RU2812283C2 - Designed il-2-fc fusion proteins - Google Patents

Abstract

FIELD: biotechnology.

SUBSTANCE: following are presented: polypeptide compositions for expansion of Tregs, one of which contains a variant human IL-2 protein, in which the said variant IL-2 protein contains one or more amino acid substitutions compared to SEQ ID NO:2, selected from the group T3A/D20N/T37R, T3A/D20N/N71K, T3A/D20N/T37R/C125S, T3A/D20N/N71K/C125S, T3A/D20N/T37R/C125A and T3A/D20N/N71K/C125A. The following are also disclosed: Tregs expansion polypeptide compositions containing a heterodimeric Tregs expansion protein complex.

EFFECT: invention provides selective and sustainable expansion of Treg.

13 cl, 145 dwg, 145 ex

Description

Priority application

This application claims priority to US Provisional Application Nos. 62/607,850, filed December 19, 2017, and 62/675,070, filed May 22, 2018, which are expressly incorporated herein by reference in their entirety, with specific reference to the figures, symbols and claims in them.

List of sequences

This application contains a sequence listing that was filed electronically in ASCII format and is incorporated herein by reference in its entirety. The ASCII copy in question, created on November 30, 2018, is named 067461-5217-WO_ST25.txt and is 605038 bytes in size.

State of the art

Homeostasis of the immune system is based on a delicate balance between different populations of immune cells, including CD8+ and CD4+ T cells (CD3+CD25-FOXP3-) and regulatory T cells (Treg; CD3+CD4+CD25+FOXP3+). Disturbances in this balance can lead to diseases such as autoimmune diseases, in which T cells remain unregulated and attack the body's own tissues. Under normal conditions, Tregs modulate T cell differentiation and effector and cytotoxic functions. Thus, the basic premise in this regard is that abnormalities in Treg cell number and/or function are contributory factors in pathological conditions. In this regard, the ability to alter the balance between cytotoxicity and regulation by fine-tuning the T cell response has great potential for the treatment of autoimmune and other diseases.

IL-2 promotes the proliferation and differentiation of B cells, T cells and NK cells. IL-2 is also important for Treg function and survival. IL-2 exerts its cellular signaling function by binding to a high-affinity trimeric receptor complex consisting of three distinct proteins: the common gamma chain (γc; CD132) and the IL-2 receptor B chain (IL-2β; CD122), which are shared with IL-15, as well as a unique alpha chain receptor (IL-2Rγ; CD25). IL-2 can also exert its cell signaling function by binding to a dimeric intermediate-affinity receptor complex consisting only of IL-2Pß and γc (IL-2Pßγ).

Due to the low concentrations of IL-2 that typically exist in tissues, IL-2 preferentially activates cells that express a high-affinity receptor complex (CD25:CD122:CD132; IL-2γßγ) and therefore has a preferential effect on FOXP3+ Tregs, which constitutively express CD25. At the same time, IL-2 can also activate and induce proliferation of FOXP3-T cells that express the intermediate-affinity receptor complex (CD122:CD132; IL-2βγ). FOXP3- T cells, such as CD4+ or CD8+ T cells, may contribute to inflammation, autoimmune reactions, organ transplant rejection, or graft-versus-host disease. Due to the ability of IL-2 to stimulate or reduce the activity of both T cells and Tregs with limited selectivity, there is a significant need in the field for the development of more selective Treg modulators. In addition, as a potential drug, IL-2 has the disadvantage of very rapid clearance, with a half-life measured in minutes, making effective administration difficult. This invention is aimed at solving both of these problems by obtaining new IL-2-Fc fusion proteins.

Accordingly, there is a need to produce useful variants of IL-2 and Fc fusion proteins.

Brief description of the invention

Accordingly, in some aspects, this disclosure relates to compositions comprising a human IL-2 protein variant (as compared to SEQ ID NO:2), wherein said IL-2 protein variant contains an amino acid substitution(s) selected from the group T3A, R38A; R38D; R38E; R38F; R38G; R38H; R38I; R38K; R38L; R38M; R38N; R38P; R38Q; R38S; R38T; R38V; R38W; R38Y; T41A; T41D; T41E; T41F; T41G; T41H; T41I; T41K; T41L; T41M; T41N; T41P; T41Q; T41R; T41S; T41V; T41W; T41Y; F42A; F42D; F42E; F42G; F42H; F42I; F42K; F42L; F42M; F42N; F42P; F42Q; F42R; F42S; F42T; F42V; F42W; F42Y; R38Q/T41K; R38Q/T41Q; R38E/T41K; R38Q/T41R; R38N/T41Q; R38Q/T41V; R38N/T41V; R38Q/T41M; R38Q/T41S; R38Q/T41L; R38N/T41M; T41I/F42Y; T41E/F42Y’ T41D/F42Y; T41M/F42Y; 41Q/F42Y; T41E/F42H; T41E/F42L; T41E/F42P; R38Q/F42Y; R38N/T41R; R38N/T41K; R38V/T41R; R38P/T41R; T41E/F42K; T41D/F42K; T41M/F42K; T41Q/F42K; R38Q/F42K; T41I/F42K; R38N/F42K; T41H/F42K; R38Q/T41K/F42Y; R38Q/T41R/F42Y; R38Q/T41Q/F42Y; R38Q/T41V/F42Y; R38N/T41K/F42K; R38Q/T41H/F42K; R38Q/T41K/F42K; R38Q/T41Q/F42K; 38Q/T41V/F42K; R38Q/T41R/F42K; Q11E; L12D; Q13E; E15Q; H16Y; L19D; D20N; N29S/Y31H/K35R/T37A/R38L/K48E/V69A/N71R/Q74P/N88D/I89V/Q126T; Q22E; K35R; T37S; K43R; F44Y; Y45F; K48R; K49E; E61Q; E62Q; K64R; E68Q; V69L; L72I; R81D; D84N; S87T; N88D; V91L; I92L; E95Q; Y107F; E116R; N119D; R120D; T123S; C125S/Q126E; C125S/S127T; C125S/I129L; C125S/S130T; C125S/T133S; T3A; F42A/Y45A/L72G; N29S/Y31H/K35R/T37A/K48E/V69A/N71R/Q74P/N88D/I89V; V69A/Q74P/I128T; N29S/Y31H/K35R/T37A/K48E/V69A/N71R/Q74P/N88D/I89V/Q126T; C125S/Q126T; N88R; R38I; L80F/R81D/L85V/I92F; L18R/L80F/R81D/L85V/I92F/Q126T; L18R/L80F/R81D/L85V/I92F/Q126T/S130R; F42A/Y45A/L72G/N88R; F42A/Y45A/L72G/Q126T; F42A/Y45A/L72G/N88R/Q126T; L19D; D20N; N88D; N88K; N88R; N88R; N88R; F42A/Y45A/L72G; N29S/Y31H/K35R/T37A/K48E/N71R/N88D/I89V; L19D/N29S/Y31H/K35R/T37A/K48E/N71R; D20N/N29S/Y31H/K35R/T37A; K48E/N71R; L19D/N29S/Y31H/K35R/T37A/K48E; D20N/N29S/Y31H/K35R/T37A; K48E; L19D K35R; L19D/T37R; D20N/T37R; L19D/N71K; D20N/N71K; D20N/R38I; D20N/T37R; 38I; D20N/R38I/N71K; D20N/N71K; D20N; D20N/T37R; D20N/R38I; D20N/T37R R38I; D20N/R38I/N71K; D20N; D20N/T37R; D20N/N71K; D20N/R38I; D20N/T37R R38I; D20N/R38I/N71K; D20N; D20N/T37R; D20N/N71K; D20N/R38I; D20N/T37R/R38I; D20N/R38I/N71K; N29S/Y31H/K35R/T37A/K48E/V69A/N71R/Q74P/N88D/I89V/Q126T, R38A/C125S; R38D/C125S; R38E/C125S; R38F/C125S; R38G/C125S; R38H/C125S; R38I/C125S; R38K/C125S; R38L/C125S; R38M/C125S; R38N/C125S; R38P/C125S; R38Q/C125S; R38S/C125S; R38T/C125S; R38V/C125S; R38W/C125S; R38Y/C125S; T41A/C125S; T41D/C125S; T41E/C125S; T41F/C125S; T41G/C125S; T41H/C125S; T41I/C125S; T41K/C125S; T41L/C125S; T41M/C125S; T41N/C125S; T41P/C125S; T41Q/C125S; T41R/C125S; T41S/C125S; T41V/C125S; T41W/C125S; T41Y/C125S; F42A/C125S; F42D/C125S; F42E/C125S; F42G/C125S; F42H/C125S; F42I/C125S; F42K/C125S; F42L/C125S; F42M/C125S; F42N/C125S; F42P/C125S; F42Q/C125S; F42R/C125S; F42S/C125S; F42T/C125S; F42V/C125S; F42W/C125S; F42Y/C125S; R38Q/T41K/C125S; R38Q/T41Q/C125S; R38E/T41K/C125S; R38Q/T41R/C125S; R38N/T41Q/C125S; R38Q/T41V/C125S; R38N/T41V/C125S; R38Q/T41M/C125S; R38Q/T41S/C125S; R38Q/T41L/C125S; R38N/T41M/C125S; T41I/F42Y/C125S; T41E/F42Y/C125S’ T41D/F42Y/C125S; T41M/F42Y/C125S; 41Q/F42Y/C125S; T41E/F42H/C125S; T41E/F42L/C125S; T41E/F42P/C125S; R38Q/F42Y/C125S; R38N/T41R/C125S; R38N/T41K/C125S; R38V/T41R/C125S; R38P/T41R/C125S; T41E/F42K/C125S; T41D/F42K/C125S; T41M/F42K/C125S; T41Q/F42K/C125S; R38Q/F42K/C125S; T41I/F42K/C125S; R38N/F42K/C125S; T41H/F42K/C125S; R38Q/T41K/F42Y/C125S; R38Q/T41R/F42Y/C125S; R38Q/T41Q/F42Y/C125S; R38Q/T41V/F42Y/C125S; R38N/T41K/F42K/C125S; R38Q/T41H/F42K/C125S; R38Q/T41K/F42K/C125S; R38Q/T41Q/F42K/C125S; 38Q/T41V/F42K/C125S; R38Q/T41R/F42K/C125S; N29S/Y31H/K35R/T37A/R38L/K48E/V69A/N71R/Q74P/N88D/I89V/C125S/Q126T; Q11E/C125S; L12D/C125S; Q13E/C125S; E15Q/C125S; H16Y/C125S; L19D/C125S; D20N/C125S; Q22E/C125S; K35R/C125S; T37S/C125S; K43R/C125S; F44Y/C125S; Y45F/C125S; K48R/C125S; K49E/C125S; E61Q/C125S; E62Q/C125S; K64R/C125S; E68Q/C125S; V69L/C125S; L72I/C125S; R81D/C125S; D84N/C125S; S87T/C125S; N88D/C125S; V91L/C125S; I92L/C125S; E95Q/C125S; Y107F/C125S; E116R/C125S; N119D/C125S; R120D/C125S; T123S/C125S; C125S/Q126E; C125S/S127T; C125S/I129L; C125S/S130T; C125S/T133S; T3A/C125S; T3A/F42A/Y45A/L72G/C125A; N29S/Y31H/K35R/T37A/K48E/V69A/N71R/Q74P/N88D/I89V/C125S; V69A/Q74P/I128T/C125S; N29S/Y31H/K35R/T37A/K48E/V69A/N71R/Q74P/N88D/I89V/C125S/Q126T; C125S/Q126T; N88R/C125S; R38I/C125S; L80F/R81D/L85V/I92F/C125S; L18R/L80F/R81D/L85V/I92F/C125S/Q126T; L18R/L80F/R81D/L85V/I92F/C125S/Q126T/S130R; T3A/F42A/Y45A/L72G/N88R/C125A; T3A/F42A/Y45A/L72G/C125A/Q126T; T3A/F42A/Y45A/L72G/N88R/C125A/Q126T; T3A/L19D/C125S; T3A/D20N/C125S; T3A/N88D/C125S; T3A/N88K/C125S; N88R/C125S; N88R/C125S; N88R/C125S; T3A/F42A/Y45A/L72G/C125A; N29S/Y31H/K35R/T37A/K48E/N71R/N88D/I89V/C125S; L19D/N29S/Y31H/K35R/T37A/K48E/N71R/C125S; D20N/N29S/Y31H/K35R/T37A; K48E/N71R/C125S; L19D/N29S/Y31H/K35R/T37A/K48E/C125S; D20N/N29S/Y31H/K35R/T37A; K48E/C125S; T3A/L19D K35R/C125S; T3A/L19D/T37R/C125S; T3A/D20N/T37R/C125S; T3A/L19D/N71K/C125S; T3A/D20N/N71K/C125S; T3A/D20N/R38I/C125S; T3A/D20N/T37R; 38I/C125S; T3A/D20N/R38I/N71K/C125S; T3A/D20N/N71K/C125S; T3A/D20N/C125S; T3A/D20N/T37R/C125S; T3A/D20N/R38I/C125S; T3A/D20N/T37R R38I/C125S; T3A/D20N/R38I/N71K/C125S; T3A/D20N/C125S; T3A/D20N/T37R/C125S; T3A/D20N/N71K/C125S; T3A/D20N/R38I/C125S; T3A/D20N/T37R R38I/C125S; T3A/D20N/R38I/N71K/C125S; T3A/D20N/C125S; T3A/D20N/T37R/C125S; T3A/D20N/N71K/C125S; T3A/D20N/R38I/C125S; T3A/D20N/T37R/R38I/C125S; T3A/D20N/R38I/N71K/C125S; and N29S/Y31H/K35R/T37A/K48E/V69A/N71R/Q74P/N88D/I89V/C125S/Q126T.

In additional aspects, this disclosure provides dimeric IL-2-Fc fusion proteins comprising: a) a first monomer comprising N- to C-terminus: i) a variant IL-2 protein according to any one of paragraphs. A1-A4; ii) a first domain linker; and iii) a first variant Fc domain; and b) a second monomer containing a second variant Fc domain.

In additional aspects, the present invention provides dimeric IL-2-Fc fusion proteins, wherein said second monomer comprises from the N- to C-terminus of: a) a variant IL-2 protein according to any one of claims. A1-A4; b) a second domain linker; and c) said second variant Fc domain.

In additional aspects, IL-2-Fc dimeric fusion proteins have first and second variant Fc domains that contain heterodimerization variants selected from the group consisting of L368D/K370S: S364K/E357Q; L368D/K370S : S364K; L368E/K370S : S364K; T411E/K360E/Q362E : D401K; and T366S/L368A/Y407V : T366W. In some cases, IL-2-Fc fusion proteins additionally contain ablative variants, including P233P/L234V/L235A/G236_/S267K. In some aspects, the fusion proteins also contain Fc domains with amino acid substitutions M428L/N434S or M428L/N434A. In some aspects, the domain linker is an IGG1 hinge, and in other aspects it may include a linker selected from (GS)n, (GSGGS)n, (GGGGS)n, and (GGGS)n, where n is an integer, component of at least one.

In additional aspects, the present invention includes a polypeptide composition comprising a variant human IL-2 protein, wherein said variant IL-2 protein contains an amino acid substitution(s) relative to SEQ ID NO:2 selected from groups T3A/D20N/T37R and T3A/D20N/N71K. In some cases, the variant IL-2 protein further contains a C125S variant or a C125A variant.

In a further aspect, the polypeptide composition is a homodimeric protein complex in which each protein monomer contains said IL-2 variant protein covalently attached to an Fc domain. In some aspects, each of these Fc domains is a variant Fc domain.

In a further aspect, the polypeptide composition is a heterodimeric protein complex comprising a first protein monomer comprising said variant IL-2 protein covalently attached to a first variant Fc domain, and a second protein monomer comprising a second variant Fc domain.

In a further aspect, the polypeptide composition has variant Fc domains that are variant human IgG1 Fc domains containing the M428L/N434S amino acid substitutions.

In a further aspect, the polypeptide composition has variant Fc domains that are human IgG1 variant Fc domains containing the amino acid substitutions E233P/L234V/L235A/G236del/S267K.

In a further aspect, the polypeptide composition has first and second variant Fc domains that include a plurality of heterodimerization variants selected from the group consisting of the variants depicted in FIG. 2. In some embodiments, the plurality of heterodimerization variants is selected from the group consisting of L368D/K370S: S364K/E357Q; L368D/K370S : S364K; L368E/K370S : S364K; T411E/K360E/Q362E : D401K; and T366S/L368A/Y407V : T366W.

In a further aspect, the polypeptide composition is selected from the group consisting of XENP27564 (SEQ ID NO:297 and 298), XENP27563 (SEQ ID NO:295 and 296), XENP26105 (SEQ ID NO:245 and 246) and XENP26109 (SEQ ID NO: 249 and 250).

Also presented are nucleic acid compositions containing: a) a first nucleic acid encoding said first protein monomer according to any one of claims. 6-13; and b) a second nucleic acid encoding said second protein monomer according to any one of claims. 6-13 respectively. In addition, expression vector compositions are provided, comprising a) a first expression vector containing said first nucleic acid; and b) a second expression vector containing said second nucleic acid, and host cells containing nucleic acid compositions and/or expression vector compositions. In addition, methods for producing a polypeptide composition are provided, including culturing a host cell of this invention under conditions under which said composition is formed, and recovering said composition.

In a further aspect, the IL-2-Fc dimeric fusion protein is selected from XENP24635; XENP24636; XENP24637; XENP24638; XENP24639; XENP24640; XENP24641; XENP24642; XENP24643; XENP24725; XENP24728; XENP24729; XENP24730; XENP24731; XENP24732; XENP25717; XENP25720; XENP25725; XENP25727; XENP25910; XENP25911; XENP25912; XENP26086; XENP26088; XENP26089; XENP26092; XENP26093; XENP26096; XENP26104; XENP26105; XENP26108; XENP26109; XENP26835; XENP26839; XENP26840; XENP26841; XENP26986; XENP26987; XENP26989; XENP26990, XENP26991, XENP25906, XENP25907; XENP25908; XENP25909; XENP26992; XENP26993; XENP26994; XENP26995; XENP26996; XENP27001; XENP27002; XENP27003; XENP27004; XENP27005; XENP27006 and XENP27007.

