IL2R common gamma chain antibodies

Abstract

Anti-CD122 and/or c antibodies and fragments thereof are disclosed. Also disclosed are compositions comprising such antibodies and fragments, and uses and methods using the same.

Claims

1. An isolated antibody or antigen binding fragment which specifically binds to common chain (c), and comprises: (i) a light chain variable region comprising the following CDRs: TABLE-US-00027 LC-CDR1: (SEQ ID NO: 68) RSSQSLLHSNGYNYLD LC-CDR2: (SEQ ID NO: 69) LGSNRDS LC-CDR3: (SEQ ID NO: 70) MQGTHWPWT; and (ii) a heavy chain variable region comprising the following CDRs: TABLE-US-00028 HC-CDR1: (SEQ ID NO: 48) GYYWS HC-CDR2: (SEQ ID NO: 49) EINHSGSTNYNPSLKS HC-CDR3: (SEQ ID NO: 77) SPGGYSGGYFQH.

2. The isolated antibody or antigen binding fragment according to claim 1, wherein: the light chain variable region sequence has at least 70% sequence identity to SEQ ID NO:67, and the heavy chain variable region sequence has at least 70% sequence identity to SEQ ID NO:76.

3. The isolated antibody or antigen binding fragment according to claim 1, wherein: the light chain variable region sequence has at least 85% sequence identity to SEQ ID NO:67, and the heavy chain variable region sequence has at least 85% sequence identity to SEQ ID NO:76.

4. The isolated antibody or antigen binding fragment according to claim 1, conjugated to a drug moiety or a detectable moiety.

5. A bispecific antibody comprising the antibody or antigen binding fragment of claim 1.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) Embodiments and experiments illustrating the principles of the invention will now be discussed with reference to the accompanying figures, in which:

(2) FIG. 1. Light chain variable domain sequences for anti-IL-2R antibody clones. CDRs are underlined and shown separately.

(3) FIG. 2. Heavy chain variable domain sequences for anti-IL-2R antibody clones. CDRs are underlined and shown separately.

(4) FIG. 3. Light chain variable domain sequences for anti-c antibody clones. CDRs are underlined and shown separately.

(5) FIG. 4. Heavy chain variable domain sequences for anti-c antibody clones. CDRs are underlined and shown separately.

(6) FIG. 5. Table showing light chain CDR sequences for anti-IL-2R antibody clones.

(7) FIG. 6. Table showing heavy chain CDR sequences for anti-IL-2R antibody clones.

(8) FIG. 7. Table showing light chain CDR sequences for anti-c antibody clones.

(9) FIG. 8. Table showing heavy chain CDR sequences for anti-c antibody clones.

(10) FIG. 9. Table showing light chain CDR sequences for P2C4-derived anti-IL-2R antibody clones.

(11) FIG. 10. Table showing heavy chain CDR sequences for P2C4-derived anti-IL-2R antibody clones.

(12) FIG. 11. Table showing light chain CDR sequences for P1A3-derived anti-c antibody clones.

(13) FIG. 12. Table showing heavy chain CDR sequences for P1A3-derived anti-c antibody clones.

(14) FIG. 13. CH2 and CH3 domain sequences for anti-IL-2R antibody clone P2C4.

(15) FIG. 14. CH2 and CH3 domain sequences for anti-c antibody clone P1A3.

(16) FIG. 15. Amino acid sequences for anti-IL-2R antibody clones. VH domains are shown underlined. (GGGS).sub.3 linkers (and variants thereof) are shown in bold. VL domains are shown double underlined. Short linkers are in italics and bold. Hinges are shown in italics. CH2 domains are shown . CH3 domains are shown .

(17) FIG. 16. Amino acid sequences for anti-c antibody clones. VH domains are shown underlined. (GGGS).sub.3 linkers (and variants thereof) are shown in bold. VL domains are shown double underlined. Short linkers are in italics and bold. Hinges are shown in italics. CH2 domains are shown . CH3 domains are shown .

