CHIMERIC ANTIGEN RECEPTOR SPECIFIC FOR HUMAN CD45RC AND USES THEREOF

Abstract

In the field of immunotherapy, a chimeric antigen receptor (CAR) specific for-human CD45RC. Also, the immune cells expressing this CAR and the use thereof as a medicament. In particular, the use of this CAR for preventing or treating CD45RC.sup.high-related diseases (including autoimmune diseases, undesired immune responses, monogenic diseases, lymphoma or cancer), or graft-versus-host disease (GVHD).

Claims

1.-34. (canceled)

35. A chimeric antigen receptor (CAR) specific for human CD45RC, wherein said CAR comprises: (a) at least one extracellular binding domain, wherein said binding domain binds to said human CD45RC, (b) optionally at least one extracellular hinge domain, (c) at least one transmembrane domain, and (d) at least one intracellular signaling domain, wherein the intracellular domain comprises at least one T cell primary signaling domain and optionally at least one T cell costimulatory signaling domain.

36. The CAR according to claim 35, wherein the extracellular binding domain comprises at least one antigen-binding fragment that binds to human CD45RC comprising: (a) a HCVR which comprises the following three CDRs: (i) V.sub.H-CDR1 of sequence SEQ ID NO: 1; (ii) V.sub.H-CDR2 with a sequence selected from the group comprising sequences SEQ ID NOs: 4, 5, 6, 8, 100, 116, 117, 118 and 119; and (iii) V.sub.H-CDR3 of sequence SEQ ID NO: 3; and (b) a LCVR which comprises the following three CDRs: (i) V.sub.L-CDR1 with a sequence selected from the group comprising sequences SEQ ID NO: 15 (SASSSVS-X.sub.12-YMH) and 18 (RASSSVS-X.sub.12-YMH), wherein Xu is absent or is selected from Asn (N), Ser (S) and Gly (G); (ii) VL-CDR2 with a sequence selected from the group comprising sequences SEQ ID NO: 16, 111, and 120; and (iii) V.sub.L-CDR3 of sequence SEQ ID NO: 17.

37. The CAR according to claim 35, wherein the extracellular binding domain comprises at least one antigen-binding fragment that binds to human CD45RC comprising: (a) a HCVR which comprises the following three CDRs: (i) V.sub.H-CDR1 of sequence SEQ ID NO: 1; (ii) V.sub.H-CDR2 with a sequence selected from the group comprising sequences SEQ ID NOs: 4 and 5; and (iii) V.sub.H-CDR3 of sequence SEQ ID NO: 3; and (b) a LCVR which comprises the following three CDRs: (i) V.sub.L-CDR1 of sequence SEQ ID NO: 15, wherein X.sub.12 is absent; (ii) V.sub.L-CDR2 of sequence SEQ ID NO: 16; and (iii) V.sub.L-CDR3 of sequence SEQ ID NO: 17.

38. The CAR according to claim 35, wherein the extracellular binding domain comprises at least one antigen-binding fragment that binds to human CD45RC comprising: (a) a HCVR which comprises the following three CDRs: (i) V.sub.Ht-CDR1 of sequence SEQ ID NO: 1; (ii) V.sub.H-CDR2 of sequence 4; and (iii) V.sub.H-CDR3 of sequence SEQ ID NO: 3; and (b) a LCVR which comprises the following three CDRs: (i) V.sub.L-CDR1 of sequence SEQ ID NO: 15, wherein X.sub.12 is absent; (ii) V.sub.L-CDR2 of sequence SEQ ID NO: 16, and (iii) V.sub.L-CDR3 of sequence SEQ ID NO: 17.

39. The CAR according to claim 35, wherein the extracellular binding domain comprises at least one antigen-binding fragment that binds to human CD45RC comprising: (a) a HCVR which comprises the following three CDRs: (i) V.sub.Ht-CDR1 of sequence SEQ ID NO: 1; (ii) V.sub.H-CDR2 with a sequence selected from the group comprising sequences SEQ ID NOs: 4, 6, and 100; and (iii) V.sub.H-CDR3 of sequence SEQ ID NO: 3; and (b) a LCVR which comprises the following three CDRs: (i) V.sub.L-CDR1 with a sequence selected from the group comprising sequences SEQ ID NOs: 15 and 18, wherein X.sub.12 is absent: (ii) V.sub.L-CDR2 with a sequence selected from the group comprising sequences SEQ ID NO: 16, 111, and 120; and (iii) V.sub.L-CDR3 of sequence SEQ ID NO: 17.

