78 041306

ICOS.33 IgGls S267E, since this variant of ICOS.4 binds preferentially to the mouse inhibitory Fc receptor. Because the tumor model is performed in a mouse expressing the mouse Fc receptor, this makes the ICOS.4 variant a good surrogate for a human antibody. The ICOS.4 mIgG2a variant is a good surrogate for the ICOS.33 IgG1 antibody because this ICOS.4 variant is more similar to human IgG1 in that it binds to mouse activating Fc receptors. In addition, these variants were particularly suitable as surrogates because no modifications were required in their variable regions, which had already cross-reacted with the murine and human ICOS protein. In contrast, anti-ICOS.1 mouse IgG1 (mIgG1) D265A does not bind to FcR. An IgG1 antibody that does not bind to the ICOS protein was used as an isotype control.

To assess antitumor activity in the Sa1N fibrosarcoma model, after treatment with chimeric surrogate monoclonal antibodies against ICOS, Sa1N cells were implanted subcutaneously in the right flanks of mice. Mice were administered mAb in five treatment groups on Days 7, 10 and 14 post-implantation:

(1) chimeric anti-ICOS.1 mouse IgG1 (mIgG1) D265A, (2) anti-ICOS.4 mIgG1 (3) anti-ICOS.4 hIgG1, (4) anti-ICOS.4 mIgG2a or (5) isotype controls IgG1, each at a dose of 10 mg/kg.

On Day 15, tumors and spleen were harvested from four mice per group for analysis by immunological monitoring. The rest of the mice were measured for body weight and tumor size twice a week until the end of the study on Day 56. If the tumors were >2000 mm ³ or appeared ulcerated, the animals were euthanized.

On Day 23 post-implantation, the last day when median tumor growth inhibition (TGI) could be calculated based on the 60% of animals in treatment groups that remain alive, the therapeutic efficacy of anti-ICOS isotypes against Sa1N tumors was evident when compared with isotype treatment. control. Median TGI values were 21% (ICOS.1 mIgG1 D265A), 55% (ICOS.4 mIgG1), 69% (ICOS.4 hIgG1) and 84% (ICOS.4 mIgG2a). No toxicity was observed in any of the treatment groups, as evidenced by mean and median body weight loss remaining below 20%.

Immunological monitoring data indicate different levels of intratumoral depletion of Treg in all anti-ICOS isotypes. In addition, increased levels of intratumoral CD8+ T cells were observed in all anti-mICOS.4 treatments.

Tumor responses partially correlated with Treg reduction at Day 15, consistent with the relative binding of these mAbs to Fc receptors. These data suggest that an anti-ICOS mAb that lowers Treg levels will be more active than one that does not.

Anticancer treatment

On Day 7 post-implantation (February 02, 2015), 70 mice were randomized into five groups of 14 mice based on tumor volume (LxWxH/2). The mean tumor volumes were approximately 134 mm ³ for each group. On Days 7, 10 and 14, the isotype control or assigned mAb was administered. Mice were dosed intraperitoneally (IP).

Immunological monitoring of T cell populations

To further investigate antitumor activity at the cellular level, immunological monitoring was performed to examine subsets of immune cells in tumor sites and determine if there was an association between antibody treatment and changes in lymphoid cell populations. On Day 15, the vivarium operator removed tumors and spleens from four animals from each treatment group. Tissues were first treated with gentleMACS Octo Dissociator™ (Miltenyi, San Diego, CA) and then stained for various T cell markers. Samples were analyzed by flow cytometry on a Fortessa cytometer (BD Biosciences, San Jose, CA).

Monitoring after treatment

Tumors and body weight of mice were measured twice a week until the end of the study. Tumors were measured in three dimensions with an electronic digital caliper and data were recorded electronically using StudyDirector software from Studylog Systems (South San Francisco, CA). The mice were checked daily for posture, coat condition and respiratory changes, as well as lethargy. Mice were sacrificed when the tumors reached an endpoint of 2000 mm ³ or appeared ulcerated.

results

Tumor response

The last day all mice in the study remained alive was Day 14 post-implantation, the last day of IP dosing. As a result, the mean tumor growth inhibition (TGI) could not be calculated. On Day 23 post-implantation, the last day on which it was possible to calculate

- 79 041306 diana TGI, the therapeutic efficacy of anti-ICOS isotypes on Sa1N tumors was evident compared to isotype control treatment. Median TGI values were 21% (ICOS.1 mIgG1 D265A), 55% (ICOS.4 mIgG1), 69% (ICOS.4 hIgG1) and 84% (ICOS.4 mIgG2a).

Tumor growth curves by treatment group are shown in FIG. 12A-E. TGI is summarized by treatment group in Table. 16. Mean and median tumor growth curves by treatment group are presented in FIG. 13A and 13B.

Table 16. Growth of tumors by treatment groups

Day 14 Day 23 Treatment group Average tumor volume (mm ³ ) TGI (%) Median tumor volume (mm ³ ) TGI (%) Isotype control mlgGl, 10 mg/kg 387 N/A 621 N/A Anti ICOS.l mlgGl D265A, 10 mg/kg 295 24 493 21 Anti ICOS.4 mlgGl, 10 mg/kg 326 16 282 55 Anti ICOS.4 MgGl, 10 mg/kg 322 17 190 69 Anti ICOS.4 m!gG2a, 10 mg/kg 264 32 101 84

Differences in potency between anti-ICOS isotypes showed the mIgG2a>hIgG1>mIgG1>IgG1 D265A hierarchy. An inert MIgG1 D265A variant that cannot bind to FcR showed some antitumor activity with a median TGI of 21% on Day 23. An unmodified mIgG1 isotype that can bind to the inhibitory Fc receptor, FcYRIIB, can enhance agonism and was shown in this study median TGI 55% at Day 23. Consistent with their higher TGI values, the mIgG2a and hlgG1 isotypes can bind to mouse activating receptors and mediate ADCC or antibody-dependent cell phagocytosis (ADCP) of ICOS-expressing Tregs. In addition, reduced levels of intratumoral Tregs have been associated with increased tumor regression in mouse tumor models. No toxicity was observed in any treatment group, as mean and median body weight changes were below 20%.

Changes in the T cell population

Treg depletion was observed on Day 15 in the mIgG2a and hIgG1 antu-ICOS.4 mAb treatment groups, as the percentage of Foxp3+ cells was significantly lower in these groups than in the isotype control group, as shown in FIG. 14A-D. The same trend was observed in the non-depleting anti-ICOS antibody (mIgG1) treated group. In addition, increased levels of CD4+ effector T cells (Teff) were observed in all treatment groups in the following order: mIgG1>hIgG1>mIgG2a. This observation suggests that some CD4+ Teffs, which are likely ICOS+, may have been depleted by the mIgG2a isotype, which has the highest debilitating potential. Elevated levels of intratumoral CD8+ T cells were also observed in all anti-mICOSA treatments.

Conclusion

As shown in Table. 17, in a stepwise Sa1N syngeneic tumor model, chimeric anti-ICOS isotypes stimulated various levels of antitumor activity ranging from 21% to 84% of the median TGI at Day 23. The antitumor activities of the isotype variants in this study were assessed as follows: mIgG2a > hIgG1 > mIgG1 > mIgG1 D265A . Tumor responses correlated in part with Treg depletion at Day 15, consistent with the relative binding of these mAbs to Fc receptors.

The results of this study indicated that isotype selection is an important determinant of anti-ICOS antibody treatment. The anti-ICOS mIgG2a isotype, which binds activating Fey receptors, equivalent to the human IgG1 isotype, was able to deplete intratumoral Tregs and showed the greatest efficacy in inhibiting tumor growth. Chimeric variants of the isotype against ICOS stimulated various levels of antitumor activity. Tumor responses correlated in part with Treg depletion at Day 15, consistent with the relative binding of these mAbs to Fc receptors. The results suggest that mg2a promotes better antitumor activity.

- 80 041306

Table 17. Pharmacological studies in vivo

Type Scheme of administration / Route of administration / Dose range Animals per study / Duration of study / Vehicle / (mcg / mouse) ^G RU ^PP U (M / F) Species / Lineage Lec. form

Antitumor I activity of antiICOS isotype variant in SalNc tumor model by immunological monitoring of subsets of immune cells immunomomtonng/A/J mice

Antibodies administered IP post-implantation on Days 7, 10 and 14, Mouse IgGl isotype control,

Anti-ICOS 1 mlgGl D265A, Anti-ICOS 4 mlgGl, Anti-ICOS 4 hlgGl, Anti-ICOS 4 mIgG2a mg/kg 14 per group, F

Example 8 Combination of Anti-ICOS Antibodies and Anti-PD-1 Antibodies

Research 1.

To assess antitumor activity in a CT26 colorectal carcinoma model, after treatment with a surrogate anti-ICOS monoclonal antibody, ICOS.4 (mouse variant of IgG1 parental hamster antibody), at various doses and/or anti-PD-1 mAb, CT26 cells were implanted subcutaneously in the right flanks of mice. When tumors reached 31 mm ³ , mice were randomized into nine treatment groups of 10-14 mice each. Each mouse on post-implantation Days 7, 10, and 14 received a mAb or an isotype control (i.e., an antibody of the same isotype but not binding to any naturally occurring mouse protein, eg, anti-KLH antibodies, diphtheria toxin, etc.).

Mice were weighed and tumors were measured twice a week until the end of the study on Day 35. If the tumors were >2000 mm ³ or appeared ulcerated, the animals were euthanized. On Day 15 post-implantation, four mice in four treatment groups were sacrificed for spleen and tumor harvesting. Tissues were processed into single cell suspensions and cells were stained using antibodies for flow cytometry to analyze T cell populations.

On Day 21 post-implantation, the last day that mean tumor growth inhibition (TGI) could be calculated relative to the isotype control antibody, the TGI values for antiICOS monotherapy were 37 and 33% at 3 mg/kg and 10 mg/kg, respectively; the TGI value for antiPD-1 monotherapy was 22%. When anti-ICOS at 10, 3, or 1 mg/kg was combined with an anti-PD-i mAb, median TGI values >54% were observed. When anti-ICOS at 0.3, 0.1, or 0.03 mg/kg was combined with an anti-PD-1 mAb, median TGI values of <40% but >20% were observed. No toxicity was observed in any treatment group.

Antibody treatment

On Day 7 after implantation of CT26 cells, 120 mice were randomized into 10 groups of 10-14 mice each depending on tumor volume. The groups had an average tumor volume of approximately 31 mm ³ . Mice were injected with antibodies on Days 7, 10 and 14.

Monitoring after treatment

The mice were checked daily for changes in body position, fur, and breathing, as well as lethargy. Animals were weighed at least twice a week and euthanized if the weight loss was >20%. The flanks of each animal were checked for the presence and size of tumors at least twice a week until death, euthanasia, or the end of the study period. Tumors were measured in three dimensions (length [L], width [W], and height [H]) with electronic digital calipers and recorded. Tumor volumes were calculated using the formula: Volume = (LxWxHx0.5). The response to treatment was measured as a function of tumor growth inhibition (TGI) and was calculated as follows: (reference, mm ³ - test sample, mm ³ )/reference, mm ³ x100. When the tumor reached a volume greater than about 2000 mm ³ or became ulcerated, the animal was sacrificed.

Immunological monitoring of T cell populations

To investigate the effect of anti-ICOS antibodies on T cell populations, tissues were collected from four mice in each of four treatment groups on Day 15 post-implantation. Spleens and tumors were homogenized on gentleMACS Octo Dissociator™ (Miltenyi, San Diego, CA). Single cell suspensions were stained for T-cell markers using fluorochrome-conjugated antibodies. Antibody fluorescence was determined by flow cytometry on a Fortessa cytometer (BD Biosciences, San Jose, CA) and the results were analyzed using FlowJo software (FlowJo, LLC, Ashland, OR).

Statistical analysis

The mean standard deviation (SD) and median tumor sizes and mean body weights were calculated. The mean value was calculated when 100% of the test animals remained alive; and the median value was calculated when at least 60% of the studied

- 81 041306 remained alive. One-way analysis of variance (ANOVA) was used to determine if there were statistically significant mean differences between treatment groups; p values <0.05 were considered significantly different. GraphPad Prism® Version 5.01 software (GraphPad Software, LaJolla, CA) was used to plot and determine statistical differences between groups. Tumor growth curves for individual mice by treatment group can be seen in FIG. 15A-J. Mean and median tumor growth curves by treatment group are shown in Fig. 16A and 16B.

results

Tumor growth inhibition

On Day 21 post-tumor implantation, the last day when median TGI could be calculated, mice treated with anti-ICO8.4 mIgG1 as monotherapy at 10 mg/kg showed 33% median TGI compared with mice treated with control mIgG1 antibody. . Mice treated with anti-ICOS.4 mIgG1 monotherapy at 3 mg/kg showed 37% TGI and those treated with anti-PD-1 4H2 mAb monotherapy showed 22% TGI. At the end of the study period (Day 35), the number of tumor free mice was 0 out of 10 in the control antibody group, 0 out of 10 in the anti-PD1 group, 1 out of 10 in the anti-ICOS group.4 mIgG1 at 10 mg/kg , and 0 out of 10 in the anti-ICOS.4 mIgG1 group at 3 mg/kg. The combination of anti-mouse PD-1 with anti-ICOS.4 mIgG1 at various doses (10, 3 or 1 mg/kg) showed antitumor activity superior to that of any monotherapy (TGI values of 54%, 60% and 66%, respectively). The number of tumor-free mice at the end of the study was the same in these groups (1 out of 10 tumor-free mice), except for the 3 mg/kg dose at which 4 out of 10 mice were tumor-free. In addition, the median TGI was also calculated after 21 days using the relative difference in area under the effect curve between control and treatment groups (Table 18).

Table 18. Tumor growth inhibition by treatment group (relative to isotopic control)

Treatment group Day 35 Day 21 Mice without tumor Median tumor volume (mm ³ ) Median TGF Median TGI (after 21 days) ^b 10 mg/kg mlgGl control mAb 0/10 558 - - 10 mg/kg anti-PD-l 4H2 nbAb 0/10 433 22% eleven% 10 mg/kg anti-ICOS.4 mlgGl mAb 0/10 376 33% 42% 3 mg/kg anti-ICOS.4 mlgGl mAb 0/10 349 37% 29% 10 mg/kg anti-ICOS.4 mlgGl mAb + 10 mg/kg anti-PD-l 4H2 mAb 1/10 256 54% 42% 3 mg/kg anti-ICOS.4 mlgGl mAb + 10 mg/kg anti-PD-l 4H2 mAb 4/10 222 60% 44% 1 mg/kg anti-ICOS.4 mlgGl mAb + 10 mg/kg anti-PD-l 4H2 mAb 1/10 190 66% 48% 0.3 mg/kg anti-ICOS.4 mlgGl mAb + 10 mg/kg anti-PD-l 4H2 mAb 0/10 377 32% 14% 0.1 mg/kg anti-ICOS.4 mlgGl mAb + 10 mg/kg anti-PD-l 4H2 mAb 2/10 333 40% 28% 0.03 mg/kg anti-ICOS.4 mlgGl mAb + 10 mg/kg anti-PD-l 4H2 mAb 0/10 452 19% 21%

^a % Median TGI calculated at Day 21 ^b % Median TGI (after 21 days) calculated using the relative difference in area under the effect curve between control and treatment groups after 21 days

Immunological monitoring

Immunological monitoring was performed on Day 15 after implantation in certain treatment groups (Fig. 17A-D). Foxp3+ Treg depletion on tumor infiltrating lymphocytes (TIL) was observed in anti-ICOS.4 mIgG1 monotherapy groups (10 mg/kg and 3 mg/kg) (FIG. 17A). Mice treated with anti-PD-1 mIgG1 showed no decrease in Treg in TIL. The groups treated with antu-PD1 mIgG1 or antu-ICOS.4 at 3 mg/kg also showed an increase in the subset of CD8+ T cells in TIL (FIG. 17B). At a dose of 10 mg/kg, anti-ICOS.4 monotherapy appeared to show similar levels of this subset compared to the control group.

Levels of Ki-67, a marker of cell proliferation, were elevated in a subset of CD4+ effector T cells following anti-ICOS.4 mIgG1 monotherapy at 3 mg/kg (Fig. 17C). It was also found that

- 82 041306 the percentage of granzyme B positive cells, a marker of cytolytic activity on CD8+ T cells, was higher in groups treated with either anti-ICOS mIgG1 at 3 mg/kg or anti-PD-1 alone (FIG. 17D).

Conclusion

In a staged syngeneic CT26 tumor model, anti-ICOS.4 mIgG1 as monotherapy showed higher TGI when anti-ICOS.4 mIgG1 was administered at 3 mg/kg (37% TGI on Day 21, 0 of 10 tumor-free mice) by compared to 10 mg/kg (33% TGI on Day 21, 0/10 tumor free mice). Immunological monitoring data showed a higher percentage of CD8+ T cells, higher levels of Ki-67 in CD4+ effectors, and higher levels of granzyme B in CD8+ T cells in the anti-ICOS mIgG1 3 mg/kg treatment group than in treatment group receiving a dose of 10 mg/kg. These data suggest that for anti-ICOS monotherapy, a dose of 3 mg/kg has a higher antitumor activity than a dose of 10 mg/kg.

Combination therapy with anti-ICOS.4 mIgG1 mAb at 10mg/kg, 3mg/kg, and 1mg/kg with anti-PD-1 mIgG1 resulted in median TGI >54%, with 1 of 10 mice free of tumor in these treatment groups, with the exception of the group receiving anti-ICOS.4 at a dose of 3 mg/kg, in which 4 out of 10 mice had no tumors. These results indicate comparable levels of antitumor activity of anti-ICOS mIgG1 in combination with anti-PD-1 mIgG1 therapies at the three highest doses.

Research 2.

This study was designed to evaluate antitumor activity in a CT26 colorectal carcinoma model following treatment with a surrogate anti-ICOS monoclonal antibody, ICOS.4 (mouse IgG1 variant of the parental hamster antibody) at various doses and/or anti-PD-1 mAb. CT26 cells were implanted subcutaneously in the right flanks of mice. When the tumors reached 45 mm ³ mice were randomized into nine treatment groups of 15-20 mice each. Each mouse on post-implantation Days 9, 12, and 15 received mAb or an irrelevant isotype control. Mice were weighed and tumors were measured twice a week until the end of the study on Day 51. If the tumors were >2000 mm ³ or appeared ulcerated, the animals were euthanized. Whole blood samples were collected from mice at various time points (Day 9, Day 15 and Day 16 after tumor implantation) for analysis. On Day 16 after tumor implantation, five mice from eight treatment groups were sacrificed for spleen and tumor removal. Tissues were processed into a single cell suspension and cells were stained using antibodies for flow cytometry to analyze T cell populations.

On Day 29 after tumor implantation, the last day that mean tumor growth inhibition (TGI) could be calculated relative to the isotype control antibody, the TGI values for anti-ICOS monotherapy were 5% at 30 mg/kg and 33% at 3 mg/kg; anti-PD-1 monotherapy showed a TGI value of 74%. When anti-ICOS at 30, 10, 3, or 1 mg/kg was combined with an anti-PD-1 mAb, mean TGI >74% were observed. No toxicity was observed in any treatment group.

Antibody treatment

On Day 9 after tumor implantation, 200 mice were randomized into 9 groups of 15-20 mice each depending on tumor volume. The groups had an average tumor volume of approximately 45 mm ³ . Mice were injected with antibodies on Days 9, 12 and 15.

Monitoring after treatment

The mice were checked daily for changes in body position, fur, and breathing, as well as lethargy. Animals were weighed at least twice a week and euthanized if the weight loss was >20%. The flanks of each animal were checked for the presence and size of tumors at least twice a week until death, euthanasia, or the end of the study period. Tumors were measured in three directions (length [L], width [W], and height [H]) with electronic digital calipers and recorded. Tumor volumes were calculated using the formula: Volume = (LxWxHx0.5). The response to treatment was measured as a function of tumor growth inhibition (TGI) and was calculated as follows: (reference, mm ³ - test sample, mm ³ )/reference, mm ³ x100. When the tumor reached a volume greater than about 2000 mm ³ or became ulcerated, the animal was sacrificed.

Immunological monitoring of T cell populations

Various methods were used to study the effect of the ICOS antibody on T and B cell populations. Whole blood samples were taken from mice at various time points (Day 9, Day 15 and Day 16) and then processed for analysis. In addition, tissue was collected from five mice in each of the eight treatment groups on Day 16 post-implantation. Spleens and tumors were homogenized on gentleMACS Octo Dissociator™ (Miltenyi, San Diego, CA). Single cell suspensions were stained for T-cell markers using fluorochrome-conjugated antibodies. Antibody fluorescence was determined by flow cytometry on a Fortessa cytometer (BD Biosciences, San Jose, CA) and the results were analyzed using FlowJo software (FlowJo, LLC, Ashland, OR).

Statistical analysis

The mean standard deviation (SD) and median tumor sizes and mean body weights were calculated. The mean value was calculated when 100% of the test animals remained alive; and the median value was calculated when at least 60% of the study animals remained alive. One-way analysis of variance (ANOVA) was used to determine if there were statistically significant mean differences between treatment groups; p<0.05 values were considered to be significantly different. GraphPad Prism® Version 5.01 software (GraphPad Software, La Jolla, CA) was used to plot and determine statistical differences between groups. Tumor growth curves for individual mice by treatment group can be seen in FIG. 18A-J. Mean and median tumor growth curves by treatment group are shown in Fig. 19A and 19B.

results

Tumor growth inhibition

On Day 29 post-tumor implantation, the last day when the mean TGI can be calculated, the therapeutic efficacy of anti-ICOS monoclonal antibody (mAb) therapies against CT26 tumors was observed both alone and in combination with an anti-PD-1 mAb (Table 1). 19). Mice treated with anti-ICOS.4 mIgG1 monotherapy at 3 mg/kg showed 33% TGI and anti-PD-I 4H2 mAb monotherapy showed 74% TGI. At the end of the study period (Day 51), the number of tumor-free mice was 0 out of 10 in the control antibody group, 8 out of 15 in the anti-PD-1 group, and 1 out of 15 at all doses of anti-ICOS.4 mIgG1 (30, 10 or 3 mg/kg). The combination of anti-PD-1 with anti-ICOS.4 mIgG1 at various doses (30, 10, 3 and 1 mg/kg) showed antitumor activity greater than or equal to the antitumor activity of monotherapy (TGI values of 74, 80, 87 and 78%, respectively ). The number of tumor-free mice at the end of the study ranged from 8 to 11 out of 15 in the four combination treatment groups.

Table 19. Tumor growth inhibition by treatment group (relative to isotype control)

Treatment group Day 29 Average tumor volume (mm3) Average % TGI 30 mg/kg mlgGl control mAb 1248 N/A 5 mg/kg anti-PD-l 4H2 nbAb 327 74% 30 mg/kg anti-ICOS.4 mlgGl SHAL 1182 5% 10 mg/kg anti-ICOS.4 mlgGl mAb N/A N/A 3 mg/kg anti-ICOS.4 mlgGl mAb 838 33% 30 mg/kg anti-ICOS.4 mlgGl mAb 328 74% + 5 mg/kg anti-PD-l 4H2 mAb 10 mg/kg anti-ICOS.4 mlgGl mAb + 5 mg/kg anti-PD-l 4H2 mAb 252 80% 3 mg/kg anti-ICOS.4 mlgGl mAb + 5 mg/kg anti-PD-l 4H2 mAb 158 87% 1 mg/kg anti-ICOS.4 mlgGl mAb + 5 mg/kg anti-PD-l 4H2 mAb 271 78%

Immunological monitoring

Immunological monitoring was performed at various time points after implantation in certain treatment groups (Fig. 20A-D). On Day 16 post tumor implantation, depletion of FoxP3+ Treg on tumor infiltrating lymphocytes (TIL) was observed in all anti-ICOS.4 mIgG1 mAb treated groups (FIGS. 20A and 20B). Tumor-bearing CT26 mice treated with anti-PD-1 mIgG1 alone showed no decrease in TIL Treg. Despite variability, CD8+ T cell counts increased by TIL in all treatment groups compared to controls (FIG. 20D).

Ki-67 protein levels, a marker of cell proliferation, increased in a subset of CD4+ effector T cells after anti-PD-1 monotherapy (moderate increase) or anti-ICOS.4 mIgG1 (high increase) (Fig. 21A-C). A further increase in Ki-67 levels was observed in the anti-PD-1 and anti-ICOS.4 mIgG1 combination treatment groups.