In additional aspects, methods are provided for activating CD25+ cells, including contacting said cells with an IL-2-Fc dimeric fusion protein of the present invention, and methods for treating an autoimmune disease, including administering to a patient in need thereof a protein composition provided herein.

Brief description of graphic materials

In fig. 1 depicts the amino acid sequences (and GenBank accession numbers) of human IL-2 and its receptors: IL-2Rα (also known as CD25), IL-2Rβ (also known as CD122), and the common gamma chain (also known as IL-2R or CD132).

In fig. 2 shows suitable pairs of Fc heterodimerization variants (including distorted and pI variants). In fig. 2 shows variants for which there are no corresponding variants of “monomer 2”; they are pI variants that can be used alone on any monomer.

In fig. 3 shows a list of constant regions of isosteric variant antibodies and their corresponding replacements. pI(-) denotes variants with a lower pI value, while pI_(+) denotes variants with a higher pI value. They may optionally and independently be combined with other heterodimerization variants of this invention (as well as other types of variants described herein).

In fig. 4 depicts useful ablative variants that eliminate FcγR binding (sometimes referred to as “knockout” or “KO” variants). Typically, ablative variants are found on both monomers, although in some cases they may be present on only one monomer.

In fig. 5 depicts particularly useful embodiments of the "non-cytokine" components of the present invention.

In fig. 6 depicts the sequences of several useful human IgG-based IL-2-Fc fusion frameworks, without cytokine sequences. Framework 1 is based on human IgG1 (allotype 356E/358M) and contains C220S on both chains, S364K/E357Q distorted variants: L368D/K370S, pI variants Q295E/N384D/Q418E/N421D on the chain with distorted L368D/K370S variants and E233P ablative variants /L234V/L235A/G236del/S267K in both circuits. Scaffold 2 is based on human IgG1 (allotype 356E/358M) and contains C220S on both chains, distorted variants S364K/E357Q: L368D/K370S, pI variants Q295E/N384D/Q418E/N421D on the chain with distorted variants L368D/K370S and ablative variants E233P /L234V/L235A/G236del/S267K in both circuits. Scaffold 3 is based on human IgG1 (allotype 356E/358M) and contains C220S on both chains, distorted variants S364K/E357Q: L368E/K370S, pI variants Q295E/N384D/Q418E/N421D on the chain with distorted variants L368E/K370S and ablative variants E233P /L234V/L235A/G236del/S267K in both circuits. Scaffold 4 is based on human IgG1 (356E/358M allotype) and contains C220S on both chains, D401K distorted variants: K360E/Q362E/T411E, pI variants Q295E/N384D/Q418E/N421D on the chain with K360E/Q362E/T411E distorted variants and ablative variants options E233P/L234V/L235A/G236del/S267K in both circuits. Scaffold 5 is based on human IgG1 (allotype 356D/358L) and contains C220S on both chains, distorted variants S364K/E357Q: L368D/K370S, pI variants Q295E/N384D/Q418E/N421D on the chain with distorted variants L368D/K370S and ablative variants E233P /L234V/L235A/G236del/S267K in both circuits. Scaffold 6 is based on human IgG1 (allotype 356E/358M) and contains C220S on both chains, S364K/E357Q distorted variants: L368D/K370S, pI variants Q295E/N384D/Q418E/N421D on the chain with distorted L368D/K370S variants and E233P ablative variants /L234V/L235A/G236del/S267K in both circuits, as well as the N297A variant in both circuits. Scaffold 7 is identical to 6 except that the mutation is N297S. Alternative scaffold formats 6 and 7 may exclude the E233P/L234V/L235A/G236del/S267K ablative variants in both strands. Framework 8 is based on human IgG4 and contains the S364K/E357Q distorted variants: L368D/K370S, the pI variants Q295E/N384D/Q418E/N421D in chain with the L368D/K370S distorted variants, as well as the S228P variant (EU numbering, i.e., S241P , as in Kabat) in both chains, which leads to the elimination of the exchange of Fab fragments, as is known in the art. Scaffold 9 is based on human IgG2 and contains distorted variants S364K/E357Q: L368D/K370S, pI variants Q295E/N384D/Q418E/N421D in chain with distorted variants L368D/K370S. Scaffold 10 is based on human IgG2 and contains the S364K/E357Q distorted variants: L368D/K370S, the pI variants Q295E/N384D/Q418E/N421D on the chain with the distorted L368D/K370S variants, and the S267K variant on both chains. Scaffold 11 is identical to scaffold 1 except that it contains the Xtend M428L/N434S mutations. Framework 12 is based on human IgG1 (allotype 356E/358M) and contains C220S in both identical chains and the ablation variants E233P/L234V/L235A/G236del/S267K in both identical chains. Scaffold 13 is based on human IgG1 (allotype 356E/358M) and contains C220S on both chains, distorted variants S364K/E357Q: L368D/K370S, pI variants P217R/P229R/N276K on the chain with distorted variants S364K/E357Q and ablative variants E233P/L234V /L235A/G236del/S267K in both circuits.

As will be appreciated by those skilled in the art and set forth below, any IL-2 variants may be included in these scaffolds in any combination. Each of these scaffolds includes sequences that are 90, 95, 98, and 99% identical (as defined herein) to the listed sequences and/or contain between 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 additional amino acid substitutions (compared to the “parent” of the above figure, which, as those skilled in the art will appreciate, already contains a number of amino acid modifications compared to the parent human IgG1 (or IgG2 or IgG4, depending on the framework). Thus, these scaffolds may contain additional amino acid modifications (typically amino acid substitutions) in addition to the distorted, pI and ablative variants contained in the scaffolds of the figure.In particular, FcRn variants such as M428L/N434S may also be included.

In fig. 7 depicts a series of exemplary variable length linkers. In some embodiments, these linkers find use by linking the C terminus of IL-2 to the N terminus of the Fc region (which in some cases includes a hinge domain).

In fig. 8A and 8B depict A) a structural model of IL-2 complexed with the high-affinity IL-2 receptor (IL-2Rαßγ), and B) the position of three IL-2 residues that are brought into contact with IL-2Rα, with the substitutions presumably weakening pH-dependent binding of IL-2 to IL-2Rα.

In fig. 9 depicts the amino acid sequences of exemplary IL-2 variants designed to reduce pH-dependent binding to IL-2PRα. It is important to note that these variants were generated using C-terminal polyhistidine tags (Hisx8 or HHHHHHHH) that were removed from the sequences depicted herein.

In fig. Figure 10 shows the association rates ( _ka ), dissociation rates ( _Kd ) and dissociation constants ( _KD ) of IL-2 variants relative to IL-2Rα at pH 6.0, as well as the fold increase in kd and KD relative to XENP14135 ( IL-2 wild type with C125S mutation). CO indicates no measurable binding.

In fig. Figure 11 shows the association rates ( _ka ), dissociation rates ( _Kd ) and dissociation constants ( _KD ) of IL-2 variants relative to IL-2Rα at pH 7.4, as well as the fold increase in kd and KD relative to XENP14135 ( IL-2 wild type with C125S mutation). CO indicates no measurable binding.

In fig. 12 depicts the fold increase in the rate of dissociation of IL-2 (k _d ) from IL-2Rα at pH 6.0, mediated by various point mutations.

In fig. 13 depicts the fold increase in the dissociation rate of IL-2 (k _d ) from IL-2Rα at pH 7.4 mediated by various point mutations.

In fig. 14 shows Biacore sensorgrams for A) XENP14135 (wild type IL-2 with the C125S mutation) and B) XENP14142 (variant IL-2 with R38I and C125S).

In fig. 15 depicts the amino acid sequences of an additional prior art variant of IL-2 (mutant 2-4 with Q126T described in WO 2009/061853 published May 14, 2009) without R38L (XENP14277) and with R38L (XENP14381) included to attenuate the pH -dependent binding to IL-2Rα. It is important to note that these variants were generated using C-terminal polyhistidine tags (Hisx8 or HHHHHHHH) that were removed from the sequences depicted herein.

In fig. 16 depicts the association rates (k _a ), dissociation rates (K _d ), and dissociation constants (K _D ) of IL-2 variants binding to IL-2Rα, with or without the R38L substitution, at pH 7.4 and pH 6.0 .

In fig. 17 depicts the amino acid sequences of exemplary IL-2 variants designed to alter binding to IL-2Pα, IL-2Pβ, γc, or IL-2Pβγ. It is important to note that these variants were generated using C-terminal polyhistidine tags (Hisx8 or HHHHHHHH) that were removed from the sequences depicted herein.

In fig. 18 depicts the normalized BLI response (relative to XENP14135) of exemplary IL-2 variants for various IL-2 receptors determined using Octet. The goal is to either increase binding to IL-2Rα, or decrease binding to IL-2Rβ and IL-2Rγ or the IL-2Rβγ interface, or both.

In fig. 19 depicts several formats of IL-2-Fc fusions of the present invention. Monovalent IL-2-Fc or "monoIL-2-Fc" (Fig. 19A) contains IL-2 recombinantly fused to the N-terminus of a heterodimeric Fc region, with the other side of the molecule being "Fc only" or "empty Fc" . Bivalent IL-2-Fc or "biIL-2-Fc" (Fig. 19B) contains IL-2 recombinantly fused to the N-terminus of both sides of the homodimeric Fc region. Monovalent IL-2-Fc with linker or "monoIL-2-Fc (with linker)" (Fig. 19C) contains IL-2 recombinantly fused to the N-terminus of the heterodimeric Fc region by a domain linker, with the other side of the molecule represents "Fc only" or "empty Fc". Bivalent IL-2-Fc with linker or "biIL-2-Fc (with linker)" (Fig. 19D) contains IL-2 recombinantly fused to the N-terminus of both sides of the homodimeric Fc region using a linker domain. Non-limiting examples of domain linkers suitable for use in the monoVIL-2-Fc (with linker) and biIL-2-Fc (with linker) formats are depicted in FIG. 7.

In fig. 20 depicts the amino acid sequences of exemplary monovIL-2-Fc fusions of the present invention. Slashes (/) indicate the boundary between IL-2 and Fc regions (in this case, the Fc region contains the hinge as well as the C220S variant).

In fig. 21 depicts the affinity (K _D), rate of association (k _a ), and rate of dissociation (k _d ) of exemplary IL-2-Fc fusions for IL-2Pα, IL-2Pß, and IL-2Pßγ determined using Octet.

In fig. 22A-22F depict the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells, CD4 ⁺ CD45RA ⁺ T cells, CD8 ⁺ CD45RA ^- T cells, CD8 ⁺ CD45RA ⁺ T cells, and Tregs by A) XENP24636. B) XENP24638, C) XENP24641, D) XENP24642, E) XENP24643 and F) XENP24731.

In fig. 23 depicts the amino acid sequences of exemplary biIL-2-Fc fusions of the present invention. Slashes (/) indicate the boundary between IL-2 and Fc regions, in this case the Fc region also contains an IgG1 hinge domain with the C220S variant.

In fig. 24A and 24B depict the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ⁻ T cells, CD4 ⁺ CD45RA ⁺ T cells, CD8 ⁺ CD45RA ⁻ T cells, CD8 ⁺ CD45RA ⁺ T cells, and Tregs by A) XENP25906 and B) XENP25907.

In fig. 25 depicts the amino acid sequences of exemplary IL-2-Fc fusions to linker domains. Slashes (/) indicate the boundary between IL-2, linkers and Fc regions, in this case the Fc region also contains an IgG1 hinge domain with the C220S variant. Linkers are underlined with a double line.

In fig. 26A-26E depict induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells, CD4 ⁺ CD45RA ⁺ T cells, CD8 ⁺ CD45RA ^- T cells, CD8 ⁺ CD45RA ⁺ T cells, and Tregs by A) XENP25908. B) XENP25909, C) XENP25910, D) XENP25911 and E) XENP25912.

In fig. 27 depicts the amino acid sequences of additional IL-2-Fc fusions to IL-2 designed to increase affinity for CD25 and decrease affinity SEQ Figure \* ARABIC for CD122. Slashes (/) indicate the boundary between IL-2 and Fc regions, in this case the Fc region also contains an IgG1 hinge domain with the C220S variant.

In fig. 28A-28T depict STAT5 phosphorylation in CD8 ⁺ and CD4 ⁺ T cells and Tregs as an indicator of activation by A) XENP24635, B) XENP24636, C) XENP24637, D) XENP24638, E) XENP24642, F) XENP25717, G) XENP25720 , H) XENP25725, I) XENP25727, J) XENP26086, K) XENP26088, L) XENP26089, M) XENP26092, N) XENP26093, O) XENP2104, XENP26) Q26P26) Q2 R) XENP26108, S) XENP2610 9 and T) recombinant IL -2 people.

In fig. 29A and 29B depict STAT5 phosphorylation in A) Tregs and B) CD4 ⁺ (CD45RA-) T cells as an indicator of activation by exemplary IL-2-Fc fusions.

In fig. 30 depicts the amino acid sequences of additional IL-2-Fc fusions. Slashes (/) indicate the boundary between IL-2, linker domains (with double underline) and Fc regions, in this case the Fc region also contains the IgG1 hinge domain with the C220S variant.

In fig. 31 depicts some preferred embodiments.

In fig. 32 depicts some preferred embodiments of embodiments constructed with A) pH switch substitutions, B) Treg selectivity substitutions, and C) a combination of pH switch substitutions and Treg selectivity substitutions.

In fig. 33 depicts the affinity (K _{D ),} rate of association ( k _a ), and rate of dissociation ( k _d ) of exemplary IL-2-Fc fusions for IL-2Pα, IL-2Pß, and IL-2Pßγ determined using Octet. CO indicates no binding.

In fig. 34 depicts the sequence of XENP27193, an Fc-IL-2 (V91K/C125A) fusion engineered to increase the ratio of Tregs to non-regulatory T cells described in WO 2014/153111. Slashes (/) indicate the boundary between IL-2, linker domains, and Fc regions (in this case, the Fc region also contains an IgG1 hinge domain with the C220S variant); and the linker is underlined with a double line.

In fig. 35A-E depicts induction of STAT5 phosphorylation in A) CD4 ⁺ CD45RA ^- T cells, B) CD8 ⁺ CD45RA ^- T cells, C) NK cells, D) γδ T cells, and E) Tregs by variant monovIL- fusions 2-Fc, XENP24638, XENP24642, XENP26105, XENP26109, XENP26835, XENP26839, XENP26991 and XENP25702. Data demonstrate that variant monovIL-2-Fc fusions preferentially induce STAT5 phosphorylation in Tregs relative to CD4 ⁺ memory T cells (CD45RA ^- ), CD8 ⁺ memory T cells (CD45RA ^- ), NK cells, and γδ T cells compared to the recombinant IL-2 and monovalent WT IL-2-Fc fusion (XENP24635), as well as the prior art bivalent IL-2-Fc fusion described in WO 2014/153111 (XENP27193).

In fig. 36 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells and Tregs by the IL-2 variant with N88R/C125S substitutions, presented in the format monoVIL-2-Fc without linker (XENP24642) and biIL-2-Fc with linker (XENP25908 ).

In fig. 37 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells and Tregs by the IL-2 variant with T3A/D20N/C125S substitutions, presented in the format monovIL-2-Fc without linker (XENP25720), biIL-2-Fc without linker (XENP26992) and biIL-2-Fc with linker (XENP27002).

In fig. 38 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells and Tregs by the IL-2 variant with T3A/D20N/T37R/C125S substitutions, presented in the format monovIL-2-Fc without linker (XENP26105), biIL-2-Fc without linker (XENP26993) and biIL-2-Fc with linker (XENP27003).

In fig. 39 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells and Tregs by the IL-2 variant with T3A/D20N/N71K/C125S substitutions, presented in the format monovIL-2-Fc without linker (XENP26109), biIL-2-Fc without linker (XENP26994) and biIL-2-Fc with linker (XENP27004).

In fig. 40 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells and Tregs by the IL-2 variant with T3A/D20N/R38I/C125S substitutions, presented in the format monovIL-2-Fc without linker (XENP26835), biIL-2-Fc without linker (XENP26995) and biIL-2-Fc with linker (XENP27005).

In fig. 41 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells and Tregs by the IL-2 variant with substitutions T3A/D20N/T37R/R38I/C125S, presented in the monoVIL-2-Fc format without a linker (XENP26839), biIL-2 -Fc without linker (XENP26996) and biIL-2-Fc with linker (XENP27006).

In fig. 42 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells and Tregs by the IL-2 variant with substitutions T3A/D20N/R38I/N71K/C125S, presented in the format monovIL-2-Fc with linker (XENP26991), biIL-2 -Fc without linker (XENP27001) and biIL-2-Fc with linker (XENP27007).

In fig. 43 depicts the amino acid sequences of exemplary bivalent IL-2-Fc fusions containing the Xtend (M428L/N434S) Fc of the present invention (in this case also containing a hinge and a C220S variant). Slashes (/) indicate the boundary between IL-2 and Fc regions.

In fig. 44 depicts the amino acid sequences of exemplary monovalent IL-2-Fc fusions containing the Xtend (M428L/N434S) Fc of the present invention. Slashes (/) indicate the boundary between IL-2 and Fc regions.

In fig. 45A-E depict the induction of STAT5 phosphorylation in various lymphocyte populations by A) XENP26105, B) XENP26109, C) XENP24635, D) XENP25908, and E) XENP27193.

In fig. 46 depicts immunophenotyping of rapamycin-treated Treg cultures treated with XENP27564 or recombinant IL-2. Data demonstrate higher CD25 expression following XENP27564 treatment.

In fig. 47 Tregs with rapamycin treated with XENP27564 or recombinant IL-2 are plotted in a histogram to assess relative CD25 expression. Data demonstrate that rapamycin-treated Treg cultures treated with IL-2-Fc XENP27564 exhibit higher CD25 expression.

In fig. 48 depicts various CD4 ^{+ compartments,} including Tregs, after incubation with rapamycin and XENP27564 or recombinant IL-2. Data demonstrate that cultures expanded with XENP27564 exhibit a larger population of effector Tregs (CD45RA ^- FoxP3 ^mid-high ) compared to cultures expanded with recombinant IL-2.