(18) FIG. 17. Nucleotide sequences for anti-IL-2R antibody clones.

(19) FIG. 18. Nucleotide sequences for anti-c antibody clones.

(20) FIG. 19. Schematic representation of the bispecific anti-IL-2R/c antibody.

(21) FIGS. 20A-B. Sensorgrams showing binding of the bispecific anti-IL-2R/c antibody to (FIG. 20A) human IL-2R and (FIG. 20B) human c as determined by surface plasmon resonance analysis.

(22) FIGS. 21A-B. Histograms showing binding of bispecific anti-IL-2R/c antibody (FIG. 21A) to cells expressing (and control cells not expressing) IL-2R/c, and (FIG. 21B) to PBMCs, as determined by flow cytometry.

(23) FIG. 22. Histograms showing induction of STAT5, Akt and ERK mediated signalling by treatment of NK92 cells with IL-2 or bispecific anti-IL-2R/c antibody in vitro, as determined by flow cytometry.

(24) FIG. 23. Graph showing percent phosphorylation of STAT5 in response to treatment with IL-2 or bispecific anti-IL-2R/c antibody in vitro for different immune cell subsets, as determined by flow cytometry.

(25) FIG. 24. Graph showing proliferation of IL-2 dependent NK92 cells in response to treatment with bispecific anti-IL-2R/c antibody or control antibodies exhibiting specificity for only IL-2R or c.

(26) FIGS. 25A-C. Graphs and schematic showing results of analysis of linker length on binding by anti-IL-2R/c antibody. (FIG. 25A) Schematic representation of antibody and scFv formats, and linkers (linkers shown in italics). (FIG. 25B) Graph showing proliferation of NK92 cells in response to treatment with bispecific anti-IL-2R/c antibody comprising linkers of different length. (FIG. 25C) Graph showing proliferation of NK92 cells in response to treatment with bispecific anti-IL-2R/c in bis-scFv format, comprising linkers of different length.

(27) FIG. 26. Histograms showing induction of STAT5 signalling in cynomolgus macaque splenocytes by treatment with IL-2 or bispecific anti-IL-2R/c antibody in vitro, as determined by flow cytometry.

(28) FIGS. 27A-D. Graphs showing T cell numbers and ratios following culture of PBMCs for 1 week in the presence of recombinant human IL-2 or the indicated amount of bispecific anti-IL-2R/c antibody (Mega2). (FIG. 27A) CD3+ cells, (FIG. 27B) CD4+ cells, (FIG. 27C) CD8+ cells, and (FIG. 27D) the ratio of CD8+ to CD4+ cells.

(29) FIG. 28. Graph showing the percentage of Tregs following culture of PBMCs for 1 week in the presence of recombinant human IL-2 or the indicated amount of bispecific anti-IL-2R/c antibody (Mega2).

(30) FIG. 29. Graph showing CD8+ T cell subsets as a percentage of CD8+ cells following culture of PBMCs for 1 week in the presence of recombinant human IL-2 or the indicated amount of bispecific anti-IL-2R/c antibody (Mega2). For each subset, from left to right, the data points are: IL-2 200 ng/ml, Mega2 3 ug/ml, Mega2 1 ug/ml, Mega2 0.3 ug/ml, Mega2 0.3 ug/ml and CD3/28.

(31) FIGS. 30A-D. Graphs showing T cell numbers and ratios following culture of PBMCs from a EBV-seropositive donor in the presence of EBV-LCLs and recombinant human IL-2 or the indicated amount of bispecific anti-IL-2R/c antibody (Mega2). (FIG. 30A) CD3+ cells, (FIG. 30B) CD4+ cells, (FIG. 30C) CD8+ cells, and (FIG. 30D) the ratio of CD8+ to CD4+ cells.