ICOS-L, a ligand for ICOS, showed higher levels of mean fluorescence intensity (MFI) on B cells after treatment with anti-ICOS.4 antibodies. ICOS-L MFI levels were also elevated in whole blood taken on Day 9, Day 15 and Day 16 post tumor implantation and in the spleen on Day 16 post tumor implantation. There appears to be a trend where the highest dose of anti-ICOS.4 mIgG1 has the highest MFI for ICOS-L (Fig. 22A-D).

When addressing ICOS levels, loss of receptor expression on CD4+ T cells was observed after antibody treatment. This was most evident in the TILS tumor (Table 20). Higher doses of anti-ICOS.4 mIgG1 correlated with lower levels of ICOS. Dosing of antibodies (isotype control at 30 mg/kg and anti-PD-1 mIgG1 D265A at 5 mg/kg; anti-ICOS.4 mAgG1 mAb at dose levels of 1, 3, 10 and 30 mg/kg) was performed by intraperitoneal injection on Days 9, 12 and 15 after implantation of CT26 cells. Whole blood was collected at various time points (day 9, day 15 and day post-implantation) and tumors were collected on day 16 post-implantation from five mice in defined treatment groups. Immunological monitoring by flow cytometry was performed on the treated samples.

Table 20 Expression of ICOS on CD4+ T cells

Analyzed tissue Day after implantation mlgGl (30 mg/kg bw) [% ICOS] ICOS.4mlgGl[%ICOS] 30 mg/kg body weight 10 mg/kg body weight 3 mg/kg body weight RVMS 9 100% 67% 75% 86% RVMS 15 100% 65% 50% 65% RVMS 16 100% - 43% 50% Tumor 16 100% - 18% 15%

Conclusion

As shown in Table. 21, in a stepwise syngeneic CT26 tumor model, treatment with an antiICOS.4 mIgG1 mAb showed antitumor activity both as a single agent and in combination with an anti-PD-1 mAb. As monotherapy, similar levels of antitumor activity were observed when anti-ICOS.4 mIgG1 was administered at 30, 10, or 3 mg/kg, although the 3 mg/kg dose provided the highest mean TGI (33%) at Day 29. Immunological data monitoring showed increased FoxP3+ Treg depletion (tumor), higher percentage of CD8+ T cells (tumor), higher levels of Ki 67 protein in CD4+ effectors (tumor), higher levels of ICOS-L in B cells (whole blood and spleen), and loss ICOS expression in CD4+ T cells (whole blood and spleen) at all doses in anti-ICOS.4 mIgG1 treated groups. These data showed that anti-ICOS monotherapy has satisfactory efficacy in this tumor model.

Anti-PD-1 mIgG1 therapy had very high activity in this experiment. Combination therapy with anti-ICOS.4 mIgG1 mAb at 30, 10 mg/kg, 3 mg/kg, and 1 mg/kg with anti-PD-1 mIgG1 resulted in mean TGI >74%, with 8-11 out of 15 mice had no tumors in these treatment groups. These results showed comparable levels of anti-tumor activity of anti-ICOS mIgG1 in combination with anti-PD-1 mIgG1 therapy at all doses. Increased antitumor efficacy in the CT26 model was observed when combining anti-ICOS and anti-PD-1 mAb. Tumor growth inhibition was >74% for each of four doses (1, 3, 10, 30 mg/kg) in anti-ICOS treatment groups in combination with anti-PD-1, with at least 8 of 15 mice having no tumors in each of these groups. The tumor growth inhibition of the anti-ICOS mAb as the sole agent was 33%, with 1 out of 15 mice having no tumor when administered at a dose of 3 mg/kg. Immunological monitoring also showed lower levels of FoxP3+ Treg, a higher percentage of CD8+ T cells, higher levels of Ki-67 in CD4+ effectors, higher levels of ICOS-L in B cells, and lower levels of ICOS receptor expression with anti-ICOS monotherapy. .

Table 21. Non-clinical pharmacological studies in vivo

Type Dosing schedule/ Route of administration/ Dose range Study animals/ Duration of study/ (mcg/mouse) ^{on GRU PP} U

Views/Line Media/ (M/F)

Dosage form

Antibodies were administered IP on Days 9, 12 and 15 after 1-30 mg/kg 15-20 at implantation. Isotype control group, F mouse IgGl, anti-ICOS 4 mlgGl, anti-PD-l clone 4H2 mlgGl

Antitumor activity of anti-ICOS mAb in combination with anti-PD-l mAb in a CT26 tumor model with immunological monitoring of immune cell subsets / Balb/c mice

Example 9 Antitumor Activity of Anti-ICOS Antibody in Combination with Anti-CTLA-4 in the CT26 Tumor Model

Short description

To assess antitumor activity in a CT26 colorectal carcinoma model following treatment with a surrogate anti-ICOS monoclonal antibody having different Fc, ICOS.4 (mouse IgG1 or IgG2 variant of the parent hamster antibody), and/or anti-CTLA-4 mAb, CT26 cells were implanted subcutaneously in right sides of mice. When tumors reached 96 mm ³ , mice were randomized into six treatment groups of 10-15 mice in each group. Each mouse was dosed with mAb on Days 13, 16, and 20 post-implantation, or an isotype control (i.e., antibodies of the same isotype but not

- 85 041306 binding any natural mouse protein, for example, antibodies against KLH, diphtheria toxin, etc.). Mice were weighed and tumors were measured twice a week until the end of the study at

Day 66. If the tumors were >2000 mm ³ or appeared ulcerated, the animals were euthanized.

At Day 30 post-implantation, the last day on which the median tumor growth inhibition (TGI) relative to the isotype control antibody can be calculated, the TGI values for anti-ICOS monotherapy were 15% and 69% with mIgG1 and mIgG2a variants (e.g., chimeric mouse antibody with V sequences). _H /V _L SEQ ID NO: 3 and 4 + IgG1 or IgG2), respectively; anti-CTLA-4 monotherapy showed a TGI value of -7%. When anti-ICOS mAb was combined with anti-CTLA-4 mAb, mean TGI values were observed to be 40% (mIgG1) and 79% (mIgG2a). No toxicity was observed in any treatment group.

Experimental Procedures Test Antibodies and Controls

The following antibodies were constructed:

(a) mouse anti-mouse ICOS IgG1 antibody - anti-ICOS.4 mAb, mouse IgG1 isotype, was expressed from Chinese hamster ovary (CHO) cell lines;

(b) mouse anti-mouse ICOS IgG2a - anti-ICOS.4 mAb, mouse IgG2a isotype, was expressed from CHO cell lines;

(c) mouse anti-mouse CTLA-4 IgG2b antibody 9D9 - anti-mouse CTLA-4 monoclonal antibody clone 9D9, mouse IgG2b isotype, was expressed from a transfected CHO cell line and formulated in PBS; and (d) mouse IgG1 isotype control - fully mouse IgG1 antibody that does not bind to ICOS; prepared at 10 mg/kg in PBS.

Obtaining tumor cells

CT26 mouse colon carcinoma cells were obtained from the American Type Culture Collection (ATCC, Catalog CRL-2638) and maintained in vitro in sterile culture containing Dulbecco's Modified Eagle's Medium (DMEM) + 10% heat-inactivated fetal bovine serum (FBS), in for less than 10 passages. Cells were confirmed to be free of viruses by testing for mouse antibody production.

Tumor implantation

CT26 cells were cultured in RPMI-1640 medium (HyClone/GE Healthcare, Logan UT, catalog number 10-040-CM, lot 16915003) supplemented with 10% fetal bovine serum (FBS) (Gibco, Life Technologies, catalog number 26140- 079, batch 1704315). Cells were separated 1:10 every three to four days. In the right flank of each mouse, 1x106 CT26 cells were implanted subcutaneously in 0.2 ml PBS using a 1 cm syringe with a 26 G1/2 needle.

Antibody treatment

On day 13 after implantation of CT26 cells, 120 mice were randomized into six groups of 10-15 mice each depending on tumor volume. The groups had an average tumor volume of approximately 96 mm ³ . Mice were injected with antibodies on Days 13, 16 and 20.

Monitoring after treatment

The mice were checked daily for changes in body position, fur, and breathing, as well as lethargy. Animals were weighed at least twice a week and euthanized if the weight loss was >20%. The flanks of each animal were checked for the presence and size of tumors at least twice a week until death, euthanasia, or the end of the study period. Tumors were measured in three directions (length [L], width [W], and height [H]) with electronic digital calipers and recorded. Tumor volumes were calculated using the formula: Volume = (LxWxHx0.5). The response to treatment was measured as a function of tumor growth inhibition (TGI) and was calculated as follows: (reference, mm ³ - test sample, mm ³ )/reference, mm ³ x100. When the tumor reached a volume greater than about 2000 mm ³ or became ulcerated, the animal was sacrificed.

results

As shown in Table. 22, on Day 30 post-tumor implantation, the last day when median TGI can be calculated, treatment efficacy of anti-ICOS mAb therapies on CT26 tumors was observed both as monotherapy and in combination with anti-CTLA-4 mAb. Mice treated with anti-ICOS.4 variants as monotherapy showed 15% median TGI (mIgG1) and 69% TGI (mIgG2a). The combination of anti-CTLA-4 mAb (-7% TGI) resulted in higher median TGI values as the anti-ICOS.4 mIgG1 variant had 40% TGI and the anti-ICOS.4 mg2a variant had 79% TGI. At the end of the study period (Day 66), the number of tumor-free mice was 0 for all groups. Tumor growth curves for individual mice by treatment group are shown in FIG. 23A-F. Doses of antibodies (isotype control at 20 mg/kg; anti-CTLA-4 mIgG2b, anti-ICOS.4 mIgG1 and anti-ICOS.4 mIgG2a at 10 mg/kg) were administered by intraperitoneal injection on Days 13, 16 and 20 after implantation of CT26 cells.

Mean and median tumor growth curves by treatment group are shown in Fig. 24A and

- 86 041306

24V. No toxicity was observed in any treatment group, as mean and median body weight changes were below 20%.

Table 22 Tumor Growth Inhibition by Treatment Group

Day 30

Treatment group Median tumor volume (mm) Median % TGI isotype control mlgGl, 20 mg/kg 1981 N/A Ahth-ICOS.4 mlgGl, 10 mg/kg 1686 15% Ahth-ICOS.4 m!gG2a, 10 mg/kg 614 69% Ahth-CTLA-4 9D9 m!gG2b, 10 mg/kg 2114 -7% Ahth-ICOS.4 mlgGl, 10 mg/kg + Ahth-CTLA-4 9D9 mIgG2b, 10 mg/kg 1195 40% Ahth-ICOS.4 mIgG2a, 10 mg/kg + Ahth-CTLA-4 9D9 mIgG2b, 10 mg/kg 410 79%

conclusions

As shown in Table. 23, in a staged CT26 syngene tumor model, both monotherapies with the anti-ICOS Fc variant (i.e., the ICG.4 variant of mouse IgG1 or IgG2 parental hamster antibody) stimulated modest antitumor activity, with anti-ICOS.4 mIgG2a showing higher efficacy than anti-ICOS.4 mIgG1 at Day 30 (69% vs. 15% median TGI). The combination of anti-ICOS.4 therapies with anti-CTLA-4 mAb increased efficacy, with median TGI increased to 79% (mIgG2a) and 40% (mIgG1). No significant changes in body weight were associated with treatment, and there were no overt signs of clinical toxicity. Anti-ICOS monotherapy stimulated antitumor activity, with anti-ICOS.4 mIgG2a showing higher antitumor efficacy at Day 30 (69% versus 15% median TGI). Antitumor efficacy increased when combined with anti-CTLA-4 mAb treatment, with the anti-ICOS.4 mIgG2a combined therapy group showing 79% TGI and anti-ICOS.4 mIgG1 showing 40% TGI.

Table 23. Pharmacological studies in vivo

Exam Type / Species / Line Scheme of administration / Method of administration / Dose range Animals Duration of study / (mcg / mouse) ^{on GRU PP} U Vehicle / (M / F) Dosage form Antitumor activity of anti-ICOS mAb in combination with CTLA-4 mAb in a CT26 / Balb/c mouse tumor model Antibodies injected IP on Days 13, 16 and 20 10-20 mg/kg 10-15 on post-implantation, group, F 66 days, ( ^females ) Isotype control mouse IgGl, anti-ICOS 4 mlgGl, anti-ICOS 4 m!gG2a , anti-CTCA-4 9D9 m!gG2b, in PBS

Example 10 Affinity, Binding, Biophysical Properties, Improved Stability and Immunogenicity

Characterization of binding properties

Human CD4+ T cells, cynomolgus PBMC, and mouse and rat splenocytes were activated by incubation with a species-specific, plate-bound anti-CD3 antibody (coated in a 6-well plate with 2 ml/well of a solution at a concentration of 4 μg/ml in PBS for 3 h at 37°C and washed twice with 1 ml of medium), and soluble species-specific anti-CD28 antibody at a concentration of 1 μg/ml in fresh medium containing 1-2x10 ⁶ cells/ml for 3-4 days. It should be noted that cynomolgus monkey PBMCs, as well as murine and rat splenocytes, become predominantly T cells after three to four days of CD3/CD28 activation. Cells were harvested, counted, centrifuged and resuspended in +100 μg/ml mouse IgG staining buffer for blocking for 15 min at room temperature. ICOS.33 IgG1f S267E was titrated from 2 μg/mL by 4-fold serial dilutions to 0.002 μg/mL in staining buffer and incubated with activated human, cynomolgus, rat, or mouse cells for 30 min at 4-8°C in a U-shaped plate . The cells were then washed twice in 150-200 µl of staining wash buffer and resuspended in 50 µl of APC-goat anti-human IgG (Fc-gamma) diluted 1:200 in staining buffer, gently vortexed and incubated for 30 min at 4-8 °C in the dark. The cells were then washed once in 200 μl of staining wash buffer, resuspended in it and analyzed on a FACS Canto flow cytometer.

As shown in FIG. 25A and 25B, ICOS.33 IgG1f S267E shows strong binding to human, cynomolgus, rat, and mouse T cells with EC50 values that do not differ significantly between the three species.

- 87 041306

In addition, the binding avidity of ICOS.33 IgG1f S257E was compared with two different competitive anti-ICOS antibodies. Briefly, antibodies were incubated with activated CD4+ T cells on ice for thirty minutes. The cells were then washed and bound antibodies were detected using an anti-human-IgG-PE secondary reagent. The signal was measured by flow cytometry and the mean fluorescence intensity was calculated. As shown in FIG. 26A-B, ICOS.33. IgG1f S267E showed higher binding avidity to activated CD4+ T cells, calculated by EC50, compared to the two competing antibodies. As discussed herein, the term EC50 in the context of an in vitro assay using an antibody or antigen-binding fragment thereof refers to the concentration of the antibody or antigen-binding fragment thereof that induces a response that is 50% of the maximum response, i. midway between the maximum response and the original response. In FIG. 26A, the EC50 value for ICOS.33 IgG1f S267E was about 0.07 µg/ml, while the EC50 value for competitive antibody 1 was about 1.4 µg/ml, and the EC50 value for competitive antibody 2 was about 0.4 µg/ml . In other words, the EC50 value for ICOS.33 IgG1f S267E was about 20 times less than the EC50 value for competitive antibody 1 and about 6 times less than the EC50 value for competitive antibody 2. FIG. 26B, the EC50 value for ICOS.33 IgG1f S267E was about 0.08 µg/ml, while the EC50 value for competitive antibody 1 was about 2.4 µg/ml, and the EC50 value for competitive antibody 2 was about 1.0 µg/ml . In other words, the EC50 value for ICOS.33 IgG1f S267E was about 30 times less than the EC50 value for competitive antibody 1, and about 12 times less than the EC50 value for competitive antibody 2.

Affinity studies

Since monomeric human ICOS was not available, experiments to determine the true affinity of ICOS.33 IgG1f S267E were performed using the Fab fragment of ICOS.33 IgG1f S267E (lot PC-1804-04) and human ICOS Fc (R&D Systems, 169-CS-050) antigen using the Biacore™ T200 instrument. Binding experiments were performed at 37° C. to achieve (or model) the affinity of the antibody for the antigen under in vivo conditions. The CM4 chip was covalently coated with an anti-hFc capture reagent from Biacore. The surface was blocked with ethylenediamine. Fc-labeled human ICOS was then captured on the CM4 chip and the ICOS.33 IgG1f S267E Fab fragment flowed through it at concentrations of 0.91, 2.7, 25, 74, 667 and 2000 nM.

Association rate constant (kon) and dissociation rate constant (koff) were plotted against time and response units (RU) using BIAevaluation software, Version 3.2. The data corresponded to the Langmuir 1:1 model. The koff/kon ratio is represented by the dissociation constant (KD) of the antibody-antigen complex. The Biacore chip was regenerated with 50 mM sodium hydroxide solution at a flow rate of 100 μl/min. The affinity of the antibody for the human ICOS antigen, measured using the ICOS.33 Fab fragment of IgG1f S267E, ranged from 52 nM to 65 nM.

Biophysical analysis

The identity of ICOS.33 IgG1f S267E was confirmed by liquid chromatography/mass spectrometry (LC-MS). For heavy and light chain mass measurements, the sample was deglycosylated, reduced and alkylated according to standard testing method and analyzed using a Waters LCT Premier ESI-TOF instrument. The light and heavy chain weights of ICOS.33 IgG1f S267E were equivalent to their predicted weight estimates of 23795 and 50161 Da, respectively, based on the amino acid sequence derived from the DNA sequence.

The identity of the antibody was determined by Edman sequence analysis. N-terminal amino acid sequencing of heavy and light chains of antibodies was performed using an automated protein sequence analyzer ABI Precise. The observed N-terminal amino acid sequences of ICOS.33 IgG1f S267E corresponded to the predicted light and heavy chain amino acid sequences.

Using the Agilent 2100 BioAnalyzer system, ICOS.33 IgG1f S267E was determined to migrate at approximately 160 kDa under non-reducing (NR) conditions. Under reducing conditions (R), the heavy chain migrated at approximately 60 kDa and the light chain migrated at approximately 25 kDa.

The purity of ICOS.33 IgG1f S267E was determined by capillary electrophoresis in the presence of sodium dodecyl sulfate (CE-SDS). Samples were analyzed using a Beckman Coulter Proteome Lab PA 800 plus system under non-reducing and reducing conditions. ICOS.33 IgG1f S267E contained 93.45% intact IgG by CE-SDS under non-reducing conditions. The following antibody fragments were detected: light chain (1.85%), heavy-light chain (0.45%), two heavy chains (0.88%), and two heavy and one light chain (3.37%). The purity of ICOS.33 IgG1f S267E was 96.51% (62.22% heavy chain + 34.29% light chain) by CE-SDS under reducing conditions.

ICOS.33 IgG1f S267E was characterized by hydrophobic interaction chromatography (HIC) to determine the level of product heterogeneity. ICOS.33 IgG1f S267E showed low heterogeneity with 98.1% of the main peak, 0.4% of the peak before the main peak, and 1.5% of the tail arm, which indicates

- 88 041306 shows low chemical or conformational heterogeneity.

Capillary isoelectric focusing (cIEF) was used to characterize ICOS.33 IgG1f S267E for charged isoforms. The sample was analyzed using an iCE analyzer model iCE3. ICOS.33 IgG1f S267E showed an isoelectric point (pI) ranging from 7.30 to 7.72 with a major peak at 7.72 (45.19%). Other peaks were observed at 7.30 (7.51%), 7.40 (16.21%) and 7.56 (31.10%). The pI range observed was within the normal range expected for IgG1 antibody samples.

Size exclusion chromatography (SEC; gel filtration) combined with multi-angle light scattering (MALS) was performed to determine the monomer content and MW distribution of the main impurities ICOS.33 IgG1f S267E. ICOS.33 IgG1f S267E was found to contain over 99.8% monomer. Determination of MW using MALS showed that the monomeric component had an MW of 144300 Da. A very small number of the aggregate had an apparent MW of 626800 Da.

Protein fingerprint identification and sequencing was performed by analysis of digested peptides by LC-MS on a Waters Acquity UPLC chromatograph with an Acquity UPLC BEN C18, 1.7 µm column (2.1x150 mm; Waters Corporation) connected to an LTQOrbitrap mass spectrometer XL. Identification of heavy and light chain sequences was 100% using MS/MS data of various digests including trypsin, chymotrypsin and pepsin. Peptide sequencing confirmed that the allotype was human IgG1 and matched the expected amino acid sequence as predicted by the DNA sequence. The only N-glycosylation site was confirmed to be N297 on the heavy chain. The disulfide bonds were found to be as expected for a human IgG1 monoclonal antibody. The S267E mutation, designed to enhance CD32b receptor binding, has also been identified in the sequence.

The oligosaccharide profile of N-linked sugars present on ICOS.33 IgG1f S267E was determined by laser-induced fluorescence capillary electrophoresis (cLIF) using a Beckman MDQ instrument. N-linked glycans present on ICOS.33 IgG1f S267E contained a mixture of asialo-biantennary Sugars without fucose, which varied depending on the level of incorporation of galactose. The main glycan structures were G0F (30.64%) and GIF (43.65%) and, to a lesser extent, G2F (19.07%).

A VP capillary differential scanning calorimeter was used to determine the thermal stability and reversibility of the antibody. Data was analyzed using Origin 7 software. Thermal stability was within the acceptable range for a typical human monoclonal antibody. In thermal scanning experiments, many antibodies showed three resolvable melting points; the first is related to the unfolding of the CH2 domain, the second is related to the unfolding of the Fab domain, and the third is related to the unfolding of the CH3 domain. ICOS.33 IgG1f S267E showed a thermogram with these three unfolding temperatures: 65.2°C (Tml), 83.2°C (Tm2), 86.3°C (Tm3). Thermal reversibility is a marker of the ability of a protein to fold back into its native conformation after a disturbance, in this case heat. Thermal reversibility experiments at 83.2° C. (second melting point) showed a reversibility of 55.2%, suggesting that the antibody has stable refolding properties.

Stability data for ICOS.33 IgG1f S267E are summarized in Table 1. 24.

Table 24 Stability of ICOS.33 IgG1f S267E

Property Way results Freeze/Thaw (1h at -80°C, 1h at RT x 6) UV, SEC F/T stability risk not identified Profile solubility/concentration UV, SEC At least 50 mg/ml in buffer (20 mM histidine, pH 6.0, 260 mM sucrose) Agitation Stability Study 350 rpm at RT in buffer (20 mM histidine, pH 6.0, 260 mM sucrose) +/- 0.05% PS80 for 7 days (50 and 10 mg/mL) Aggregation was not observed

Example 11 Binding Affinity of ICOS.33 IgG1f S267E to Human FcyR by Surface Plasma Resonance

The binding of human FcyR to ICOS.33 IgG1f S267E was studied by surface plasmon resonance (SPR) and compared with a control anti-ICOS IgG1f antibody. Antibodies were captured on the protein A sensor surface and a series of FcyR titrations were injected as analytes.

For these studies, protein A was immobilized on flow cells 1-4 of the CM5 sensor chip using standard ethyl(dimethylaminopropyl)carbodiimide (EDC)/Nhydroxysuccinimide (NHS) chemistry with ethanolamine blocking in running buffer containing 10 mM HEPES, pH 7.4, 150 mM NaCl, 3 mM EDTA, 0.05% surfactant p20, up to a density of 3000 RU. ICOS.33 IgG1f S267E (3 µg/mL) or control antibody hIgG1f (3 µg/mL) were captured on the surface of protein A to a density of -400-500 RU and FcyR binding of the analytes was tested in running buffer consisting of 10

- 89 041306 mM NaPO4, 130 mM NaCl, 0.05% p20, pH 7.1 buffer (PBS-T) at 25°C using an association time of 120 s and a dissociation time of 180 s at a flow rate of 30 µl/min. A series of FcyR concentrations (3:1 dilution) from 1 μM to 0.15 nM (CD64 proteins) or from 10 μM to 1.5 nM (all other FcyRs) were tested to determine binding kinetics and affinity. The kinetic data corresponded to either a 1:1 Langmuir model or an equilibrium model using the Biacore T200 evaluation software.