In fig. 49A and B depict the suppression of proliferation of A) CD8 responder T cells and B) CD4 responder T cells by rapamycin-treated Treg cultures expanded with either XENP27564 or recombinant IL-2. Data suggest that Tregs expanded with XENP27564 may have enhanced suppressor function.

In fig. 50 depicts CD25 expression on Tregs in the suppressor assay depicted in FIG. X indicated by A) CD25 CIF per Treg and B) CD25 ⁺ Treg percentage.

In fig. 51A and B depict CD127 expression on Tregs in the suppressor assay depicted in FIG. X indicated by A) CD127 CIF per Tregs and B) percentage of CD127+ Tregs.

In fig. 52A-E depict proliferation (determined by CFSE or Tag-it Violet dilution) of various lymphocyte populations after incubation or PBMCs and Tregs with A) XENP27563, B) XENP27564, C) XENP24635, D) recombinant IL-2, and E) recombinant IL-15. Data demonstrate that XENP27563 and XENP27564 exhibit Treg selectivity.

In fig. 53A-B depict CD8+ T cell proliferation depicted by A) proliferating cells (determined by CFSE or Tag-it Violet dilution) and B) total cells after incubation or PBMCs and Tregs by XENP27563, XENP27564, XENP24635, recombinant IL-2, recombinant IL-15 and negative control anti-RSV mAb XENP15074. Data demonstrate that XENP27563 and XENP27564 are much less effective in inducing CD8+ T cell proliferation compared to recombinant IL-2 and IL-15, as well as IL-2-Fc fusion containing IL-2 WT (with the C125S mutation).

In fig. 54 depicts CD4+ T cell proliferation depicted by A) proliferating cells (determined by CFSE or Tag-it Violet dilution) and B) total cells after incubation or PBMCs and Tregs by XENP27563, XENP27564, XENP24635, recombinant IL- 2, recombinant IL-15 and negative control anti-RSV mAb XENP15074. Data demonstrate that XENP27563 and XENP27564 are much less effective in inducing CD4+ T cell proliferation compared to recombinant IL-2 and IL-15, as well as IL-2-Fc fusion containing IL-2 WT (with the C125S mutation).

In fig. 55 depicts proliferating CD8 ⁺ T cells (demonstrated by the percentage of cells expressing Ki67) after incubation of PBMCs with the indicated concentrations of the indicated study drugs and A) 5 ng/ml, B) 10 ng/ml, or C) 20 ng/ml associated anti-CD3 tablet (OCT3). Data demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) are impaired in inducing CD8 ⁺ T cell proliferation.

In fig. 56 depicts proliferating CD4 ⁺ T cells (demonstrated by the percentage of cells expressing Ki67) after incubation of PBMCs with the indicated concentrations of the indicated study drugs and A) 5 ng/ml, B) 10 ng/ml, or C) 20 ng/ml associated anti-CD3 tablet (OCT3). Data demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) are impaired in inducing CD4 ⁺ T cell proliferation.

In fig. 57 depicts proliferating CD45RA ^- CD8 ⁺ T cells (demonstrated by the percentage of cells expressing Ki67) after incubation of PBMCs with the indicated concentrations of the indicated study drugs and A) 5 ng/ml, B) 10 ng/ml, or C) 20 ng/ml associated with anti-CD3 tablet (OCT3). Data demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) are impaired in inducing CD8 ⁺ CD45RA ⁻ T cell proliferation.

In fig. 58 depicts proliferating CD8 ⁺ CD45RA ⁺ T cells (demonstrated by the percentage of cells expressing Ki67) after incubation of PBMCs with the indicated concentrations of the indicated study drugs and A) 5 ng/ml, B) 10 ng/ml, or C) 20 ng/ml associated with anti-CD3 tablet (OCT3). Data demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) are impaired in inducing CD8 ⁺ CD45RA ⁺ T cell proliferation.

In fig. 59 depicts proliferating CD4 ⁺ CD45RA ⁻ T cells (demonstrated by the percentage of cells expressing Ki67) after incubation of PBMCs with the indicated concentrations of the indicated study drugs and A) 5 ng/ml, B) 10 ng/ml, or C) 20 ng/ml associated with anti-CD3 tablet (OCT3). Data demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) are impaired in inducing CD4 ⁺ CD45RA ⁻ T cell proliferation.

In fig. 60 depicts proliferating CD4 ⁺ CD45RA ⁺ T cells (demonstrated by the percentage of cells expressing Ki67) after incubation of PBMCs with the indicated concentrations of the indicated study drugs and A) 5 ng/ml, B) 10 ng/ml, or C) 20 ng/ml associated with anti-CD3 tablet (OCT3). Data demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) are impaired in inducing CD4 ⁺ CD45RA ⁺ T cell proliferation.

In fig. 61 depicts proliferating NK cells (demonstrated by the percentage of cells expressing Ki67) after incubation of PBMCs with the indicated concentrations of the indicated test drugs and A) 5 ng/ml, B) 10 ng/ml, or C) 20 ng/ml plate-bound anti -CD3 (OCT3). Data demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) are impaired in inducing NK cell proliferation.

In fig. 62 depicts proliferating Tregs (demonstrated by the percentage of cells expressing Ki67) after incubation of PBMCs with the indicated concentrations of the indicated study drugs and A) 5 ng/ml, B) 10 ng/ml, or C) 20 ng/ml plate-bound anti-CD3 (OKT3). Data demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) induce Treg proliferation.

In fig. 63 depicts proliferating CD8 ⁺ T cells, CD8 ⁺ CD45RA ^- T cells, CD8 ⁺ CD45RA ⁺ T cells, CD4 ⁺ T cells, CD4 ⁺ CD45RA ^- T cells, CD4 ⁺ CD45RA ⁺ T cells, NK cells and Tregs (indicated as the percentage of cells expressing Ki67) after treatment with the indicated concentrations of A) XENP27563, B) XENP27564, C) XENP24635, D) IL-2 and E) IL-15 and 5 ng/ml plate-bound anti-CD3 (OKT3).

In fig. 64 depicts proliferating CD8 ⁺ T cells, CD8 ⁺ CD45RA ^- T cells, CD8 ⁺ CD45RA ⁺ T cells, CD4 ⁺ T cells, CD4 ⁺ CD45RA ^- T cells, CD4 ⁺ CD45RA ⁺ T cells, NK cells and Tregs (indicated as the percentage of cells expressing Ki67) after treatment with the indicated concentrations of A) XENP27563, B) XENP27564, C) XENP24635, D) IL-2 and E) IL-15 and 10 ng/ml plate-bound anti-CD3 (OKT3).

In fig. 65 depicts proliferating CD8 ⁺ T cells, CD8 ⁺ CD45RA ^- T cells, CD8 ⁺ CD45RA ⁺ T cells, CD4 ⁺ T cells, CD4 ⁺ CD45RA ^- T cells, CD4 ⁺ CD45RA ⁺ T cells, NK cells and Tregs (indicated as the percentage of cells expressing Ki67) after treatment with the indicated concentrations of A) XENP27563, B) XENP27564, C) XENP24635, D) IL-2 and E) IL-15 and 20 ng/ml plate-bound anti-CD3 (OKT3).

In fig. 66 depicts the expansion of A) CD4 ⁺ CD45RA ^- T cells, B) CD8 ⁺ CD45RA ^- T cells, C) CD8α ^- CD16 ⁺ NK cells, and D) FoxP3 ⁺ Tregs in cynomolgus monkeys treated with 3× dose of XENP27563 and 3 × dose of XENP27564. The data demonstrate that both XENP27563 and XENP27564 lead to selective expansion of Tregs and that the two investigational drugs induce similar pharmacological activity.

In fig. 67 depicts changes in serum concentration levels over time in cynomolgus monkeys administered A) 3x dose of XENP27563 and B) 3x dose of XENP27564. Data indicate that the two study drugs exhibit similar pharmacokinetic profiles.

In fig. 68 depicts changes in serum albumin concentrations in cynomolgus monkeys administered 3× dose of XENP27563 and 3× dose of XENP27564. Data demonstrate that in one animal administered XENP27563, a long-lasting decrease in albumin levels was detected after the first and second administration; in addition, in one animal administered XENP27564, a transient decrease in albumin levels was detected only after the second administration, with a rapid recovery to baseline levels.

In fig. 69 depicts telemetry blood pressure data for A) the first cynomolgus monkey treated with XENP27563 on day 0 and day 15, B) the second cynomolgus monkey treated with XENP27563 on day 0 and day 15, and C) a third cynomolgus monkey treated with XENP27564 on day 0 and day 15. The data show a sharp decrease in blood pressure in the first monkey one day after the 2nd dose and a sharp decrease in blood pressure in the second monkey one day after the 1st dose, while in the third monkey the blood pressure remained stable throughout the duration of the study.

In fig. 70 depicts the expansion of A) CD4 ⁺ CD45RA ^- T cells, B) CD8 ⁺ CD45RA ^- T cells, C) CD8α ^- CD16 ⁺ NK cells, and D) Tregs over time in cynomolgus monkeys following 1x, 2x administration and 3× doses of XENP27564.

In fig. 71 depicts Treg expansion on days 7 and 14 in a cynomolgus monkey treated with 1× and 3× doses of XENP27564. Data demonstrate that 1× and 3× doses produce similar pharmacological activity in monkeys and suggest that maximum effect is achieved at lower doses.

In fig. 72 depicts changes in serum albumin concentration (as an indicator of vascular leakage) in cynomolgus monkeys following administration of 1×, 3×, and 10× doses of XENP27564. Data demonstrate that higher doses of XENP27564 increased toxicity.

In fig. 73 depicts the change in serum C-reactive protein concentration (as an indicator of vascular leakage) in cynomolgus monkeys following administration of 1×, 3×, and 10× doses of XENP27564. The data demonstrate that higher doses of XENP27564 significantly increased toxicity.

In fig. 74 depicts changes in A) sodium concentration, B) chlorine concentration, C) eosinophil content, and D) basophil content in cynomolgus monkeys administered 1×, 2×, and 3× doses of XENP27564. Data demonstrate that higher doses of XENP27564 increased toxicity, while lower doses were well tolerated in cynomolgus monkeys.

In fig. 75 depicts changes in serum concentration levels over time in cynomolgus monkeys administered 1x, 2x, and 3x doses of XENP27564. Data demonstrate sustained pharmacokinetics over several days in cynomolgus monkeys.

In fig. 76 depicts the change in A) eosinophil content and B) basophil content in cynomolgus monkeys administered 3× dose of XENP27563 or 3× dose of XENP27563 on days 0 and 15. Data demonstrate that repeated administration of XENP27564 was well tolerated in cynomolgus monkeys.

In fig. 77 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD44 ^hi cells, CD8 ⁺ CD44 ^hi cells, and Tregs (in splenocytes from B6 mice) by A) XENP26105, B) XENP26109, and C) recombinant human IL-2. Data indicate that engineered IL-2-Fc fusions are equally selective and effective for Tregs in mice, making them suitable for use in preclinical mouse models for the study of autoimmune diseases.

In fig. 78 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells, CD4 ⁺ CD45RA ⁺ T cells, CD8 ⁺ CD45RA ^- T cells, CD8 ⁺ CD45RA ⁺ T cells, CD56 ⁺ NK cells, γδ T cells, and also human Tregs, by A) XENP27563 and B) XENP27564.

In fig. 79 depicts the induction of STAT5 phosphorylation in CD4 ⁺ CD45RA ^- T cells, CD4 ⁺ CD45RA ⁺ T cells, CD8 ⁺ CD45RA ^- T cells, CD8 ⁺ CD45RA ⁺ T cells, CD16 ⁺ NK cells, CD56 ⁺ NK cells, γδ T cells as well as cynomolgus Tregs by A) XENP27563 and B) XENP27564. Data indicate that engineered IL-2-Fc fusions are equally selective and effective for Tregs in cynomolgus monkeys, making them suitable for use in preclinical models in cynomolgus monkeys.

Detailed description of the invention

Introduction

This invention is directed to compositions and methods for engineered IL-2-Fc fusions for the treatment of autoimmune diseases. Autoimmune diseases can be treated using mechanisms that preferentially activate regulatory T cells, commonly referred to as “Treg cells” or “Tregs.” Tregs are a subpopulation of immune T cells that are immunosuppressive and modulate the immune system by maintaining tolerance to self-antigens to prevent autoimmune diseases. Tregs generally suppress the proliferation of effector T cells. Tregs express biomarkers including CD4, FOXP3, and CD25 (CD25 is also known as IL-2Rα protein).

Tregs can be regulated by IL-2, which is required for Tregs to function and survive. Due to the ability of IL-2 to stimulate or reduce the activity of both T cells and Tregs with limited selectivity, there is a significant need in the field for the development of more selective Treg modulators. In addition, as a potential drug, IL-2 has the disadvantage of very rapid clearance, with a half-life measured in minutes, making effective administration difficult. This invention is aimed at solving both of these problems by obtaining new IL-2-Fc fusion proteins.

Thus, the present invention provides IL-2 proteins that are constructed in two different ways. The first is that the IL-2 variants of the present invention preferentially activate CD25+ cells such as Tregs over other T cells that are CD25- to provide increased selectivity for Tregs over other T cells and thus thus leading to the formation of compositions for suppressing immune function and thus facilitating the treatment of autoimmune diseases. Typically, this is accomplished either by increasing binding to IL-2Rα, or by decreasing binding to IL-2Rβ (and/or IL-2Rγ) or the IL-2Rβγ interface, or both.

In addition to the development of selectivity described above, this invention also provides IL-2 proteins that have an increased serum half-life, which is achieved using Fc fusions. In this case, the addition of an Fc domain will result in an increase in the half-life of the IL-2 molecule, as is well known in the art. At the same time, this invention provides two additional methods for increasing serum half-life.

The first involves the FcRn receptor. In IgG, a site in the Fc between the Cγ2 and Cγ3 domains mediates interaction with the neonatal FcRn receptor. Binding to FcRn results in recycling of endocytosed antibody from the endosome back into the bloodstream (Raghavan et al., 1996, Annu Rev Cell Dev Biol 12:181-220; Ghetie et al., 2000, Annu Rev Immunol 18:739-766, both completely incorporated by reference). This process, combined with the elimination of renal filtration due to the large size of the full-length molecule, results in preferential serum half-lives of antibodies ranging from one to three weeks. In order to increase the retention of Fc proteins in vivo, the increase in binding affinity should occur at a level around pH 6, while the maintenance of lower affinity should occur at a level around pH 7.4. Although still under study, Fc regions are thought to have longer half-lives in vivo because binding to FcRn at pH 6 in the endosome results in Fc sequestration (Ghetie and Ward, 1997 Immunol Today. 18( 12):592-598, incorporated by reference in its entirety). The endosomal compartment then recycles Fc to the cell surface. Once the compartment opens to the extracellular space, the higher pH, ~7.4, promotes the release of Fc back into the bloodstream.

Accordingly, increased serum half-life may favor the use of Fc variants, which increase binding to FcRn and, in many cases, result in increased half-life.

An additional method for increasing the serum half-life of an IL-2-Fc fusion molecule is based on pH engineering for recycling from the endosomal sorting pathway. As is known in the art, endocytosis of cytokines such as IL-2 into an endosome results in endocytotic sorting, whereby the cytokine is either degraded or returned back into the bloodstream (see Fallon et al., JBC 275(10):6790, 2000, hereby incorporated herein by reference in its entirety). After internalization into the endosome, IL-2, IL-2Rß and γc are degraded, while IL-2R is constitutively recycled to the cell surface. Since blood pH is approximately 7.2 to 7.4 and endosomal pH is approximately 6, by engineering IL-2 to increase IL-2Rα ligand binding at pH 6, IL-2/IL-2Rα is recycled and does not decompose, which leads to an increase in half-life in the blood serum.

In addition, the activity of the IL-2-Fc fusion molecules of the present invention may also depend on other factors. For example, the present invention provides bivalent IL-2 constructs such as those depicted in FIG. 19B, which produces homodimers of IL-2-Fc fusion variants, thereby resulting in bivalent receptor binding. Alternatively, the present invention provides monovalent IL-2 constructs such as those depicted in FIG. 19A, which produces heterodimers in which one of the monomers is an IL-2-Fc fusion variant and the other is an "empty fragment" Fc monomer. In addition, the presence of additional flexible linkers may in some cases result in increased activity, for example as shown in FIG. 19C for monovalent structures and FIG. 19D for bivalent structures.

Accordingly, the present invention provides engineered IL-2 variants, as well as engineered IL-2-Fc fusion proteins, that exhibit preferential activation of CD25+ cells, such as Tregs, over CD25− T cells, and that exhibit increased half-life in blood serum.

Definitions

For the purpose of a more complete understanding of this application, several definitions are provided below. Such definitions are intended to cover grammatical equivalents.

In this context, “ablation” refers to the reduction or elimination of activity. Thus, for example, “abolishing FcγR binding” means that an amino acid variant of the Fc region exhibits less than 50% of the initial binding compared to an Fc region not containing the specific variant, with a loss of less than 70-80-90-95-98% of activity, which is preferred, and generally has activity below the level of detectable binding in the Biacore assay. Particularly useful in eliminating FcγR binding are those shown in FIG. 4.

The term "ADCC" or "antibody-dependent cell-mediated cytotoxicity" as used herein refers to a cell-mediated reaction in which nonspecific cytotoxic cells that express FcγR recognize bound antibody on a target cell and subsequently cause lysis of the target cell. ADCC correlates with Fc region binding to FcγRIIIa; increased binding to FcγRIIIa leads to increased ADCC activity. As discussed herein, many embodiments of the present invention completely eliminate ADCC activity.

The term "modification" as used herein refers to an amino acid substitution, insertion and/or deletion in a polypeptide sequence, or a change in a moiety chemically linked to a protein. For example, the modification may be an altered carbohydrate or PEG structure attached to the protein. The term "amino acid modification" as used herein refers to an amino acid substitution, insertion and/or deletion in a polypeptide sequence. For clarity, unless otherwise noted, an amino acid modification always refers to the amino acid encoded by DNA, for example, the 20 amino acids that have codons in DNA and RNA.