(32) FIGS. 31A-C. Graphs showing T cell subsets following culture of PBMCs from a EBV-seropositive donor in the presence of EBV-LCLs and recombinant human IL-2 or the indicated amount of bispecific anti-IL-2R/c antibody (Mega2). (FIG. 31A) CD8+ T cell subsets as a percentage of CD8+ cells. For each subset, from left to right, the data points are: IL-2 200 ng/ml, Mega2 3 ug/ml, Mega2 1 ug/ml, Mega2 0.3 ug/ml, Mega2 0.3 ug/ml and CD3/28. (FIG. 31B) CD8+PD1+ cells as a percentage of CD8+ cells, and (FIG. 31C) Tregs as a percentage of CD4+ cells.

(33) FIG. 32. Graph showing CTL cytotoxicity following culture of PBMCs from a EBV-seropositive donor in the presence of EBV-LCLs and recombinant human IL-2 or the indicated amount of bispecific anti-IL-2R/c antibody (Mega2).

(34) FIGS. 33A-B. Bar charts showing thermostability of P1A3 family clones. Binding of P1A3 and the mutated clones (FIG. 33A) B4 and E9 and (FIG. 33B) B3 and E8 to c before and after heat treatment. MeanSD on duplicates is shown.

(35) FIGS. 34A-B. Bar charts showing thermostability of P2C4 family clones. Binding of P2C4 and the mutated clones (FIG. 34A) A9, B1, B5, B6, B8, C4, C7, C12, E2, E3, E7, E8, E9, G2, G11, H1, H2, and H3, and (FIG. 34B) A4, B12, C1, D10, E6, F8, F11 and C1D10 to IL2-R before and after heat treatment. MeanSD on duplicates is shown.

(36) FIGS. 35A-B. Graphs showing binding of (FIG. 35A) P2C4_FW2 single chain antibody to IL-2R, and (FIG. 35B) P1A3_FW2 single chain antibody to c.

(37) FIGS. 36A-B. Graphs showing binding of the bispecific antibody clone P2C4/P1A3 to (FIG. 36A) IL-2R, and (FIG. 36B) c, for antibodies having the NSGAGTAAA (SEQ ID NO:157) or GGGGSAAA (SEQ ID NO:158) short linkers.

(38) FIGS. 37A-B. Graphs showing binding of bispecific engineered antibody clones to (FIG. 37A) IL-2R, and (FIG. 37B) c.

(39) FIGS. 38A-B. Graphs showing in vitro response of antigen-specific CD8+ T cells to bispecific anti-IL-2R/c antibody exposure, as measured by flow cytometry. (FIG. 38A) Bispecific anti-IL-2R/c antibody-dependent expansion of CD8+ T cells. (FIG. 38B) CD8:CD4+ T cell ratio following exposure to the antibody relative to IL-2, following autologous LCL co-culture. *p value <0.05.

(40) FIGS. 39A-B. Graphs showing in vitro response of Treg cells to bispecific anti-IL-2R/c antibody exposure, as measured by flow cytometry, in (FIG. 39A) an antigen-specific, or (FIG. 39B) non-specific setting. *p value <0.05.

(41) FIGS. 40A-E. Bar charts showing levels of cytokines as measured by Luminex analysis (FIG. 40A) IFN, (FIG. 40B) IL-15, (FIG. 40C) IL-1, (FIG. 40D) IL-6, and (FIG. 40E) TNF in the plasma of non-human primates, before and after administration of anti-IL-2R/c antibody.

(42) FIGS. 41A-C. Bar charts showing in vivo response of T cell subsets to bispecific anti-IL-2R/c antibody injection, as measured by flow cytometry. (FIG. 41A) T cells as a proportion of the total leukocyte population, (FIG. 41B) Ki-67+ positive CD8+ cells as a proportion of the total CD8+ T cell population. (FIG. 41C) Ki-67+ positive CD4+ cells as a proportion of the total CD4+ T cell population. Bispecific antibody dependent expansion is indicated by the increase in T cells relative to the total leukocyte population.