Sensorgram data that showed very high association and dissociation rates with steady state binding were fitted to a 1:1 affinity affinity model at steady state, while those that showed slower kinetics were fitted to a 1:1 Langmuir model. Data at single analyte concentrations (hCD64 at 0.11 µM and hCD32a-H131, hCD32a-R131, hCD32b, hCD16a-V158, hCD16a-F158, hCD16b-NA1 and hCD16b-NA21 at 1.1 µM) were compared for control anti-ICOS antibodies and ICOS.33 IgG1f S267E, while ICOS.33 IgG1f S267E showed higher binding and slower dissociation rate for several of the FcyRs, with hCD32a-R131 and hCD32b having the most marked increases in binding and slower dissociation rates.

The kinetic and affine values of the best fit are given in Table. 25. These data quantitatively demonstrated that the S267E mutation altered the binding affinity for several FcRs compared to the hIgG1f control antibody. For example, binding to hCD32a-R131 improved from a KD value of 1500 nM (control hIgG1f) to 34 nM (ICOS.33 IgG1f S267E), which corresponded to an improvement of more than 40 times, and binding to hCD32b improved from a K _D value of more than 5000 nM (control hIgG1f) to 170 nM (ICOS.33 IgG1f S267E), which corresponded to an improvement of at least 29 times. Binding to cyno CD32a and CD32b was lower than that observed for human CD32a and CD32b.

Table 25. Kinetic and affinity data for ICOS.33 IgG1f S267E binding to

Human FcyR

ligand FcyR ka (1/Ms) kd (1/s) Kd(hM) Model Anti-ICOS hCD64 ^{1.1x106 6.0x10'5} 0D 1:1 Appr. Langmuir control hCD32a-H131 1100 Aff-Th in static state. antibody hCD32a-R131 1500 Aff-Th in static state. hCD32b >5000 Aff-Th in static state. hCD16a-V158 560 Aff-Th in static state. hCD16a-F158 >5000 Aff-Th in static state. hCD16B-NAl 3500 Aff-Th in static state. hCD16B-NA2 >5000 Aff-Th in static state. ICOS.33 hCD64 ^{1.6x106 5.1x10'5} 0.03 1:1 Appr. Langmuir IgGlfS267E hCD32a-H131 840 Aff-Th in static state. hCD32a-R131 34 Aff-Th in static state. hCD32b 170 Aff-Th in static state. hCD16a-V158 1400 Aff-Th in static state. hCD16a-F158 >5000 Aff-Th in static state. hCD16B-NAl 1400 Aff-Th in static state. hCD16B-NA2 >5000 Aff-Th in static state. cyCD64 .88 1:1 Appr. Langmuir cyCD32a 4000 Aff-Th in static state. cyCD32b 2300 Aff-Th in static state. cyCD16 1700 Aff-Th in static state.

Example 12 Pharmacokinetic parameters (PK) ICOS.33 IgG1f S267E

PK parameters derived from the single dose PK/PD and tolerability study using ICOS.33 IgG1f S267E are summarized in Table 1. 26. Exposure was dose proportional over the range of 1 mg/kg to 10 mg/kg, with an elimination half-life of 13-14 days. Anti-drug antibodies (ADA) were detected seven days post-dose in three of four cynomolgus monkeys in the 1 mg/kg dose group and continued to increase up to 42 days post-dose. The increase in ADA signal was consistent with rapid antibody clearance in these monkeys, and this part of the ADA-dependent data was not included in the PK data analysis.

- 90 041306

Table 26 Pharmacokinetic parameters of ICOS.33 IgGlf S267E after intravenous (IV) administration to cynomolgus monkeys

Study monkey number Dose (mg/kg) Area under the curve (AUC)(0INF) (µM x d) TU ₂ (half-life) (days) Clearance (CLT) (ml/day/kg) Volume in steady state (Vss) (ml/kg) DT15107 4 (2 females / 2 males) 4 (2 females / 2 males) 1 10 2.2±0.4 23±3.8 13±2.8 14±3.3 3.1±0.46 2.9±0.48 57±3.9 66 ± 11

Pharmacokinetics of mouse surrogate antibody in mice

The pharmacokinetics of a surrogate anti-mouse ICOS mAb (ICOS.4, a variant of the mouse IgG1 parental hamster antibody) after a single intravenous dose of 1 mg/kg and a single intraperitoneal injection (at a dose of 0.1, 1 and 10 mg/kg) was evaluated in tumor-free mice. BALB/c, which is a laboratory strain of albino domestic mice. The antibody showed a more than dose-proportional increase in exposure in the dose range of 1 mg/kg-10 mg/kg, as shown in table. 27. The half-life ranged from 0.53 days at the lowest dose of 1 mg/kg to 1.5 days at the highest dose of 10 mg/g. The non-linear PK in mice appears to be due, at least in part, to target-mediated drug distribution.

Table 27. Pharmacokinetic parameters of ICOS.4 mouse IgG1 after intraperitoneal administration to mice

Method of administration Dose (mg/kg) Stach (nM) AUC(O-INF) (µM x d) Т1/2 (day) CLT (ml/day/kg) Vss (ml/kg) IV 1 0.15 0.65 43.2 101.4 IP 0.1 4.1 0.005 0.55 IP 1 51 0.187 0.53 IP 10 703 3.3 1.5

Example 13 Cross-Reactivity and Tissue Staining

Binding assays demonstrated that the affinity (EC50) of ICOS.33 IgG1f S267E for activated CD4+ T cells was similar in mice, rats, cynomolgus monkeys and humans.

In a tissue cross-reactivity assay, FITC-conjugated ICOS.33 IgG1f S267E was applied to frozen sections (fixed with acetone or acetone/formalin) from 20 normal human tissues, each from one to four donors. Specific staining was observed in subsets of lymphocytes in lymphoid tissues (thymus, tonsils, and spleen) and in rich lymphoid tissues (stomach and small intestine), as well as scattered very rare mononuclear cells (MNC) in several tissues (thyroid, skin, lung, uterus). and testis), which are largely related to the underlying inflammation. Positive labeling was not observed in the brain, cerebellum, heart, liver, kidney, pancreas, peripheral nerves, colon, pituitary and prostate. In lymphoid tissues, positive lymphocytes are mainly distributed in the thymus medulla, in the light zone of the terminal center and the interfollicular region of the tonsil. These results are consistent with previous immunohistochemistry (IHC) using the parent antibody ICOS.4.

ICOS.33 IgG1f S267E-FITC staining was also assessed by immunohistochemistry (IHC) on frozen sections from 10 normal cynomolgus monkey tissues, including brain, heart, liver, lung, kidney, spleen, thymus, tonsils, skin, and testis. In general, staining patterns were similar to human tissues. Specific staining was observed in subsets of lymphocytes in lymphoid tissues (tonsils, spleen, and thymus). No unpredictable staining was found in the examined tissues.

Example 14 Cytokine Release, Complement Activation and Tolerance

Release of cytokines in human whole blood treated with ICOS. 33 IgG1fS267E

This assay was designed to evaluate cytokine responses in human peripheral blood cells following ICOS.33 IgG1f S267E treatment on fresh whole blood samples.

Fresh normal sodium heparinized whole blood (100 µl) was added to 96 well round bottom plates. 100 μl of ICOS.33 IgG1f S267E or ICOS.33 diluted in AIM-V serum-free medium, isotype control anti-KLH-hIgG1-2F5 mAb or anti-CD28 mAb TGN (5.11A1) was added to the wells to obtain a final antibody concentration of 10 μg /ml per well and a final volume of 200 μl per well. SEB (100 μl) diluted in AIM-V medium was added to the wells to give a final concentration of 100 ng/ml SEB and a final volume of 200 μl per well. CD3-CD28 (100 μl) diluted in AIM-V medium was added to the wells to give a final concentration of 1 μg/ml CD3-CD28 and a final volume of 200 μl per well. LPS (100 μl) diluted in AIM-V medium was added to the wells to give a final concentration of 10 μg/ml LPS and a final volume of 200 μl per well. The plates were incubated in an atmospheric incubator with 5-7% CO ₂ for 20 hours at 37°C. The plasma cell culture supernatants from each well were collected after 20 h and stored.

- 91 041306 at -20°C. Samples were sent to BMS Lawrenceville, NJ (LVL) in a dry ice container for analysis.

To evaluate cytokine secretion, 12 μl of premixed standards, controls, and samples were transferred to assay plates. Magnetic beads (6 μl) were added to each 384 well plate, then sealed and incubated for two hours at room temperature on a plate shaker. After two hours of incubation, the magnetic beads were washed twice and 6 μl of detection antibodies were added to each well. The plates were sealed again and incubated at room temperature on a plate shaker. Streptavidin-phycoerythrin (6 μl) was added to each well containing detection antibodies, then incubated for 30 minutes at room temperature and washed twice using a plate washer. Running fluid (80 μl) was added to each well and the beads were resuspended for five minutes on a plate shaker. Plate samples were read using the Bioplex 3D Instrument Array System. Raw data was measured as mean fluorescence intensity (MFI). Concentration (pg/ml) was calculated using Xponent software.

A panel of 75 cytokines was evaluated in blood from eight normal human donors for ICOS.33 IgG1f S267E mediated cytokine release. The addition of ICOS.33 IgG1f S267E to donated whole blood did not mediate cytokine secretion compared to isotype control. These data showed that treatment with ICOS.33 IgG1f S267E did not lead to cytokine release syndrome (CRS) in whole blood.

Interval IV Dosing Toxicity Study in Monkeys

This study was performed to determine the potential toxicity and biological activity of ICOS.33 IgG1f S267E administered intravenously to monkeys either once a week or once every three weeks for a period of one month, to assess the reversibility of any observed changes, to determine the systemic effects of ICOS .33 IgG1f S267E, evaluation of immune responses and data to support the use of ICOS.33 IgG1f S267E in humans. ICOS.33 IgG1f S267E was given intravenously as a slow bolus injection at doses of 0 (vehicle, once a week on Days 1.8, 15, 22 and 29), 1.5 mg/kg (once every 3 weeks on Days 1 and 22), 15 mg/kg (once a week) or 75 mg/kg (once a week) in groups of five female monkeys and five male monkeys. All doses were administered at 2 ml/kg in a vehicle/vehicle consisting of 20 mM histidine, 260 mM sucrose, 50 μM diethylenetriaminepentaacetic acid and 0.05% (w/v) polysorbate 80 (pH 6.0). As potential pharmacodynamic indicators, at least in part, all monkeys were immunized with keyhole limpet hemocyanin (KLH, an immunogen for stimulating the primary response), the viral vectors Adenovirus-5 (Ad5)-Gag, and Ad5-Nef (an immunogen for stimulating antigen-specific CD8 T-cell response) and tetanus toxoid (an immunogen to stimulate a secondary response) on Day 1. For example, immunization with tetanus toxoid increases the number of tetanus toxoid-specific T cells and allows assessment of the PK/PD of an antigen-specific population.

Evaluation criteria included survival, toxicokinetics, clinical observations (including eating behaviors), body weight, physical (including respiratory, cardiovascular, and neurological) and ophthalmic examinations, clinical pathology assessments, immunogenicity (anti-ICOS. 33 IgG1f S267E antibodies; ADA), immunotoxicological and pharmacological assessments (including receptor occupancy and receptor expression on CD4 helper T cells, T cell-dependent antibody response (TDAR) to KLH or tetanus toxoid, peripheral blood lymphocyte phenotyping, T cell activation, phenotyping of antigen-specific T cells and ex vivo secondary response to KLH, Gag or Nef peptides), organ mass, and macro- and microscopic pathology analysis. Scheduled autopsies were performed at 1 month (three/group/gender) and after an 8-week recovery period (two/group/gender).

After repeated dosing, mean systemic exposures to ICOS.33 IgG1f S267E (AUC [0-T]) increased approximately dose-proportional from 15 mg/kg to 75 mg/kg (once weekly) with no significant sex-related differences observed at all doses . After repeated dosing at a dose of 1.5 mg/kg (once every three weeks), the mean systemic exposures of ICOS.33 IgG1f S267E (AUC [0-504 h]) were lower (0.4x) than after dosing on Day 1, whereas AUC values (0-168 h) after repeated dosing at 15 and 75 mg/kg (QW) were slightly higher (2.1-2.6-fold) than after dosing on Day 1, suggesting accumulation.

Treatment-induced ADA responses were found in 8 and 2 of 10 monkeys per group at 1.5 mg/kg (once every three weeks) and 15 mg/kg (once a week), respectively, on or after Day 8 . During the recovery phase, ADA was only detected in monkeys at a dose of 1.5 mg/kg. After repeated dosing, serum concentrations of ICOS.33 IgG1f S267E in monkeys with ADA were generally unmeasurable (i.e., <lower limit of quantitation; LLOQ) or lower than in monkeys without ADA at 1.5 and 15 mg/day. kg, and the presence of ADA contributed to a decrease in the mean AUC at 1.5 mg/kg.

Toxicokinetic summary data for ICOS.33 IgG1f S267E are presented in table. 28.

- 92 041306

Table 28. Toxicokinetic summary data - mean values, pooled by sex

Parameter Period 1.5 mg/kg (Q3W) ICOS.33IgGlfS267E 15 mg/kg (QW) 75 mg/kg (QW) Stakh Day 1 42.6 377 2.019 Day 22 39.2/47.4 ^a 680/733 ^a 3,790 AUC(0-168 h) (mcg.h/ml) Day 1 3.520 34,200 176,000 Day 22 2.860 ^b / 4.610 ^a 71.800/84.600 ^a 452,000 AUC(0-504h) (mcg.h/ml) Day 1 6.240 ^e NA NA Day 2 ^2.510d / ^NAa NA NA

a Values were calculated by including/excluding data from animals with detectable treatment-induced ADA on Day 8 and/or after Day 8 (168 hours after the first dose).

^b Mean systemic exposure averaged from individual AUC(0-72 h) and AUC(0-168 h) ^c Mean systemic exposure averaged from individual AUC(0-168 h), AUC(0-336 h) and AUC (0-504 h) ^d Mean systemic exposure averaged from individual AUC(0-72 h), AUC(0-168 h), AUC(0-336 h) and AUC(0-504 h)

NA = Not applicable

ICOS.33 IgG1f S267E was well tolerated at all doses with no ICOS.33 IgG1f S267E associated clinical observations or effects on body weight, physical (including respiratory, cardiovascular, and neurological) and ophthalmic evaluations, hematology, coagulation, serum chemistry, urinalysis, organ mass and macro- or microscopic pathology. In addition, no ICOS-related effects were observed. -cells, effector memory CD4 T cells, CD25+ activated CD4 T cells, HLA-DR+ activated CD4 T cells, naive CD8 T cells, effector memory CD8 T cells, CD25+ activated CD8 T cells and HLA-DR+ activated CD8 T cells), CD8+ T cell proliferation, and ex vivo secondary responses at any dose tested.

Signs of ICOS.33 IgG1f S267E-mediated effects were noted at all doses. Expression of the ICOS receptor on CD4 helper T cells was close to 0% at 4 hours post dosing on Day 1 at all doses, indicative of downregulation and/or internalization of the ICOS receptor, and generally remained low during the dosing and recovery period at >15 mg/kg. The low expression of the ICOS receptor prevented the evaluation of the occupancy of the ICOS receptor per se. At a dose of 1.5 mg/kg administered once every three weeks, ICOS receptor expression on CD4 helper T cells began to recover after Day 8, increased to 41% before dosing on Day 22, decreased to 4% after dosing on Day 22 and increased to 42% on Day 29. Complete recovery of receptor expression was observed by Day 43 (91%). Receptor occupancy generally correlated with receptor expression (eg, 71% and 85% RO on Days 22 [pre-dosing] and 29). Generally, ICOS receptor expression levels were inversely correlated with serum ICOS.33 IgG1f S267E concentrations. This is consistent with the conclusion that the ICOS.33 IgG1f S267E antibody caused the loss of the receptor.

There was a dose-dependent suppression of keyhole limpet hemocyanin (KLH) specific IgM (up to 52% on Day 8) and IgG (up to 78% on Day 29) responses compared to vehicle control. Suppression of the T cell-dependent antibody response to KLH by ICOS.33 IgG1f S267E may represent an alternative mode of action and was observed in a previous study in cynomolgus monkeys. Although not linked to any mechanism, suppression of TDAR by an agonist of the ICOS bone-stimulating pathway may be associated with impaired T-helper cell agonism resulting from early and sustained down-regulation of ICOS expression.

Other ICOS.33 IgG1f S267E related effects at all or some dose levels during the dosing and/or recovery period included reduction in the mean absolute number of total T cells and CD4 helper T cells, percentage of CD4 regulatory T cells, percentage of CD4+ Central memory T cells, percentage of CD8+ central memory T cells, percentage of Ki67+ CD4+ T cells, and percentage of Gag+ and Nef+ CD8 T cells.

In conclusion, ICOS.33 IgG1f S267E was clinically tolerated by monkeys for one month at intravenous doses <75 mg/kg administered once a week. ICOS.33 IgG1f S267E-related effects were noted at all doses, as demonstrated by changes in ICOS receptor expression and occupancy, suppression of the T cell-dependent antibody response to KLH, reduced levels of certain T cell subsets, decreased CD4 T cell activation, and a reduced percentage of antigen-specific CD8 T cells. Many of these changes were still evident by the end of the recovery period at >15 mg/kg QW, consistent with continued exposure.

- 93 041306

ICOS.33 IgGlf S267E during the recovery period and subsequent sustained downregulation of ICOS receptor expression at these doses. A lower dose of 1.5 mg/kg, given once every three weeks, resulted in lower serum concentrations of ICOS.33 IgG1f S267E after the first dose, and restored the cell surface receptor before the second dose. There were no adverse findings associated with ICOS.33 IgG1f S267E. Thus, the non-observed adverse effect level (NOAEL) was considered to be 75 mg/kg (mean AUC [0-168 h] 452,000 µg.h/ml). In addition, for the potential determination of the maximum recommended starting dose for humans, the dose of 75 mg/kg was also considered the highest non-hazardous toxic dose (HNSTD).

Pharmacokinetic study of intravenous single doses and receptor occupancy in monkeys

The pharmacokinetics of ICOS.33 IgG1f S267E were evaluated in monkeys not exposed to the protein. All monkeys were immunized intramuscularly with 2.5 mg keyhole limpet hemocyanin (KLH). After immunization, monkeys were injected intravenously with ICOS.33 IgG1f S267E in 20 mM histidine (pH 6.0), 250 mM sucrose buffer, 50 μM penthetic acid (DPTA), and 0.05% polysorbate 80 at doses of 0.1 mg/kg or 10 mg/kg (N = 2/sex for vehicle and groups 1 and 10 mg/kg) via femoral vein. Serial blood samples (about 0.5 ml) were collected before dosing and 6, 24, 72, 168, 240, 336, 408, 504, 672, 840 and 1008 hours after dosing. Blood samples were left to clot and then centrifuged at 4°C (1500-2000xg) to obtain serum. Serum samples were stored at -20°C and placed on dry ice for analysis. Samples not analyzed on the day of receipt were stored frozen in a freezer set to <-70° C. until analysis.

Serum samples from cynomolgus monkeys were analyzed using the appropriate Gyros® immunoassay for detection of ICOS.33 IgG1f S267E. Biotinylated human ICOS mG1 (batch #22Oct2015-Biotin) was used as the capture molecule for ICOS.33 IgG1f S267E. Samples, standards and QCs were adjusted to a final matrix concentration of 10% cynomolgus serum and loaded into Gyrolab. Used the Wash 2 V2 Wizard method with Gyrolab Bioaffy 200 CD. After the final washing steps, captured ICOS.33 IgG1f S267E was detected using Alexa 647-labeled mouse anti-Hi IgG Fc-specific monoclonal antibody, clone 10C7 (Lot No. 15C3483473-10C7A) as the detection molecule. ICOS.33 IgG1f S267E concentration in cynomolgus monkey serum samples was calculated from fluorescence intensity measured by Gyrolab using a 4-parameter logistic (4-PL) calibration curve generated from ICOS.33 IgG1f S267E calibrators.

The range of the calibration curve for ICOS.33 IgG1f S267E ranged from 3 to 30,000 ng/ml in cynomolgus monkey serum. The upper and lower limits of quantitation were 30,000 ng/mL and 3 ng/mL, respectively (ie, ULOQ 30,000 ng/mL, LLOQ 3 ng/mL). Quality control samples were prepared at 20, 200, 2000, and 20,000 ng/ml in cynomolgus monkey serum and analyzed on each CD to ensure acceptable assay results. Calibrators, QCs and samples were diluted 10-fold in PTB. The results of the analysis were in an acceptable range: the % CV of the standards was below 25%, and the QC recovery was within ± 30% of the nominal values.

Monkey sera samples were analyzed by ABO/BAS, Lawrenceville, New Jersey, using the appropriate Meso Scale Discovery (MSD) electrochemiluminescent immunoassay for the presence of anti-ICOS.33 IgG1f S267E ADA. The supernatant of the cellular anti-idiotypic mouse antibody ICG.33 IgG1f S267E was used to prepare a positive control (PC). Biotinylated ICOS.33 IgG1f S267E was used as the capture molecule and ICOS.33 IgG1f S267E labeled with Sulfo Tag was used as detection reagent. Biotinylated ICOS.33 IgG1f S267E and Sulfo Tag ICOS.33 IgG1f S267E were diluted in PTB and combined to form a master mix with a final concentration of biotinylated ICOS.33 IgG1f S267E 1000 ng/mL and 1000 ng/mL Sulfo Tag-labeled ICOS.33 IgG1f S267E. Samples were diluted at 10% minimum required dilution (MRD) in the master mix and incubated at 22° C. for 2 hours. The master mix was then transferred to streptavidin-coated MSD plate at 50 μl/well. After another 1 hour incubation at 22° C., the plate was washed and MSD reading buffer was added. The plate was then immediately read using an MSD Sector Imager 6000. The presence of detectable anti-ICOS.33 IgG1f S267E antibodies in monkey serum samples was determined using the ratio of sample signal to sample negative signal.

Monkey serum samples were analyzed by BAR, Lawrenceville, New Jersey. Cynomolgus monkey serum samples were analyzed for total ICOS.33 IgG1f S267E using reverse phase liquid chromatography with direct trypsin digestion with tandem mass spectrometry (LC/MS/MS). Monkey serum samples were also analyzed for deamidated and unmodified ICOS.33 IgG1f S267E at position N329 using immunoaffinity enrichment in combination with LC/MS/MS target capture. Standard curves defining the range of analysis were prepared in commercially obtained cynomolgus monkey serum and analyzed with samples for

- 94 041306 studies as a complete analytical set. Concentrations for total ICOS.33 IgGlf

S267E was presented in μg/ml via an Excel spreadsheet for toxicokinetic and pharmacokinetic interpretation.

PK parameter values were calculated using a non-compartmental assay method (Phoenix WinNonlin 6.4, Certara, Princeton, New Jersey). Exposure values below the lower limit of quantitation (LLOQ: <10 ng/mL (0.07 nM) for ICOS.33 IgG1f S267E) were not used in the assay. The area under the curve from time 0 to the time the last sample was taken (AUC(0-T)) was calculated using a combination of linear and log trapezoidal methods.

PK parameters of ICOS.33 IgG1f S267E after a single intravenous dose of 1 mg/kg and 10 mg/kg for cynomolgus monkeys are shown in Table. 29. Following intravenous administration, plasma concentrations of ICOS.33 IgG1f S267E showed a bi-exponential decline. Accelerated clearance was observed in three of the four monkeys in the 1 mg/kg group after Day 7. As a result, only concentration-time data up to Day 14 were used for analysis for all animals in the 1 mg/kg group, and AUC (0-14d) was reported to abolish the effects of ADA.

Immunogenicity testing of plasma samples showed that five of the eight monkeys included in the study developed ADA, and that monkeys with higher levels of ADA showed faster clearance at the 1 mg/kg dose. AUC (0-42d) was reported for the 10 mg/kg group. ICOS.33 IgG1f S267E showed a dose-proportional increase in Cmax and AUC (0-T) and AUC (0-INF). When the dose was increased at a ratio of 1:10, the Cmax in male and female monkeys increased at a ratio of 1:8 and 1:8, respectively; and the AUC value (0-T) increased at a ratio of 1:16 and 1:11, respectively, and the AUC value (0-INF) increased at a ratio of 1:11 and 1:11, respectively.

The concentrations of intact ICOS.33 IgG1f S267E and deamidated product after an intravenous dose of 10 mg/kg were quantified using LCMS/MS. Deamidated product concentrations ranged from 0.5 to 8% of total ICOS.33 IgG1f S267E at all time points measured. The AUC value (0-42 days) for the deamidated product was 2.9% of total ICOS exposure.33 IgG1f S267E.