The term "amino acid substitution" or "substitution" as used herein means the replacement of an amino acid at a specific position in the parent polypeptide sequence with another amino acid. In particular, in some embodiments of the present invention, the substitution is an amino acid that does not occur naturally at the particular position, or does not occur naturally in the body, or in any organism. For example, the S364K substitution refers to a variant polypeptide, in this case an Fc variant, in which the serine at position 364 is replaced by a lysine. The numbering is based on the numbering of the parent polypeptide, for example, R38W in the context of IL-2 numbering. To be clear, a protein that has been engineered to change the nucleic acid coding sequence but not to replace the original amino acid (for example, exchanging CGG (encoding arginine) for CGA (still encoding arginine) to increase expression level in the host) is not "amino acid replacement"; that is, despite the creation of a new gene encoding the same protein, if the protein has the same amino acid at the specific position at which it began, it is not an amino acid substitution.

The term "amino acid insertion" or "insertion" as used herein means the addition of an amino acid sequence at a specific position in the sequence of the parent polypeptide. For example, -233E or 233E indicates a glutamic acid insertion after position 233 and before position 234. Additionally, -233ADE or A233ADE indicates an AlaAspGlu insertion after position 233 and before position 234.

The term "amino acid deletion" or "deletion" as used herein means the removal of an amino acid sequence at a specific position in the sequence of the parent polypeptide. For example, E233- or E233#, E233() or E233del denotes a deletion of glutamic acid at position 233. Additionally, EDA233- or EDA233# denotes a deletion of the GluAspAla sequence that begins at position 233.

The term “protein” as used herein refers to at least two covalently linked amino acids and includes proteins, polypeptides, oligopeptides and peptides. The peptidyl group may contain naturally occurring amino acids and peptide bonds or synthetic peptidomimetic structures, that is, "analogues" such as peptoids (see Simon et al., PNAS USA 89(20):9367 (1992), which is incorporated herein by reference in full). Amino acids can be either naturally occurring or synthetic (eg, not an amino acid encoded by DNA); as will be appreciated by those skilled in the art. Typically, naturally occurring amino acids are used in the proteins of this invention. In addition, variant polypeptides may include synthetic derivatization of one or more side chains or ends, glycosylation, PEGylation, cyclic rearrangement, cyclization, linkers with other molecules, fusion to proteins or domains of proteins, and the addition of peptide tags or tags.

The term "residue" as used herein refers to the position in the protein and its associated amino acid identity. For example, arginine 38 (also called Arg38 or R38) is the residue at position 38 (numbered from the mature sequence) in the human IL-2 protein.

The term “variant protein” or “protein variant” or “variant” as used herein means a protein that differs from the parent protein by at least one amino acid modification. A protein variant may refer to the protein itself, a composition containing the protein, or the amino acid sequence that encodes it. Preferably, the protein variant has at least one amino acid modification compared to the parent protein, for example, from about one to about seventy amino acid modifications, and preferably from about one to about five amino acid modifications compared to the parent protein. As described below, in some embodiments of the present invention, the parent polypeptide, such as the parent Fc polypeptide, is a human wild-type sequence, such as a human IgG1, IgG2, IgG3, or IgG4 Fc region. In the context of IL-2 variants, the parent polypeptide is human IL-2, the mature sequence of which is shown in FIG. 1. The protein sequence variant herein will preferably have at least about 80% sequence identity to the parent protein, and most preferably at least about 90% identity, more preferably at least about 95-98- 99% identical. A protein variant may refer to the protein variant itself, compositions containing the protein variant, or the DNA sequence that encodes it.

The term "Fc" or "Fc region" or "Fc domain" as used herein refers to a polypeptide comprising the constant region of an IgG antibody, excluding the first constant domain of the immunoglobulin domain and, in some cases, all or part of the hinge. In IgG, the Fc domain contains the immunoglobulin domains Cγ2 and Cγ3 (CH2 and CH3) and the hinge region between Cγ1 (CH1) and Cγ2 (CH2). In the context of IgG antibodies, each IgG isotype has three CH regions. Accordingly, the “CH” domains in the context of IgG are as follows: “CH1” refers to positions 118-220 according to the EU index as per Kabat. "CH2" refers to positions 237-340 according to the EU index as per Kabat, and "CH3" refers to provisions 341-447 according to the EU index as per Kabat. Unless otherwise indicated, the Fc domains of the present invention contain a hinge starting at position 216 (EU numbering) and ending at the C-terminus of the CH3 domain at position 447; this is called the “hinge-CH2-CH3” IgG. In some cases, such as in the case of Fc fusions herein, the hinge acts as a linker to the domain described herein. In some embodiments, as more fully described below, amino acid modifications are made to the Fc region, for example, to alter binding to one or more FcγR receptors or to an FcRn receptor or to facilitate heterodimerization of Fc domains.

Accordingly, the term “Fc variant” or “variant Fc” as used herein refers to a protein containing an amino acid modification in the Fc domain. The Fc variants of this invention are defined according to the amino acid modifications that comprise them. Thus, for example, N434S or 434S is an Fc variant with a substituted serine at position 434 relative to the parent Fc polypeptide, numbering corresponding to the EU index. Similarly, M428L/N434S defines an Fc variant with substitutions M428L and N434S relative to the original Fc polypeptide. The identity of the DT amino acid may be uncertain, in which case the above variant is designated 428L/434S. It should be noted that the order in which the substitutions are presented is arbitrary, i.e., 428L/434S is the same Fc variant as M428L/N434S, etc. For all provisions discussed in this invention that are associated with antibodies, unless otherwise indicated, the numbering of amino acid positions corresponds to the EU index. The EU index or EU-index, as in Kabat or the EU numbering scheme, refers to the numbering of EU antibodies (Edelman et al., 1969, Proc Natl Acad Sci USA 63:78-85, which is incorporated herein by reference in its entirety) . The modification may be an addition, deletion or substitution. Substitutions may include naturally occurring amino acids and, in some cases, synthetic amino acids. Examples include US Pat. No. 6,586,207; WO 98/48032; WO 03/073238; US2004-0214988A1; WO 05/35727A2; WO 05/74524A2; J. W. Chin et al., (2002), Journal of the American Chemical Society 124:9026-9027; J. W. Chin, & P. G. Schultz, (2002), ChemBioChem 11:1135-1137; J. W. Chin, et al., (2002), PICAS United States of America 99:11020-11024; and L. Wang & P. G. Schultz (2002), Chem. 1-10, all of which are incorporated herein by reference in their entirety.

The term "interleukin-2" or "IL-2" as used herein refers to human IL-2 having the sequence shown in FIG. 1.

The term “IL-2 variant” or “IL-2 variant” as used herein refers to a protein containing an amino acid modification in the mature human IL-2 sequence depicted in FIG. 1. The IL-2 variants of the present invention are defined according to the amino acid modifications that compose them, as set forth above, using human mature form numbering.

The term "Fc fusion protein" or "immunoadhesin" as used herein means a protein containing an Fc region, typically linked (optionally by a linker moiety as described herein, which may be the hinge region of an IgG such as IgG1 ), with another protein such as IL-2. Thus, an IL-2-Fc fusion protein is a protein comprising IL-2 (in this case, variant IL-2) and the Fc domains set forth herein (in this case also typically Fc variants). They typically have the structure IL-2-hinge-CH2-CH3. As will be appreciated in the art, the two Fc domains will self-assemble to form the dimeric Fc fusion proteins set forth herein.

The term "position" as used herein refers to a location in the sequence of a protein. The provisions may be numbered sequentially or according to a specified format, such as the EU suffix for antibody numbering.

The term “non-naturally occurring modification” as used herein refers to an amino acid modification that is not isotypic. For example, since none of the IgGs contain a serine at position 434, the 434S substitution in IgG1, IgG2, IgG3, or IgG4 (or their hybrids) is considered a non-naturally occurring modification.

The term “amino acid” and “amino acid identity” as used herein refers to one of the 20 naturally occurring amino acids that are encoded by DNA and RNA.

The term "effector function" as used herein refers to a biochemical phenomenon that results from the interaction of the Fc region of an antibody with an Fc receptor or ligand. Effector functions include, but are not limited to, ADCC, ADCP and CDC.

The term "Fc gamma receptor", "FcγR" or "Fc gamma R" as used herein refers to any member of the family of proteins that binds the Fc region of an IgG antibody and is encoded by the FcγR gene. In humans, this family includes, but is not limited to, FcγRI (CD64), including the FcγRIa, FcγRIb, and FcγRIc isoforms; FcγRII (CD32), including isoforms FcγRIIa (including allotypes H131 and R131), FcγRIIb (including FcγRIIb-1 and FcγRIIb-2) and FcγRIIc; and FcγRIII (CD16), including isoforms FcγRIIIa (including allotypes V158 and F158) and FcγRIIIb (including allotypes FcγRIIIb-NA1 and FcγRIIIb-NA2) (Jefferis et al., 2002, Immunol Lett 82: 57-65 , which is incorporated herein by reference in its entirety), as well as any unknown isoforms or allotypes of FcγRs or human FcγRs. FcγR can be derived from any organism, including, but not limited to, humans, mice, rats, rabbits and monkeys. Mouse FcγRs include, but are not limited to, FcγRI (CD64), FcγRII (CD32), FcγRIII (CD16), and FcγRIII-2 (CD16-2), as well as any unknown FcγR or mouse FcγR isoforms or allotypes.

The term "FcRn" or "neonatal Fc receptor" as used herein refers to a protein that binds the Fc region of an IgG antibody and is encoded, at least in part, by the FcRn gene.

The term "parent polypeptide" as used herein refers to the original polypeptide that is subsequently modified to produce a variant. The parent polypeptide may be a naturally occurring polypeptide or a variant, or an engineered variant of a naturally occurring polypeptide. Parent polypeptide may refer to the polypeptide itself, compositions that contain the parent polypeptide, or the amino acid sequence that encodes it. Accordingly, the term “parental IL-2” as used herein means an unmodified human IL-2 protein that has been modified to produce a variant, and the term “parental Fc” or “parental Fc domain” as used herein means refers to an unmodified human IgG Fc domain that has been modified to produce a variant Fc domain.

By “stranded” in the context of heterodimeric Fc fusion monomers of the present invention, it is meant herein that, like two DNA strands that are “matched,” heterodimerization variants are included in each monomer for the purpose of maintaining the ability to “match” to form heterodimers. For example, if some pI variants are constructed into monomer A (for example, making the pI higher), then steric variants, which are "charge pairs" that can also be used, do not affect the pI variants, for example, charge variants that make the pI higher , are placed on the same "chain" or "monomer" in order to retain both functionality. Likewise, for "skewed" variants that comprise pairs of aggregates, as discussed in more detail below, one skilled in the art will consider the pI when deciding which chain or monomer that comprises one aggregate of the pair to be directed to in order to make dividing pI by maximum also using pI distorted variants.

The term “wild type or WT” as used herein means an amino acid sequence or nucleotide sequence that occurs in nature, including allelic variations. The WT protein has an amino acid sequence or nucleotide sequence that has not been intentionally modified.

The proteins of this invention are typically isolated or recombinant. The term “isolated,” when used to describe the various polypeptides described herein, means a polypeptide that has been identified and isolated and/or recovered from the cell or cell culture from which it was expressed. Typically, the isolated polypeptide will be obtained through at least one purification step. The term "recombinant" means that antibodies are produced using recombinant nucleic acid techniques in exogenous host cells.

"Percentage (%) amino acid sequence identity" with respect to a protein sequence is defined as the percentage of amino acid residues in a candidate sequence that are identical to amino acid residues in a particular (parent) sequence after sequence alignment and introduction of gaps, if necessary, to achieve the maximum percentage of sequence identity , and without considering any conservative substitutions as part of the sequence identity. Alignment for the purpose of determining percent amino acid sequence identity can be achieved in a variety of ways known to one skilled in the art, for example, using publicly available computer software such as BLAST, BLAST-2, ALIGN or Megalign (DNASTAR) software. Those skilled in the art can determine appropriate parameters for measuring alignment, including any algorithms necessary to achieve maximum alignment across the sequences to be compared. One particular program is the ALIGN-2 program described in paragraphs [0279]-[0280] of US Publication No. 20160244525, incorporated herein by reference.

The degree of identity between an amino acid sequence of the present invention ("sequence of this invention") and the parent amino acid sequence is calculated as the number of exact matches in the alignment of the two sequences divided by the length of the "sequence of this invention" or the length of the parent sequence, whichever which one is the shortest. The result is expressed as percent identity.

In some embodiments, two or more amino acid sequences are at least 50%, 60%, 70%, 80%, or 90% identical. In some embodiments, two or more amino acid sequences are at least 95%, 97%, 98%, 99%, or even 100% identical.

The term “linker” as used herein refers to a protein linker that is used to connect two other protein domains (eg, a variant IL-2 domain and a variant Fc domain). In some cases, the linker is a "domain linker" used to link any two domains, as defined herein, together. Although any suitable linker can be used, in many embodiments a glycine-serine polymer is used, including, for example, (GS)n, (GSGGS)n, (GGGGS)n and (GGGS)n, where n is an integer of at least one (and typically from 3 to 4 to 5), and any peptide sequence that allows for the recombinant joining of two domains with sufficient length and flexibility so that each domain retained its biological function. In some cases and with attention to "chain presence" as outlined below, charged domain linkers may be used. In addition, the hinge domain of a human IgG1 protein may also constitute a domain linker.

The term “regulatory T cells” or “Treg” as used herein refers to T cells that are CD3+/CD4+/CD8-/CD25+/FOXP3+.

IL-2-Fc fusion proteins of this invention

The present invention provides the (IL-2)-(Fc domain) fusion proteins shown herein and generally in FIG. 19. As will be appreciated by those skilled in the art, the fusion proteins of this invention are actually two different polypeptides that are subject to self-assembly due to the presence of Fc regions or into homodimeric proteins of FIG. 19B, or into heterodimeric proteins (Fig. 19A). The proteins of this invention typically have three distinct domains: an Fc domain, one or more linker domains, and an IL-2 domain.

IL-2 domains of this invention

The IL-2-Fc fusion proteins of the present invention contain IL-2 domains that are variant domains of human IL-2. As discussed herein, these domains are designed to include specificity variants that result in increased activation of CD25+ T cells, such as Tregs, compared to CD25− T cell populations, and optionally also contain amino acids substitutions designed to increase the binding of IL-2 to IL-2Rα at pH 6 so that the IL-2 variant Fc fusion proteins are recycled through the endocytic pathway rather than degraded.

Expressive options

Preliminary IL-2 variants of the present invention also contain a C125S variant, which has previously been shown to increase human IL-2 expression. Thus, unless otherwise indicated, all variants described herein contain the C125S variant; in some cases the C125A variant can also be used.

In addition, in some cases, IL-2 variants of the present invention contain a T3A variant that results in the removal of an O-glycosylation site to reduce complexity.

In addition, the IL-2 variants of the present invention contain additional mutations.

Specificity Options

Accordingly, the present invention provides variant IL-2 proteins with increased specificity for CD25+ T cells, including Tregs. Typically, this is accomplished either by increasing binding to IL-2Rα, or by decreasing binding to IL-2Rβ and IL-2Rγ or the IL-2Rβγ interface, or both.

In one embodiment, the variant IL-2 contains the D20N amino acid substitution, in addition to the C125S expression variant, and thus has the D20N/C125S amino acid set. Of note, the D20N variant has previously been reported to result in loss of binding to both the high-affinity receptor (IL-2Rαβ) and the intermediate-affinity receptor (IL-2Rβ); see Collins et al., PNAS USA 85:7709-7713 (1988), suggesting that “differential binding or activation between IL-2Rβ or IL-2Rαβγ is not achieved by substitution of Asp at position 20,” see US Pat. No. 6955807, Description of the prior art.

In one embodiment, the variant IL-2 contains the T3A amino acid variant, in addition to the D20N amino acid substitution and the C125S expression variant, and thus has the D20N/C125S amino acid set.

In one embodiment, the variant IL-2 contains the T37R amino acid variant, in addition to the T3A amino acid variant, the D20N amino acid substitution, and the C125S expression variant, and thus has the T3A/D20N/T37R/C125S amino acid set.

In one embodiment, the variant IL-2 contains the amino acid variant N71K, in addition to the amino acid variant T3A, the amino acid substitution D20N and the expression variant C125S, and thus has the amino acid set T3A/D20N/N71K/C125S.

In one embodiment, the variant IL-2 contains the amino acid variants N29S/Y31H/K35R/T37A/K48E/N71R/N88D/I89V, in addition to the C125S expression variant, and thus has the amino acid set N29S/Y31H/K35R/T37A /K48E/N71R/N88D/I89V/C125S.

Treg specificity assays

As is known in the art, activation of STAT5 proteins (STAT5a and STAT5b) by phosphorylation is one of the earliest signaling events mediated by IL-2. Thus, by looking at STAT5 phosphorylation in different T cell populations using the constructs of this invention, specificity can be assessed.

Typically, as described in the Examples, STAT5 phosphorylation assays are performed using the methods outlined in the Examples section. As described in the Examples section, typically 5 different cell types are examined, including CD4+/CD45RA+, CD4+/CD45RA-, CD8+CD45RA+, CD8+CD45RA-, and Tregs (CD3+/CD4+/CD8-/CD25+/FOXP3+) for ensure sampling of other T cell types (e.g. CD45RA is expressed on memory T cells rather than naïve T cells).

Typically, there is increased activity compared to wild type IL-2.

pH options

In addition, this invention provides variant IL-2 proteins with increased pH specificity, in which binding at pH 6 (endosomal pH) is increased.

In this embodiment, the IL-2 variant may have one or more amino acid substitutions selected from R38A, R38D, R38E, R38F, R38G, R38H, R38I, R38K, R38L, R38M, R38N, R38P, R38Q, R38S, R38T, R38V, R38W, R38Y, T41A, T41D, T41E, T41F, T41G, T41H, T41I, T41K, T41L, T41M, T41N, T41P, T41Q, T41R, T41S, T41V, T41W, T41Y, F42A, F42D, F42E, F42G, F42G , F42I, F42K, F42L, F42M, F42N, F42P, F42Q, F42R, F42S, F42T, F42V, F42W and F42Y. In addition, these mutations can be combined with C125S.