40. The CAR according to claim 35, wherein the extracellular binding domain comprises at least one antigen-binding fragment that binds to human CD45RC comprising: 1) a HCVR of sequence SEQ ID NO: 61 and a LCVR of sequence SEQ ID NO: 81; 2) a HCVR of sequence SEQ ID NO: 62 and a LCVR of sequence SEQ ID NO: 82; 3) a HCVR of sequence SEQ ID NO: 62 and a LCVR of sequence SEQ ID NO: 83; 4) a HCVR of sequence SEQ ID NO: 62 and a LCVR of sequence SEQ ID NO: 84: 5) a HCVR of sequence SEQ ID NO: 63 and a LCVR of sequence SEQ ID NO: 82; 6) a HCVR of sequence SEQ ID NO: 63 and a LCVR of sequence SEQ ID NO: 83; 7) a HCVR of sequence SEQ ID NO: 63 and a LCVR of sequence SEQ ID NO: 84: 8) a HCVR of sequence SEQ ID NO: 64 and a LCVR of sequence SEQ ID NO: 82; 9) a HCVR of sequence SEQ ID NO: 64 and a LCVR of sequence SEQ ID NO: 83; 10) a HCVR of sequence SEQ ID NO: 64 and a LCVR of sequence SEQ ID NO: 84; 11) a HCVR of sequence SEQ ID NO: 101 and a LCVR of sequence SEQ ID NO: 85; 12) a HCVR of sequence SEQ ID NO: 101 and a LCVR of sequence SEQ ID NO: 103; 13) a HCVR of sequence SEQ ID NO: 65 and a LCVR of sequence SEQ ID NO: 85; 14) a HCVR of sequence SEQ ID NO: 65 and a LCVR of sequence SEQ ID NO: 103; 15) a HCVR of sequence SEQ ID NO: 62 and a LCVR of sequence SEQ ID NO: 85; 16) a HCVR of sequence SEQ ID NO: 101 and a LCVR of sequence SEQ ID NO: 82; 17) a HCVR of sequence SEQ ID NO: 121 and a LCVR of sequence SEQ ID NO: 85; 18) a HCVR of sequence SEQ ID NO: 122 and a LCVR of sequence SEQ ID NO: 85; 19) a HCVR of sequence SEQ ID NO: 123 and a LCVR of sequence SEQ ID NO: 85; 20) a HCVR of sequence SEQ ID NO: 124 and a LCVR of sequence SEQ ID NO: 85; 21) a HCVR of sequence SEQ ID NO: 63 and a LCVR of sequence SEQ ID NO: 85; 22) a HCVR of sequence SEQ ID NO: 67 and a LCVR of sequence SEQ ID NO: 85; 23) a HCVR of sequence SEQ ID NO: 67 and a LCVR of sequence SEQ ID NO: 103; 24) a HCVR of sequence SEQ ID NO: 61 and a LCVR of sequence SEQ ID NO: 113; 25) a HCVR of sequence SEQ ID NO: 61 and a LCVR of sequence SEQ ID NO: 126; or 26) a HCVR and a LCVR comprising a sequence of the non-CDR regions sharing at least 70% of identity with the sequence of the non-CDR regions of the HCVR and LCVR according to 1) to 23).

41. The CAR according to claim 35, wherein the extracellular binding domain comprises at least one antigen-binding fragment that binds to human CD45RC comprising: (a) a HCVR which comprises the following three CDRs: (i) V.sub.H-CDR1 of sequence SEQ ID NO: 1; (ii) VH-CDR2 with a sequence selected from the group comprising sequences SEQ ID NOs: 4, 5, 6, 8, 100, 116, 117, 118 and 119; and (iii) V.sub.H-CDR3 of sequence SEQ ID NO: 3; and (b) a LCVR which comprises the following three CDRs: (i) V.sub.L-CDR1 with a sequence selected from the group comprising sequences SEQ ID NOs: 15 and 18, wherein X.sub.12 in SEQ ID NOs: 15 and 18 is selected from Asn (N), Ser (S) and Gly (G); (ii) V.sub.L-CDR2 of sequence SEQ ID NO: 16; and (iii) V.sub.L-CDR3 of sequence SEQ ID NO: 17.

42. The CAR according to claim 41, wherein the amino acid residue at Kabat position L71 of the LCVR is Phe (F).

43. The CAR according to claim 35, comprising: (i) an anti-human CD45RC scFv, (ii) a hinge domain derived from CD8α, (iii) a human CD8α transmembrane domain, and (iv) an intracellular signaling domain comprising a human CD28 signaling domain and a human CD3 zeta signaling domain.

44. The CAR according to claim 43, comprising: (i) an anti-human CD45RC scFv comprising a HCVR having the sequence of SEQ ID NO: 61 and a LCVR having the sequence of SEQ ID NO: 81, linked by a linker having the sequence of SEQ ID NO: 134, (ii) a hinge domain derived from CD8α having the sequence of SEQ ID NO: 145, (iii) a human CD8α transmembrane domain having the sequence of SEQ ID NO: 153, and (iv) an intracellular signaling domain comprising a human CD28 signaling domain having the sequence of SEQ ID NO: 167 and a human CD3 zeta signaling domain having the sequence of SEQ ID NO: 157.