Table 29 Summary of pharmacokinetic parameters

Parameter ICOS.33IgGlfS267E 1 mg/kg ^{a, c} 10 mg/kg ^{b, s} M F Average M F Average Stach (µM) 0.168±0.006 0.161±0.006 0.165±0.006 1.4±0.019 1.3±0.046 1.38±0.02 Thalf (days) 14±3 and ±0.7 13±2.7 14±5.2 14±2.3 14±2.7 AUC(O-T) (µMday) 1.28±0.2 1.5±0.7 1.2±0.1 21±0.5 16.9±0.5 19±0.5 AUC(0INF) (µMday) 2.3± ^0.5d 2± ^0.09d 2.17± ^0.2d 25.9±2.5 20.4 ± 2.3 23±3.8

a AUC(0-T) is limited to 14 days in the 1 mg/kg dose group ^b AUC(0-42d) is given for animals in the 10 mg/kg dose group ^c Number of monkeys = 2/gender ^d >20% AUC extrapolated

Although increased C1q binding was observed in vitro, the absence of overt clinical signs in the single dose study in monkeys and the absence of hemodynamic effects in the cardiovascular instrumented monkey model suggest a low risk of complement activation. ICOS.33 IgG1f S267E was well tolerated when administered intravenously as a single dose of 1 mg/kg or 10 mg/kg to cynomolgus monkeys with a dose-proportional increase in exposure. No adverse clinical laboratory findings were observed.

Example 15 Competition of Anti-ICOS Antibodies for Binding

Epitope-specific sorting experiments were performed to determine which anti-ICOS antibodies compete with which others for binding to huICOS. Epitope-specific sorting is a process that uses a competitive immunoassay to test antibodies in a paired combinatorial manner, and antibodies that compete for the same binding site, i.e. the same or overlapping antigen epitope, are grouped into families (bins ). Pair competition between anti-huICOS antibodies was determined as follows. Reference antibody (i.e. ICOS.33, 20H4, 27B9, 23B6, 12D10, 23A10, 15B7, 12F3, 13B4, 17H9, 26E11, 23H5, 6D1, 12A9, 5C4, 10B10, 17C4, 1D7, 21E1, , 25B10, 8A10, 4D11, 6D5, 7C6, 26E9, 3E8, 16H4, 25E4, or 2644) was bound to the surface of the sensor chip, the test antibody was pre-incubated with the huICOS polypeptide construct in the mixture, and the pre-incubated mixture was passed over

- 95 041306 sensor chip to determine the extent to which the test antibody prevents binding of the huICOS polypeptide construct to the reference antibody on the surface of the chip. Competitive experiments were performed using the BIACORE® Surface Plasmon Resonance (SPR) instrument. Specifically, a reference anti-huICOS antibody was immobilized on the surfaces of the Sensor Chip CM5 chip (S Series, GE Healthcare, catalog number BR-1005-30), flow cell 2, flow cell 3, and flow cell 4 (5000 resonant units, RU ) and flow cell 1 was used as a negative control. Test antibody (i.e. ICOS.33, 20H4, 27B9, 23B6, 12D10, 23A10, 15B7, 12F3, 13B4, 17H9, 26E11, 23H5, 6D1, 12A9, 5C4, 10B10, 17C4, 1D7, 21E1, 1 , 25B10, 8A10, 4D11, 6D5, 7C6, 26E9, 3E8, 16H4, 25E4 or 2644) were diluted to 120 μg/ml (2X) at initial concentration. Test antibody dilution series were performed by diluting 1:3 antibody concentration with buffer for seven different concentrations and a control sample (with 0 μg/ml) to obtain a titration curve. The series of concentrations of each antibody was divided in half. In the first half of the concentration series, 40 nM (2X) human ICOS antigen (eg, huICOS/Fc) was added to give a final concentration series (60 μg/mL - 0.0 μg/mL) and 20 nM final antigen concentration per well. In the second half of the concentration series, a buffer was added in place of the antigen to obtain an antibody diluted to the same concentration, and this half was treated as a control. The complexes of the tested anti-ICOS antibodies and huICOS/Fc were incubated for 2 hours. 40 μl of the complexes were injected onto the surface coated with the reference antibody at a rate of 30 μl/min. Used BIACORE® T200 SPR instrument and running buffer in HBE-EP, GE Healthcare, part number BR-1001-88, filtered, degassed, 0.01 M HEPES, pH 7.4, 0.15 NaCl, 3 mM EDTA, 0.005% surfactant P20. The surface was regenerated with 25 mM NaOH (order code: BR-1003-58, GE Healthcare) at 100 μl/min for 5 s. The data was analyzed using Microsoft Excel, where the concentration of test antibodies was plotted against the corresponding response unit to obtain titration curves.

Competitive binding experiments showed that antibodies ICOS.33, 20H4, 27B9, 23B6, 12D10, 23A10, 15B7, 12F3, 13B4, 17H9, 26E11, 23H5, 6D1, 12A9, 5C4, 10B10, 17C4, 1D7, 21E1 15H11 25B10, 8A10, 4D11, 6D5, 7C6 and 26E9 cross-compete with each other and block ligand binding (B7-H2; ICOS-L) to huICOS. Antibodies 3E8, 16H4 and 25E4 cross-compete with each other but do not block ligand binding to huICOS. In contrast, although the 2644 antibody was found to cross-compete with the 3E8 antibody, it was also able to block ligand binding to ICOS.

Example 16 Anti-ICOS Antibody Epitope Mapping

Epitope Mapping of Anti-ICOS Antibodies with Yeast Display

Epitopes for anti-huICOS antibodies 3E8 and ICOS.4 were determined by displaying randomly mutagenized human ICOS extracellular domain variants (NP_036224.1 residues 21-134 provided as SEQ ID NO: 173) on yeast (Saccharomyces cerevisiae) cells, and sorting of these yeast cells based on their binding or inability to bind to certain antibodies. Selected yeast cells were amplified and subjected to additional rounds of selection based on their ability to bind to the tested anti-ICOS antibodies (See, for example, Chao et al. (2004) J. Mol. Biol. 342:539). Sequences for huICOS variants were determined from the obtained yeast and the effects of each residue on antibody binding were analyzed. A binding epitope for antibodies of the present invention is defined as a locus within the huICOS sequence where single amino acid mutations impair binding to anti-huICOS antibodies of the present invention.

Briefly, reduced fidelity PCR was used to clone DNA encoding human ICOS (coding for residues 21-134 of SEQ ID NO: 1) into constructs allowing expression of huICOS variants as amino-terminal fusion protein regions, further containing the c-Tuc tag sequence and the protein Aga1p of the yeast cell wall. Such constructs, when expressed in yeast (Saccharomyces cerevisiae), display huICOS polypeptide variants on the yeast cell surface, attached to the cell surface by an Agalp polypeptide. The stus label was used as a positive control to display huICOS fusion proteins on this yeast cell. These yeast cells were then further sorted into those that expressed correctly folded huICOS fusion proteins (as determined by binding of a control mouse anti-huICOS antibody detectable by allophycocyanin (APC) labeled goat anti-mouse secondary IgG) but did not bind to the antibodies of the present invention ( as determined by detection with phycoerythrin (PE) labeled goat anti-human IgG as secondary antibody). These selected yeast cells were pooled, amplified and used in the subsequent round of selection. The huICOS sequence was determined from the yeast constructs remaining after selection.

Yeast populations that bind to ICOS.4 and 3E8 show different mutation patterns, indicating that these two antibodies recognize different epitopes. Similar experiments were carried out with the 9D5 antibody, which blocks the binding of the ICOS ligand to ICOS and which competes

- 96 041306 Works with ICOS.4. For experiments with 9D5, a 3D ICOS molecular model based on the crystal structure of the CTLA-4/B7-2 complex (e.g., Stamper et al. (2001) Nature 410:608) was used to distinguish which amino acid residues are masked and which are exposed on surfaces to determine which of the selected mutations are most likely antibody-specific contact residues (ie, epitope residues) rather than simple structure-breaking mutations. Epitopes identified by yeast display for ICOS.4, 3E8 and 9D5 are presented in Table 1. 30. Epitopes in table. 30 are listed as residues in huICOS SEQ ID NO: 1, which includes a 20 amino acid signal sequence. Accordingly, the residue numbers for mature huICOS (ie, ICOS protein without signal sequence) will be the residues shown in Table 1. 30 minus 20 (eg V48 with signal sequences or V28 without signal sequence).

Table 30 Anti-ICOS mAb epitopes

Clone ICOS residues (SEQ ID NO: 1) ICOS.4 V48, Q50, G70, S71, G72, F114, DI 15, Pl 16, Pl 17, Pl 18, L123 3E8 D64, K78, S79, L80, K81, F82, S85 9D5 P45,147,P117,P118,K120

Similar yeast display experiments were performed with ICOS-L (B7-H2) instead of anti-ICOS mAb to determine which residues on ICOS are critical for ICOS/ICOS-L interaction, ie. binding site for ICOS-L on ICOS. The ICOS-L binding site was determined to be at ICOS residues Q50, K52, F114, P117, P118, and F119, as indicated in SEQ ID NO: 1. Epitope testing for anti-ICOS mAb in Table 1. 30 suggests that ICOS.4 and 9D5 should block ICOS-L binding, while 3E8 may not, consistent with what was observed experimentally (as discussed in Example 15 above).

Epitope Mapping of Anti-ICOS Antibodies with HDX-MS

Deuterium exchange experiments with antibodies ICOS.4 and 9D5 confirmed that the region from S112 and L123 contacts when ICOS binds to ICOS-L, indicating a functional epitope region of ICOS residues 112-123 (SEQ ID NO: 1), or SIFDPPPFKVTL (SEQ ID NO: 203). This region overlaps with the C-terminal portion of the epitope as determined by yeast display and represents the largest group of residues along the primary sequence.

Epitopes for anti-huICOS antibodies ICOS.4 and 9D5 were determined using hydrogen-deuterium exchange mass spectrometry (HDX-MS), see, for example, Epitope Mapping Protocols in Methods in Molecular Biology, Vol. 66, G. E. Morris, Ed. (1996) as described herein. ICOS-Fc was mixed with mAb in a 1:1 ratio and HDX-MS was run for 1, 10 min, 4 h in duplicate.

The results show that ICOS.4 and 9D5 bind to the same discontinuous epitope as shown below (epitope underlined) and in FIG. 1.

MKSGLWYFFLFCLRIKVLTGEINGSANYEMFIFHNGGVQILCKYPDIVQQFKMQLLKGGQ 60

ILCDLTKTKGSGNTVSIKSLKFCHSQLSNNSVSFFLYNLDHSHANYYFCNLSIFDPPPFK 120 VTLTGGYLHIYESQLCCQLKFWLPIGCAAFVVVCILGCILICWLTKKKYSSSVHDPNGEY 180 MFMRAVNTAKKSRLTDVTL 199 (SEQ ID NO:1)

Example 17 ICOS Expression in Peripheral Blood and Tumor-Infiltrating Lymphocytes from Lung, Kidney and Colon Cancer Patients

Understanding the expression of ICOS on tumor-infiltrating lymphocytes (TIL) in various tumor types and patient populations has helped to identify the appropriate disease features and patient population for effective therapy for ICOS.33 IgG1f S267E, especially when combined with anti-PD-1 agents such as nivolumab. The frequency and magnitude of ICOS and PD-1 expression in peripheral blood cells and TIL (CD8+ and CD4+ T cells) were profiled in non-small cell lung cancer, renal cell and colorectal carcinoma (CRC) samples.

Fresh tumor tissues and corresponding peripheral blood samples were obtained from patients with lung cancer, kidney cancer or CRC (ConversantBio, MT Group, Benaroya) and immersed overnight at 4°C in Hypothermosol FRS (Biolife Solutions) and ACD Solution A (BD Biosciences) , respectively. All samples were processed and stained within 24 h after surgery. Tumor tissues were weighed and dissociated using a Miltenyi dissociation kit (Miltenyi, cat. no. 130-095929), with peripheral blood cells isolated after red blood cell (RBC) lysis in RBC lysis buffer (BioLegend, cat. no. 420301). Cell suspensions (from tumor tissue or peripheral blood) were washed twice with HBSS (no Ca, no Mg), stained with NIR Viability Dye (Molecular Probes by Life Technologies, Catalog L34976), blocked with human AB serum in Dulbecco's phosphate buffered saline (dPBS). ) and added to wells containing antibody mixtures (Table 31) for incubation on ice in the dark for 45 min. Cells were then washed twice with dPBS/BSA/Na azide, fixed and permeabilized using FOXP3 buffer kit (BioLegend, catalog number 421403). Fluorescence minus one (FMO) controls were prepared

- 97 041306 for all antibodies and used to determine positive cell populations. Samples were analyzed on a Fortessa flow cytometer (BD Biosciences) and data was analyzed using Flow Jo software (Tree Star).

As shown in Table. 31, a multi-marker expression panel was designed, and ICOS expression on CD8+ and CD4+ T cells was analyzed.

Table 31. Antibodies used for immunofluorescence staining for subsets of T cells

Marker Clone Fluorophore Provider Catalog number CD3 SK7 BUV 395 B.D. Bioscientcs 564001 CD4 OKT4 BV 785 BioLegend 317442 FOXP3 206D AF647 BioLegcnd 320114 CD25 4E3 PE-e610 cBiosciencc 61-0257-42 CD152 BN 13 BV421 BD Biosciences 562743 CD45 new AF700 BD Biosciences 560566 Viability - Near 1R ThermoFisher Scientific L10119 2B4 C1.7 AF488 BioLegend 329506 CD8a SKI BV605 BD Biosciences 564116 ICOS C398.4A BV510 BioLegend 313525 CD56 NCAM16.2 BV650 BioLegend 318343 PD-1 EH12.1 PE BioLegend 560795

For Treg, Teff, B cells and NK cell staining, fresh head and neck, lung, CRC, and endometrial cancers were placed in a 6-well plate on ice, immersed in 1-2 ml of dissociation medium. The tumors were cut into small pieces and the tumor solution was placed in a Dounce homogenizer for dissociation. Tumor solutions were filtered through a 70 μm filter with additional dissociation medium and centrifuged. The resulting cells were resuspended in staining buffer. A fresh tissue sample from a metastatic omental tumor was dissociated using a Miltenyi Dissociation Kit (Miltenyi, catalog number 130-095-929). Frozen tumor samples were thawed and DNase (2 ml DNase solution) was added dropwise. Thawed medium (8 ml, heated in a bath at 37° C.) was added to the tumor and DNase solution and filtered through a 70 μm filter. Cells were centrifuged and resuspended in staining buffer.

Expression of ICOS on TIL was assessed using FACS analysis. Tumor-derived cell suspensions were blocked with staining buffer containing Near-IR Dead Cell Stain. Cell population surface markers were stained with antibodies (as shown in Table 32) to identify positive cell populations, followed by FOXP3 intracellular staining after fixation and permeabilization. Flow cytometry data were collected using a Fortessa X-20 flow cytometer. After gating the parameters of the FSC-SSC-Live/Dead markers to exclude debris and dead cells, the frequency of ICOS+ cells was determined for subsets of CD4+ Teff, Treg, CD8+ T cells, B cells and NK cells. Cd4+ analysis by flow cytometry was performed using FlowJo software.

Table 32. Antibodies used for immunofluorescent staining of B cells, NK cells and subsets of T cells

conjugate Target Company Catalog Clone Alexa Fluor® 700 CD8 BioLegcnd 301028 RPA-T8 BL1V395 CD8 BD Biosciences 563795 RPA-T8 BUV737 CD14 BD Biosciences 564444 M5E2 BUV805 CD4 B.D. Bioscientcs 564910 SK3 BV421 ICOS BioLegend 313524 C398.4A BV510 CD4S BioLegend 304036 H130 BV510 CD45RA BioLegcnd 304142 I110O BV605 CDllc BioLegcnd 301636 3.9 BV650 CD15 BioLegend 323034 W6D3 BV650 CD45 BD Biosciences 563717 shzo BV711 PD1 BD Biosciences 564017 EH12 1 BV786 CD3 B.D. Bioscientcs 563800 SK7 PE/Cy5 CD19 BioLegcnd 302210 H1B19 PE/Cy7 FOXP3 cBiosciencc 25-4777-42 236A7E7 PE-eFluor® 610 CD56 cBiosciencc 61-0567-42 CMSSB PcrCP/CyS 5 HLADR B.D. Bioscientcs 560652 G46-6

ICOS expression was assessed using anti-ICOS clone C398.4a in whole blood samples from 16 healthy donors and 14 lung cancer patients, 22 RCC patients, and 14 CRC patients. Compared with healthy donors, the rates of ICOS+ CD4+ T cells derived from patients with cancer were higher (Table 33, P<0.001 for all groups of patients with cancer compared with healthy donors, Mann-Whitney test) . Rates of ICOS+ CD8+ T cells from RCC patients were significantly higher compared to healthy donors (Table 33, P<0.01, Mann-Whitney test). In blood samples from patients with lung cancer and CRC, the percentage of ICOS+ CD8+ T cells was also higher than in samples from healthy donors, without reaching statistical significance (Table 33).

Because higher rates were observed than reported in the literature, and because clone C398.4a positively stained more T cells than another commonly used clone, ISA-3,6,7,8,9,10, the rates of cells expressing high ICOS levels (ICOShi) were also analyzed. In accordance with the works on this topic, CD8+ T cells expressed minimal amounts of ICOShi, comparable to the background ones (Table 33). However, for CD4+ T cells, ICOShi cell rates ranged on average from 3.0% in healthy donors to 4.9% in RCC patients (Table 33, RCC versus healthy: P<0.05, Mann test - Whitney).

ICOS expression was then evaluated in TIL of 11 lung cancer patients, 21 RCC patients, and 8 CRC patients. The rates of ICOS+ CD4+ and ICOS+ CD8+ TIL were similar for different types of tumors (Table 34). As in peripheral blood, high expression of ICOS by CD4+ and CD8+ TIL by T cells was measured. On average, a greater percentage of CD4+ T cells expressed high levels of ICOS than CD8+ T cells (Table 34). Co-expression of ICOS and PD-1 in TIL was also measured. High levels of PD-1 (PD-1hi) were expressed by ICOShi CD4+ TIL with great inter-patient variability (Table 34). Compared to ICOShi CD4+ TIL, a higher proportion of ICOShi CD8+ T cells co-expressed high levels of PD-1 (Table 34).

Table 33 Mean Frequency ±SD ICOS+ and ICOShi CD4+ and CD8+ T Cells in Peripheral Blood Samples from Healthy Donors and Cancer Patients

Healthy (#16) NSCLC (#14) RCC (#22) CRC (#14) % 1COS+ SEI+T cells 21±7 43±15 39±16 42±19 % ICOS+ SG>8+t_ _cells 14±8 23±14 25±12 23±14 % !COS ^h ' CD4+T cells 3.0±1.8 3.5±1.8 4.9±2.9 3.7±2.4 % !COS ^h ' CD8± T cells 0.9±0.7 0.9±0.8 1.3±0.9 0.9±0.7

Abbreviations:

ICOShi: Cells expressing high levels of ICOS; N: Number of samples; SD: Standard deviation; NSCLC: Non-small cell lung cancer; RCC: Renal cell carcinoma; CRC: Colorectal carcinoma

Table 34 Mean Rates ±SD of ICOS+, ICOShi, and PD-lhi ICOShi CD4+ and CD8+ T Cells in TIL from Cancer Patients

NSCLC (#11) RCC (#21) CRC (#8) % ICOS± CO4±T- _cells 59±21 53±19 63±14 % ICOS± CDS+Τ-κ tap holes 35±19 35±20 27±21 % ICOS ^hl CO4±T cells 28±15 19±20 29±12 % ICOS ^hl CO8 + y_ _cells 8.3±7.2 6.5±8.5 4.8±4.5 % PD-l ^bl lCOS ^hi CO4 + T cells 43±18 48±23 33±17 % PD-l ^hi ICOS* ¹ ' CO8±T-cells 63±16 71±29 62±23

Abbreviations:

SD: Standard deviation; ICOShi: Cells expressing high levels of ICOS; PD-1: Cells expressing high levels of PD-1; TIL: Tumor-filtering lymphocytes; NSCLC: Non-small cell lung cancer; RCC: Renal cell carcinoma; CRC: Colorectal carcinoma; N: Number of samples

Human tumor samples from patients with two lung adenocarcinomas, one endometrial adenocarcinoma, one sero-papillary carcinoma omental metastasis, one colorectal adenocarcinoma liver metastasis, and one head and neck squamous cell carcinoma were dissociated and stained for flow cytometric analysis of ICOS expression on different lymphocyte populations. Of five different lymphocyte populations (CD4+ Teff, Treg, CD8+ T cells, B cells, NK cells), CD4+ Teff and Treg expressed ICOS at the highest frequency. In cell populations that expressed ICOS (CD4 Teff, Tregs, CD8 T cells and NK cells), Tregs expressed more ICOS per cell compared to other cell types.

Thus, ICOS was expressed at higher levels on CD4+ T cells than on peripheral blood CD8+ T cells and TIL. ICOS expression was variable across patients and was similar across the three tumor types tested. On average, 33-48% of ICOShi CD4+ TIL and 62-71% of ICOShi CD8+ TIL co-expressed high levels of PD-1. In addition, Tregs expressed higher levels of ICOS than CD4+ Teffs, CD8+ T cells, NK cells, and B cells in the human tumor microenvironment.

Example 18 Dose Escalation and Combination Cohort Study to Evaluate Safety and Tolerability, Pharmacokinetics and Efficacy of ICOS.33 IgG1f S267E Alone or in Combination with One or More Anti-PD-1 Antibodies, One or More Anti-PD-L1 antibodies and/or one or more anti-CTLA-4 antibodies in patients with advanced solid tumors

- 99 041306

Phase 1/2 open-label study ICOS.33 IgG1f S267E administered alone or in combination with an anti-PD-1 antibody, an anti-PD-L1 antibody, and/or an anti-CTLA-4 antibody (e.g., nivolumab and/or ipilimumab ) was performed in participants with advanced solid tumors.

The study included the following parts:

1) monotherapy with dose escalation (Preliminary Safety Cohorts and Part A);

2) combination therapy with dose escalation using nivolumab (Part B) or ipilimumab (Part C); And

3) the maximum tolerated dose expansion phase with nivolumab (Part D) or ipilimumab (Part E).

Research objectives

The primary objective of this study is to evaluate the safety and tolerability of ICOS.33 IgG1f S267E, administered alone and in combination with nivolumab or ipilimumab, in participants with advanced solid tumors.

Secondary objectives include investigating the preliminary efficacy of ICOS.33 IgG1f S267E, administered alone and in combination with nivolumab or ipilimumab, in participants with advanced solid tumors; PK characterization of ICOS.33 IgG1f S267E administered alone and in combination with nivolumab or ipilimumab in participants with advanced solid tumors; characterization of the immunogenicity of ICOS.33 IgG1f S267E administered alone and in combination with nivolumab or ipilimumab in participants with advanced solid tumors; and monitoring of binding to the ICOS.33 target of IgG1f S267E, administered alone and in combination with nivolumab or ipilimumab, in participants with advanced solid tumors.

In addition, exploratory goals include studying the relationship between antitumor activity and measurements of specific biomarkers in tumor tissue and peripheral blood before and after treatment with ICOS.33 IgG1f S267E alone and in combination with nivolumab or ipilimumab; characterizing the relationship(s) between the PK of ICOS.33 IgG1f S267E, alone and in combination with the PK of nivolumab or the PK of ipilimumab, and safety, efficacy and/or clinical biomarkers; assessment of overall survival (OSR) in participants treated with ICOS.33 IgG1f S267E alone and in combination with nivolumab or ipilimumab; characterization of the PK and immunogenicity of nivolumab and ipilimumab when administered in combination with ICOS.33 IgG1f S267E; characterization of the immunogenicity of nivolumab and ipilimumab when administered in combination with ICOS.33 IgG1f S267E; evaluation of the potential effect of ICOS.33 IgG1f S267E on the corrected QT interval (QTc); and studying the associations between selected peripheral blood biomarkers and the incidence of adverse events (AEs) and serious adverse events (SAEs).

General design

A schematic of the study design is shown in Fig. 27.

Monotherapy consists of two different cohorts:

Preliminary Safety Cohorts: ICOS.33 IgG1f S267E was administered as monotherapy at a dose of 2 mg and 8 mg once every four weeks for 24 weeks.