In this embodiment, the IL-2 variant may have amino acid substitution(s) selected from R38Q/T41K, R38Q/41Q, R38E/T41K, R38Q/T41R, R38N/T41Q, R38Q/T41V, R38N/ T41V, R38Q/T41M, R38Q/T41S, R38Q/T41L, R38N/T41M, T41I/F42Y, T41E/F42Y, T41D/F42Y, T41M/F42Y, T41Q/F42Y, T41E/F42H, T41E/F42L, T41E/F42P , R38Q/F42Y, R38N/T41R, R38N/T41K, R38V/T41R, R38P/T41R, T41E/F42K, T41D/F42K, T41M/F42K, T41Q/F42K, R38Q/F42K, T41I/F42K, R38N/F42K, T41H / F42K, R38Q/T41K/F42Y, R38Q/T41R/F42Y, R38Q/T41Q/F42Y, R38Q/T41V/F42Y, R38N/T41K/F42K, R38Q/T41H/F42K, R38Q/T41K/F42K, R38Q/T41Q/F42K , R38Q/T41V/F42K and R38Q/T41R/F42K. In addition, these mutations can be combined with C125S.

Suitable variants of IL-2

The present invention provides a number of particularly useful IL-2 variants that have the desired activities either alone or when fused to Fc domains, including both wild-type Fc domains and variant Fc domains set forth herein . In addition, these IL-2 variants can be used in monovalent constructs (eg, Fig. 19A) or bivalent constructs (eg, Fig. 19B).

In one embodiment, the variant IL-2 domain has the amino acid substitutions R38I/C125S and is used in a bivalent construct. In this embodiment, the variant IL-2 domain may be fused to a wild-type Fc domain, for example from IgG2 or IgG4. Alternatively, the variant IL-2 domain can be fused to a variant Fc domain, for example, a domain that contains ablative variants and FcRn variants.

In one embodiment, the variant IL-2 domain has the amino acid substitutions R38I/C125S and is used in a monovalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions R38L/C125S and is used in a monovalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions R38L/C125S and is used in a bivalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions D20N/C125S and is used in a bivalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions D20N/C125S and is used in a monovalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/C125S and is used in a bivalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/C125S and is used in a monovalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions N29S/Y31H/K35R/T37A/K48E/N71R/N88D/I89V/C125S and is used in a bivalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions N29S/Y31H/K35R/T37A/K48E/N71R/N88D/I89V/C125S and is used in a monovalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/T37R/C125S and is used in a bivalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/T37R/C125S and is used in a monovalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/N71K/C125S and is used in a bivalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/N71K/C125S and is used in a monovalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/T37R/R38I/C125S and is used in a bivalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/T37R/R38I/C125S and is used in a monovalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/R38I/N71K/C125S and is used in a bivalent construct.

In one embodiment, the variant IL-2 domain has the amino acid substitutions T3A/D20N/R38I/N71K/C125S and is used in a monovalent construct.

Particularly preferred proteins include XENP14142, XENP14144, XENP23833, XENP25720, XENP26086, XENP26105, XENP26987, XENP27003, XENP26109, XENP26994, XENP26841, XENP27004, XENP26839, XEN P26996, XENP26990, XENP27006, XENP26840, XENP27001, XENP26991, XENP27007, XENP27563, XENP26105, XENP27564 and XENP26109.

Fc-2 domains according to this invention

As discussed herein, the present invention provides Fc fusion proteins containing two Fc domains, wherein at least one of the Fc domains containing the IL-2 fusion optionally contains a domain linker. As shown in FIG. 19, the dimeric proteins of the present invention can be configured to have a single variant of IL-2, sometimes referred to herein as "monovalent IL-2", as shown in FIG. 19A, where one of the Fc domains is covalently linked to the IL-2 protein variant and the other is "empty" or "Fc only". This embodiment is based on heterodimeric Fc domains, as discussed below. Alternatively, "bivalent" IL-2 constructs are used, such as those shown in FIG. 19B, where each Fc domain is fused to an IL-2 variant; in these embodiments, homodimeric Fc domains are used, as discussed above.

In both embodiments, whether a homo- or heterodimeric Fc fusion construct is used, the Fc domains typically contain certain specific amino acid variants for multiple functions.

Fc options for additional functionality

In addition to pI amino acid variants, there are a number of useful Fc amino acid modifications that can be made for a variety of reasons, including, but not limited to, altering binding to one or more FcγR receptors, altering binding to FcRn receptors, etc. d.

Accordingly, the proteins of this invention may contain amino acid modifications, including heterodimerization variants set forth herein, which include pI variants and steric variants. Each set of variants may be independently and optionally included or excluded from any particular heterodimeric protein.

FcγR variants

Accordingly, there are a number of useful Fc substitutions that can be made to alter binding to one or more FcγR receptors. Substitutions that result in increased binding as well as decreased binding may be suitable. For example, it is known that increased binding to FcγRIIIa leads to increased ADCC (antibody-dependent cell-mediated cytotoxicity; a cell-mediated reaction in which nonspecific cytotoxic cells that express FcγR recognize bound antibody on a target cell and subsequently cause lysis of the target cell) . Likewise, in some cases, reducing binding to FcγRIIb (inhibitory receptor) may also be useful. Amino acid substitutions that find use in this invention include those listed in USSN 11/124620 (especially in FIG. 41), 11/174287, 11/396495, 11/538406, all of which are expressly incorporated herein by reference in full and specifically for the options described therein. Specific embodiments that find use include, but are not limited to, 236A, 239D, 239E, 332E, 332D, 239D/332E, 267D, 267E, 328F, 267E/328F, 236A/332E, 239D/332E/330Y, 239D, 332E/330L, 243A, 243L, 264A, 264V and 299T.

In addition, there are additional Fc substitutions that benefit from increased FcRn receptor binding and increased serum half-life, as specifically disclosed in USSN 12/341769, incorporated herein by reference in its entirety, including but not limited to limited to 434S, 434A, 428L, 308F, 259I, 428L/434S, 428L/434A, 259I/308F, 436I/428L, 436I or V/434S, 436V/428L and 259I/308F/428L.

Ablative options

Likewise, another category of functional variants are “FcγR ablative variants” or “Fc knockout variants (FcKO or KO).” In these embodiments, for certain therapeutic applications, it is desirable to reduce or eliminate the normal binding of the Fc domain to one or more or all Fcγ receptors (eg, FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) to avoid additional mechanisms of action. That is, for example, it is desirable to eliminate binding to FcγRIIIa to avoid or significantly reduce ADCC activity such that one of the Fc domains contains one or more ablative Fcγ receptor variants. These ablative options are depicted in FIG. 4, and each may be independently and optionally included or excluded, with preferred aspects employing ablative options selected from the group consisting of G236R/L328R, E233P/L234V/L235A/G236del/S239K, E233P/L234V/L235A/G236del /S267K, E233P/L234V/L235A/G236del/S239K/A327G, E233P/L234V/L235A/G236del/S267K/A327G and E233P/L234V/L235A/G236del. It should be noted that the ablative variants mentioned herein eliminate FcγR binding, but generally not FcRn binding.

Homodimeric Fc domains

In some embodiments, the present invention provides bivalent homodimeric proteins containing homodimeric Fc domains, as generally depicted in FIG. 19V. In this embodiment, each monomer is identical and typically contains a variant IL-2 linker-Fc domain, where the linker is typically an IgG1 hinge.

In this embodiment, the Fc domains may have ablative variants, typically depicted in FIG. 4. Suitable ablative options are depicted in FIG. 4, which typically lead to the elimination of binding to FcRI, FcγRIIb and FcγRIIIa. Particularly useful in this embodiment of IgG1 is the ablative amino acid set E233P/L234V/L235A/G236_/S267K (“G236_” is the deletion described herein).

In addition, there are additional Fc substitutions that benefit from increased FcRn receptor binding and increased serum half-life, as specifically disclosed in USSN 12/341769, incorporated herein by reference in its entirety, including but not limited to limited to 434S, 434A, 428L, 308F, 259I, 428L/434S, 428L/434A, 259I/308F, 436I/428L, 436I or V/434S, 436V/428L and 259I/308F/428L.

Heterodimeric Fc domains

As a complement to homodimeric bivalent IL-2 fusion proteins, alternative embodiments employ monovalent IL-2 fusion proteins in which one of the Fc domains is "empty" and the present invention relies on heterodimerization variants to combine the two Fc domains depicted in fig. 19V. These embodiments rely on the use of two different variant Fc sequences that will result in self-assembly to form heterodimeric Fc domains and heterodimeric Fc fusion proteins.

Heterodimeric protein constructs rely on the self-assembling nature of the two Fc domains of antibody heavy chains, for example, two “monomers” that assemble into a “dimer.” Heterodimeric proteins are produced by changing the amino acid sequence of each monomer, as discussed in more detail below. Thus, the present invention is generally directed to the creation of heterodimeric Fc fusion proteins that are based on amino acid variants in constant regions that differ on each chain to promote the formation of heterodimers and/or to allow easier purification of heterodimers compared to homodimers.

There are a number of mechanisms that can be used to create the heterodimers of this invention. In addition, as those skilled in the art will appreciate, these mechanisms can be combined to provide high heterodimerization. Thus, amino acid variants that result in heterodimers are called “heterodimerization variants.” As discussed below, heterodimerization variants can include steric variants (e.g., "knobs and valleys" or "distorted" variants, described below, and "charge pair" variants, described below) as well as "pI variants", allowing the purification of homodimers from heterodimers. As described generally in WO2014/145806, incorporated herein by reference in its entirety, and specifically as described below for the discussion of “heterodimerization options,” suitable heterodimerization mechanisms include “knobs and valleys” (“KIH”; sometimes referred to in herein as "distorted" variants (see discussion in WO2014/145806), "electrostatic interaction" or "charge pairs" described in WO2014/145806, pI variants described in WO2014/145806, and general additional Fc variants described in WO2014/145806 and below.

In this invention, there are several basic mechanisms that can lead to easier purification of heterodimeric proteins; each is based on the use of pI variants such that each monomer has a different pI value, facilitating isoelectric purification of dimeric proteins A-A, A-B and B-B. In addition, the heterodimeric proteins of the present invention also promote size-based separation. As further discussed below, it is also possible to “distort” the formation of heterodimers compared to homodimers. Thus, the combination of steric heterodimerization variants and pI variants or charge pair variants finds particular use in the present invention.

In general, embodiments that are particularly used in this invention are based on sets of variants that include distorted variants that promote heterodimerization formation over homodimerization formation, in combination with pI variants that increase the pI difference between the two monomers.

In addition, as discussed in more detail herein, pI variants may be contained either within constant domains and/or Fc domains of the monomer, or charged linkers, such as domain linkers, may be used.

In the present invention, in which the pI value is used as a separation mechanism for the purification of heterodimeric proteins, amino acid variants can be introduced into one or both monomeric polypeptides; i.e., the pI value of one of the monomers (referred to herein as “monomer A” for simplicity) may be set to be away from monomer B, or both monomer A and monomer B may be changed, with this increases the pI value of monomer A and decreases the pI value of monomer B. As discussed, changes in the pI of one or both monomers can be made by removing or adding a charged residue (e.g., a neutral amino acid is replaced by a positively or negatively charged amino acid residue, e.g., glycine is replaced by glutamine acid), resulting in a change in a charged residue from a positive or negative charge to the opposite charge (e.g., aspartic acid is replaced by lysine) or a change in a charged residue to a neutral residue (e.g., loss of charge; lysine is replaced by serine). Several of these options are depicted in the shapes section.

Accordingly, this embodiment of the present invention provides for making a sufficient pI change in at least one of the monomers such that heterodimers can be separated from homodimers. As will be appreciated by those skilled in the art and as discussed below, this can be accomplished by using a "wild type" heavy chain constant region and a variant region that has been designed to increase or decrease its pI value (wild type A-+B or A - -B wild type), or by increasing it in one area and decreasing it in another area (A+ -B- or A- B+).

Thus, in general, a component of some embodiments of the present invention are amino acid variants in Fc domains that are directed to alter the isoelectric point (pI) of at least one, if not both, monomers of a dimeric protein by including amino acid substitutions (“pI variants” or “pI substitutions”) into one or both monomers. As shown herein, separation of heterodimers from two homodimers can be achieved if the pI values of the two monomers differ by only 0.1 pH unit, all of 0.2, 0.3, 0.4 and 0.5 or more find use in the present invention.

Heterodimerization variants

The present invention provides heterodimeric proteins, including heterodimeric antibodies in various formats, which use heterodimeric variants to promote heterodimer formation and/or purification from homodimers. Several heterodimerization variants are depicted in FIG. 2.

There are a number of suitable pairs of heterodimerization distorted heterodimerization variant sets. These options are included in “pairs” of “sets”. Those. one set of the pair is included in the first monomer, and the other set of the pair is included in the second monomer. It should be noted that these sets do not necessarily behave as peaks-in-valleys variants, with a one-to-one correspondence between a residue in one monomer and a residue in another; those. These pairs of sets form an interface between the two monomers that promotes the formation of heterodimers and discourages the formation of homodimers, causing the percentage of heterodimers that form spontaneously under biological conditions to be greater than 90% rather than the expected 50% (25% homodimer A/ A:50% heterodimer A/B:25% homodimer B/B).

Steric variations

In some embodiments, the formation of heterodimers can be facilitated by the addition of steric variations. Those. As a result of the variation of amino acids in each heavy chain, different heavy chains are more likely to be associated with the formation of a heterodimeric structure than with the formation of homodimers with the same Fc amino acid sequences. Suitable steric variations are included in the figure section.

One mechanism is commonly referred to in the art as “knobs in trenches,” referring to the design of amino acids that results in the creation of steric influences that promote the formation of heterodimers and prevent the formation of homodimers; it is sometimes called "ridges in valleys" as described in USSN 61/596846, Ridgway et al., Protein Engineering 9(7):617 (1996); Atwell et al., J. Mol. Biol. 1997 270:26; US Pat. No. 8,216,805, all of which are incorporated herein by reference in their entirety. The figure section identifies several monomer A–monomer B pairs that are based on the ridge-in-groove mechanism. In addition, as described in Merchant et al., Nature Biotech. 16:677 (1998), these ridge-in-hole mutations can combine with disulfide bonds to form distorted heterodimerization variants.

An additional mechanism that finds use in the creation of heterodimers is sometimes called “electrostatic influence”, as described in Gunasekaran et al., J. Biol. Chem. 285(25):19637 (2010), incorporated herein by reference in its entirety. These are sometimes referred to herein as "charge pairs". In this embodiment, electrostatic forces are used to distort the formation of variants towards heterodimerization. Those skilled in the art will appreciate that these can also affect pI and therefore purification, and thus may also be considered pI variants in some cases. However, because they were designed to enhance heterodimerization and were not used as purification tools, they are classified as “steric variants.” These include, but are not limited to, D221E/P228E/L368E paired with D221R/P228R/K409R (e.g., they are "monomer matching sets") and C220E/P228E/368E paired with C220R/E224R/P228R/K409R .

Additional variants of monomers A and monomers B, which can be combined with other variants, optionally and independently in any amount, such as pI variants described herein, or other steric variants, which are shown in Fig. 37 US 2012/0149876, the figure and legend and SEQ ID NO of which are expressly incorporated herein by reference.

In some embodiments, the steric variants set forth herein may be optionally and independently incorporated with any pI variant (or other variants, such as Fc variants, FcRn variants, etc.) into one or both monomers and may be independently and optionally included or excluded from the proteins of this invention.

A list of suitable distorted variants is shown in FIG. 2, which depicts some pairs particularly suitable for use in many embodiments. Particularly useful in many embodiments are pairs of sets, including, but not limited to, S364K/E357Q: L368D/K370S; L368D/K370S : S364K; L368E/K370S : S364K; T411T/E360E/Q362E : D401K; L368D/K370S : S364K/E357L, K370S : S364K/E357Q and T366S/L368A/Y407V : T366W (optionally containing a disulfide linker, T366S/L368A/Y407V/Y349C : T366W/S354C). In terms of nomenclature, the pair "S364K/E357Q: L368D/K370S" means that one of the monomers has a set with the double variant S364K/E357Q, and the other has a set with the double variant L368D/K370S; as stated above, the "chain presence" of the specified pairs depends on the initial pI value.

pI options (options with isoelectric point) for heterodimers

In general, as will be appreciated by those skilled in the art, there are two general categories of pI variants: variants that result in an increase in the pI value of the protein (basic changes), and variants that result in a decrease in the pI value of the protein (acidic changes). As described herein, all combinations of these options can be performed: one monomer can be wild type or a variant that does not exhibit a significantly different pI value from the wild type, and the other can be either more basic or more acidic. Alternatively, each monomer is modified, one to be more basic and one to be more acidic.

Preferred combinations of pI variants are depicted in FIG. 5. As set forth herein and depicted in the figure section, these changes are shown relative to IgG1, but all isotypes can be changed in the same manner as isotype hybrids. In the case where the heavy chain constant domain is derived from IgG2-4, R133E and R133Q can also be used.

In one embodiment, a preferred combination of pI variants has one monomer (negative side) containing the 295E/384D/418E/421D variants (Q295E/N384D/Q418E/N421D in relation to human IgG1), and a second monomer (positive side) containing the positively charged scFv linker containing (GKPGS) ₄ .

Isotypic variants

In addition, some embodiments of the present invention rely on the "import" of pI amino acids at certain positions from one IgG isotype to another, thereby reducing or eliminating the possibility of introducing unwanted immunogenicity into these embodiments. Some of them are shown in Fig. 21 US Publication No. 2014/0370013, incorporated herein by reference. Those. IgG1 is a common isotype for therapeutic antibodies for a number of reasons, including its high effector function. At the same time, the constant region of the heavy chain of IgG1 has a higher pI value than that of IgG2 (8.10 compared to 7.31). By introducing IgG2 residues at specific positions into the IgG1 scaffold, the pI value of the resulting monomer is reduced (or increased) and further exhibits a longer serum half-life. For example, IgG1 has glycine (pI 5.97) at position 137, and IgG2 has glutamic acid (pI 3.22); import of glutamic acid will influence the pI value of the resulting protein. As described below, several amino acid substitutions are typically required to significantly affect the pI of a variant antibody. At the same time, it should be noted that, as discussed below, even changes in IgG2 molecules contribute to an increase in serum half-life.

In other embodiments, non-isotypic amino acid substitutions are made to either reduce the overall charged state of the resulting protein (e.g., by changing an amino acid with a higher pI value to an amino acid with a lower pI value), or to provide accommodations in structure for the purpose of stability, etc. etc., as described in more detail below.