45. A nucleic acid encoding the CAR according to claim 35.

46. An expression vector comprising the nucleic acid according to claim 45.

47. An immune cell population, engineered to express at the cell surface a CAR according to claim 35.

48. The immune cell population according to claim 47, wherein said immune cell population is a regulatory T cell population.

49. The immune cell population according to claim 48, wherein said regulatory T cell population is selected from the group consisting of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ Treg, Tr1 cells, TGF-β secreting Th3 cells, regulatory NKT cells, regulatory γδ T cells, regulatory CD8.sup.+ T cells, and double negative regulatory T cells.

50. A composition comprising at least one immune cell population engineered to express at the cell surface a CAR according to claim 35.

51. The composition according to claim 50, wherein said composition is a pharmaceutical composition further comprising at least one pharmaceutically acceptable excipient or carrier.

52. A method of treating a subject in need thereof comprising administering to said subject a therapeutically effective amount of the immune cell population according to claim 35, or a composition comprising at least one of said immune cell population.

53. A method of inducing immune tolerance, or of preventing or reducing transplant rejection, or of preventing or treating graft-versus-host disease (GVHD), or of preventing, reducing and/or treating a CD45RChigh-related condition selected from the group consisting of an autoimmune disease, an undesired immune response, a monogenic disease, lymphoma and cancer, in a subject in need thereof, said method comprising administering to said subject a therapeutically effective amount of the immune cell population according to claim 35, or a composition comprising at least one of said immune cell population.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[1700] FIG. 1 A1(1)-B(2) shows the expression level of CD45RC (as detected by ABIS-45RC or by the commercially available MT2 antibody) on different leukocyte types in human blood. Staining with ABIS-45RC or MT2 was realized on different cell types in total blood (EDTA) from healthy volunteers. Red blood cells were then lysed (Versalyse, Beckman Coulter) before cytometer analysis (Navios, Beckman Coulter). Cells were first gated on morphology, doublet cells and lived cells. (A(1) through A(3)) Representative dot plot analysis of CD45RC expression detected by ABIS-45RC or MT2 from one out of three healthy volunteers analyzed on different leukocyte types. ABIS-45RC is shown on left panels and MT2 on right panels. x-axis shows the fluorescence intensity of cell lineage markers labelling for each type of leukocytes as indicated; y-axis represents the fluorescence intensity of anti-CD45RC antibody labelling. The horizontal lines define cells with high, intermediate/low and negative levels of CD45RC expression as indicated in the upper left dot plot and numbers represent the percentage of cells in each category. (B(1) and B(2)) Mean expression+/− SEM of CD45RC.sup.high, CD45RC.sup.low and CD45RC.sup.− on different leukocyte types of three donors labelled with ABIS-45RC (B(1)) or MT2 (B(2)).

[1701] FIG. 2 shows that both ABIS-45RC and the commercial anti-CD45RC MT2 antibody compete for the same epitope. PBMCs were isolated from blood of healthy volunteers and T cells were labeled with an anti-CD3 labeled mAb, with chimeric ABIS-45RC (at the indicated concentrations) and anti-CD45RC (mouse clone MT2)-FITC labeled at 1.33 mg/mL. ABIS-45RC reactivity was revealed using a biotin donkey anti-human IgG+Strepta PerCP-Cy 5.5 secondary antibody. Numbers in the windows of the dot plots of the upper row represent the percentage of cells that were co-labeled by both antibodies.

[1702] FIG. 3 A-F shows that cytotoxicity induced by ABIS-45RC is higher compared to commercial anti-CD45RC MT2. PBMCs from healthy volunteers (n=3) were incubated at 37° C. with medium, isotype negative control (2.5 or 10 mg/mL), ABIS-CD45RC (2.5 or 10 mg/mL) or dexamethasone (10 mg/mL) as a positive control for the indicated time points and then cells were labeled with an anti-CD3-FITC mouse antibody and apoptotic cells by labeling with Annexin-V-PE. (A-E) Graphs indicate % of Annexin V.sup.+ cells in indicated cell populations. (F) Representative dot plots of Annexin V.sup.+ (early apoptotic) and DAPI.sup.+ (late apoptotic) cells. Numbers indicate the percentage of cells in each category.

[1703] FIG. 4 A-C shows the use of ABIS-45RC to treat GVHD in immune humanized NSG immunodeficient mice. (A) Experimental procedure showing that peripheral blood mononuclear cells (PBMCs) from healthy donor volunteers were infused intravenously (iv) (day 0) into previously (day −1) sublethaly (2 Gy) irradiated NSG immunodeficient mice. ABIS-45RC was administered intraperitoneally (ip) with the indicated protocol between day 0 to 20. Isotype control were human IgG (IVIg preparation) and was administered using the same protocol as ABIS-45RC. (B-C) Survival curves for NSG mice and statistics were analyzed using a Kaplan-Meier analysis (* p<0.01, ** p<0.001).