Part A: ICOS.33 IgG1f S267E was administered at 25, 80, 200, 400 and 800 mg once every four weeks for 24 weeks.

Parts B and C consist of cohorts receiving various combination therapies, including:

B1: ICOS.33 IgG1f S267E administered once every 12 weeks + nivolumab 480 mg once every 4 weeks at baseline dose level of ICOS.33 IgG1f S267E, recommended Bayesian Logistic Regression Model (BLRM)-Copula and available PK/PD data from Part A.

B2: ICOS.33 IgG1f S267E administered once every 4 weeks + nivolumab 480 mg once every 4 weeks at the ICOS.33 IgG1f S267E dose level recommended by the Bayesian Logistic Regression Model (BLRM)-Copula (BLRM-RD) and available PK data /PD from part A.

C1: ICOS.33 IgG1f S267E administered once every 12 weeks + ipilimumab 3 mg/kg once every 4 weeks at baseline dose level of ICOS.33 IgG1f S267E, recommended Bayesian Logistic Regression Model (BLRM)-Copula and available PK/PD data from part A.

C2: ICOS.33 IgG1f S267E administered once every 4 weeks + ipilimumab 3 mg/kg once every 4 weeks at the ICOS.33 IgG1f S267E dose level recommended by the Bayesian Logistic Regression Model (BLRM)-Copula and available PK/PD data from parts A.

Recruitment in parts B1 and C1 was carried out simultaneously. Recruitment in Parts B2 and C2 was performed only if additional data on safety, PK or PD were required for dose optimization and/or regimen selection.

ICOS.33 IgG1f S267E doses for Parts B and C (combination with nivolumab or ipilimumab) were determined using all available data on safety (clinical and laboratory), PK, and target-binding biomarkers/pharmacodynamic biomarkers, as well as modeling recommendations within the modeling Bayesian hierarchical networks, i.e. 100 041306 BLRM-Copula model to include single agent toxicity profiles of both ICOS.33 IgG1f S267E (Pre-Safety Cohorts and Part A) and nivolumab or ipilimumab, and any available combination toxicity profiles from Parts B and C (for subsequent doses of ICOS.33 IgG1f S267E in Parts B and C), modeling PK/PD, and did not exceed the maximum administered dose (MAD) of ICOS.33 IgG1f S267E monotherapy in Pre-Safety Cohorts and Part A. Dose level of ICOS.33 IgG1f S267E recommended BLRM-Copula model, i.e. BLRM-RD is defined as a general concept such that the BLRM-RD for any cohort is always based on all available and most recent information.

At no time did the dose of ICOS.33 IgG1f S267E given in combination with nivolumab or ipilimumab (Parts B and C) exceed the dose of ICOS.33 IgG1f S267E determined to be safe in the dose escalation monotherapy group (Part A), or at no time during combination therapy in parts B and C did the dose of ICOS.33 IgG1f S267E exceed the highest dose determined to be tolerated in the dose escalation monotherapy group (part A). In addition, the starting dose level of ICOS.33 IgG1f S267E used in combination with nivolumab or ipilimumab (Parts B and C) was one dose level lower than that of monotherapy (Part A) that had passed the DLT period.

Parts B1 and C1 consist of a PK/pharmacodynamic substudy aimed at studying the downregulation kinetics and re-expression of the ICOS receptor (and/or changes in selected target-binding biomarkers/pharmacodynamic biomarkers) following administration of ICOS.33 IgG1f S267E in the presence of multiple doses of nivolumab alone once every 4 weeks (Part B1) or ipilimumab once every 4 weeks (Part C1).

Various doses of ICOS.33 IgG1f S267E were administered in parts B1 and C1:

Doses that induce or are predicted to induce various levels of downregulation of the ICOS receptor, including at least one dose that induces nearly complete downregulation of the receptor (and/or alteration of selected target binding biomarkers/pharmacodynamic biomarkers) for at least at least 4 weeks. These dose levels allow characterization of the kinetics of ICOS receptor re-expression following near-complete down-regulation for a period of time equal to, less than, and/or greater than the dosing intervals used in Part A. Understanding the kinetics of the ICOS receptor can provide information on the dosing intervals of the ICOS under test.33 IgG1f S267E for future study(s).

BLRM-RD: Dose levels were determined based on all available safety data (clinical and laboratory) and PK, as well as changes in peripheral blood target-binding markers (eg, downregulation of ICOS on T cells and ICOS+ B cells) from previous studies. and completed parts of current cohorts, and/or the BLRM/BLRM-Copula model, when applicable.

After 24 weeks of treatment as monotherapy or two years of combination therapy, the participant may be eligible for retreatment. For Part A, scans are collected centrally and can be reviewed by Blinded Independent Central Assessment (BICR) at a later time or at any time during the course of the study. For Parts B and C, scans are collected centrally for real-time revision by BICR.

Physical examinations, vital signs measurements, 12-lead ECG, and clinical-instrumental assessments were performed at specific times throughout dosing intervals.

Participants were closely monitored for AE throughout the study. Blood was collected for PK analyzes during follow-up visits on days 30, 60 and 100 after study drug administration.

Participants completed four phases of the study: screening, treatment, safety follow-up, and response/survival follow-up, as described below. The total duration of participation in the study was about 2 years.

Tetanus vaccine

All patients in Parts A, B, and C received an approved tetanus vaccine. Administration of a powerful sensitizing antigen, such as tetanus toxoid, primes the immune system, induces an immune response, and promotes a more immunogenic state.

The ability of ICOS.33 IgG1f S267E to enhance the secondary immune response will be determined by monitoring anti-tetanus antibodies and CD4+ T cell proliferative and cytokine responses following tetanus vaccination. Approximately 70% of the general population has protective antibodies to tetanus. However, cellular immune responses are usually detected in the peripheral blood one month after tetanus vaccination. Tetanus has been used as a reporter antigen in cancer patients receiving vaccine immunotherapy and can be easily controlled. Therefore, tetanus vaccination may provide a powerful secondary immune response with ICOS.33 IgG1f S267E alone and in combination with nivolumab or ipilimumab.

- 101 041306

Screening

The screening phase lasted up to 28 days and ran until the first administration of study drug. During the screening phase, the participant's initial eligibility is established and an informed consent form is signed. Tumor biopsies are collected from all participants, centrally assessed for ICOS expression by immunohistochemistry, and the results are evaluated prior to the first dose of study drug. Participants are recruited using interactive communications technology (IRT).

Treatment phase

The treatment phase in the Pre-Safety Cohort and Part A consists of six four-week treatment cycles (1 cycle = 28 days). In the Pre-Safety and Part A Cohort, each treatment cycle consisted of ICOS.33 IgG1f S267E monotherapy for a total of 24 weeks.

Dose levels for parts B and C were determined based on all available safety data (clinical and laboratory) and PK, as well as changes in peripheral blood target-binding markers (eg, down-regulation of ICOS on T cells and ICOS+ B cells) from previous and completed parts of the current cohorts, and guided by the BLRM/BLRM-Copula model, as needed.

Parts B1 and C1 used four-week cycles, ie ICOS.33 IgG1f S267E + nivolumab or ipilimumab was administered starting on Day 1 of Cycle 1. Nivolumab and ipilimumab were administered on Day 1 of each cycle. ICOS.33 IgG1f S267E was administered once every 12 weeks or on Day 1 of every third cycle (Day 1 of Cycle 1, Day 1 of Cycle 4, Day 1 of Cycle 7, etc.). Participants in Parts B1 and C1 continued treatment for up to a total of 2 years.

The treatment phase in parts B2 and C2 consisted of ICOS.33 IgG1f S267E + nivolumab or ipilimumab administered on Day 1 of each cycle for up to 2 years in total, and participants were recruited only if additional safety data were required, PK or PD to optimize the dose and/or mode of administration.

After each treatment cycle, the decision to treat the participant with additional cycles of study drug was based on tumor evaluations performed every 12 weeks (once every 12 weeks ± 1 week) and completed before the first dose in the next cycle. Tumor progression or response endpoints were assessed according to Response Evaluation Criteria In Solid Tumors (RECIST) v1.1 or Prostate Cancer Working Group 3 (PCGW3) guidelines. (Scher et al., 2016. Trial Design and Objectives for Castration-Resistant Prostate Cancer: Updated Recommendations From the Prostate Cancer Clinical Trials Working Group 3. Clin Oncol. 34(12): 1402-1418).

Post-progression treatment with additional cycles of study drug was allowed up to a maximum of 24 weeks for Part A and two years for Parts B, C, D, and E in selected participants with initially defined PD according to RECIST v1.1 or PCGW3 (prostate only) after discussion and agreement between the clinical trial director and the clinical supervisor/principal that the benefit/risk assessment supports continued administration of the investigational product (e.g., participants continue to experience clinical benefit as assessed by the investigator, tolerate treatment, and meet other specific criteria).

Participants with an unconfirmed progressive disease (PD), stable disease (SD), partial response (PR), or complete response (CR) response at the end of a given cycle progressed to the next treatment cycle. Participants were generally allowed to continue treatment within the study until one of the following occurred:

1) completion of the maximum number of cycles,

2) PD confirmed,

3) clinical deterioration, suggesting no likelihood of further benefit from treatment,

4) intolerance to therapy or

5) a participant who meets the study's treatment discontinuation criteria.

Individual participants with confirmed CR were given the option of discontinuing study treatment on a case-by-case basis, after specific consultation and agreement between the clinical trial director and the clinical supervisor/principal, in situations where the benefit/risk justified discontinuation of study treatment.

Follow-up to assess safety

Upon completion of 24 weeks of study treatment for Part A (or up to a maximum of 48 weeks, if required) or two years for Parts B, C, D, and E (or up to a maximum of four years, if necessary), a decision was made to terminate the participant's participation in the study (eg, after completion of treatment [EOT]), and all participants entered a follow-up period to assess safety.

For participants who have completed all scheduled treatment cycles, the EOT visit is the same as the last scheduled and end-of-treatment visit and the start of the safety follow-up visit at week 1. For participants who have not completed all scheduled treatment cycles within the study, the EOT visit is the most recent visit of the treatment period (with all available safety and response data) and was considered the start of the follow-up visit to assess safety.

After the EOT visit, all participants were assessed for new AEs for at least 100 days after the last treatment dose in the study. AE follow-up visits occurred on Days 30, 60, and 100 after the last dose or the day of discontinuation (±7 days). All participants must complete 3 safety follow-up visits, whether they started a new cancer treatment or not, except for those participants who withdrew their consent to participate in the study.

Follow-up to assess survival

After completion of follow-up visits to assess safety, all participants who received monotherapy and combination therapy entered into a follow-up period to assess survival. Participants were followed up approximately every 3 months (12 weeks) until death, loss of vision, withdrawal of informed consent, or completion of the study, whichever came first. The duration of this phase was up to two years after the first treatment dose in the study, although in individual cases, if evidence of effectiveness is evident, a longer follow-up period is considered. Follow-up to assess response

After completion of the safety follow-up period, participants with a current SD, PR, or CR at the time of the EOT visit proceeded to the response follow-up period. This period occurred concurrently with the follow-up period to assess the survival of the mentioned participants. Participants continued to have x-ray and clinical tumor evaluations approximately every 3 months (12 weeks) until death, loss of vision, withdrawal of informed consent, or completion of the study, whichever came first. Radiological assessments of tumors for participants with ongoing clinical improvement continued after participants completed the survival phase of the study. Participants who had disease progression after the initial course of study treatment were not assessed for response after the EOT visit and were allowed to receive other tumor-targeted therapy as needed. If the participant interrupted treatment for any reason other than PD, radiological follow-up was performed until the participant received additional treatment. Treatment with additional cycles after 24 weeks

All participants received treatment for 24 weeks as monotherapy or combination therapy if the criteria for discontinuation of treatment were not previously achieved. All participants completing treatment with ongoing disease control (CR, PR, or SD) or unconfirmed PD were eligible for an additional 24 weeks of treatment in the Part A study or a total of two years for combination therapy on a case-by-case basis after careful evaluation and discussion. with the BMS medical monitor/principal to determine if the risk/benefit ratio supports further study treatment. After the completion of the additional treatment period in the study, all participants entered a follow-up period to assess safety.

Treatment after progression

Post-progression treatment was permitted in selected participants with initial PD as defined by RECIST v1.1 or PCGW3 (prostate only) after discussion and agreement with the BMS medical supervisor/principal that a benefit/risk assessment favors continued administration investigational drug (for example, participants continue to experience clinical benefit as assessed by the investigator, tolerate treatment, and meet other criteria).

Participants re-consented with an Informed Consent Form (ICF) to continue treatment after progression. Treatment after progression requires continued tumor evaluations. Retreatment during follow-up to assess response

Retreatment was permitted in this study if disease progressed during the follow-up period. Participants who complete approximately 24 weeks of study treatment (or up to a maximum of 48 weeks, if necessary) for Part A and approximately two years of study treatment (or up to a maximum of 4 years, if necessary) for Parts B, C, D, and E or less, in case of interruption of treatment due to CR, which entered the follow-up period to assess response with continued disease control (CR, PR or SD) and without any significant toxicity, are eligible for re-treatment after subsequent confirmed disease progression for 12 months after the last dose of study drug on a case-by-case basis, after careful evaluation and discussion with the BMS medical monitor/principal to determine whether the risk/benefit ratio supports the addition of study treatment, and whether the participant continues to meet inclusion criteria for treatment with the study drug.

Participants eligible for retreatment are treated with the original monotherapy or combination therapy regimen (e.g., same dose and dosing regimen administered for the first 24 weeks), unless it is subsequently determined that the dose(s) and regimen exceed the last BLRM-RD score, in which case the participant is receiving BLRM-RD treatment. Participants entering this phase follow a schedule of procedures. Samples for PK and pharmacodynamics are collected less frequently (before dosing in each treatment cycle). During retreatment, blood-derived pharmacodynamic biomarker samples are collected.

Participant type and target disease characteristics

a) Participants must be at least 18 years of age and have histological or cytological evidence of metastatic and/or unresectable colorectal cancer (CRC), head and neck squamous cell carcinoma (HNSCC), non-small cell lung cancer (NSCLC), prostate adenocarcinoma (PRC) and urothelial cancer (UCC) with measurable disease according to RECIST v1.1 or PCGW3 (prostate only) and have at least 1 biopsyable lesion in addition to the target lesion.

b) Presence of at least 1 lesion with measurable disease as defined in RECIST v1.1 or PCGW3 (prostate only) for solid tumors to assess response. Participants with lesions in a previously exposed area as the only site of measurable disease are eligible for inclusion provided that the lesion(s) has shown clear progression and can be measured.

c) Participants must have received and then progressed or tolerated at least 1 standard treatment regimen in late stage or metastatic settings, if such therapy exists, and were considered for all other potentially effective therapies prior to inclusion.

d) Participants previously treated with any agent specifically targeting checkpoint pathway inhibition (such as αntu-PD-1, anti-PD-L1, or anti-CTLA-4) were admitted after a washout period of any time greater than 4 weeks since the last treatment.

Types of tumors

a) Colorectal cancer (CRC)

i) Histologically confirmed CRC that is metastatic or recurrent with documented disease progression;

ii) Documented microsatellite instability, mismatch repair, KRAS and BRAF status if known;

iii) Prior Therapy Requirement: Participants must have received at least 1 but no more than 3 prior therapies for metastatic and/or unresectable disease (or have progressed within 6 months of adjuvant therapy);

iv) The participant must have incurable metastatic disease (i.e., patients with disease that is potentially treatable by surgical resection are not eligible for treatment).

b) Head and neck squamous cell carcinoma (HNSCC) (oral cavity, pharynx, larynx)

i) Histologically proven incurable locally advanced, recurrent, or metastatic HNSCC (oral, pharyngeal, laryngeal), stage III or IV, and not amenable to topical therapy to cure (surgery or radiotherapy with or without chemotherapy).

ii) Must have documented HPV status and subtype, specifically HPV16 and HPV18.

iii) Participants must have received and then progressed to or were intolerant or refractory to at least 1 but no more than 2 previous regimens of systemic therapy (eg, platinum-based chemotherapy) for the treatment of advanced metastatic or locally advanced unresectable disease.

iv) Prior curative radiation therapy must have been completed at least 4 weeks prior to study drug administration. Prior focal palliative radiation therapy must have been completed at least 2 weeks prior to study drug administration.

c) Non-small cell lung cancer (NSCLC)

i) Participants must have histological or cytological confirmation of NSCLC (according to the seventh International Association for the Study of Lung Cancer [IASLC]) with squamous or non-squamous histology that is at an advanced stage (metastatic and/or inoperable).

- 104 041306 (1) Participants must have had at least 1, but not more than 2, prior NSCLC systemic therapies Maintenance, adjuvant, or neoadjuvant (chemotherapy or chemoradiation) therapy is not considered an additional line of treatment.

(2) Participants were to be offered platinum-based chemotherapy for NSCLC. Platinum-based chemotherapy could be administered under adjuvant, neoadjuvant, or chemoradiation settings. Participants with relapsing/metastatic disease that relapses within 6 months of completing such treatment are considered eligible for treatment within the treatment. Prior adjuvant or neoadjuvant chemotherapy is permitted if the last administration of the previous regimen occurred at least 4 weeks prior to study entry.

(3) Previous radical chemoradiation of a locally advanced disease is also allowed if the last chemotherapy or radiotherapy (depending on which therapy was last administered) was at least 4 weeks prior to enrollment in the study.

(4) . Participants with known EGFR mutations or ALK rearrangements should receive EGFR or ALK inhibitors, respectively. The mutational status of EGFR, ALK, KRAS and ROS1 should be documented if known.

d) Adenocarcinoma of the prostate (PRC)

i) Histological or cytological confirmation of prostate adenocarcinoma.

ii) Participants have received orchiectomy treatment or are receiving a luteinizing hormone releasing hormone analogue and have testosterone levels <50 ng/dl.

iii) Metastatic process in any 1 of the following forms: computed tomography (CT), magnetic resonance imaging (MRI) and bone scan.

f) Urothelial carcinoma (UCC)

i) Histological or cytological confirmation of metastatic or surgically inoperable transitional urothelial cell carcinoma involving the bladder, urethra, ureter, or renal pelvis. Dose limiting toxicity (DLT) ICOS.33 IgG1fS267E.

To allow dose escalation, DLT was determined based on the frequency, intensity, and duration of adverse events (AEs) for which no clear alternative cause was identified. The DLT period is 35 days (5 weeks).

In the Pre-Safety Cohorts, participants who received 1 dose of ICOS.33 IgGlf S267E and completed the study, or who discontinued treatment due to DLT in the 4-week DLT period, are considered to be eligible for DLT for ICOS.33 IgG1f S267E monotherapy.

In Part A, participants who received 2 doses of ICOS.33 IgG1f S267E and completed the study, or who discontinued treatment due to DLT in the 5-week DLT period, are considered participants being evaluated for DLT for ICOS.33 IgG1fS267E monotherapy.

In Parts B, C, D, and E, participants receiving 1 dose of ICOS.33 IgG1f S267E or 2 doses of either nivolumab or ipilimumab, or participants who discontinued treatment due to DLT during the 5-week DLT evaluation period following combination treatment, are considered to be eligible for DLT participants for combination treatment. Participants who left the study during the DLT evaluation period or received less than 2 doses for reasons other than DLT in monotherapy (Part A) or 1 dose in combination therapy (Parts B, C, D, E) are not considered DLT-evaluable participants and are not replaced by a new participant at the same dose level. Participants in Part A who had a dose delay during the DLT evaluation period for reasons other than DLT are considered DLT-evaluated participants if they received at least 2 doses of therapy.

For the purposes of participant management, any AE that meets the criteria for DLT, regardless of the cycle in which it occurs, results in discontinuation of study treatment. Participants who left the study during the 5-week DLT evaluation period for reasons other than DLT may be replaced with a new participant at the same dose level. The frequency of occurrence of DLT during the 5-week DLT evaluation period is taken into account when deciding whether to increase the dose and to determine BLRM-RD. Adverse events (AEs) occurring after the DLT period are considered for the purpose of determining the BLRM-RD by agreement between the sponsor, medical observer/principal, and investigators.

Participants experiencing DLT enter a follow-up period to assess safety. DLTs occurring after a DLT follow-up period of 4 weeks for Pre-Safety Cohorts or after a DLT follow-up period of 5 weeks for parts A, B and C are taken into account when determining the maximum administered dose (MAD) for the combination part.

This study showed that administered anti-ICOS antibodies are safe and effective in the treatment of cancer.

Example 19 Combined Effects of Increasing Anti-ICOS Antibody Doses on Tumor Growth

Effect of increasing doses of anti-ICOS agonist antibody, ICOS.33 IgG1f S267E, in combination

- 105 041306 with anti-PD-l antibody was evaluated for inhibition of tumor growth in a mouse model. As shown in FIG. 28, the combination showed reduced efficacy at higher doses, ie. a hook effect in which near-saturating or saturating concentrations of an antibody result in a decrease in potency compared to that of an antibody at lower concentrations, i.e. concentrations that do not lead to saturation.

Briefly, mice (averaging about 20 mg) with established CT26 tumors were treated with either anti-PD-1 monotherapy or in combination with ICOS.33 IgG1f S267E. Dose escalation of anti-ICOS was started at 0.1 mg/kg with a three-fold increase to 10 mg/kg (or a maximum dose of approximately 200 μg/mouse, fixed dose). The anti-PD-1 antibody was administered at a dose of 10 mg/kg (or a maximum dose of approximately 200 μg/mouse, fixed dose). Anti-ICOS and anti-PD1 antibody were administered at the same schedule (ie every 4 days starting on Day 7) after tumor implantation.

As shown in FIG. 28, maximal tumor growth inhibition (TGI) with anti-ICOS and anti-PD1 combination therapy was observed at a lower dose of anti-ICOS antibody (3mg/kg) than at the highest dose tested (10mg/kg), demonstrating a reduction in TGI. at doses exceeding 3 mg/kg, i.e. maximum efficiency is achieved at subsaturating doses.