In addition, by pI engineering of the constant domains of both heavy and light chains, significant changes can be observed in each monomer of the heterodimer. As discussed herein, having pI values of two monomers that differ by at least 0.5 can facilitate separation by ion exchange chromatography or isoelectric focusing, or other isoelectric point sensitive techniques.

Calculation of pI value

The pI value of each monomer may depend on the pI domain of the variant IL-2 and Fc domain set forth herein. Thus, in some embodiments, the pI change is calculated based on the variant heavy chain constant domain using the diagram in FIG. 19 US Publication 2014/0370013. As discussed herein, the monomer to be designed is typically determined by the intrinsic pI value of the Fv and framework regions. Alternatively, the pI value of each monomer can be compared.

Combination of heterodimeric variants and Fc variants

As will be appreciated by those skilled in the art, for use in heterodimeric Fc domains, all of the listed heterodimerization variants (including distorted and/or pI variants) can optionally and independently be combined in any manner, provided that they retain their "presence" chains" or "separation of monomers". In addition, all of these options can be combined into any of the heterodimerization formats.

In the case of pI variants, although embodiments of particular use are depicted in the figure section, other combinations can be created based on the basic rule of changing the pI difference between two monomers to facilitate purification.

In addition, any of heterodimerization variants, distorted variants, and pI variants are also independently and optionally combined with ablative Fc variants, Fc variants, FcRn variants, as generally set forth herein.

In addition to ablative variants, Fc domains typically also include the C220S variant (e.g., since the Fc domains of this invention contain a hinge region beginning at position 216, EU numbering), which is excluded because it does not include light chains, and said cysteine is used to form the light chain disulfide.

In addition, the Fc domains of the fusion proteins of the present invention may optionally contain amino acid substitutions with half-life extension.

It has recently been suggested that antibodies with variable regions that have lower isoelectric point values may also have a longer serum half-life (Igawa et al., 2010 PEDS. 23(5): 385-392, included in full volume by reference). At the same time, the mechanism of this still remains unclear. In addition, the variable regions differ from antibody to antibody. Variants of the constant regions with a reduced pI value and an increased half-life will provide a more modular approach to improve the pharmacokinetic properties of the antibodies described herein.

Suitable structures of this invention

As set forth herein, the present invention provides a number of useful monovalent and bivalent constructs.

Heterodimeric monovalent constructs

In some embodiments, the IL-2-Fc fusion proteins of the present invention are heterodimeric monovalent constructs such as those depicted in FIG. 19A and 19C. In this embodiment, the variant IL-2 domain is typically fused to the Fc domain of the human IgG1 variant, either using a hinge as a domain linker (typically containing the C220S variant) or an additional linker attached to the hinge, wherein a different Fc -domain (containing the hinge) remains “empty”.

In some embodiments, the variant IL-2 domain is attached to the "+" side of the monomer (see Fig. 5A), which contains the Fc domain of the variant human IgG1 (including the hinge with the C220S variant), the "S364K/E357Q" distorted variants. and ablative variants E233P/L234V/L235A/G236del/S267K, and the “empty Fc side” is the Fc domain of variant human IgG1 (including the hinge with the C220S variant), “distorted variants” L368D/K370S and ablative variants E233P/ L234V/L235A/G236del/S267K. In this embodiment, preferred constructs comprise a variant IL-2 domain with amino acid substitutions selected from the group consisting of T3A/D20N/T37R, T3A/D20N/T37R/C125S, T3A/D20N/T37R/C125A, T3A/D20N/N71K, T3A/D20N/N71K/C125S and T3A/D20N/N71K/C125A.

In some embodiments, the variant IL-2 domain is attached to the "+" side of the monomer (see Fig. 5A), which contains the Fc domain of the variant human IgG1 (including the hinge with the C220S variant), the "S364K/E357Q" distorted variants. and ablative variants E233P/L234V/L235A/G236del/S267K, as well as FcRn variants M428L/N434S, and the "empty Fc side" represents the Fc domain of variant human IgG1 (including the hinge with the C220S variant), "distorted variants" L368D/K370S and ablative variants E233P/L234V/L235A/G236del/S267K, as well as FcRn variants M428L/N434S. In this embodiment, preferred constructs comprise a variant IL-2 domain with amino acid substitutions selected from the group consisting of T3A/D20N/T37R, T3A/D20N/T37R/C125S, T3A/D20N/T37R/C125A, T3A/D20N/N71K, T3A/D20N/N71K/C125S and T3A/D20N/N71K/C125A.

In one embodiment, "monomer 1" comprises a variant IL-2 domain containing amino acid substitutions T3A/D20N/N71K/C125S (compared to wild-type IL-2, SEQ IN NO: 2), and "Fc backbone monomer 1" , selected from those in FIG. 6, including SEQ ID NO: 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31 or 33. In this embodiment, "monomer 2" contains an "empty Fc" selected from "Fc-framework monomer 2" from SEQ ID NOs: 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32 and 34.

In one embodiment, "monomer 1" comprises a variant IL-2 domain containing amino acid substitutions T3A/D20N/N71K/C125S (compared to wild-type IL-2, SEQ IN NO: 2), and "Fc backbone monomer 1" , selected from those in FIG. 6, including SEQ ID NO: 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31 or 33, but with FcRn variants M428L/N434S. In this embodiment, "monomer 2" contains an "empty Fc" selected from "monomer 2 Fc frameworks" from SEQ ID NO: 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32 and 34, but with FcRn M428L/N434S variants.

In one embodiment, "monomer 1" contains a variant IL-2 domain containing amino acid substitutions T3A/D20N/T37R/C125S (compared to wild-type IL-2, SEQ IN NO: 2), and "Fc backbone monomer 1" , selected from those in FIG. 6, including SEQ ID NO: 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31 or 33. In this embodiment, "monomer 2" contains an "empty Fc" selected from "Fc-framework monomer 2" from SEQ ID NOs: 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32 and 34.

In one embodiment, it contains a variant IL-2 domain containing amino acid substitutions T3A/D20N/T37R/C125S (compared to wild-type IL-2, SEQ IN NO: 2), and an "Fc framework" selected from those in FIG. . 6, including SEQ ID NO: 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31 or 33, but with FcRn variants M428L/N434S. In this embodiment, "monomer 2" contains an "empty Fc" selected from "monomer 2 Fc frameworks" from SEQ ID NO: 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32 and 34, but with FcRn M428L/N434S variants.

In these embodiments, preferred designs include XENP26105, XENP27563, XENP26109 and XENP27564.

Homodimeric bivalent constructs

In some embodiments, the IL-2-Fc fusion proteins of the present invention are homodimeric bivalent constructs such as those depicted in FIG. 19B and 19D. In this embodiment, each of the domains of the variant IL-2 is typically fused to the Fc domain of the human IgG1 variant, either by using a hinge as a domain linker (typically containing the C220S variant) or by using an additional linker attached to the hinge.

In some embodiments, the IL-2 variant domains are fused to the human IgG1 variant Fc domain (including the hinge with the C220S variant) and the E233P/L234V/L235A/G236del/S267K ablative variants. In this embodiment, preferred constructs comprise a variant IL-2 domain with amino acid substitutions selected from the group consisting of T3A/D20N/T37R, T3A/D20N/T37R/C125S, T3A/D20N/T37R/C125A, T3A/D20N/N71K, T3A/D20N/N71K/C125S and T3A/D20N/N71K/C125A.

In some embodiments, the IL-2 variant domains are fused to the human IgG1 variant Fc domain (including the hinge with the C220S variant), the E233P/L234V/L235A/G236del/S267K ablative variants, and the M428L/N434S FcRn variants. In this embodiment, preferred constructs comprise a variant IL-2 domain with amino acid substitutions selected from the group consisting of T3A/D20N/T37R, T3A/D20N/T37R/C125S, T3A/D20N/T37R/C125A, T3A/D20N/N71K, T3A/D20N/N71K/C125S and T3A/D20N/N71K/C125A.

Nucleic acids according to this invention

The present invention further provides nucleic acid compositions encoding homodimeric bivalent IL-2-Fc fusion proteins and heterodimeric monovalent IL-2 fusion proteins.

As will be appreciated by those skilled in the art, nucleic acid compositions will depend on the format of the heterodimeric protein. Thus, for example, when a format requires two amino acid sequences, such as in the case of the heterodimeric monovalent formats of FIG. 19A, the two nucleic acid sequences may be included in one or more expression vectors for expression.

Alternatively, when using the homodimeric bivalent formats depicted in FIG. 19B, one nucleic acid construct and one expression vector are used.

As is known in the art, nucleic acids encoding components of the present invention can be incorporated into expression vectors known in the art, depending on the host cells used to produce the heterodimeric Fc fusion proteins of the present invention. As a rule, nucleic acids are functionally linked to any number of regulatory elements (promoters, replication points, selectable markers, ribosome binding sites, inducers, etc.). Expression vectors can be extrachromosomal or integration vectors.

The nucleic acids and/or expression vectors of the present invention are then transformed into any number of different types of host cells, as is well known in the art, including mammalian, bacterial, yeast, insect and/or fungal cells, wherein mammalian cells ( for example, CHO cells) find use in many embodiments.

In some embodiments, each of the nucleic acids encoding each monomer, depending on the format, is contained in a single expression vector, typically under the control of different or the same promoters. In embodiments with particular applicability in this invention, each of these two or three nucleic acids is contained in different expression vectors.

The heterodimeric Fc fusion protein of this invention is prepared by culturing host cells containing expression vector(s) as is well known in the art. Once produced, conventional purification steps of the fusion proteins or antibodies are performed, including an ion exchange chromatography step. As discussed herein, having pI values of two monomers that differ by at least 0.5 can facilitate separation by ion exchange chromatography or isoelectric focusing, or other isoelectric point sensitive techniques. Those. inclusion of pI substitutions that change the isoelectric point (pI) of each monomer such that each monomer has a different pI value and the heterodimer also has a different pI value, thus facilitating isoelectric purification of the heterodimer (eg, anion exchange columns, cation exchange columns). These substitutions also facilitate the detection and monitoring of subsequent purification of any contaminating homodimers (e.g., IEF gels, cIEF gels, and IEX analytical columns).

Compositions

The formulations of the compositions used in accordance with this invention are prepared for storage by mixing fusion proteins having the required purity with optional pharmaceutically acceptable carriers, excipients or stabilizers (as generally set forth in Remington's Pharmaceutical Sciences 16th edition, Osol, A. Ed. . [1980]), in the form of lyophilized formulations or aqueous solutions.

Treatment

The IL-2-Fc fusion protein compositions of the present invention find use in the treatment of an autoimmune disease, for example, by activating a patient's CD25+ cells using the dimeric proteins of the present invention.

Examples

Example 1: Design of IL-2 to Increase Half-Life

As discussed above, IL-2 has the disadvantage of very rapid clearance. This rapid clearance is due in part to the internalization of the IL-2:IL-2R complex. After internalization into the endosome, IL-2, IL-2Rß and γc are degraded, while IL-2R is constitutively recycled to the cell surface. The 2D1 variant of IL-2, which has the L18M/L19S mutations, demonstrated a longer half-life than wild-type IL-2. Fallon et al. reported that wild-type IL-2 has lower affinity for IL-2R at endosomal pH, indicating a pH-dependent binding between IL-2 and IL-2Rα. In addition, this group found that the increased half-life of 2D1 results from recycling with IL-2Rα due to the fact that 2D1 has a higher affinity for IL-2Rα at endosomal pH than IL-2 WT. Notably, residues L18 and L19 in IL-2 have been described to bind to IL-2β and γc. In contrast, it was hypothesized that modifying residues at the IL-2:IL-2Rα interface would be more suitable to attenuate pH-dependent binding, increase recycling of IL-2 from IL-2Rα, and thereby increase half-life.

Engineering IL-2 at the IL-2/IL-2Rα interface to attenuate pH-dependent binding

By examining the crystal structure of the IL-2/IL-2Rα interface (PDB code 2ERJ), a likely natural “pH switch” at the IL-2/IL-2Rα interface was discovered, consisting of an interaction between Arg38, Thr41 and/or Phe42 IL-2 and His120 IL-2Rα (see Fig. 8). It was hypothesized that at low pH in the endosome, following internalization of the IL-2:IL-2Rαßγ complex, His20 would become protonated, leading to the release of IL-2 from IL-2Rα and subsequent lysosomal degradation of the remaining IL-2:IL-2 complex. 2Рßγ. IL-2R is known to recycle to the cell surface, and increasing the binding affinity of IL-2 to IL-2Rα at pH 6.0 by replacing one or more of these interacting residues may increase IL-2 recycling and prolong the half-life. Computational predictions using Protein Design Automation technology (see, for example, WO 1998/047089, published October 22, 1998) generated saturation options for these three contacting residues.

The plasmid encoding IL-2 was constructed using standard gene synthesis followed by subcloning into the expression vector pTT5. IL-2 contained a C-terminal polyhistidine tag (8xHis) for purification and a C125S substitution for increased expression. Substitutions predicted above were introduced using standard mutagenesis techniques. Proteins were produced by transient transfection in HEK293E cells and purified by Ni-NTA chromatography. Sequences for exemplary embodiments are depicted in FIG. 9, in which case the polyhistidine tag was removed.

The affinity of IL-2 for IL-2Rα at pH 7.4 and pH 6.0 was determined using Biacore, a surface plasmon resonance (SPR) technology. Experimental steps for Biacore typically included the following: immobilization (capturing the ligand on the sensor chip); association (passing different concentrations of the analyte through the sensor chip); and dissociation (passing a buffer through sensor chips) to determine the affinity of the drugs under study. The buffer-only reference stream was also included in the background correction method for data processing. In this particular screen, human CD25 (IL-2Rα) was first captured on the sensor chips, and then different concentrations of IL-2 variants were passed through the sensor chips. Separate experiments were performed using buffer at pH 7.4 to simulate cell surface conditions and using buffer at pH 6.0 to simulate endosomal conditions. The resulting dissociation constants ( _KD ), association rates ( _ka ) and dissociation rates ( _kd ) are depicted in FIG. 10 and 11. Binding affinities and kinetic rate constants were obtained by analyzing the processed data using a 1:1 binding model. In fig. 12 shows the fold increase in dissociation rate at pH 6.0, while FIG. Figure 13 shows the fold increase in the dissociation rate at pH 7.4.

A large number of variants with different substitutions at R38, T41 and F42 had worse or similar dissociation rates at pH 6.0 compared to IL-2 WT; however, some variants with significantly increased dissociation rates have been identified (i.e., IL-2 variants that are more likely to recycle with IL-2Rα). Preferred options include XENP14142 (R38I) and XENP14144 (R38L). In fig. 14 shows Biacore sensorgrams for XENP14135 (wild-type IL-2 with the C125S mutation) and XENP14142 (variant IL-2 with R38I and C125S). Similar dissociation curves for XENP14142 at pH 7.4 and pH 6.0 demonstrate successful attenuation of pH-dependent binding.

Attenuation of pH-dependent binding in the context of other IL-2 variants

R38L was then combined with a prior art IL-2 variant (mutant 2-4 with Q126T, described in WO 2009/061853 published May 14, 2009) to examine whether dissociation rates were reduced in the context of other IL-2 variants.

As stated above, substitutions were introduced using standard mutagenesis techniques. Proteins were produced by transient transfection in HEK293E cells and purified by Ni-NTA chromatography. Sequences for the prior art variant described with and without R38L are depicted in FIG. 15, in which case the polyhistidine tag was removed.

The affinity of IL-2 for IL-2Rα at pH 7.4 and pH 6.0 was determined using Biacore as described in Example 1A. The resulting dissociation constants ( _KD ), association rates ( _ka ) and dissociation rates ( _kd ), as well as the ratio of the affinity at pH 7.4 to the affinity at pH 6.0 are depicted in FIG. 16. The data demonstrate that the affinity ratio increased with the inclusion of the R38L substitution, indicating successful attenuation of pH-dependent binding.

Example 2: Design of IL-2 for Treg Selectivity

Proliferative signaling by IL-2 is mediated by IL-2Rß and γc as part of the intermediate-affinity IL-2 receptor complex (IL-2Rßγ) or as part of the high-affinity IL-2 receptor complex (IL-2Rαßγ). CD25 provides high binding affinity for the IL-2Rαβγ complex to IL-2, but is otherwise insufficient for signaling on its own. As a result of its high binding affinity to the IL-2Rαβ complex, IL-2 has a positive effect on Tregs that constitutively express IL-2Rα. Thus, it was hypothesized that the increased affinity of IL-2 for IL-2Rα could lead to an additional binding bias in favor of the IL-2Rαßγ complex on Tregs. Alternatively, decreased affinity of IL-2 for IL-2Pß, γc, or IL-2Pßγ could result in distorted binding to CD25-negative T cells and NK cells.

By studying the crystal structure of the interface between IL-2 and its receptors, as well as by modeling using MOE software, residues that could be substituted were predicted to increase the affinity of IL-2 for IL-2Rα or decrease the affinity of IL-2 for IL-2Rα. IL-2Рß, γc and/or IL-2Рßγ.

Plasmids encoding IL-2 were constructed using standard gene synthesis followed by subcloning into the pTT5 expression vector. IL-2 contained a C-terminal polyhistidine tag (8xHis) for purification and a C125S substitution for increased expression. Substitutions predicted above were introduced using standard mutagenesis techniques. Proteins were produced by transient transfection in HEK293E cells and purified by Ni-NTA chromatography. Sequences for exemplary embodiments are depicted in FIG. 17, in which case the polyhistidine tag was removed.

The binding of IL-2 to its receptor components was determined using Octet, a method based on biolayer interferometry (BLI). Experimental steps for Octet typically included the following: immobilization (capturing the ligand into the biosensor); association (immersion of ligand-coated biosensors into wells containing serial dilutions of the analyte); and dissociation (returning the biosensors to a well containing a buffer) to determine the affinity of the drugs under study. A reference well containing only buffer was also included in the background correction method for data processing. Specifically, anti-human Fc (AHC) biosensors have been used to capture the bivalent CD25(IL-2Pα)-Fc fusion, the bivalent CD122 (IL-2Pß)-Fc fusion, or the heterodimeric CD122:CD132 (IL-2Pßγ)-Fc fusion and immersed in several concentrations of IL-2 variants. The resulting BLI response for IL-2 variants was normalized to the BLI response for XENP14135 (wild-type IL-2 with C125S) and plotted in FIG. 18. Notably, some substitutions at the IL-2:IL-2β interface, such as D20N and N88D, significantly reduced or eliminated the binding of IL-2 to IL-2β.