(43) FIGS. 42A-B. Bar charts showing in vivo response of NK cells to bispecific anti-IL-2R/c antibody injection, as measured by flow cytometry. (FIG. 42A) NK cells as a proportion of the pre-dose total leukocyte population, (FIG. 42B) Ki-67+ positive NK cells as a proportion of the total NK cell population.

EXAMPLES

(44) In the following Examples, the inventors describe the isolation of anti-IL-2R and anti-c antibodies, construction, engineering and in vitro and in vivo functional characterisation of bispecific anti-IL-2R/c antibodies.

Example 1: Isolation of Anti-Human IL-2R and Anti-Human c Antibodies

(45) Anti-IL-2R and anti-c antibodies were isolated from a human antibody phage display library via in vitro selection. Specific Fab antibodies were originally identified by ELISA using recombinant IL-2R and c proteins as antigens.

Example 2: Construction of a Bispecific Antibody Targeting the Medium Affinity IL-2R-c

(46) Clones showing a strong binding in ELISA (Example 1) were selected and used to construct a knob-in-hole monovalent, bispecific human antibody based on a single chain variable fragment (scFv) linked to a IgG1 backbone Fc region as schematised in FIG. 19.

(47) The knob-in-hole format prevents homodimerisation and formation of bivalent, monospecific antibodies.

(48) A LALA mutation (substitution of leucine residues 234 and 235 in wild type heavy chain constant domain 2 by alanine) was introduced in the Fc portion of the antibody to abrogate binding to Fc receptor.

(49) The size of the linker between the scFv domain and the Fc domain has no effect on the function of the construct (see Example 6.2 and FIGS. 25A-C).

(50) Bispecific scFv (bis-scFv) Format:

(51) P1A3 and P2C4 scFv were tied with a linker to form a bispecific antibody composed of two single chain variable domains connected by a linker (FIG. 25A, right). Different linker sizes were tested (FIG. 25C) and activity was tested by measuring NK92 cell growth.

(52) The bis-scFv was effective in maintaining proliferation of NK92 cells in the absence of IL-2. The linker size between the two single chain fragments did not affect the bispecific compound activity (FIG. 25C).

Example 3: Analysis of Binding to IL-2R Chains

(53) The binding of the bispecific antibody to either IL-2R or c was analysed by flow cytometry.

(54) Antibodies were incubated with HEK-293.6E cells that had previously been transfected with constructs encoding either IL-2R or c.

(55) Binding to the cells was detected using a fluorescent-conjugated secondary antibody. An isotype IgG1 was used as a negative control. Bispecific constructs with specificity for either IL-2R or c and for an irrelevant target were also tested.

(56) The anti-IL-2R/c antibody was shown to bind to cells expressing its targets (FIG. 21A).

Example 4: Analysis of Affinity for IL-2R Chains

(57) Association/dissociation of the anti-IL-2R/c bispecific antibody to/from the receptor chains was measured in Surface Plasmon Resonance using recombinant IL-2R or c chains immobilised on a chip, and flowing various concentrations of the antibody over the surface.

(58) The antibody showed a very high affinity for IL-2R or c chains, with a very rapid binding and a very slow dissociation (FIGS. 20A-B).

(59) Affinity was measured for bispecific anti-IL-2R/c antibody P2C4/P1A3, and other bispecific antibodies shown in Table 1.

(60) TABLE-US-00026 TABLE 1 Bispecific Ab K.sub.D (M) anti-IL-2R clone anti-c clone for IL-2R for c P2C4 P1A3 1.43 10.sup.7 2.09 10.sup.8 P2H7 P2B9 1.01 10.sup.7 7.98 10.sup.8 P2D12 P1A3 1.81 10.sup.7 7.87 10.sup.8 P1G11 P1A3 1.28 10.sup.7 3.37 10.sup.7

Example 5: Binding to PBMC Subsets

(61) The bispecific IgG anti-IL-2R/c antibody was tested on PBMCs isolated from healthy donors to check which cell subsets it binds to. Antibody or isotype IgG control were added to PBMCs and detected with a fluorescently-conjugated secondary anti-human IgG antibody in flow cytometry assays.