Table 35

Summary information on the sequence listing

SE QID NO Sequence name Subsequence 1 Human ICOS (NP_03 6224.1) MKSGLWYFFL FCLRIKVLTG EINGSANYEM FIFHNGGVQI 40 LCKYPDIVQQ FKMQLLKGGQ ILCDLTKTKG SGNTVSIKSL 80 KFCHSQLSNN SVSFFLYNLD HSHANYYFCN LSIFDPPPFK 120 VTLTGGYLHI YESQLCCQLK FWLPIGCAAF VVVCILGCIL 160 ICWLTKKKYS SSVHDPNGEY MFMRAVNTAK KSRLTDVTL 199 2 Human ICOS-L (NP 001269979 •O MRLGSPGLLF LLFSSLRADT QEKEVRAMVG SDVELSCACP 40 EGSRFDLNDV YVYWQTSESK TVVTYHIPQN SSLENVDSRY 80 RNRALMSPAG MLRGDFSLRL FNVTPQDEQK FHCLVLSQSL 120 GFQEVLSVEV TLHVAANFSV PVVSAPHSPS QDELTFTTCTS 160 INGYPRPNVY WINKTDNSLL DQALQNDTVF LNMRGLYDVV 200 SVLRIARTPS VNIGCCIENV LLQQNLTVGS QTGNDIGERD 240 KITENPVSTG EKNAATWSIL AVLCLLVVVA VAIGWVCRDR 280 CLQHSYAGAW AVSPETELTE SWNLLLLLS 309 3 heavy chain EVQLVESGGG LVKPAGSLTL SCVASGFTFS DYFMHWVRQA 40

- 106 041306

parental hamster antibody PGKGLEWVAV IDTKSFNYAT YYSDLVKGRF TVSRDDSQGM 80 VYLQMNNLRK EDTATYYCTA TIAVPYYFDY WGQGTMVTVS 120 SATTTAPSVY PLAPACDSTT STTNTVTLGC LVKGYFPEPV 160 TVSWNSGALT SGVHTFPSVL HSGLYSLSSS VTVPSSTWPS 200 QTVTCNVAHP ASSTKVDKKI VPGDGSGCKP CTCPGPEVSS 240 VFIFPPKPKD VLTISLSPKV TCVVVDISQD DPEVQFSWFI 280 DGKEVHTAVT QPREEQFNST YRMVSVLPIL HQDWLNGKEF 320 KCKVNSPAFP VPIEKTISKR RGQLQVPQVY TMPPPKEQLT 360 QSQVSLTCMI KGFYPEDIDV AWQKNGQPEQ SFKNTPPVLD 400 TDETYFLYSK LDVKKDDWEK GDTFTCSVVH EALHNHHTEK 440 TLSQRPGK 448 4 Parental hamster antibody light chain DIQMTQSPSS LPASLGDRVT INCQASQDIS NYLSWYQQKP 40 GKAPKLLIYY TNLLADGVPS RFSGSGSGRD YSFTISSLES 80 EDIGSYYCQQ YYNYRTFGPG TKLEIKRADA KPTVSIFPS 120 SEQLGTGSAT LVCFVNNFYP KDINVKWKVD GSEKRDGVLQ 160 SVTDQDSKDS TYSLSSTLSL TKADYERHNL YTCEVTHKTS 200 TAAIVKTLNR NEC 213 5 Heavy chain variable domain ICOS.33 IgGlf S267E EVQLVESGGG LVKPGGSLRL SCAASGFTFS DYFMHWVRQA 40 PGKGLEWVGV IDTKSFNYAT YYSDLVKGRF TISRDDSKNT 80 LYLQMNSLKT EDTAVYYCTA TIAVPYYFDY WGQGTLVTVS 120 S 121 6 Light chain variable domain ICOS.33 IgGlf S267E DIQMTQSPSS LSASVGDRVT ITCQASQDIS NYLSWYQQKP 40 GKAPKLLIYY TNLLAEGVPS RFSGSGSGTD FTFTISSLQP 80 EDIATYYCQQ YYNYRTFGPG TKVDIK 106 7 heavy chain ICOS.33 IgGlf S267E EVQLVESGGG LVKPGGSLRL SCAASGFTFS DYFMHWVRQA 40 PGKGLEWVGV IDTKSFNYAT YYSDLVKGRF TISRDDSKNT 80 LYLQMNSLKT EDTAVYYCTA TIAVPYYFDY WGQGTLVTVS 120 SASTKGPSVF PLAPSSKSTS GGTAALGCLV KDYFPEPVTV 160 SWNSGALTSG VHTFPAVLQS SGLYSLSSVV TVPSSSLGTQ 200 TYICNVNHKP SNTKVDKRVE PKSCDKTHTC PPCPAPELLG 240 GPSVFLFPPK PKDTLMISRT PEVTCVVVDV EHEDPEVKFN 280 WYVDGVEVHN AKTKPREEQY NSTYRVVSVL TVLHQDWLNG 320 KEYKCKVSNK ALPAPIEKTI SKAKGQPREP QVYTLPPSRE 360 EMTKNQVSLT CLVKGFYPSD IAVEWESNGQ PENNYKTTPP 400 VLDSDGSFFL YSKLTVDKSR WQQGNVFSCS VMHEALHNHY 440 TQKSLSLSPG 450 8 Light chain ICOS.33 IgGlf S267E DIQMTQSPSS LSASVGDRVT ITCQASQDIS NYLSWYQQKP 40 GKAPKLLIYY TNLLAEGVPS RFSGSGSGTD FTFTISSLQP 80 EDIATYYCQQ YYNYRTFGPG TKVDIKRTVA APSVFIFPPS 120 DEQLKSGTAS VVCLLNNFYP REAKVQWKVD NALQSGNSQE 160 SVTEQDSKDS TYSLSSTLTL SKADYEKHKV YACEVTHQGL 200 SSPVTKSFNR GEC 213 9 ICOS.33 IgGlf S267E CDRH1 DYFMH 5 10 ICOS.33 IgGlf S267E CDRH2 VIDTKSFNYA TYYSDLVKG 19 eleven ICOS.33 IgGlf S267E CDRH3 TIAVPYYFDY 10 12 ICOS.33 IgGlf QASQDISNYL S 11

- 107 041306

S267E CDRL1 13 CDRL2 parental hamster antibody YTNLLAD 7 14 ICOS.33 IgGlf S267E CDRL2 YTNLLA 7 15 ICOS.33 IgGlf S267E CDRL3 QQYYNYRT 8 16 Heavy chain variable domain 17C4 MDILCSTLLL LTVPSWVLSQ VTLRESGPAL VKPTQTLTLT 40 CTFSGFSLST SGMCVSWIRQ PPGKALEWLA LIDWDDDKFY 80 STSLKTRLTI SKDTSKNQVV LTMTNMDPVD TATYYCARMS 120 TPTYYGLDVW GQGTTVTVSS 140 17 Light chain variable domain 17C4 MRVLAQLLGL LLLCFPGARC DIQMTQSPSS LSASVGDRVT 40 ITCRASQGIS SWLAWYQQKP EKAPKSLIYA ASSLQSGVPS 80 RFSGSGSGTD FTLTISSLQP EDFATYYCQQ YNSYPLTFGG 120 GTKVEIK 127 18 17С4 CDRH1 TSGMCVS 7 19 17С4 CDRH2 LIDWDDDKFY STSLKT 16 20 17С4 CDRH3 MSTPTYYGLD V 11 21 17С4 CDRL1 RASQGISSWL A 11 22 17С4 CDRL2 AASSLQS 7 23 17С4 CDRL3 QQYNSYPLT 9 24 Heavy chain variable domain 9D5 MDILCSTLLL LTIPSWVLSQ ITLKESGPTL VKPTQTLTLT 40 CTFSGFSLGT SGLGVGWIRQ PPGKALEWLA FIYWDDDKRY 80 SPSLKSRLTI TKDTSKNQVV LTMTNMDPVD TATYYCAHRR 120 GFFDYWGQGT LVTVSS 136 25 Light chain variable domain 9D5 MRVLAQLLGL LLLCFPGARC DIQMTQSPSS LSASVGDRVT 40 ITCRASQGIS SWLAWYQQKP EKAPKSLIYA ASSLQSGVPS 80 RFSGSGSGTD FTLTISSLQP EDFATYYCQQ YNSYPLTFGG 120 GTKVEIK 127 26 9D5 CDRH1 TSGLGVG7 27 9D5 CDRH2 FIYWDDDKRY SPSLKS 16 28 9D5 CDRH3 RRGFFDY 7 29 9D5 CDRL1 RASQGISSWL A 11 thirty 9D5 CDRL2 AASSLQS 7 31 9D5 CDRL3 QQYNSYPLT 9 32 Variable domain of heavy MEFGLTWVFL VALLRGVQCQ VQLVESGGGV VQPGMSLRLS 40 CAASGFTFST YGMQWVRQAP GKGLEWVTVI WHDGSHKDYA 80

- 108 041306

3E8 chains DSVKGRFTIS RDNSKNTMYL QMNSLRAEDT AVYYCARDRQ 120 TGEGYFDFWG QGTLVTVSS 139 33 Light chain variable domain 3E8 MRVLAQLLGL LLLCFPGARC DIQMTQSPSS LSASVGDRVT 40 ITCRASQGIS SWLAWYQQKP EKAPKSLIYA ASSLQSGVPS 80 RFSGSGSGTD FTLTISSLQP EDFATYYCQQ YNSYPYTFGQ 120 GTKLEIK 127 34 3E8 CDRH1 TYGMQ 5 35 3E8 CDRH2 VIWHDGSHKD YADSVKG 17 36 3E8 CDRH3 DRQTGEGYFD F 11 37 3E8 CDRL1 RASQGISSWL A 11 38 3E8 CDRL2 AASSLQS 7 39 3E8 CDRL3 QQYNSYPYT 9 40 Heavy chain variable domain 1D7 MDTLCSTLLL LTIPSWVLSQ ITLKESGPTL VKPTQTLTLT 40 CTFSGFSLGS NGLGVGWIRQ PPGKALEWLA LIYWDDDKRY 80 SPSLKSRLTI TKDSSKNQVV LTMTNMDPVD TATYYCAHRN 120 SGFDYWGQGI LVTVSS 136 41 Light cspi-a variable domain 1D7 MRVLAQLLGL LLLCFPGARC DIQMTQSPSS LSASVGDRVT 40 ITCRASQGFS SWLAWYQQKP EKAPKSLIYA ASSLQSGVPS 80 RFSGSGSGTD FTLTISSLQP EDFATYYCQQ YNSYPYTFGQ 120 GTKLEIK 127 42 1D7 CDRH1 SNGLGVG 7 43 1D7 CDRH2 LIYWDDDKRY SPSLKS 16 44 1D7 CDRH3 RNSGFDY 7 45 1D7 CDRLl-a RASQGFSSWL A 11 46 1D7 CDRL2-a AASSLQS 7 47 1D7 CDRL3-a QQYNSYPYT 9 48 Light chain variable domain-b 1D7 MRVLAQLLGL LLLCFPGARC DIQMTQSPSS LSASVGDRVT 40 ITCRASQGIS SWLAWYQQKP EKAPKSLIYA ASSLQSGVPS 80 RFSGSGSGTD FTLTISSLQP EDFATYYCQQ YNSYPLTFGG 120 GTKVEIK 127 49 1D7 CDRLl-b RASQGISSWL A 11 50 1D7 CDRL2-b AASSLQS 7 51 1D7 CDRL3-b QQYNSYPLT 9 52 Heavy constant domain ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80

- 109 041306

huIgGlf chains YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 53 huIgGlf S267E heavy chain constant domain (“SE”) ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVE HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 54 Heavy chain constant domain huIgGlf S267E/L328F (“SELF”) ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVE HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA FPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 55 Heavy chain constant domain huIgGlf P238D ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 56 Heavy chain constant domain huIgGlf P238D/P271G (“V4”) ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDGEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 57 huIgGlf heavy chain constant domain P238D/P271G (“V4”) D270E ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEEGEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280

- 110 041306

LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 58 Heavy chain constant domain huIgGlf E233D/P238D/P 271G/A330R (“V7”) ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPDLLGG 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDGEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPRPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 59 Heavy chain constant domain huIgGlf G237D/P238D/ H268D//P271G (“V8”) ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGD 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS DEDGEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 60 Heavy chain constant domain huIgGlf G237D/P238D/P271G/A330R (“V9”) ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGD 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDGEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPRPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 61 Heavy chain constant domain huIgGlf G237D/P238D/P271G/A330R (“V9”) D270E ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGD 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEEGEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPRPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 62 Heavy chain constant domain huIgGlf G237D/P238D/ H268D/P271G/ A330R (“VII”) ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGD 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS DEDGEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPRPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 63 Heavy constant domain ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80

- 111 041306

huIgGlf chains E233D/G237D/ P238D/H268D/ P271G/A330R (“V12”) YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPDLLGD 120 DSVFLFPPKP KDTLMISRTP EVTCVVVDVS DEDGEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPRPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 64 huKappa light chain constant domain RTVAAPSVFI FPPSDEQLKS GTASVVCLLN NFYPREAKVQ 40 WKVDNALQSG NSQESVTEQD SKDSTYSLSS TLTLSKADYE 80 KHKVYACEVT HQGLSSPVTK SFNRGEC 107 65 Signal sequence MRAWIFFLLC LAGRALA 17 66 C-terminal CHI IgGl (same for IgG3 (17-15-1515), igG3 (1715-15), IgG3 (17-15), IgG3 (15-15-15), IgG3 (15) and IgG4 VDKRV 5 67 Upper Hinge IgGl EPKSCDKTHT 10 68 Middle hinge IgGl SRRSR 5 69 IgGl lower hinge (same for ₁ oz OZ(1715-15-15), IgG3 (17-15-15), IgG3 (17-15), IgG3 (15-15-15), IgG3 (15) and IgG4) APELLGG 7 70 C-terminal CHI IgG2 VDKTV 5 71 Middle hinge IgG2 CCVECPPCP 9 72 Inferior hinge IgG2 APPVAG 6 73 Upper Hinge IgG3 (17-15-15-15) (same for _{1 oz} OZ(1715-15) and IgG3 (P-15)) ELKTPLGDTT NT 12 74 Middle hinge IgG3 (17-15-15-15) CPRCPEPKSC DTPPPCPRCP EPKSCDTPPP CPRCPEPKSC 40 DTPPPCPRCP 50

- 112 041306

75 Middle hinge IgG3 (17-15-15) CPRCPEPKSC DTPPPCPRCP EPKSCDTPPP CPRCP 35 76 Middle hinge IgG3 (17-15) CPRCPEPKSC DTPPPCPRCP 20 77 Upper hinge IgG3 (15-15-15) (same for IgG3(15)) EPKS 4 78 Middle hinge IgG3 (15-15-15) CDTPPPCPRC PEPKSCDTPP PCPRCPEPKS CDTPPPCPRC 40 P 41 79 Middle Hinge IgG3 (15) CDTPPPCPRC P 11 80 Upper hinge IgG4 ESKYGPP 7 81 Middle hinge IgG4 CPSCP 5 82 Inferior Hinge IgG4 APEFLGG 7 83 wild-type human IgGl CH1 ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKKV 98 84 CH1 human IgG2 wild type ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTV 98 85 CH2 human wild-type IgGl PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 40 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 80 EYKCKVSNKA LPAPIEKTIS KAK 103 86 CH2 human IgGl with A330S/P331S PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 40 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 80 EYKCKVSNKA LPSSIEKTIS KAK 103 87 SNZ human IgGl wild type GQPREPQVYT LPPSREEMTK NQVSLTCLVK GFYPSDIAVE 40 WESNGQPENN YKTTPPVLDS DGSFFLYSKL TVDKSRWQQG 80 NVFSCSVMHE ALHNHYTQKS LSLSPG 106 88 IgGl-IgG2IgGlf ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKKVER KCCVECPPCP APELLGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVKFNWYVD 160 GVEVHNAKTK PREQYNSTY RVVSVLTVLH QDWLNGKEYK 200 CKVSNKALPA PIEKTISKAK GQPREPQVYT LPPSREEMTK 240 NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPVLDS 280 DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPG 326 89 IgGl-IgG2CS- ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40

- 113 041306

IgGlf WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKKVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPG 325 90 IgGl-IgG2- IgGl If ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKKVER KCCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPSS IEKTISKAKG QPREPQVYTL PPSRREMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPG 325 91 IgGl-IgG2CSIgGl If ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKKVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPSS IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPG 325 92 IgGl-IgG2lgGlf2 ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKKVER KCCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPG 325 93 IgGlIgG2(C219S)IgGlf2 ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKKVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPG 325 94 IgG2-IgGlf2 ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280

- 114 041306

GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPG 325 95 IgG2(C219S)- IgGlf2 ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPG 325 96 Human IgG2 wt hinge ERKCCVECPP CPAPPVAG 18 97 Human IgG2cC219S hinge ERKSCVECPP CPAPPVAG 18 98 Hinge IgG2/IgGl ERKCCVECPP CPAPELLGG 19 99 IgG2 (C219S)/IgGl hinge ERKSCVECPP CPAPELLGG 19 100 Wild type human IgGl hinge EPKSCDKTHT CPPCPAPELL GG 22 101 IgG hinge 1.1 (L234A/L235E/ G237A) EPKSCDKTHT CPPCPAPEAE GA 22 102 CH2 human IgG2 wild type PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVQFNW 40 YVDGVEVHNA KTKPREEQFN STFRVVSVLT VVHQDWLNGK 80 EYKCKVSNKG LPAPIEKTIS KTK 103 103 CNS human IgG2 wild type GQPREPQVYT LPPSREEMTK NQVSLTCLVK GFYPSDIAVE 40 WESNGQPENN YKTTPPMLDS DGSFFLYSKL TVDKSRWQQG 80 NVFSCSVMHE ALHNHYTQKS LSLSPGK 107 104 IgGlfcCterminal K ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 105 IgG2.3 ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240

- 115 041306

QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 106 IgG2.3Gl-AY ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APELLGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVKFNWYVD 160 GVEVHNAKTK PREQYNSTY RVVSVLTVLH QDWLNGKEYK 200 CKVSNKALPA PIEKTISKAK GQPREPQVYT LPPSREEMTK 240 NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPVLDS 280 DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPGK 327 107 IgG2.3Gl-KH ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 108 IgG2.5 ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 109 IgGl. If ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPEAEGA 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPSSIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 BY IgG2.3Gl.lfKH ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPSS IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 111 IgGl-deltaTHT ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40

- 116 041306

WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKCPPCP APELLGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVKFNWYVD 160 GVEVHNAKTK PREQYNSTY RVVSVLTVLH QDWLNGKEYK 200 CKVSNKALPA PIEKTISKAK GQPREPQVYT LPPSREEMTK 240 NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPVLDS 280 DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPGK 327 112 IgG2.3plusTHT ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVETHTCP PCPAPPVAGP 120 SVFLFPPKPK DTLMISRTPE VTCVVVDVSH EDPEVQFNWY 160 VDGVEVHNAK TKPREEQFNS TFRVVSVLTV VHQDWLNGKE 200 YKCKVSNKGL PAPIEKTISK TKGQPREPQV YTLPPSREEM 240 TKNQVSLTCL VKGFYPSDIA VEWESNGQPE NNYKTTPPML 280 DSDGSFFLYS KLTVDKSRWQ QGNVFSCSVM HEALHNHYTQ 320 KSLSLSPGK 329 113 IgG2.3plusGGG ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVEGGGCP PCPAPPVAGP 120 SVFLFPPKPK DTLMISRTPE VTCVVVDVSH EDPEVQFNWY 160 VDGVEVHNAK TKPREEQFNS TFRVVSVLTV VHQDWLNGKE 200 YKCKVSNKGL PAPIEKTISK TKGQPREPQV YTLPPSREEM 240 TKNQVSLTCL VKGFYPSDIA VEWESNGQPE NNYKTTPPML 280 DSDGSFFLYS KLTVDKSRWQ QGNVFSCSVM HEALHNHYTQ 320 KSLSLSPGK 329 114 IgG2.5Gl.lf- KH ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPSS IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 115 IgG2.5Gl-AY ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCCVECPPCP APELLGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVKFNWYVD 160 GVEVHNAKTK PREQYNSTY RVVSVLTVLH QDWLNGKEYK 200 CKVSNKALPA PIEKTISKAK GQPREPQVYT LPPSREEMTK 240 NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPVLDS 280 DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPGK 327 116 IgG2.5Gl-KH ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240

- 117 041306

QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 117 IgG2.5plusTHT ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSWT VPSSNFGTQT 8 0 YTCNVDHKPS NTKVDKTVER KCCVETHTCP PCPAPPVAGP 120 SVFLFPPKPK DTLMISRTPE VTCVVVDVSH EDPEVQFNWY 160 VDGVEVHNAK TKPREEQFNS TFRVVSVLTV VHQDWLNGKE 200 YKCKVSNKGL PAPIEKTISK TKGQPREPQV YTLPPSREEM 240 TKNQVSLTCL VKGFYPSDIA VEWESNGQPE NNYKTTPPML 280 DSDGSFFLYS KLTVDKSRWQ QGNVFSCSVM HEALHNHYTQ 320 KSLSLSPGK 329 118 IgGl-G2.3GlAY ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSWT VPSSSLGTQT 8 0 YICNVNHKPS NTKVDKRVER KSCVECPPCP APELLGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVKFNWYVD 160 GVEVHNAKTK PREQYNSTY RVVSVLTVLH QDWLNGKEYK 200 CKVSNKALPA PIEKTISKAK GQPREPQVYT LPPSREEMTK 240 NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPVLDS 280 DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPGK 327 119 IgGl-G2.3Gl- KH ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSWT VPSSSLGTQT 8 0 YICNVNHKPS NTKVDKRVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 120 G2-G1-G1-G1 ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSWT VPSSNFGTQT 8 0 YTCNVDHKPS NTKVDKTVER KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 121 G2.5-G1-G1-G1 ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSWT VPSSNFGTQT 8 0 YTCNVDHKPS NTKVDKTVER KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 122 G1-G2.3-G2-G2 ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40

- 118 041306

WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 123 GlKRGEGSSNLF ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 124 Gl-KRGEGS ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 125 Gl-SNLF ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 126 IgGlITNDRTPR ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YTCNVDHKPS NTKVDKTVER KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 127 Gl-SNLFPR ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YICNVNHKPS NTKVDKRVER KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240

- 119 041306

MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 128 G2RKEGSGNSFL ASTKGPSVFP LAPCSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 129 G2-RKEGSG ASTKGPSVFP LAPCSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 130 G2-NSFL ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 131 IgG2TIDNTRRP ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YICNVNHKPS NTKVDKRVEP KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 132 G2-NSFLRP ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YTCNVDHKPS NTKVDKTVEP KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 133 G1-G1-G2-G1- ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40

- 120 041306

AY WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVQFNW 160 YVDGVEVHNA KTKPREEQFN STFRVVSVLT VVHQDWLNGK 200 EYKCKVSNKG LPAPIEKTIS KTKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 134 G1-G1-G2-G1KH ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPPVAGP 120 SVFLFPPKPK DTLMISRTPE VTCVVVDVSH EDPEVQFNWY 160 VDGVEVHNAK TKPREEQFNS TFRVVSVLTV VHQDWLNGKE 200 YKCKVSNKGL PAPIEKTISK TKGQPREPQV YTLPPSREEM 240 TKNQVSLTCL VKGFYPSDIA VEWESNGQPE NNYKTTPPVL 280 DSDGSFFLYS KLTVDKSRWQ QGNVFSCSVM HEALHNHYTQ 320 KSLSLSPGK 329 135 G2-G2 3-G1- G2-KH ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 136 G2 5-G2 3-G1G2-KH ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 137 G2-G2 3-G1- G2-AY ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APELLGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVKFNWYVD 160 GVEVHNAKTK PREQYNSTY RVVSVLTVLH QDWLNGKEYK 200 CKVSNKALPA PIEKTISKAK GQPREPQVYT LPPSREEMTK 240 NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPMLDS 280 DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPG 326 138 G2 5-G2 3-G1G2-AY ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APELLGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVKFNWYVD 160 GVEVHNAKTK PREQYNSTY RVVSVLTVLH QDWLNGKEYK 200 CKVSNKALPA PIEKTISKAK GQPREPQVYT LPPSREEMTK 240

- 121 041306

NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPMLDS 280 DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPGK 327 139 G1-G2.3-G1- Gl-KH ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVKFNWYVDG 160 VEVHNAKTKP REEQYNSTYR VVSVLTVLHQ DWLNGKEYKC 200 KVSNKALPAP IEKTISKAKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPVLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 140 G2-G1-G2-G2AY ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVQFNW 160 YVDGVEVHNA KTKPREEQFN STFRVVSVLT VVHQDWLNGK 200 EYKCKVSNKG LPAPIEKTIS KTKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPM 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 141 G2.5-G1-G2G2-AY ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVQFNW 160 YVDGVEVHNA KTKPREEQFN STFRVVSVLT VVHQDWLNGK 200 EYKCKVSNKG LPAPIEKTIS KTKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPM 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 142 G1-G2-G1-G1AY ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCVECPPCP APELLGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVKFNWYVD 160 GVEVHNAKTK PREQYNSTY RVVSVLTVLH QDWLNGKEYK 200 CKVSNKALPA PIEKTISKAK GQPREPQVYT LPPSREEMTK 240 NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPVLDS 280 DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPGK 327 143 G2-G1-G2-G2KH ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCDKTHTCP PCPAPPVAGP 120 SVFLFPPKPK DTLMISRTPE VTCVVVDVSH EDPEVQFNWY 160 VDGVEVHNAK TKPREEQFNS TFRVVSVLTV VHQDWLNGKE 200 YKCKVSNKGL PAPIEKTISK TKGQPREPQV YTLPPSREEM 240 TKNQVSLTCL VKGFYPSDIA VEWESNGQPE NNYKTTPPML 280 DSDGSFFLYS KLTVDKSRWQ QGNVFSCSVM HEALHNHYTQ 320 KSLSLSPG 328 144 G2.5-G1-G2G2-KH ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80