Example 3 IL-2-Fc Fusion Proteins

To further explore the issue of the short half-life of IL-2, IL-2 was prepared as Fc fusions (hereinafter referred to as IL-2-Fc fusions) to facilitate production and promote FcRn-mediated recycling of the complex and prolong the half-life.

Preparation of IL-2-Fc fusions

The plasmid encoding IL-2 was constructed using standard gene synthesis followed by subcloning into the expression vector pTT5 containing Fc fusion partners (eg, constant regions depicted in Fig. 6). IL-2 may contain a C125S substitution to increase expression and a T3A substitution to remove the O-glycosylation site. Schemes of exemplary IL-2-Fc fusion formats are depicted in FIG. 19. Selected substitutions were introduced using standard mutagenesis techniques.

The monovalent format of IL-2-Fc or "monovIL-2-Fc" (Fig. 19A) contains IL-2 fused to the N-terminus of the first heterodimeric Fc region (see, for example, IL-2-Fc framework 1 - monomer 2, in Fig. 6), while the other side of the molecule is a heterodimeric Fc of the “Fc only” type or the “empty Fc” type (see, for example, framework 1 IL-2-Fc 1 - monomer 1, in Fig. 6 ). Sequences for exemplary monovIL-2-Fc fusions are depicted in FIG. 20. The bivalent format of IL-2-Fc or "biIL-2-Fc" (Fig. 19B) contains IL-2 fused to the N-terminus of the homodimeric Fc region (see, for example, IL-2-Fc framework 12, in Fig. 6). Sequences for exemplary biIL-2-Fc fusions are depicted in FIG. 23. MonovIL-2-Fc fusions and biIL-2-Fc fusions may have a variable length linker between the C terminus of IL-2 and the N terminus of the Fc region (see FIG. 7 for non-limiting examples of domain linkers that may find use in IL-2-Fc fusions, and image formats in Fig. 19C-D). Sequences for exemplary IL-2-Fc fusions with variable length linkers are depicted in FIG. 25.

Proteins were produced by transient transfection in HEK293E cells and purified using a two-step purification method involving protein A chromatography and anion exchange chromatography.

Investigation of Prior Art IL-2 Variants Constructed as MonovIL-2-Fc Fusions

To study the reliability and effectiveness of the monovIL-2-Fc fusion format, a number of prior art IL-2 variants were prepared in this format. These IL-2-Fc fusions include XENP24637 (based on the variant described in WO 2012/107417 published on August 16, 2012), XENP24638 (based on the 2-4 mutant described in WO 2005/007121 published on January 27, 2005 ). , XENP24642 (based on the variant described in WO 1999/060128 published November 25, 1999), XENP24728 (based on H9-RET described by Mitra et al. 2015) and XENP24729 (based on H9-RETR described by Mitra et al. . 2015). Additional variants, such as XENP24641, XENP24730, XENP24731 and XENP24732, were based on individual substitutions or combinations of substitutions described in the prior art. Exemplary sequences are depicted in FIG. 20.

Affinity screening of monovIL-2-Fc fusions

The affinity of the monovIL-2-Fc fusions described above for various IL-2 receptors was determined using Octet as generally described in Example 2. Specifically, to determine the affinity for IL-2Rα of the CD25(IL-2Rα)-Fc fusion ( R&D Systems, Minneapolis, MN) were loaded onto AR2G biosensors and immersed in multiple concentrations of IL-2-Fc fusions. To determine the affinity for IL-2Pß and IL-2Pßγ, bivalent CD122(IL-2Pß)-Fc-His fusions or heterodimeric CD122:CD132(IL-2Pßγ)-Fc-His fusions were loaded onto HIS1K biosensors and immersed in several concentrations of IL fusions -2-Fc. The resulting dissociation constants ( _KD ), association rates ( _ka ) and dissociation rates ( _kd ) are depicted in FIG. 21.

Induction of STAT5 phosphorylation by monovIL-2-Fc fusion

Once cytokines bind to their receptors, Janus kinases (JAKs) associate with receptors for phosphorylating STAT proteins, which then translocate into the nucleus to further regulate downstream processes. Thus, phosphorylation of STAT proteins (specifically STAT5, which includes STAT5a and STAT5b) represents one of the earliest signaling events triggered by the binding of IL-2 to high-affinity IL-2 receptors or to intermediate-affinity IL-2 receptors ( Lin and Leonard (2000); Wuest et al. (2008)). Accordingly, the ability of the monovIL-2-Fc fusions described above to induce STAT5 phosphorylation in various cell types, such as CD8 ⁺ and CD4 ⁺ T cells, as well as Tregs, was examined.

Freshly collected PBMCs were incubated with the indicated IL-2-Fc study drugs at the indicated concentrations for 15 minutes. After incubation, PBMCs were stained with anti-CD3-BV510 (UCHT1), anti-CD4-BV605 (RPA-T4), and anti-CD8-Alexa700 (SK1) for 30–45 min at room temperature. Cells were washed and incubated with pre-cooled (-20°C) 90% methanol for 20-60 minutes. After incubation with methanol, cells were washed again and stained with anti-CD25-BV421 (M-A251), anti-CD45RA-PE (HI100), anti-FOXP3-Alexa488 (259D) and anti-pSTAT5-Alexa647 (pY687) to label different cell populations and STAT5 phosphorylation. Data depicting the induction of STAT5 phosphorylation in CD8 ⁺ T cells (CD3 ⁺ CD8 ⁺ CD25 ⁻ ), CD4 ⁺ T cells (CD3 ⁺ CD4 ⁺ CD25 ⁻ ), and Tregs (CD3 ⁺ CD4 ⁺ CD25 ⁺ FOXP3 ⁺ ) are shown in Fig. 22 Notably, IL‐2‐Fc fusions XENP24638 and XENP24642 were potent Treg activators (as indicated by STAT5 phosphorylation) with minimal activation of CD8 ⁺ and CD4 ⁺ T cells, consistent with the activity reported for IL‐2 variants , which constitute IL-2-Fc fusions.

Study of prior art IL-2 variants constructed as biIL-2-Fc fusions

To study the reliability and effectiveness of the biIL-2-Fc fusion format, a prior art IL-2 variant (described in WO 1999/060128, published November 25, 1999) as well as a control IL-2 (with C125S and T3A substitutions) were generated in specified format, the sequences for which are shown in FIG. 23.

Induction of STAT5 phosphorylation by biIL-2-Fc fusions

Freshly collected PBMCs were incubated with the indicated IL-2-Fc study drugs at the indicated concentrations for 15 minutes. After incubation, PBMCs were stained with anti-CD3-BUV395 (UCHT1), anti-CD4-BV605 (RPA-T4), and anti-CD8-Alexa700 (SK1) for 30–45 min at room temperature. Cells were washed and incubated with pre-cooled (-20°C) 90% methanol for 20-60 minutes. After incubation with methanol, cells were washed again and stained with anti-CD25-BV510 (M-A251), anti-CD45RA-PE (HI100), anti-FOXP3-Alexa488 (259D) and anti-pSTAT5-Alexa647 (pY694) to label different cell populations and STAT5 phosphorylation. Data depicting the induction of STAT5 phosphorylation in CD8 ⁺ T cells, CD4 ⁺ T cells, and Tregs are shown in Fig. 24.

Study of prior art IL-2 variants constructed as IL-2-Fc fusions with domain linkers

To study the effect of including linkers between IL-2 and the Fc region, prior art IL-2 variants (as described in WO 1999/060128, published November 25, 1999, and WO 2012/107417, published August 16, 2012) were prepared as monoVIL-2-Fc fusions or biIL-2-Fc fusions with Gly-Ser linkers. Sequences for these IL-2-Fc fusions are depicted in FIG. 25.

Induction of STAT5 phosphorylation by IL-2-Fc fusions to domain linkers

Freshly collected PBMCs were incubated with the indicated IL-2-Fc study drugs at the indicated concentrations for 15 minutes. After incubation, PBMCs were stained with anti-CD3-BUV395 (UCHT1), anti-CD4-BV605 (RPA-T4), and anti-CD8-Alexa700 (SK1) for 30–45 min at room temperature. Cells were washed and incubated with pre-cooled (-20°C) 90% methanol for 20-60 minutes. After incubation with methanol, cells were washed again and stained with anti-CD25-BV510 (M-A251), anti-CD45RA-PE (HI100), anti-FOXP3-Alexa488 (259D) and anti-pSTAT5-Alexa647 (pY694) to label different cell populations and STAT5 phosphorylation. Data depicting the induction of STAT5 phosphorylation in CD8 ⁺ T cells, CD4 ⁺ T cells, and Tregs are shown in Fig. 26.

Example 4: Construction of variant IL-2-Fc fusions with increased affinity for CD25 and decreased affinity for CD122

As discussed in Example 2, increasing the affinity of IL-2 for IL-2Rα could lead to an additional binding bias in favor of the IL-2Rαßγ complex on Tregs, while decreasing the affinity of IL-2 for IL-2Rß, γc, or IL-2. 2Pßγ could lead to distorted binding to CD25-negative T cells and NK cells. Here, substitutions that increased binding to IL-2Rα were combined with substitutions that decreased binding to IL-2Rβ in the context of IL-2-Fc fusions to increase Treg selectivity.

The plasmid encoding IL-2 was constructed using standard gene synthesis followed by subcloning into the expression vector pTT5 containing Fc fusion partners (eg, constant regions depicted in Fig. 6). IL-2 contained a C125S substitution to increase expression and a T3A substitution to remove the O-glycosylation site. Selected substitutions described in Example 2 were introduced using standard mutagenesis techniques. Proteins were produced by transient transfection in HEK293E cells and purified using a two-step purification method involving protein A chromatography and anion exchange chromatography. Sequences for exemplary IL-2-Fc fusions designed to increase affinity for CD25 and/or decrease affinity for CD122 are depicted in FIG. 27.

Affinity screening of variant IL-2-Fc fusions engineered to increase affinity for CD25 and decrease affinity for CD122

The affinities of the variant IL-2-Fc fusions described above for various IL-2 receptors were determined using Octet as generally described in Example 3B(a). Specifically, to determine affinity for IL-2Rα, CD25(IL-2Rα)-Fc fusions (R&D Systems, Minneapolis, MN) were loaded onto AR2G biosensors and dipped into several concentrations of IL-2-Fc fusions. To determine the affinity for IL-2Pß and IL-2Pßγ, bivalent CD122(IL-2Pß)-Fc-His fusions or heterodimeric CD122:CD132(IL-2Pßγ)-Fc-His fusions were loaded onto HIS1K biosensors and immersed in several concentrations of IL fusions -2-Fc. The resulting dissociation constants ( _KD ), association rates ( _ka ) and dissociation rates ( _kd ) are depicted in FIG. 33.

Phosphorylation of STAT5 by variant IL-2-Fc fusions engineered to increase affinity for CD25 and decrease affinity for CD122 in different cell populations

Freshly collected PBMCs were incubated with the indicated IL-2-Fc study drugs at the indicated concentrations for 15 minutes. After incubation, PBMCs were stained with anti-CD3-BV396 (UCHT1), anti-CD4-BV605 (RPA-T4), and anti-CD8-Alexa700 (SK1) for 30–45 min at room temperature. Cells were washed and incubated with pre-cooled (-20°C) 90% methanol for 20-60 minutes. After incubation with methanol, cells were washed again and stained with anti-CD25-BV421 (M-A251), anti-CD45RA-PE (HI100), anti-FOXP3-Alexa488 (259D) and anti-pSTAT5-Alexa647 (pY687) to label different cell populations and STAT5 phosphorylation. Data depicting the induction of STAT5 phosphorylation in different cell populations are depicted in FIG. 28-29.

The data demonstrate that, compared to XENP24635 (C125S-only IL-2-Fc) and recombinant human IL-2 (rhIL-2) controls, many exemplary variants were potent Treg activators while having minimal induction of STAT5 phosphorylation at CD25- negative T cells (CD8 ⁺ and CD4 ⁺ ).

Example 5: Construction of Additional IL-2-Fc Fusion Variants

Additional IL-2-Fc fusions were designed combining the characteristics described in the previous examples, including valency, linker domains, pH switch, and Treg selectivity, and were prepared as generally described in Example 2. Exemplary sequences are depicted in FIG. 30. In addition, an Fc-IL-2 fusion (V91K/C125A) engineered to increase the ratio of Tregs to non-regulatory T cells described in WO 2014/153111 was generated as a comparator (herein referred to as XENP27193; sequence depicted in Fig. 34).

Induction of STAT5 phosphorylation by additional monovalent IL-2-Fc fusions

Freshly collected PBMCs were incubated with the indicated IL-2-Fc study drugs at the indicated concentrations for 15 minutes at 37°C. After incubation, PBMCs were first stained with anti-CD3-BUV395 (UCHT1), anti-CD4-BV605 (RPA-T4), anti-CD8-AF700 (SK1), and anti-CD56-PE antibodies. After the first staining, cells were permeabilized using PerFix EXPOSE (Beckman Coulter, Indianapolis, IN). After permeabilization, cells were stained with anti-CD25-BV421 (M-A251), anti-CD45RA-BV510 (HI100), anti-FoxP3-AF488 (259D), and anti-pSTAT5-AF647 (47/Stat5 (pY694)) antibodies. After the second staining, cells were analyzed by flow cytometry to examine STAT5 phosphorylation in different lymphocyte populations. Data depicting pFIAT5 SIF for various lymphocyte populations indicating signaling by IL-2-Fc fusions involving IL-2 receptors are presented in FIG. 35.

The data demonstrate that each variant induced STAT5 phosphorylation in Tregs. Notably, variant IL-2-Fc fusions preferentially induced Tregs over CD4 ⁺ memory T cells (CD45RA ⁻ ), CD8 ⁺ memory T cells (CD45RA ⁻ ), NK cells, and γδ T cells compared with both recombinant IL-2 and monovalent IL-2-Fc WT fusion (XENP24635). In comparison, the prior art Fc-IL-2 variant fusion XENP27193 was comparatively less selective for Tregs.

Bivalent IL-2-Fc fusions are more potent than monovalent counterparts.

The induction of STAT5 phosphorylation by various IL-2-Fc study drugs was studied as described in Example 5A. Data depicting pSTAT5 SIF in Tregs and memory CD4 ⁺ T cells (CD45RA ⁻ ) indicating signaling by IL-2-Fc fusions involving IL-2 receptors are presented in FIG. 36-42.

The data demonstrate that for each IL-2 variant, bivalent versions of IL-2-Fc were more effective in inducing STAT5 phosphorylation in Tregs than the corresponding monovalent versions of IL-2-Fc. In particular, the construction of domain linkers between the IL-2 component and the Fc component (eg, in XENP27002, XENP27003, XENP27004, XENP27005, XENP27006, and XENP27007) results in further enhancement of the activity of bivalent IL-2-Fc fusions. Notably, each of the bivalent IL-2-Fc fusion constructs (with and without linkers) retained selectivity for Tregs over other lymphocyte populations such as CD4 ⁺ CD45RA ⁻ T cells.

Example 6: Maximizing the Impact of IL-2-Fc Fusions

Enabling Xtend Fc

The IL-2-Fc fusions described above were engineered with Xtend Fc (M428L/N434S) with increased binding to FcRn to further stimulate FcRn-mediated recycling of the fusions and subsequently prolong the circulating half-life. Sequences for exemplary bivalent IL-2-Fc fusions with Xtend Fc are depicted in FIG. 43, and sequences for exemplary monovalent IL-2-Fc fusions with Xtend Fc are depicted in FIG. 44.

Selection of IL-2-Fc fusions with balance of selectivity and Treg activity

It has also been shown that IL-2-Fc fusion with lower activity will reduce antigen-dependent clearance and therefore increase half-life in the bloodstream. In view of the data in Example 5B demonstrating that bivalent IL-2-Fc fusions and IL-2-Fc fusions having domain linkers have increased potency, there was particular interest in monovalent IL-2-Fc fusions that lacked domain linkers . Accordingly, in order to identify IL-2-Fc fusions with an optimal balance between selectivity and potency, the in vitro potency (as demonstrated by induction of STAT5 phosphorylation in different lymphocyte populations) was compared to monovalent IL-2 fusions (XENP26105 and XENP26109, respectively, containing variants D20N/T37R and D20N/N71K) with XENP24635 (a monovalent IL-2-Fc fusion with the C125S mutation), as well as XENP25908 and XENP27193 (which are comparator IL-2-Fc fusions demonstrated for the treatment of autoimmune diseases), in STAT5 phosphorylation assay, data for which are depicted in FIG. 45.

Notably, both XENP26105 and XENP26109 were less potent than XENP24635 and XENP27193 (EC50 1 nM and 5 nM, respectively, versus 0.02 nM and 0.01 nM), but were able to achieve similar levels of activity in Tregs at more high doses, while maintaining selectivity for Tregs. Although XENP26105 had comparable potency to XENP25908 (corresponding to an EC50 of 1 nM versus 0.7 nM), the data demonstrate that XENP26105, as well as XENP26109, were able to achieve much higher levels of activity in Tregs than XENP25908. The reduced potency and selectivity observed for XENP26105 and XENP26109 suggest that they will be suitable for selective and sustained Treg expansion in the clinical setting. Accordingly, the potential of XENP27563 and XENP27564 and the Xtend Fc analogues XENP26105 and XENP26109 were further explored.

Example 7 In Vitro Characterization of XENP27563 and XENP27564

Treg cultures treated with CD25-selective IL-2-Fc fusions exhibit higher CD25 expression

It was previously reported that rapamycin promotes the proliferation of CD4+CD25+FOXP3+ Tregs in vitro, and the resulting expanded Tregs suppress the proliferation of CD4+ and CD8+ T cells (see, for example, Battaglia et al. (2006). Rapamycin promotes expansion of functional CD4 +CD25+FOXP3+ regulatory T cells of both healthy subjects and type 1 diabetic patients. J Immunol. 177(12) 8338-8347; and Strauss et al. (2007) Selective survival of naturally occurring human CD4+CD25+Foxp3+ regulatory T cells cultured with rapamycin. J Immunol. 178(1) 320-329).