(62) The anti-IL-2R/c bispecific antibody did not show high binding for CD4+ or CD8+ T cells. However, the antibody bound efficiently to CD56+NK cells, CD19+ B cells and CD14+/CD16+ monocytes (FIG. 21B).

Example 6: Activity/IL-2 Agonistic Effects of Anti-IL-2R/c Bispecific Antibody

(63) 6.1 Signalling Pathway Phosphorylation

(64) IL-2 is known to trigger intracellular signalling via STAT5, ERK and Akt pathways. The anti-IL-2R/c bispecific antibody was tested for its ability to induce signalling through these pathways.

(65) IL-2-sensitive NK92 cells were deprived from serum and then stimulated either with IL-2 (100 U/ml, i.e. 0.5 nM) or the anti-IL-2R/c antibody (10 g/ml, i.e. 95 nM.sup.2) for 30 minutes, and phosphorylation of STAT5, Akt and ERK was detected using fluorescent antibodies in flow cytometry assays.

(66) The anti-IL-2R/c antibody induced STAT5 and Akt phosphorylation, although in a milder way than IL-2 (FIG. 22). In this assay, the IL-2R/c antibody did not trigger phosphorylation of ERK (FIG. 22).

(67) One of the biggest obstacles to therapeutic use of IL-2 is the preferential stimulation of cells expressing the high affinity heterotrimeric receptor CD25, e.g. regulatory T cells (Tregs), activated T cells, activated B cells, some myeloid precursor cells, and epithelial cells.

(68) Phosphorylation of STAT5 in the presence of IL-2 or anti-IL-2R/c antibody was measured by flow cytometry in Tregs, CD8+ T cells and NK cells obtained from healthy donors.

(69) Small amounts of IL-2 were sufficient to activate NK or T cells, but even at low levels of IL-2 Tregs were preferentially and strongly activated. At concentrations giving less than 20% activation of the STAT5 signaling pathway in NK or CD8+ T cells, Tregs already showed 100% activation (FIG. 23).

(70) By contrast, the bispecific antibody showed a different activation profile with a lower preferential activation of Tregs. At concentrations resulting in a 20% activation of NK and CD8+ T cells, Tregs showed between 39 and 49% STAT5 phosphorylation. At concentrations giving 50% of activation in NK and CD8+ T cells, the Treg population was still not completely activated, with 73-78% STAT5 phosphorylation (FIG. 23).

(71) 6.2 Proliferation of IL-2-Dependent Cells

(72) Viability and growth of NK92 cells was measured with Alamar blue dye in the absence of IL-2.

(73) The anti-IL-2R/c antibody was able to maintain proliferation of NK92 cells in the absence of IL-2, whilst the antibody constructs binding to only one chain of the IL-2 receptor did not show any effect (FIG. 24).

(74) To assess whether the length of the linker had an effect on the functionality of the antibody, the same assay was conducted using antibodies with different linker sizes.

(75) Growth of NK92 cells was not affected by linker size (FIG. 25B). The data in FIG. 25B were obtained using the shortest and the longest linkers (between 5 and 23 amino-acids).

(76) Linkers of different length were analysed in the bispecific antibody format, or bispecific scFv format, as represented schematically in FIG. 25A. Briefly, P1A3 and P2C4 scFv were linked with linkers of different size, and the activity was tested by measuring NK92 cell growth.

(77) The results are shown in FIGS. 25B and 25C. The bis-scFv is effective in maintaining proliferation of NK92 cells in the absence of IL-2, and the size of the linker between the two scFv fragments does not affect activity.