- 122 041306

YTCNVDHKPS NTKVDKTVER KSCDKTHTCP PCPAPPVAGP 120 SVFLFPPKPK DTLMISRTPE VTCVVVDVSH EDPEVQFNWY 160 VDGVEVHNAK TKPREEQFNS TFRVVSVLTV VHQDWLNGKE 200 YKCKVSNKGL PAPIEKTISK TKGQPREPQV YTLPPSREEM 240 TKNQVSLTCL VKGFYPSDIA VEWESNGQPE NNYKTTPPML 280 DSDGSFFLYS KLTVDKSRWQ QGNVFSCSVM HEALHNHYTQ 320 KSLSLSPGK 329 145 IgGldeltaHinge ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KCPPCPAPEL LGGPSVFLFP 120 PKPKDTLMIS RTPEVTCVVV DVSHEDPEVK FNWYVDGVEV 160 HNAKTKPREE QYNSTYRVVS VLTVLHQDWL NGKEYKCKVS 200 NKALPAPIEK TISKAKGQPR EPQVYTLPPS REEMTKNQVS 240 LTCLVKGFYP SDIAVEWESN GQPENNYKTT PPVLDSDGSF 280 FLYSKLTVDK SRWQQGNVFS CSVMHEALHN HYTQKSLSLS 320 PGK 323 146 IgG2deltaHinge ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCPPCPAPPV AGPSVFLFPP 120 KPKDTLMISR TPEVTCVVVD VSHEDPEVQF NWYVDGVEVH 160 NAKTKPREEQ FNSTFRVVSV LTVVHQDWLN GKEYKCKVSN 200 KGLPAPIEKT ISKTKGQPRE PQVYTLPPSR EEMTKNQVSL 240 TCLVKGFYPS DIAVEWESNG QPENNYKTTP PMLDSDGSFF 280 LYSKLTVDKS RWQQGNVFSC SVMHEALHNH YTQKSLSLSP 320 GK 322 147 IgG2.5deltaHinge ASTKGPSVFP LAPSSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCPPCPAPPV AGPSVFLFPP 120 KPKDTLMISR TPEVTCVVVD VSHEDPEVQF NWYVDGVEVH 160 NAKTKPREEQ FNSTFRVVSV LTVVHQDWLN GKEYKCKVSN 200 KGLPAPIEKT ISKTKGQPRE PQVYTLPPSR EEMTKNQVSL 240 TCLVKGFYPS DIAVEWESNG QPENNYKTTP PMLDSDGSFF 280 LYSKLTVDKS RWQQGNVFSC SVMHEALHNH YTQKSLSLSP 320 GK 322 148 IgGl-deltaG237 ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGP 120 SVFLFPPKPK DTLMISRTPE VTCVVVDVSH EDPEVKFNWY 160 VDGVEVHNAK TKPREEQYNS TYRVVSVLTV LHQDWLNGKE 200 YKCKVSNKAL PAPIEKTISK AKGQPREPQV YTLPPSREEM 240 TKNQVSLTCL VKGFYPSDIA VEWESNGQPE NNYKTTPPVL 280 DSDGSFFLYS KLTVDKSRWQ QGNVFSCSVM HEALHNHYTQ 320 KSLSLSPG 328 149 IgG2-plusG237 ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSCVECPPCP APPVAGGPSV 120 FLFPPKPKDT LMISRTPEVT CVVVDVSHED PEVQFNWYVD 160 GVEVHNAKTK PREEQFNSTF RVVSVLTVVH QDWLNGKEYK 200 CKVSNKGLPA PIEKTISKTK GQPREPQVYT LPPSREEMTK 240 NQVSLTCLVK GFYPSDIAVE WESNGQPENN YKTTPPMLDS 280

- 123 041306

DGSFFLYSKL TVDKSRWQQG NVFSCSVMHE ALHNHYTQKS 320 LSLSPGK 327 150 IgG2.4 ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCSVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 151 IgG2.3/4 ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KSSVECPPCP APPVAGPSVF 120 LFPPKPKDTL MISRTPEVTC VVVDVSHEDP EVQFNWYVDG 160 VEVHNAKTKP REEQFNSTFR VVSVLTVVHQ DWLNGKEYKC 200 KVSNKGLPAP IEKTISKTKG QPREPQVYTL PPSREEMTKN 240 QVSLTCLVKG FYPSDIAVEW ESNGQPENNY KTTPPMLDSD 280 GSFFLYSKLT VDKSRWQQGN VFSCSVMHEA LHNHYTQKSL 320 SLSPGK 326 152 C220S hinge IgG2 ERKCSVECPP CPAPPVAG 18 153 C220S hybrid hinge IgG2/IgGl ERKCSVECPP CPAPELLGG 19 154 IgG2 wild-type hinge ERKCCVECPP CPAP 14 155 C219S Hinge IgG2 ERKSCVECPP CPAP 14 156 C220S Hinge IgG2 ERKCSVECPP CPAP 14 157 C219X hinge part IgG2 ERKXCVECPP CPAP 14 158 C220X hinge part IgG2 ERKCXVECPP CPAP 14 159 IgG2 CH1+ hinge IgG2 (wild type) ASTKGPSVFP LAPCSRSTSE STAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSNFGTQT 80 YTCNVDHKPS NTKVDKTVER KCCVECPPCP APPVAG 116 160 IgG2cC219X ERKXCVECPP CPAPPVAG 18 161 IgG2 with C220X ERKCXVECPP CPAPPVAG 18 162 Hybrid IgG2/IgGl with C219X ERKXCVECPP CPAPELLGG 19

- 124 041306

163 Hybrid IgG2/IgGl with S220X ERKCXVECPP CPAPELLGG 19 164 Hybrid IgG2/IgGl deltaG ERKCCVECPP CPAPELLG 18 165 Hybrid IgG2/IgGl with C219S deltaG ERKSCVECPP CPAPELLG 18 166 Hybrid IgG2/IgGl with C220S deltaG ERKCSVECPP CPAPELLG 18 167 Hybrid IgG2/IgGl with C219X deltaG ERKXCVECPP CPAPELLG 18 168 Hybrid IgG2/IgGl with C220X deltaG ERKCXVECPP CPAPELLG 18 169 Hinge IgG2 PVAG 4 170 Hinge IgGl SCDKTHT 7 171 Hinge 1 IgGl ELLG 4 172 Hinge 2 IgGl ELLGG 5 173 Extracellular domain of mature huICOS protein (21 - 134 from NP 036224.1) EINGSANYEM FIFHNGGVQI LCKYPDIVQQ FKMQLLKGGQ 40 ILCDLTKTKG SGNTVSIKSL KFCHSQLSNN SVSFFLYNLD 80 HSHANYYFCN LSIFDPPPFK VTLTGGYLHI YESQ 114 174 Nucleotide sequence of the 3E8 heavy chain variable domain ATG GAG TTT GGG CTG ACC TGG GTT TTC CTC GTT GCT 36 CTT TTA AGA GGT GTC CAG TGT CAG GTG CAG CTG GTG 72 GAG TCT GGG GGA GGC GTG GTC CAG CCT GGG ATG TCC 108 CTG AGA CTC TCC TGT GCA GCG TCT GGA TTC ACC TTC 144 AGT ACC TAT GGC ATG CAG TGG GTC CGC CAG GCT CCA 180 GGC AAG GGG CTG GAG TGG GTG ACA GTT ATA TGG CAT 216 GAT GGA AGT CAT AAA GAC TAT GCA GAC TCC GTG AAG 252 GGC CGA TTC ACC ATC TCC AGA GAC AAT TCC AAG AAC 288 ACG ATG TAT CTG CAA ATG AAC AGC CTG AGA GCC GAG 324 GAC ACG GCT GTG TAT TAG TGT GCG AGA GAT CGG CAA 360 ACT GGG GAG GGC TAG TTT GAC TTC TGG GGC CAG GGA 396 ACC CTG GTC ACC GTC TCC TCA 417 175 Nucleotide sequence of the 3E8 light chain variable domain ATG AGG GTC CTC GCT CAG CTC CTG GGG CTC CTG CTG 36 CTC TGT TTC CCA GGT GCC AGA TGT GAC ATC CAG ATG 72 ACC CAG TCT CCA TCC TCA CTG TCT GCA TCT GTA GGA 108 GAC AGA GTC ACC ATC ACT TGT CGG GCG AGT CAG GGT 144 ATT AGC AGC TGG TTA GCC TGG TAT CAG CAG AAA CCA 180 GAG AAA GCC CCT AAG TCC CTG ATC TAT GCT GCA TCC 216 AGT TTG CAA AGT GGG GTC CCA TCA AGG TTC AGC GGC 252 AGT GGA TCT GGG ACA GAT TTC ACT CTC ACC ATC AGC 288 AGC CTG CAG CCT GAA GAT TTT GCA ACT TAT TAG TGC 324 CAA CAG TAT AAT AGT TAG CCG TAG ACT TTT GGC CAG 360 GGG ACC AAG CTG GAG ATC AAA 381

- 125 041306

176 Nucleotide sequence of the 17C4 heavy chain variable domain ATG GAG ATA CTT TGT TCC ACG CTC CTG STA CTG ACT 36 GTC CCG TCC TGG GTC TTA TCC CAG GTC ACC TTG AGG 72 GAG TCT GGT CCT GCG CTG GTG AAA CCC ACA CAG ACC 108 STS ACA CTG ACC TGC ACC TTC TCT GGG TTC TCA CTC 144 AGC ACT AGT GGA ATG TGT GTG AGC TGG ATS CGT CAG 180 CCC CCA GGG AAG GCC CTG GAG TGG CTT GCA CTC ATT 216 GAT TGG GAT GAT GAT AAA TTC TAG AGC ACA TCT CTG 252 AAG ACC AGG CTC ACC ATC TCC AAG GAC ACC TCC AAA 288 AAC CAG GTG GTC CTT ACA ATG ACC AAC ATG GAC CCT 324 GTG GAC ACA GCC ACG TAT TAG TGT GCA CGG ATG TCA 360 ACA CCT ACC TAG TAG GGT TTG GAC GTC TGG GGC CAA 396 GGG ACC ACG GTC ACC GTC TCC TCA 420 177 Nucleotide sequence of the 17C4 light chain variable domain ATG AGG GTC CTC GCT CAG CTC CTG GGG CTC CTG CTG 36 CTC TGT TTC CCA GGT GCC AGA TGT GAC ATC CAG ATG 72 ACC CAG TCT CCA TCC TCA CTG TCT GCA TCT GTA GGA 108 GAC AGA GTC ACC ATC ACT TGT CGG GCG AGT CAG GGT 144 ATT AGC AGC TGG TTA GCC TGG TAT CAG CAG AAA CCA 180 GAG AAA GCC CCT AAG TCC CTG ATC TAT GCT GCA TCC 216 AGT TTG CAA AGT GGG GTC CCA TCA AGG TTC AGC GGC 252 AGT GGA TCT GGG ACA GAT TTC ACT CTC ACC ATC AGC 288 AGC CTG CAG CCT GAA GAT TTT GCA ACT TAT TAG TGC 324 CAA CAG TAT AAT AGT TAG CCT CTC ACT TTC GGC GGA 360 GGG ACC AAG GTG GAG ATC AAA 381 178 Nucleotide sequence of heavy chain variable domain 1D7 ATG GAC ACA CTT TGC TCC ACG CTC CTG CTG CTG ACC 36 ATC CCT TCA TGG GTC TTG TCC CAG ATC ACC TTG AAG 72 GAG TCT GGT CCT ACG CTG GTG AAA CCC ACA CAG ACC 108 CTC ACG CTG ACC TGC ACC TTC TCT GGG TTC TCA CTC 144 GGC TCT AAT GGA CTG GGT GTG GGC TGG ATC CGT CAG 180 CCC CCA GGA AAG GCC CTG GAG TGG CTT GCA CTC ATT 216 TAT TGG GAT GAT GAT AAG CGC TAG AGT CCA TCT CTG 252 AAG AGC AGG CTC ACC ATC ACC AAG GAC TCC TCC AAA 288 AAC CAG GTG GTC CTT ACA ATG ACC AAC ATG GAC CCT 324 GTG GAC АСА GCC ACG TAT TAG TGT GCA CAC AGG AAC 360 AGT GGC TTT GAC TAG TGG GGC CAG GGA ATS CTG GTC 396 ACC GTC TCC TCA 408 179 Nucleotide sequence of light chain variable domain-a 1D7 ATG AGG GTC CTC GCT CAG CTC CTG GGG CTC CTG CTG 36 CTC TGT TTC CCA GGT GCC AGA TGT GAC ATC CAG ATG 72 ACC CAG TCT CCA TCC TCA CTG TCT GCA TCT GTA GGA 108 GAC AGA GTC ACC ATC ACT TGT CGG GCG AGT CAG GGT 144 TTT AGC AGC TGG TTA GCC TGG TAT CAG CAG AAA CCA 180 GAG AAA GCC CCT AAG TCC CTG ATC TAT GCT GCA TCC 216 AGT TTG CAA AGT GGG GTC CCA TCA AGG TTC AGC GGC 252 AGT GGA TCT GGG ACA GAT TTC ACT CTC ACC ATC AGC 288 AGC CTG CAG CCT GAA GAT TTT GCA ACT TAT TAG TGC 324 CAA CAG TAT AAT AGT TAG CCT TAG ACT TTT GGC CAG 360 GGG ACC AAG CTG GAG ATC AAA 381 180 The nucleotide sequence of the variable ATG AGG GTC CTC GCT CAG CTC CTG GGG CTC CTG CTG 36 CTC TGT TTC CCA GGT GCC AGA TGT GAC ATC CAG ATG 72 ACC CAG TCT CCA TCC TCA CTG TCT GCA TCT GTA GGA 108 GAC AGA GTC ACC ATC ACT TGT CGG GCG AGT CAG GGT 144

- 126 041306

light chain-b domain 1D7 ATT AGC AGO TGG TTA GCC TGG TAT GAG GAG AAA CCA 180 GAG AAA GCC CCT AAG TCC CTG ATC TAT GCT GCA TCC 216 AGT TTG CAA AGT GGG GTC CCA TCA AGG TTC AGC GGC 252 AGT GGA TCT GGG ACA GAT TTC ACT CTC ACC ATC AGC 288 AGC CTG CAG CCT GAA GAT TTT GCA ACT TAT TAC TGC 324 CAA CAG TAT AAT AGT TAC CCT CTC ACT TTC GGC GGA 360 GGG ACC AAG GTG GAG ATC AAA 381 181 Nucleotide sequence of the 9D5 heavy chain variable domain ATG GAC ACA CTT TGC TCC ACG CTC CTG CTG CTG ACC 36 ATC CCT TCA TGG GTC TTG TCC CAG ATC ACC TTG AAG 72 GAG TCT GGT CCT ACG CTG GTG AAA CCC ACA CAG ACC 108 CTC ACG CTG ACC TGC ACC TTC TCT GGG TTC TCA CTC 144 GGC ACT AGT GGA CTG GGT GTG GGC TGG ATC CGT CAG 180 CCC CCA GGA AAG GCC CTG GAG TGG CTT GCA TTC ATT 216 TAT TGG GAT GAT GAT AAG CGC TAG AGC CCA TCT CTG 252 AAG AGC AGG CTC ACC ATC ACC AAG GAC ACC TCC AAA 288 AAC CAG GTG GTC CTT ACA ATG ACC AAC ATG GAC CCT 324 GTG GAC ACA GCC ACA TAT TAG TGT GCA CAC AGA CGG 360 GGC TTT TTT GAC TAC TGG GGC CAG GGA ACC CTG GTC 396 ACC GTC TCC TCA 408 182 Nucleotide sequence of the 9D5 light chain variable domain ATG AGG GTC CTC GCT CAG CTC CTG GGG CTC CTG CTG 36 CTC TGT TTC CCA GGT GCC AGA TGT GAC ATC CAG ATG 72 ACC CAG TCT CCA TCC TCA CTG TCT GCA TCT GTA GGA 108 GAC AGA GTC ACC ATC ACT TGT CGG GCG AGT CAG GGT 144 ATT AGC AGC TGG TTA GCC TGG TAT CAG CAG AAA CCA 180 GAG AAA GCC CCT AAG TCC CTG ATC TAT GCT GCA TCC 216 AGT TTG CAA AGT GGG GTC CCA TCA AGG TTC AGC GGC 252 AGT GGA TCT GGG ACA GAT TTC ACT CTC ACC ATC AGC 288 AGC CTG CAG CCT GAA GAT TTT GCA ACT TAT TAC TGC 324 CAA CAG TAT AAT AGT TAC CCG CTC ACT TTC GGC GGA 360 GGG ACC AAG GTG GAG ATC AAA 381 183 Nucleotide sequence of ICOS.33 kappa GAC ATC CAG ATG ACC CAG TCT CCA TCC TCC CTG TCT 36 GCA TCT GTA GGA GAC AGA GTC ACC ATC ACT TGC CAG 72 GCC AGT CAG GAC ATT AGC AAT TAT TTA AGC TGG TAT 108 CAG CAG AAA CCA GGG AAA GCC CCT AAG CTC CTG АТС 144 TAC TAT ACA AAT STA TTG GCA GAA GGG GTC CCA TCA 180 AGG TTC AGT GGA AGT GGA TCT GGG ACA GAT TTT ACT 216 TTC ACC ATC AGC AGC CTG CAG CCT GAA GAT ATT GCA 252 ACA TAT TAC TGT CAA CAG TAT TAT AAC TAT CGG ACG 288 TTC GGC CCT GGG ACC AAA GTG GAT ATC AAA CGT ACG 324 GTG GCT GCA CCA TCT GTC TTC ATC TTC CCG CCA TCT 360 GAT GAG CAG TTG AAA TCT GGA ACT GCC TCT GTT GTG 396 TGC CTG CTG AAT AAC TTC TAT CCC AGA GAG GCC AAA 432 GTA CAG TGG AAG GTG GAT AAC GCC CTC CAA TCG GGT 468 AAC TCC CAG GAG AGT GTC ACA GAG CAG GAC AGC AAG 504 GAC AGC ACC TAC AGC CTC AGC AGC ACC CTG ACG CTG 540 AGC AAA GCA GAC TAC GAG AAA CAC AAA GTC TAC GCC 576 TGC GAA GTC ACC CAT CAG GGC CTG AGC TCG CCC GTC 612 ACA AAG AGC TTC AAC AGG GGA GAG TGT TAG 642 184 Nucleotide sequence of ICOS.33- GAG GTG CAG CTG GTG GAG TCT GGG GGA GGC TTG GTA 36 AAG CCT GGG GGG TCC CTT AGA CTC TCC TGT GCA GCC 72 TCT GGA TTC ACT TTC AGT GAC TAT TTC ATG CAC TGG 108

- 127 041306

glf-S267E GTC CGC CAG GCT CCA GGG AAG GGG CTG GAG TGG GTT 144 GGC GTC ATA GAC ACT AAA AGT TTT AAT TAT GCA ACC 180 TAT TAC TCT GAT TTG GTG AAA GGC AGA TTC ACC ATC 216 TCA AGA GAT GAT TCA AAA AAC ACG CTG TAT CTG CAA 252 ATG AAC AGC CTG AAA ACC GAG GAC ACA GCC GTG TAT 288 TAG TGT ACC GCA ACC ATC GCT GTC CCA TAT TAG TTC 324 GAT TAG TGG GGC CAG GGA ACC CTG GTC ACC GTC TCC 360 TCA GCT AGC ACC AAG GGC CCA TCG GTC TTC CCC CTG 396 GCA CCC TCC TCC AAG AGC ACC TCT GGG GGC ACA GCG 432 GCC CTG GGC TGC CTG GTC AAG GAC TAG TTC CCC GAA 468 CCG GTG ACG GTG TCG TGG AAC TCA GGC GCC CTG ACC 504 AGC GGC GTG CAC ACC TTC CCG GCT GTC STA CAG TCC 540 TCA GGA CTC TAC TCC CTC AGC AGC GTG GTG ACC GTG 576 CCC TCC AGC AGC TTG GGC ACC CAG ACC TAC ATC TGC 612 AAC GTG AAT CAC AAG CCC AGC AAC ACC AAG GTG GAC 648 AAG AGA GTT GAG CCC AAA TCT TGT GAC AAA ACT CAC 684 ACA TGC CCA CCG TGC CCA GCA CCT GAA CTC CTG GGG 720 GGA CCG TCA GTC TTC CTC TTC CCC CCA AAA CCC AAG 756 GAC ACC CTC ATG ATC TCC CGG ACC CCT GAG GTC ACA 792 TGC GTG GTG GTG GAC GTG GAG CAC GAA GAC CCT GAG 828 GTC AAG TTC AAC TGG TAG GTG GAC GGC GTG GAG GTG 864 CAT AAT GCC AAG АСА AAG CCG CGG GAG GAG CAG TAG 900 AAC AGC ACG TAG CGT GTG GTC AGC GTC CTC ACC GTC 936 CTG CAC CAG GAC TGG CTG AAT GGC AAG GAG TAG AAG 972 TGC AAG GTC TCC AAC AAA GCC CTC CCA GCC CCC ATC 1008 GAG AAA ACC ATC TCC AAA GCC AAA GGG CAG CCC CGA 1044 GAA CCA CAG GTG TAG ACC CTG CCC CCA TCC CGG GAG 1080 GAG ATG ACC AAG AAC CAG GTC AGC CTG ACC TGC CTG 1116 GTC AAA GGC TTC TAT CCC AGC GAC ATC GCC GTG GAG 1152 TGG GAG AGC AAT GGG CAG CCG GAG AAC AAC TAG AAG 1188 ACC ACG CCT CCC GTG CTG GAC TCC GAC GGC TCC TTC 1224 TTC CTC TAT AGC AAG CTC ACC GTG GAC AAG AGC AGG 1260 TGG CAG CAG GGG AAC GTC TTC TCA TGC TCC GTG ATG 1296 CAT GAG GCT CTG CAC AAC CAC TAG ACG CAG AAG AGC 1332 CTC TCC CTG TCC CCG GGT TGA 1353 185 Amino acid sequence of the heavy chain of IgG-2644 EVQLLESGGG LVQPGGSLRL SCEASGFIFK YYAMSWVRQA 40 PGKGLEWVSG ISGSGGSTYY ADSVKGRFTI SRDNSKHTLY 80 LQMNSLRAED TAVYYCAKDG DFDWIHYYYG MDVWGQGTTV 120 TVSSASTKGP SVFPLAPSSK STSGGTAALG CLVKDYFPEP 160 VTVSWNSGAL TSGVHTFPAV LQSSGLYSLS SVVTVPSSSL 200 GTQTYICNVN HKPSNTKVDK RVEPKSCDKT HTCPPCPAPE 240 LLGGPSVFLF PPKPKDTLMI SRTPEVTCVV VDVSHEDPEV 280 KFNWYVDGVE VHNAKTKPRE EQYNSTYRVV SVLTVLHQDW 320 LNGKEYKCKV SNKALPAPIE KTISKAKGQP REPQVYTLPP 360 SREEMTKNQV SLTCLVKGFY PSDIAVEWES NGQPENNYKT 400 TPPVLDSDGS FFLYSKLTVD KSRWQQGNVF SCSVMHEALH 440 NHYTQKSLSL SPG 453 186 Amino acid sequence of the variable region EVQLLESGGG LVQPGGSLRL SCEASGFIFK YYAMSWVRQA 40 PGKGLEWVSG ISGSGGSTYY ADSVKGRFTI SRDNSKHTLY 80 LQMNSLRAED TAVYYCAKDG DFDWIHYYYG MDVWGQGTTV 120 TVSS 124

- 128 041306

heavy chain IgG-2644 187 Amino acid sequence of the IgG-2644 heavy chain constant region ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPG 329 188 Light chain amino acid sequence of IgG-2644 AIQLTQSPSS LSASVGDRVT ITCRASQGIS SALAWYQQKP 40 GKAPKLLIYD ASSLESGVPS RFSGSGSGTD FTLTISSLQP 80 EDFATYYCQQ FNSYPHTFGG GTKVEIKRTV AAPSVFIFPP 120 SDEQLKSGTA SVVCLLNNFY PREAKVQWKV DNALQSGNSQ 160 ESVTEQDSKD STYSLSSTLT LSKADYEKHK VYACEVTHQG 200 LSSPVTKSFN RGEC 214 189 Amino acid sequence of the light chain variable region of IgG-2644 AIQLTQSPSS LSASVGDRVT ITCRASQGIS SALAWYQQKP 40 GKAPKLLIYD ASSLESGVPS RFSGSGSGTD FTLTISSLQP 80 EDFATYYCQQ FNSYPHTFGG GTKVEIK 107 190 Amino acid sequence of the light chain constant region of IgG-2644 RTVAAPSVFI FPPSDEQLKS GTASVVCLLN NFYPREAKVQ 40 WKVDNALQSG NSQESVTEQD SKDSTYSLSS TLTLSKADYE 80 KHKVYACEVT HQGLSSPVTK SFNRGEC 107 191 IgG-2644 CDRH1 Amino Acid Sequence YYAMS 5 192 Amino acid sequence of IgG-2644 CDRH2 GISGSGGSTY YADSVKG 17 193 Amino acid sequence of IgG-2644 CDRH3 DGDFDWIHYY YGMDV 15 194 Amino acid sequence of IgG-2644 CDRL1 RASQGISSAL A 11 195 Amino acid sequence of IgG-2644 DASSLES 7