CD4+ T cells were enriched from human PBMCs by negative selection using the EasySep™ Human CD4+ T Cell Enrichment Kit (STEMCELL Technologies, Vancouver, Canada). Tregs were expanded using Human Treg Expander Dynabeads™ (Thermo Fisher Scientific, Waltham, MA) in RPMI1640 + 10% fetal bovine serum + 0.1 μg/ml rapamycin + 500 U/ml IL-2 for 1–4 days. Tregs were transferred to T75 flasks coated with 0.5 μg/ml anti-CD3 (OKT3, Biolegend, San Diego, CA) and cultured with RPMI1640 + 10% fetal bovine serum + 0.1 μg/ml rapamycin + 100 U/ml IL-2 + 0.5 μg/ml anti-CD28 mAb. Experiments were performed at least 8 days after initial enrichment of CD4+ T cells from PBMCs. Tregs enriched and cultured in this manner are hereafter referred to as rapamycin-enhanced Tregs.

Tregs with rapamycin were further cultured with 0.5 μg/ml plate-bound anti-CD3 (OKT3) in RPMI1640 medium containing 10% FBS, 0.5 μg/ml anti-CD28 mAb, 100 ng/ml rapamycin and 10 ng ml of recombinant IL-2 or 10 μg/ml XENP27564 (IL-2-Fc fusion with IL-2 variant (D20N/N71K/C125S)). 14 days after culture, Tregs were stained with anti-CD25-FITC (M-A251), anti-FoxP3-PE (PCH101), anti-CTLA-4-PE-Dazzle594 (L3D10), anti-PD-1-BB700 (EH12. 1), anti-GITR-PE-Cy7 (108-17), anti-Ki67-Alexa647, anti-ICOS-Alexa700 (C398.4a), anti-TIGIT-BV421 (A15153G), anti-LAG-3 (11C3C65) , anti-CCR4-BV605 (L291H4), anti-CD8-BV650 (SK1), anti-CD39-BV711 (A1), anti-TIM-3-BV785 (F38-2E2), anti-CD40BUV396 (SK3), anti- CD3-BUV496 (UCHT1), anti-CD45-BUV805 (HI30), anti-CD45RA-BUV737 (HI100), and Zombie NIR (APC-Cy7) and analyzed by flow cytometry, the data of which are depicted in Fig. 46-47. Data demonstrate that Tregs treated with CD25-selective XENP27564 exhibit higher CD25 expression. Next, as shown in FIG. 48, XENP27564 demonstrated higher expansion of the effector Treg population (CD45RA-FoxP3 ^mid-high ).

7B: Treg cultures treated with CD25-selective IL-2-Fc fusions exhibit higher suppressive function

Rapamycin-treated Tregs further cultured with either IL-2 or XENP27564 as described in Example 7A were assessed for their suppressive function at day 15. 1 x 10 ⁵ CFSE-labeled PBMCs were incubated with the indicated number of Tag-it labeled Violet Tregs for 4 days and the expansion of CD8+ responders and CD4+ responders was determined by CFSE dilution. Lymphocyte populations were stained as follows: anti-CD8-PerCp-By5.5 (SK1), anti-CD3-PE-Cy7 (OKT3), anti-CD127-APC (A019D5), anti-CD25-APC-Fire750 (M-A251 ), anti-CD45RO-Alexa700 (UCHL1), anti-CD16-BV605 (3G6), anti-CD56-BV605 (HCD56), anti-CD45RA-BV785 (HI100), anti-CD4-BUV395 (SK3) and Zombie Aqua ( BV510). It is noteworthy that the data depicted in FIG. 49 indicate that Tregs expanded using Treg-selective IL-2-Fc fusions may have enhanced suppressive function.

In addition, the expression of CD25 and CD127 on Tregs was examined in a suppressor assay, the data for which are depicted in FIG. 50-51. Consistent with the above data, XENP27564-expanded Tregs showed higher levels of CD25 expression. Notably, XENP27564-expanded Tregs exhibited lower expression of CD127, a marker previously found to be inversely correlated with suppressive function of Tregs (Liu et al. (2006) CD127 expression inversely correlates with FoxP3 and suppressive function of human CD4+ T reg cells J Exp Med 203(7): 1701–1711). This may explain the increased suppressive function observed for XENP27564-expanded Tregs.

CD25-selective IL-2-Fc fusions demonstrate Treg selectivity and suppressive effects on other lymphocyte populations following anti-CD3 activation

In the suppressor assay, CFSE-labeled PBMCs and rapamycin-labeled Tag-it Violet Tregs were incubated with the indicated concentrations of the indicated study drugs for 4 days with 100 ng/ml plate-bound anti-CD3 (OKT3). Lymphocyte populations were stained as follows: anti-CD8-PerCp-Cy5.5 (SK1), anti-CD3-PE-Cy7 (OKT3), anti-CD25-APC-Cy7 (M-A251), anti-CD45RO-Alexa700 (UCHL1 ), anti-CD16-BV605 (3G6), anti-CD56-BV605 (HCD56), anti-CD45RA-BV785 (HI100), anti-CD4-BUV395 (SK3) and Zombie Aqua (BV510). The proliferation of different lymphocyte populations (determined by CFSE or Tag-it Violet dilution; Zombie dye used to exclude dead cells) after treatment with study drugs is depicted in FIG. 52 Data demonstrate that CD25-selective IL-2-Fc fusions XENP27563 and XENP27564 resulted in selective expansion of Tregs compared with XENP24635 (IL-2-Fc with C125S only), recombinant IL-2, and recombinant IL-15. In fact, the data depicted in FIG. 53 and 54 demonstrate that CD25-selective IL-2-Fc fusions had a significantly reduced ability to induce CD8+ and CD4+ T cell proliferation compared to XENP24635, recombinant IL-2, and recombinant IL-15.

In the proliferation assay, PBMCs were incubated with the indicated concentrations of the indicated study drugs with the indicated concentration of plate-bound anti-CD3 mAb (OCT3). Lymphocyte populations were stained as follows: anti-FoxP3-PE (PCH101), anti-CD8-PerCP-Cy5.5 (SK1), anti-CD3-PE-Cy7 (OKT3), anti-Ki67-APC, anti-CD45RO-Alexa700 (UCHL1), anti-CD25-BV421 (M-A251), anti-CD16-BV605 (3G6), anti-CD56-BV605 (HCD56), anti-CD45RA-BV785 (HI100), anti-CD4-BUV396 (SK3) and Zombie NIR (APC-Cy7). Proliferation of different lymphocyte populations was determined based on the percentage expression of Ki67, a proliferation marker for which data are depicted in FIG. 55-65. Consistent with the data presented in the suppressor assay above, the proliferation assay data depicted in FIG. 55-61 demonstrate that XENP27563 and XENP27564 (IL-2-Fc fusions engineered for CD25 selectivity) have impaired function in inducing proliferation of CD8+ T cells, CD8+CD45RA- T cells, CD8+CD45RA+ T cells, CD4+ T cells, CD4+CD45RA- T cells, CD4+CD45RA+ T cells, NK cells; and the data shown in Fig. 63–65 demonstrate that XENP27563 and XENP27564 lead to selective expansion of Tregs over other lymphocyte populations.

Example 8: IL-2-Fc Fusions Activate Selective and Sustained Treg Expansion in Cynomolgus Macaques

To explore the clinical potential of XENP27563 and XENP27564, their activity in cynomolgus monkeys was studied. Before drug administration to animals, the activity of IL-2-Fc fusions against cynomolgus macaque lymphocytes was confirmed. Two analyzes were performed, described below.

In the first assay, human PBMCs were stimulated with varying concentrations of XENP27563 or XENP27564 for 15 minutes at 37°C. PBMCs were then stained with anti-CD3-BUV395 (UCHT1), anti-CD4-BV605 (RPA-T4), anti-CD8-BV711 (RPA-T8), anti-CD25-BV421 (M-A251), anti-CD45RA-BV510 (HI100) and anti-CD56-PE. Cells were then permeabilized using PerFix EXPOSE (Beckman Coulter, Indianapolis, IN). After permeabilization, cells were stained with anti-CD16-AF700 (DJ130C), anti-FoxP3-AF488 (259D) and pSTAT5 (pY694) and analyzed by flow cytometry for STAT5 phosphorylation in different lymphocyte populations, data for which are depicted in Fig. 78.

In the second assay, cynomolgus monkey PBMCs were stimulated with varying concentrations of XENP27563 or XENP27564 for 15 minutes at 37°C. PBMCs were then stained with anti-CD3-BV421 (SP34), anti-CD4-BV785 (OKT4), anti-CD8-BUV395 (RPA-T8), anti-CD25-BV510 (M-A251), anti-CD45RA-APC/H7 (HI100) and anti-CD56-PE. Cells were then permeabilized using PerFix EXPOSE (Beckman Coulter, Indianapolis, IN). After permeabilization, cells were stained with anti-CD16-AF700 (DJ130C), anti-FoxP3-AF488 (259D) and pSTAT5 (pY694) and analyzed by flow cytometry for STAT5 phosphorylation in different lymphocyte populations, data for which are depicted in Fig. 79.

Data demonstrate that XENP27563 and XENP27564 are equally selective and effective against both human Tregs and cynomolgus Tregs.

8A: Comparison of PD and FC for XENP27563 and XENP27564

In the first study in cynomolgus monkeys, animals (n = 2) were intravenously administered either a 3× dose of XENP27563 or a 3× dose of XENP27564 on days 0 and 15. Blood was collected over time to study the expansion of different lymphocyte populations and to study serum concentrations of study drugs. Serum albumin concentrations and blood pressure of the animals were also taken into account to study the tolerability of the study drugs.

In fig. Figure 66 depicts the expansion of different lymphocyte populations over time. The data demonstrate that both study drugs were capable of Treg expansion while maintaining near baseline levels of CD8+CD45RA- T cells, CD4+CD45RA- T cells, and CD16+ NK cells. In addition, the data demonstrate that the two study drugs produce similar pharmacological activity in cynomolgus monkeys. In fig. 67 shows the concentration of the studied drugs in the blood serum. The data demonstrate that the two study drugs exhibited similar pharmacokinetic profiles, with XENP27564 having a half-life of 1.5 days. Taken together, this suggests that IL-2-Fc fusions engineered to be CD25-selective and attenuated promote selective and sustained Treg expansion.

Capillary leak syndrome is a characteristic toxic side effect associated with treatment with cytokines such as IL-2. One of the signs of fluid leakage from blood vessels is hypoalbuminemia, a decrease in serum albumin concentration. Accordingly, the change in serum albumin concentration in animals was studied, the data for which are depicted in FIG. 68. Notably, in one animal administered XENP27563, a long-lasting decrease in albumin levels was detected after both the first and second administration. In one animal administered XENP27564, a prolonged decrease in albumin levels was detected after the second administration, but concentrations quickly returned to baseline levels. These data suggest that XENP27564, which is less potent than XENP27563, may contribute to a higher tolerability and therapeutic index.

Another indicator of fluid leakage from blood vessels is a significant decrease in blood pressure. Accordingly, blood pressure was recorded for animals on days 0, 1, 3, 5, 9, 16, 18, 20 and 24, the data for which is shown in fig. 69. Notably, the first monkey administered XENP27563 experienced a decrease in blood pressure on day 1 (day 1 after 1st dose), and the second monkey administered XENP27563 experienced a decrease in blood pressure on day 16 (day 1 after 2nd dose), while stable blood pressure was observed on all days for the XENP27564-administered monkey (telemetry data was impaired for the 2nd XENP27564-administered monkey). This also suggests that IL-2-Fc fusions with lower potency may contribute to higher tolerability and therapeutic index.

Finally, eosinophil and basophil levels were also studied as additional indicators of tolerability, the data for which are depicted in FIG. 76 Taken together, the data demonstrate the reintroduction of XENP27564.

8B: XENP27564 Dose Escalation Study

In the first study in cynomolgus monkeys, animals (n = 3) were intravenously administered either a 1×, 3×, or 10× dose of XENP27564. Blood was collected over time to study the expansion of different lymphocyte populations, as well as serum albumin and C-reactive protein (CRP) concentrations.

The expansion of different lymphocyte populations is depicted in FIG. 70-71. Consistent with data from the first study in cynomolgus monkeys, XENP27564 mediates selective and sustained Treg expansion. Additionally, the data demonstrate that 1× and 3× doses produce similar pharmacological activity (as demonstrated by Treg expansion) in monkeys. Notably, a higher dose (10× dose) of XENP27564 did not result in enhanced pharmacodynamics. According to the data presented above, FIG. 75 depicts steady-state pharmacokinetics over several days in cynomolgus monkeys at all doses of XENP27564 studied.

As with the first study, albumin reduction was examined as an indicator of vascular leakage and tolerance, for which data are depicted in FIG. 72. In addition, serum concentrations of CRP, an acute phase protein associated with inflammation, were studied as another measure of tolerability, for which data are depicted in FIG. 73. Sodium concentration, chlorine concentration, eosinophil content and basophil content were also studied as additional indicators of tolerability (data for which are depicted in Fig. 74). Notably, the data demonstrate that higher doses of XENP27564 resulted in increased toxicity, as evidenced by both decreased albumin levels and increased serum CRP concentrations (as well as sodium concentration, chloride concentration, eosinophil content, and basophil content), while meanwhile, lower doses that still resulted in significant increases in Tregs were more tolerable.

Example 9: IL-2-Fc Fusions Are Equally Selective for Tregs in Mice

Splenocytes from B6 mice were incubated with IL-2-Fc fusions and recombinant IL-2 for 15 minutes. After incubation, cells were stained with anti-CD4-PE (GK1.5), anti-CD25-BV605 (PC61), and anti-CD44-BV510 (IM7). Cells were then permeabilized using PerFix EXPOSE (Beckman Coulter, Indianapolis, IN). After permeabilization, cells were stained with anti-CD3-AF700 (2C11), anti-CD8-AF488 (53-6.7), anti-FoxP3-eF450 (FJK-16S) and anti-pSTAT5 (pY694) and analyzed by flow cytometry for phosphorylation STAT5 in various lymphocyte populations, data for which are depicted in FIG. 77 Data indicate that engineered IL‐2‐Fc fusions are uniformly selective and potent for Tregs in mice, making them suitable for use in preclinical mouse models to study autoimmune diseases.

Claims (17)

1. A polypeptide composition for the expansion of Tregs, characterized in that it contains a variant human IL-2 protein, wherein said variant IL-2 protein contains one or more amino acid substitutions compared to SEQ ID NO:2, selected from the group T3A/D20N/T37R , T3A/D20N/N71K, T3A/D20N/T37R/C125S, T3A/D20N/N71K/C125S, T3A/D20N/T37R/C125A and T3A/D20N/N71K/C125A. 2. A polypeptide composition comprising a homodimeric protein complex for expansion of Tregs, wherein the homodimeric protein complex contains first and second protein monomers, wherein each protein monomer contains a variant IL-2 protein covalently attached to an Fc domain, and wherein the variant IL-2 protein contains one or more amino acid substitutions compared to SEQ ID NO:2, selected from the group T3A/D20N/T37R, T3A/D20N/N71K, T3A/D20N/T37R/C125S, T3A/D20N/N71K/C125S, T3A/D20N/ T37R/C125A and T3A/D20N/N71K/C125A. 3. The polypeptide composition according to claim 2, characterized in that each of said Fc domains is a variant Fc domain. 4. A polypeptide composition comprising a heterodimeric protein complex for Tregs expansion, wherein the heterodimeric protein complex comprises a first protein monomer and a second protein monomer, wherein the first protein monomer comprises a variant IL-2 protein covalently attached to a first variant Fc domain, and wherein the variant protein IL-2 contains one or more amino acid substitutions compared to SEQ ID NO:2, selected from the group T3A/D20N/T37R, T3A/D20N/N71K, T3A/D20N/T37R/C125S, T3A/D20N/N71K/C125S, T3A /D20N/T37R/C125A and T3A/D20N/N71K/C125A, and where the second protein monomer contains a second variant Fc domain. 5. The polypeptide composition according to claim 3 or 4, characterized in that said variant Fc domains, the first variant Fc domain and/or the second variant Fc domain, are variant Fc domains of human IgG1 containing amino acid substitutions M428L/N434S . 6. Polypeptide composition according to claim 3 or 4, characterized in that said variant Fc domains, the first variant Fc domain and/or the second variant Fc domain, are variant Fc domains of human IgG1 containing amino acid substitutions E233P/L234V /L235A/G236del/S267K. 7. Polypeptide composition according to claims. 4, 5 or 6, characterized in that said first and second variant Fc domains, which include a set of heterodimerization variants selected from the group consisting of the variants depicted in FIG. 2. 8. The polypeptide composition according to claim 7, characterized in that the specified set of heterodimerization options is selected from the group consisting of L368D/K370S: S364K/E357Q; L368D/K370S : S364K; L368E/K370S : S364K; T411E/K360E/Q362E : D401K; and T366S/L368A/Y407V : T366W. 9. Polypeptide composition according to claims. 4, 5, 6, 7, or 8, characterized in that one of said variant Fc domains includes a set of pI variants selected from the group consisting of the variants depicted in FIG. 3. 10. The polypeptide composition according to claim 9, characterized in that said set of pI variants contains Q295D/N384D/Q418E/N421D. 11. The polypeptide composition according to claim 4, characterized in that said polypeptide composition is selected from the group consisting of XENP27564 (SEQ ID NO: 297 and 298), XENP27563 (SEQ ID NO: 295 and 296), XENP26105 (SEQ ID NO: 245 and 246) and XENP26109 (SEQ ID NO: 249 and 250). 12. Polypeptide composition for expansion of Tregs, containing a heterodimeric protein complex for expansion of Tregs, where the heterodimeric protein complex contains: a first monomer containing the amino acid sequence shown in SEQ ID NO: 297; And a second monomer containing the amino acid sequence set forth in SEQ ID NO: 298. 13. Polypeptide composition for expansion of Tregs, containing a heterodimeric protein complex for expansion of Tregs, where the heterodimeric protein complex contains: a first monomer containing the amino acid sequence shown in SEQ ID NO: 295; And a second monomer containing the amino acid sequence set forth in SEQ ID NO: 296.

Images