(78) 6.3 Cross-Reactivity with Cynomolgus Monkey Cells

(79) The anti-IL-2R/c antibody was also tested on Non-Human Primate cells. Briefly, Cynomolgus splenocytes were incubated in the presence of human IL-2 or the bispecific antibody and STAT5 phosphorylation was measured. The antibody was found to be cross-reactive with Cynomolgus IL-2R, and triggered phosphorylation of STAT5 as efficiently as human IL-2 (FIG. 26).

(80) 6.4 Conclusion

(81) Taken together, these data show that the anti-IL-2R/c bispecific antibody has some agonist effects to IL-2, and that these effects are not highly preferentially directed towards CD25-expressing cells.

Example 7: Modulation of the Immune Response: Control of T Cell Expansion in a Non-Specific Stimulation Setting

(82) Peripheral blood mononuclear cells (PBMCs) were isolated from a volunteer donor and cultured for 1 week in the presence of recombinant human IL-2 (200 ng/ml), the anti-IL-2R/c bispecific antibody (3, 1, 0.3, 0.1, or 0.03 g/ml) or anti-CD3/CD28 beads as a positive control. After 1 week, cell expansion was assessed by measuring absolute cell counts; cell subset proportions were measured by FACS.

(83) 7.1 Expansion of T Cells

(84) At comparable concentrations (IL-2 200 ng/ml12 nM; Bispecific anti-IL-2R/c antibody 3 g/ml20 nM), the antibody triggers T cell proliferation to a lesser extent than IL-2 (FIG. 27A to 27C). The bispecific antibody shows a dose-dependent effect on T cell proliferation (FIG. 27A to 27D). In a non-specific stimulation setting, the CD8:CD4 cell ratio was not significantly different in the presence of the anti-IL-2R/c antibody as compared to when cells were cultured with IL-2 (FIG. 27D).

(85) 7.2 Stimulation of Regulatory T Cells

(86) Regulatory T cells (Tregs) express the high affinity IL-2 receptor sub-chain IL-2R. In a non-specific stimulation setting, IL-2 preferentially stimulates regulatory T cells amongst CD3+CD4+ T cells; such Treg expansion was not triggered by the bispecific anti-IL-2R/c antibody (FIG. 28).

(87) 7.3 Stimulation of Effector Vs Memory Cells

(88) With respect to memory CD8+ lymphocytes, the bispecific anti-IL-2R/c antibody triggers greater expansion of the effector memory CD8+ T cell subset, whilst triggering less expansion of the central memory and nave CD8+ T cell subsets as compared to expansion in response to stimulation with IL-2 (FIG. 29).

Example 8: Modulation of the Immune Response: Control of T Cell Expansion in a Specific Stimulation Setting

(89) PBMCs from an Epstein-Barr virus (EBV) seropositive volunteer donor were infected with EBV to make lymphoblastoid cell lines (LCLs). LCLs were sorted and -irradiated in order to inhibit their subsequent proliferation. Irradiated LCLs were co-cultured at a density of 110.sup.5 cells/ml with 210.sup.6 autologous PBMCs/ml for 2 weeks, in the presence of IL-2, the anti-IL-2R/c bispecific antibody or anti-CD3/CD28 beads (positive control). Cells were then analysed for proliferation and the proportions of different cell subsets.

(90) A cytotoxic killing assay was performed using a fluorescent peptide substrate for granzyme B and capsase 8. Expanded T-cells were co-incubated with live LCLs at a ratio of 2:1 for one hour. Killing was measured by analysis of peptide-fluorescent positive cells by flow cytometry, which indicated that cells were undergoing CTL-induced programmed cell death.

(91) 8.1 Expansion of T Cells

(92) The bispecific antibody triggers expansion of T cells, even at low concentrations. Anti-IL-2R/c bispecific antibody-mediated T cell expansion is slightly greater than expansion observed following stimulation with IL-2 (FIG. 30A). Whilst the antibody does not significantly influence the number of CD4+ T cells (FIG. 30B), the antibody elicits an increase in the number of CD8+ T cells to a greater extent than IL-2 (FIG. 30C), and hence increased the CD8:CD4 cell ratio (FIG. 30D).