- 129 041306

CDRL2 196 IgG-2644 CDRL3 Amino Acid Sequence QQFNSYPHT 9 197 IgG-2644 heavy chain nucleotide sequence GAG GTG GAG CTG TTG GAG TCT GGG GGA GGC TTG GTA 36 GAG CCT GGG GGG TGC CTG AGA CTC TCC TGT GAA GCC 72 TCT GGA TTC ATC TTT AAA TAG TAT GCC ATG AGC TGG 108 GTC CGC CAG GCT CCA GGG AAG GGG CTG GAG TGG GTC 144 TCA GGT ATT AGT GGT AGT GGT GGT AGC ACA TAS TAC 180 GCA GAC TCC GTG AAG GGC CGG TTC ACC ATC TCC AGA 216 GAC AAT TCC AAG CAC ACG CTG TAT CTG CAA ATG AAC 252 AGC CTG AGA GCC GAG GAC ACG GCC GTT TAT TAG TGT 288 GCG AAA GAT GGG GAT TTT GAC TGG ATC CAC TAT TAG 324 TAT GGT ATG GAC GTC TGG GGC CAA GGG ACC ACG GTC 360 ACC GTC TCC TCA GCG TCG ACC AAG GGC CCA TCC GTC 396 TTC CCC CTG GCA CCC TCC TCC AAG AGC ACC TCT GGG 432 GGC ACA GCG GCC CTG GGC TGC CTG GTC AAG GAC TAG 468 TTC CCC GAA CCG GTG ACG GTG TCG TGG AAC TCA GGC 504 GCC CTG ACC AGC GGC GTG CAC ACC TTC CCG GCT GTC 540 STA CAG TCC TCA GGA CTC TAC TCC CTC AGC AGC GTG 576 GTG ACC GTG CCC TCC AGC AGC TTG GGC ACC CAG ACC 612 TAS ATC TGC AAC GTG AAT CAC AAG CCC AGC AAC ACC 648 AAG GTG GAC AAG AGA GTT GAG CCC AAA TCT TGT GAC 684 AAA ACT CAC ACA TGC CCA CCG TGC CCA GCA CCT GAA 720 CTC CTG GGG GGA CCG TCA GTC TTC CTC TTC CCC CCA 756 AAA CCC AAG GAC ACC CTC ATG ATC TCC CGG ACC CCT 792 GAG GTC ACA TGC GTG GTG GTG GAC GTG AGC CAC GAA 828 GAC CCT GAG GTC AAG TTC AAC TGG TAG GTG GAC GGC 864 GTG GAG GTG CAT AAT GCC AAG ACA AAG CCG CGG GAG 900 GAG CAG TAG AAC AGC ACG TAG CGT GTG GTC AGC GTC 936 CTC ACC GTC CTG CAC CAG GAC TGG CTG AAT GGC AAG 972 GAG TAG AAG TGC AAG GTC TCC AAC AAA GCC CTC CCA 1008 GCC CCC ATC GAG AAA ACC ATC TCC AAA GCC AAA GGG 1044 CAG CCC CGA GAA CCA CAG GTG TAG ACC CTG CCC CCA 1080 TCC CGG GAG GAG ATG ACC AAG AAC CAG GTC AGC CTG 1116 ACC TGC CTG GTC AAA GGC TTC TAT CCC AGC GAC ATC 1152 GCC GTG GAG TGG GAG AGC AAT GGG CAG CCG GAG AAC 1188 AAC TAG AAG ACC ACG CCT CCC GTG CTG GAC TCC GAC 1224 GGC TCC TTC TTC CTC TAT AGC AAG CTC ACC GTG GAC 1260 AAG AGC AGG TGG CAG CAG GGG AAC GTC TTC TCA TGC 1296 TCC GTG ATG CAT GAG GCT CTG CAC AAC CAC TAG ACG 1332 CAG AAG AGC CTC TCC CTG TCC CCG GGT 1359 198 Nucleotide sequence of the IgG-2644 heavy chain variable domain GAG GTG CAG CTG TTG GAG TCT GGG GGA GGC TTG GTA 36 CAG CCT GGG GGG TCC CTG AGA CTC TCC TGT GAA GCC 72 TCT GGA TTC ATC TTT AAA TAG TAT GCC ATG AGC TGG 108 GTC CGC CAG GCT CCA GGG AAG GGG CTG GAG TGG GTC 144 TCA GGT ATT AGT GGT AGT GGT GGT AGC ACA TAS TAC 180 GCA GAC TCC GTG AAG GGC CGG TTC ACC ATC TCC AGA 216 GAC AAT TCC AAG CAC ACG CTG TAT CTG CAA ATG AAC 252 AGC CTG AGA GCC GAG GAC ACG GCC GTT TAT TAG TGT 288 GCG AAA GAT GGG GAT TTT GAC TGG ATC CAC TAT TAG 324 TAT GGT ATG GAC GTC TGG GGC CAA GGG ACC ACG GTC 360

- 130 041306

ACC GTC TCC TCA 372 199 Nucleotide sequence of IgG-2644 heavy chain constant region GCG TCG ACC AAG GGC CCA TCC GTC TTC CCC CTG GCA 36 CCC TCC TCC AAG AGC ACC TCT GGG GGC ACA GCG GCC 72 CTG GGC TGC CTG GTC AAG GAC TAC TTC CCC GAA CCG 108 GTG ACG GTG TCG TGG AAC TCA GGC GCC CTG ACC AGC 144 GGC GTG CAC ACC TTC CCG GCT GTC STA CAG TCC TCA 180 GGA STS TAS TSS STS AGC AGC GTG GTG ACC GTG ССС 216 ТСС AGC AGC TTG GGC ACC CAG АСС ТАС АТС TGC ААС 252 GTG AAT CAC AAG CCC AGC AAC ACC AAG GTG GAC AAG 288 AGA GTT GAG CCC AAA TCT TGT GAC AAA ACT CAC ACA 324 TGC CCA CCG TGC CCA GCA CCT GAA CTC CTG GGG GGA 360 CCG TCA GTC TTC CTC TTC CCC CCA AAA CCC AAG GAC 396 ACC CTC ATG ATC TCC CGG ACC CCT GAG GTC ACA TGC 432 GTG GTG GTG GAC GTG AGC CAC GAA GAC CCT GAG GTC 468 AAG TTC AAC TGG TAG GTG GAC GGC GTG GAG GTG CAT 504 AAT GCC AAG ACA AAG CCG CGG GAG GAG CAG TAG AAC 540 AGC ACG TAG CGT GTG GTC AGC GTC CTC ACC GTC CTG 576 CAC CAG GAC TGG CTG AAT GGC AAG GAG TAG AAG TGC 612 AAG GTC TCC AAC AAA GCC CTC CCA GCC CCC ATC GAG 648 AAA ACC ATC TCC AAA GCC AAA GGG CAG CCC CGA GAA 684 CCA CAG GTG TAG ACC CTG CCC CCA TCC CGG GAG GAG 720 ATG ACC AAG AAC CAG GTC AGC CTG ACC TGC CTG GTC 756 AAA GGC TTC TAT CCC AGC GAC ATC GCC GTG GAG TGG 792 GAG AGC AAT GGG CAG CCG GAG AAC AAC TAG AAG ACC 828 ACG CCT CCC GTG CTG GAC TCC GAC GGC TCC TTC TTC 864 CTC TAT AGC AAG CTC ACC GTG GAC AAG AGC AGG TGG 900 CAG CAG GGG AAC GTC TTC TCA TGC TCC GTG ATG CAT 936 GAG GCT CTG CAC AAC CAC TAG ACG CAG AAG AGC CTC 972 TCC CTG TCC CCG GGT 987 200 Nucleotide sequence of the light chain of IgG-2644 GCC ATC CAG TTG ACC CAG TCT CCA TCC TCC CTG TCT 36 GCA TCT GTA GGA GAC AGA GTC ACC ATC ACT TGC CGG 72 GCA AGT CAG GGC ATT AGC AGT GCT TTA GCC TGG TAT 108 CAG CAG AAA CCA GGG AAA GCT CCT AAG CTC CTG АТС 144 TAT GAT GCC TCC AGT TTG GAA AGT GGG GTC CCA TCA 180 AGG TTC AGC GGC AGT GGA TCT GGG ACA GAT TTC ACT 216 CTC ACC ATC AGC AGC CTG CAG CCT GAA GAT TTT GCA 252 ACT TAT TAG TGT CAA CAG TTT AAT AGT TAG CCT CAC 288 ACT TTC GGC GGA GGG ACC AAG GTG GAG ATC AAA CGT 324 ACG GTG GCT GCA CCA TCT GTC TTC ATC TTC CCG CCA 360 TCT GAT GAG CAG TTG AAA TCT GGA ACT GCC TCT GTT 396 GTG TGC CTG CTG AAT AAC TTC TAT CCC AGA GAG GCC 432 AAA GTA CAG TGG AAG GTG GAT AAC GCC CTC CAA TCG 468 GGT AAC TCC CAG GAG AGT GTC ACA GAG CAG GAC AGC 504 AAG GAC AGC ACC TAG AGC CTC AGC AGC ACC CTG ACG 540 CTG AGC AAA GCA GAC TAG GAG AAA CAC AAA GTC TAG 576 GCC TGC GAA GTC ACC CAT CAG GGC CTG AGC TCG CCC 612 GTC ACA AAG AGC TTC AAC AGG GGA GAG TGT 642 201 Nucleotide sequence of the variable domain of the light GCC ATC CAG TTG ACC CAG TCT CCA TCC TCC CTG TCT 36 GCA TCT GTA GGA GAC AGA GTC ACC ATC ACT TGC CGG 72 GCA AGT CAG GGC ATT AGC AGT GCT TTA GCC TGG TAT 108 CAG CAG AAA CCA GGG AAA GCT CCT AAG CTC CTG АТС 144 TAT GAT GCC TCC AGT TTG GAA AGT GGG GTC CCA TCA 180

- 131 041306

IgG-2644 chains AGG TTC AGC GGC AGT GGA TTC GGG ACA GAT TTC ACT 216 CTC ACC ATC AGC AGC CTG CAG CCT GAA GAT TTT GCA 252 ACT TAT TAC TGT CAA CAG TTT AAT AGT TAC CST CAC 288 ACT TTC GGC GGA GGG ACC AAG GTG GAG ATC AAA 321 202 Nucleotide sequence of IgG-2644 light chain constant domain CGT ACG GTG GCT GCA CCA TCT GTC TTC ATC TTC CCG 36 CCA TCT GAT GAG CAG TTG AAA TCT GGA ACT GCC TCT 72 GTT GTG TGC CTG CTG AAT AAC TTC TAT CCC AGA GAG 108 GCC AAA GTA CAG TGG AAG GTG GAT AAC GCC CTC CAA 144 TCG GGT AAC TCC CAG GAG AGT GTC ACA GAG CAG GAC 180 AGC AAG GAC AGC ACC TAG AGC CTC AGC AGC ACC CTG 216 ACG CTG AGC AAA GCA GAC TAG GAG AAA CAC AAA GTC 252 TAG GCC TGC GAA GTC ACC CAT CAG GGC CTG AGC TCG 288 CCC GTC ACA AAG AGC TTC AAC AGG GGA GAG TGT 321 203 Amino acid sequence of an epitope ICOS.4 SIFDPPPFKV TL 12 204 huIgGlf heavy chain constant domain with C-terminal lysine ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKRVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE 240 MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 205 Isoform 2 (Q9Y6W8-2) MKSGLWYFFL FCLRIKVLTG EINGSANYEM FIFHNGGVQI 40 LCKYPDIVQQ FKMQLLKGGQ ILCDLTKTKG SGNTVSIKSL 80 KFCHSQLSNN SVSFFLYNLD HSHANYYFCN LSIFDPPPFK 120 VTLTGGYLHI YESQLCCQLK FWLPIGCAAF VVVCILGCIL 160 ICWLTKKM 168 206 Human IgGl (P018571) ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS 40 WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT 80 YICNVNHKPS NTKVDKKVEP KSCDKTHTCP PCPAPELLGG 120 PSVFLFPPKP KDTLMISRTP EVTCVVVDVS HEDPEVKFNW 160 YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK 200 EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSRDE 240 LTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV 280 LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT 320 QKSLSLSPGK 330 207 VKIO18 DIQMTQSPSS LSASVGDRVT ITCQASQDIS NYLNWYQQKP 40 GKAPKLLIYD ASNLETGVPS RFSGSGSGTD FTFTISSLQP 80 EDIATYYCQQ YDNLP 95 208 JK3 FTFGPGTKVD IK 12 209 VH3-15 EVQLVESGGG LVKPGGSLRL SCAASGFTFS NAWMSWVRQA 40 PGKGLEWVGR IKSKTDGGTT DYAAPVKGRF TISRDDSKNT 80 LYLQMNSLKT EDTAVYYCTT 100

- 132 041306

210 JH4 YFDYWGQGTL VTVSS 15 211 heavy chain mICOS 1-mGl EVDLVETGGG LVQPGGSLKL SCVASGFTFS RYWMFWIRQA 40 PGKGLEWVSS VSTTDGRSTYY PDSVQGRFTI SRNDAENTVY 80 LQMNSLRSED TATYYCAKEG YYDGSYYAYY FDYWGQGVTV 120 TVSSAKTTPP SVYPLAPGSA AQTNSMVTLG CLVKGYFPEP 160 VTVTWNSGSL SSGVHTFPAV LQSDLYTLSS SVTVPSSTWP 200 SETVTCNVAH PASSTKVDKK IVPRDCGCKP CICTVPEVSS 240 VFIFPPKPKD VLTITLTPKV TCVVVDISKD DPEVQFSWFV 280 DDVEVHTAQT QPREEQFNST FRSVSELPIM HQDWLNGKEF 320 KCRVNSAAFP APIEKTISKT KGRPKAPQVY TIPPPKEQMA 360 KDKVSLTCMI TDFFPEDITV EWQWNGQPAE NYKNTQPIMD 400 TDGSYFVYSK LNVQKSNWEA GNTFTCSVLH EGLHNHHTEK 440 SLSHSPGK 448 212 Light chain mICOS 1-mGl DVQMAQSPSS LAASPGESVS INCKASKSIS KYLAWYQQKP 40 GKANKLLIYS GSTLQSGTPS RFSGSGSGTD FTLTIRNLEP 80 EDFGLYYCQQ HNAYPPTFGT GTKLELKRAD AAPTVSIFPP 120 SSEQLTSGGA SVVCFLNNFY PKDINVKWKI DGSERQNGVL 160 NSWTDQDSKD STYSMSSTLT LTKDEYERHN SYTCEATHKT 200 STSPIVKSFN RNEC 214 213 heavy chain ICOS 4-mGl EVQLVESGGG LVKPAGSLTL SCVASGFTFS DYFMHWVRQA 40 PGKGLEWVAV IDTKSFNYAT YYSDLVKGRF TVSRDDSQGM 80 VYLQMNNLRK EDTATYYCTA TIAVPYYFDY WGQGTMVTVS 120 SAKTTPPSVY PLAPGSAAQT NSMVTLGCLV KGYFPEPVTV 160 TWNSGSLSSG VHTFPAVLQS DLYTLSSSVT VPSSTWPSET 200 VTCNVAHPAS STKVDKKIVP RDCGCKPCIC TVPEVSSVFI 240 FPPKPKDVLT ITLTPKVTCV VVDISKDDPE VQFSWFVDDV 280 EVHTAQTQPR EEQFNSTFRS VSELPIMHQD WLNGKEFKCR 320 VNSAAFPAPI EKTISKTKGR PKAPQVYTIP PPKEQMAKDK 360 VSLTCMITDF FPEDITVEWQ WNGQPAENYK NTQPIMDTDG 400 SYFVYSKLNV QKSNWEAGNT FTCSVLHEGL HNHHTEKSLS 440 HSPGK 445 214 light chain ICOS 4-mGl DIQMTQSPSS LPASLGDRVT INCQASQDIS NYLSWYQQKP 40 GKAPKLLIYY TNLLADGVPS RFSGSGSGRD YSFTISSLES 80 EDIGSYYCQQ YYNYRTFGPG TKLEIKRADA APTVSIFPPS 120 SEQLTSGGAS VVCFLNNFYP KDINVKWKID GSERQNGVLN 160 SWTDQDSKDS TYSMSSTLTL TKDEYERHNS YTCEATHKTS 200 TSPIVKSFNR NEC 213 215 heavy chain ICOS 34-Glf EVQLVESGGG LVKPGGSLRL SCAASGFTFS DYFMHWVRQA 40 PGKGLEWVGV IDTKSFNYAT YYSDLVKGRF TISRDDSKNT 80 LYLQMNSLKT EDTAVYYCTT TIAVPYYFDY WGQGTLVTVS 120 SASTKGPSVF PLAPSSKSTS GGTAALGCLV KDYFPEPVTV 160 SWNSGALTSG VHTFPAVLQS SGLYSLSSVV TVPSSSLGTQ 200 TYICNVNHKP SNTKVDKRVE PKSCDKTHTC PPCPAPELLG 240 GPSVFLFPPK PKDTLMISRT PEVTCVVVDV SHEDPEVKFN 280 WYVDGVEVHN AKTKPREEQY NSTYRVVSVL TVLHQDWLNG 320 KEYKCKVSNK ALPAPIEKTI SKAKGQPREP QVYTLPPSRE 360 EMTKNQVSLT CLVKGFYPSD IAVEWESNGQ PENNYKTTPP 400 VLDSDGSFFL YSKLTVDKSR WQQGNVFSCS VMHEALHNHY 440 TQKSLSLSPG 450 216 Light chain ICOS 34-Glf DIQMTQSPSS LSASVGDRVT ITCQASQDIS NYLSWYQQKP 40 GKAPKLLIYY TNLLADGVPS RFSGSGSGTD FTFTISSLQP 80 EDIATYYCQQ YYNYRTFGPG TKVDIKRTVA APSVFIFPPS 120 DEQLKSGTAS VVCLLNNFYP REAKVQWKVD NALQSGNSQE 160 SVTEQDSKDS TYSLSSTLTL SKADYEKHKV YACEVTHQGL 200 SSPVTKSFNR GEC 213 217 heavy chain ICOS 35-Glf EVQLVESGGG LVKPGGSLRL SCAASGFTFS DYFMHWVRQA 40 PGKGLEWVGV IDTKSFNYAT YYSDLVKGRF TISRDDSKNT 80 LYLQMNSLKT EDTAVYYCTA TIAVPYYFDY WGQGTLVTVS 120 SASTKGPSVF PLAPSSKSTS GGTAALGCLV KDYFPEPVTV 160 SWNSGALTSG VHTFPAVLQS SGLYSLSSVV TVPSSSLGTQ 200 TYICNVNHKP SNTKVDKRVE PKSCDKTHTC PPCPAPELLG 240 GPSVFLFPPK PKDTLMISRT PEVTCVVVDV SHEDPEVKFN 280 WYVDGVEVHN AKTKPREEQY NSTYRVVSVL TVLHQDWLNG 320 KEYKCKVSNK ALPAPIEKTI SKAKGQPREP QVYTLPPSRE 360 EMTKNQVSLT CLVKGFYPSD IAVEWESNGQ PENNYKTTPP 400 VLDSDGSFFL YSKLTVDKSR WQQGNVFSCS VMHEALHNHY 440 TQKSLSLSPG 450 218 Light chain ICOS 35-Glf DIQMTQSPSS LSASVGDRVT ITCQASQDIS NYLSWYQQKP 40 GKAPKLLIYY TNLLADGVPS RFSGSGSGTD FTFTISSLQP 80 EDIATYYCQQ YYNYRTFGPG TKVDIKRTVA APSVFIFPPS 120 DEQLKSGTAS VVCLLNNFYP REAKVQWKVD NALQSGNSQE 160 SVTEQDSKDS TYSLSSTLTL SKADYEKHKV YACEVTHQGL 200 SSPVTKSFNR GEC 213 219 IL-2 pathway agonist NKTR-214 PTSSSTKKTQ LQLEHLLLDL QMILNGINNY KNPKLTRMLT 40 FKFYMPKKAT ELKHLQCLEE ELKPLEEVLN LAQSKNFHLR 80 PRDLISNINV IVLELKGSET TFMCEYADET ATIVEFLNRW 120 ITFSQSIIST LT 132

- 133 -

Claims (15)

CLAIM 1. An isolated monoclonal antibody that binds to a human inducible costimulatory (ICOS) molecule, wherein the antibody contains a heavy chain variable domain containing CDR1, CDR2, and CDR3 regions having the amino acid sequences shown in SEQ ID NOs: 9, 10, and 11, respectively. , and a light chain variable domain containing CDR1, CDR2 and CDR3 regions having the amino acid sequences shown in SEQ ID NOs: 12, 14 and 15, respectively. 2. An antibody according to claim 1, characterized in that the antibody contains heavy and light chain variable domains containing the amino acid sequences indicated in SEQ ID NOs: 5 and 6, respectively. 3. An antibody according to claim 1 or 2, characterized in that the antibody contains heavy and light chains containing the amino acid sequences indicated in SEQ ID NO: 7 and 8, respectively. 4. An antibody according to claim 1 or 2, wherein the antibody is a humanized antibody. 5. An antibody according to claim 1 or 2, characterized in that the antibody is an IgG1 or IgG2a antibody. 6. An antibody according to claim 5, characterized in that the antibody contains one amino acid substitution in the Fc region compared to the amino acid sequence of human IgG1, as indicated in SEQ ID NO: 206, and the amino acid substitution in the Fc region is S267E. 7. An antibody according to claim 6, characterized in that the substitution increases the affinity of the antibody for FcyRIIb and/or reduces antibody-dependent cell-mediated cytotoxic (ADCC) activity compared to the amino acid sequence of human IgG1 as indicated in SEQ ID NO: 206. 8. An antibody according to any one of claims 1 to 7, characterized in that the antibody blocks the binding and/or interaction of the ICOS ligand (ICOS-L) with human ICOS. 9. An antibody according to any one of claims 1 to 8, characterized in that the antibody binds to human, cynomolgus, mouse and rat ICOS. 10. An antibody according to any one of claims 1 to 9, characterized in that the antibody has one or more of the following properties: (a) binds to human T cells with an EC50 value of about 0.7 nM and cynomolgus monkey T cells with an EC50 value of about 0.3 nM; (b) binds to human activated CD4+ T cells; (c) does not bind to human CD28 or human CTLA-4; (d) activates at least one primary T lymphocyte, such as a CD4+ effector T cell (Teff), a follicular T helper cell (Tfh), and a regulatory T cell (Treg); (e) induces protein kinase B (pAkt) phosphorylation in an in vitro primary T cell signaling assay with an EC50 value of about 30 nM; (f) induces the production of interleukin-10 (IL-10) in response to staphylococcal enterotoxin B in a co-culture assay of Tfh and naïve B cells; (g) induces a greater increase in CD3-stimulated Teff proliferation compared to CD45RA+ Tregs and CD45RO+ Tregs in an in vitro assay; (h) reduces Teff suppression by Treg cells; (i) does not increase cytokine production in whole blood assay at 10 µg/mL; (j) increases secretion of at least one of IL-10 and IFN-g by Tfh cells in vitro; (k) stimulates ICOS-mediated signaling; (l) has increased affinity for CD32B and/or CD32A and/or (m) has reduced affinity for CD16. 11. An isolated nucleic acid encoding a heavy chain variable region or a full-length heavy chain of an antibody according to claim 1, wherein the heavy chain variable domain contains the amino acid sequence shown in SEQ ID NO: 5 and the full-length heavy chain contains the amino acid sequence shown in SEQ ID NO: 7. 12. An isolated nucleic acid encoding a light chain variable region or full length light chain of an antibody according to claim 1, wherein the light chain variable domain contains the amino acid sequence shown in SEQ ID NO: 6 and the full length light chain contains the amino acid sequence shown in SEQ ID NO: 8. 13. An isolated nucleic acid encoding the heavy and light chain variable regions of the antibody of claim 1, wherein the heavy and light chain variable domains comprise the amino acid sequences set forth in SEQ ID NOs: 5 and 6, respectively. 14. An isolated nucleic acid encoding the full length heavy and light chains of the antibody of claim 1, wherein the heavy and light chains comprise the amino acid sequences set forth in SEQ ID NOs: 7 and 8, respectively. 15. A method for producing an antibody according to claim 1, including the expression of one or more nucleic acids - 134 -