(93) 8.2 Effects on T Cell Subsets

(94) At the highest concentration (1 g/ml), the anti-IL-2R/c antibody favours the expansion of effector CD8+ T cells over CD8+ memory cells as compared to stimulation with IL-2 (FIG. 31A). Compared to IL-2 stimulation, the anti-IL-2R/c antibody also triggers an increase in the CD8+PD-1+ subset (FIG. 31B), whilst decreasing the proportion of Tregs (FIG. 31C).

(95) 8.3 Cytotoxic T Lymphocyte-Mediated Killing

(96) The anti-IL-2R/c bispecific antibody is able to elicit CTL cytotoxicity. At comparable molarity (12 nM (200 ng/ml) for IL-2 vs. 7 nM (1 g/mL) for the antibody), the antibody-mediated cytotoxicity is lower than cytotoxicity triggered by IL-2 (FIG. 32).

(97) 8.4 Conclusion

(98) Taken together, the data suggest that the bispecific anti-IL-2R/c antibody triggers a different mechanism of action than that of IL-2. The antibody preferentially elicits expansion of effector CD8+ T cells. The antibody allows stimulation of cytotoxic T cells but does not preferentially stimulate Tregs as IL-2 does.

Example 9: Sequence Engineering to Improve Stability

(99) One of the greatest challenges whilst constructing bispecific antibodies is the stability of the heterogenic construct. Unlike monospecific IgGs, the present bispecific anti-IL-2R/c antibody is an artificial assembling of two different pairs of light/heavy chains.

(100) In order to improve the general stability of constructs, original antibody clones P2C4 and P1A3 were engineered to increase their thermostability.

(101) 9.1 Thermostable Clones

(102) Libraries of randomly mutagenised clones were built from the parent clones P2C4 and P1A3 and mutants were screened for binding to the respective targets in a two-round panning followed by ELISA. Binders were then subjected to heating to 55 C. The mutants still binding after heating were sequenced, and unique clones were identified.

(103) Thermostability of the clones was assessed after heating for 4 hours between 45 C. and 65 C., by measuring binding to their respective target, either c (FIGS. 33A and 33B) or IL-2R (FIGS. 34A and 34B), in ELISA. The mutated clones showed higher thermal stability than the parent clones.

(104) 9.1 Engraftment of Highly Stable Framework

(105) In order to further increase stability of the antibody, clones were engrafted in frameworks of antibodies that were known to be highly stable.

(106) P2C4 and P1A3 were engrafted into frameworks of antibodies known to have high stability. ELISA experiments were conducted to ensure that the new clones retained the ability to bind to IL-2R and c.

(107) Both P2C4_FW2 and P1A3_FW2 showed a dose-dependent binding profile to IL-2R and c respectively (FIGS. 35A and 35B).

Example 10: Binding of New Bispecific Constructs to IL-2R/c

(108) 10.1 Short Linker Between Variable and Constant Domains

(109) Bispecific antibody constructs were prepared including one of the following short linkers between the scFv and constant domain (antibody format: VH domain-linker-VL domain-short linker-hinge-CH2 domain (+LALA)-CH3 domain (+knob/hole+ cys)): NSGAGTAAA (SEQ ID NO:157) or GGGGSAAA (SEQ ID NO:158).

(110) Bispecific constructs with NSGAGTAAA (SEQ ID NO:157) or GGGGSAAA (SEQ ID NO:158) short linkers were generated and tested for binding to IL-2R and c by ELISA.

(111) The bispecific antibodies were found to bind to with similar affinity irrespective of the identity of the short linker (FIGS. 36A and 36B).

(112) Bispecific antibodies were constructed with the new sequences and binding was assessed by ELISA on either IL-2R or IL-2Rc. The constructs were found to bind similar or better affinity than the parent bispecific antibodies to IL-2R (FIGS. 37A and 37B).

Example 11: Effects on T Cell Expansion and Polarisation

(113) Assays using T cells were performed in order to measure the effect of the anti-IL-2R/c bispecific antibody on T cell expansion in vitro and its impact on antigen specific and non-specific qualitative polarisation and subset specificity. Peripheral blood from EBV-positive individuals was used to generate both EBV-transformed lymphoblastoid B-cell lines (LCLs) and EBV-specific CTL lines.

(114) Briefly, to generate LCLs, PBMCs were cultured for 1 week in the presence of cyclosporine and EBV, and for 2 additional weeks in refreshed media with cyclosporine but without EBV. After culture, cells were transferred to a G-Rex column and growth was monitored. For the generation of CTLs, LCLs were irradiated to act as an antigen source for CTLs. PBMCs were co-cultured with LCLs at an effector to stimulator (E:S) ratio of 40:1. Cells were stimulated by addition of IL-2, the anti-IL-2R/c bispecific antibody, or CD3/28 beads.

(115) After 7 days, cells underwent a media change and additional stimulations. At day 10, cells were analysed for lymphocyte expansion and phenotype by flow cytometry.

(116) Addition of the bispecific antibody was found to result in a significant increase in antigen-specific CD8+ T cell expansion as compared to expansion in response to stimulation with IL-2 (FIG. 38A). Furthermore, in vitro cultures showed improved CD8:CD4 ratios following antibody stimulation (FIG. 38B).

(117) The impact of the bispecific antibody on the expansion of regulatory T cells (Tregs) was then measured and compared to expansion of Tregs in response to stimulation with IL-2, in the antigen specific (autologous LCL co-culture) and non-specific (anti-CD3/CD28 microbead) settings. Addition of the bispecific antibody results in significantly reduced expansion of Tregs as compared to Treg expansion in response to IL-2, in both the non-specific (FIG. 39A) and antigen-specific stimulation settings (FIG. 39B).

Example 12: In Vivo Data in Non-Human Primates

(118) A dose escalating experiment was established in Cynomolgus macaques in order to measure the effects of intravenous (iv) injection of the anti-IL-2R/c bispecific antibody, its ability to drive proliferation of T cells and NK cells, and its potential toxicity through cytokine storm.

(119) Three macaques were administered a single dose of the anti-IL-2R/c antibody, intravenously through the femoral artery; macaque A received 1 mg/kg, macaque B received 5 mg/kg, and macaque C received 10 mg/kg. Blood was collected before antibody injection and at 1 h, 24 h, 72 h and 120 h post-injection.

(120) Vital signs and physical examinations were performed throughout the study and then for a further 3 weeks. PBMCs were isolated at all time points, leukocyte subsets were analysed by immune-staining and flow cytometry, and cell expansion was assessed by analysis of Ki-67 expression. Plasma cytokine levels were measured by Luminex at all time points.

(121) Veterinary physical examination indicated no abnormalities in general appearance, mucosal membranes, cardiovascular, respiratory, integumentary, alimentary, musculoskeletal, nervous, urogenital, auditory, or ocular systems. Animals displayed no clinical findings of febrile illness or depression. One animal (macaque B) showed mild weight loss from which he recovered during the course of the study. Animals showed no overt signs of toxicity commonly associated with IL-2 administration (PMID: 1418698 and 8454416).

(122) Consistent with these observations, cytokine analysis demonstrated only mild increases in inflammatory mediators post injection (FIGS. 40A to 40E). Flow cytometric analysis indicated a marked proliferation of CD4+ and CD8+ T cell populations (FIGS. 41A to 41C).

(123) NK cell proliferation was also observed in response to the antibody treatment (FIGS. 42A and 42B). It should be noted that this expansion was observed after a single dose of antibody, compared to continuous infusion or repeated doses required for IL-2, suggesting that anti-IL-2R/c bispecific antibody has a longer half-life than IL-2.