[1704] FIG. 5 A-J is a combination of flow cytometry dot plots, showing the reactivity against human T cells of humanized ABIS-45RC antibodies variants and of the murine ABIS-45RC antibody, at two concentrations (2 μg/mL on the left panel, 1 μg/mL on the right panel). (A) humanized ABIS-45RC variant A; (B) humanized ABIS-45RC variant B; (C) humanized ABIS-45RC variant C; (D) humanized ABIS-45RC variant D; (E) humanized ABIS-45RC variant E; (F) humanized ABIS-45RC variant F; (G) humanized ABIS-45RC variant G; (H) humanized ABIS-45RC variant H; (I) humanized ABIS-45RC variant I; (J) murine ABIS-45RC.

[1705] FIG. 6 shows two dot plots of flow cytometry showing that both ABIS-45RC (left panel) and engineered Asn/Phe ABIS-45RC (right panel) have an equivalent pattern of reactivity against human T cells.

[1706] FIG. 7 A-C shows the expression level of CD45RC on CD3.sup.+ leukocyte in human blood from three healthy volunteers. Cells were first gated on morphology, doublet cells and lived cells. (A) representative dot plot analysis of CD45RC expression detected by murine ABIS-45RC from one out of three healthy volunteers analyzed; (B) representative dot plot analysis of CD45RC expression detected by humanized ABIS-45RC variant A1 from one out of three healthy volunteers analyzed; (C) representative dot plot analysis of CD45RC expression detected by humanized ABIS-45RC variant A3 from one out of three healthy volunteers analyzed. x-axis shows the FSC; y-axis represents the fluorescence intensity of anti-CD45RC antibody labelling. The squares define cells with high, intermediate/low and negative levels of CD45RC expression as indicated and numbers represent the percentage of cells in each category.

[1707] FIG. 8 A-B shows that apoptosis induced by ABIS-45RC or by the humanized variants A1 and A3 is comparable. PBMCs from healthy volunteers were incubated at 37° C. with an isotype negative control (10 μg/mL), with murine ABIS-CD45RC (10 μg/mL), with the humanized variant A1 (10 μg/mL) or with the humanized variant A3 (10 μg/mL) for the indicated time points, and cells were then labeled with anti-CD3 and anti-CD45RA antibodies and apoptotic cells by labeling with Annexin-V-PE. The graphs indicate fold apoptosis in CD3.sup.+ CD45RA.sup.hi cells (A) and in CD3.sup.− cells (B), compared to the isotype control condition.

[1708] FIG. 9 shows the skin graft survival of treated humanized mice with anti-human CD45RC treatment. NSG mice transferred with total human PBMCs to induce human skin rejection were treated with murine ABIS-45RC or humanized variant A1, together with rapamycin (Rapa). Results are expressed in skin graft survival score.

[1709] FIG. 10 depicts a schematic view of a CAR structure and of an exemplary structure of a CD45RC-CAR. The CAR comprises an extracellular domain (e.g. a scFv CD45RC), optionally a hinge domain (e.g. derived from CD8α), a transmembrane (TM) domain (e.g. derived from CD8), a costimulatory intracellular domain (e.g. derived from CD28) and a primary signaling domain (e.g. derived from CD3ζ). In the same plasmid, GFP coding sequences may optionally be present immediately 3′ of the CD45RC-CAR sequences separated from them by a T2A self-splicing sequence (not drawn). Thus, GFP may be used as a surrogate marker of CAR-CD45RC expression.

[1710] FIG. 11 shows that HEK (human embryonic kidney 293 cells) cells transfected with the plasmid encoding CD45RC-CAR and GFP express GFP 3 days after transfection (left panel). As a positive control of transfection, HEK cells were transfected with a plasmid encoding for GFP only (right panel).

[1711] FIG. 12 shows that Jurkat cells can be transduced with the CD45RC-CAR and GFP encoding lentiviral vector after 6 days of culture in comparison with non-transduced Jurkat cells.

[1712] FIG. 13 shows that Jurkat cells expressed at cell surface the transduced CD45RC-CAR (as shown by the protein L staining) and that expression of the CAR correlate with expression of the GFP.

[1713] FIG. 14 shows that Jurkat cells transduced with the CD45RC-CAR lentiviral vector can induce apoptosis in human T cells. Human T cells were cultured in different cell ratio conditions with Jurkat cells transduced with CD45RC-CAR or Ctrl-CAR (a control CAR having different antigenic specificity). Apoptosis was evaluated by flow cytometry after 18h of culture. The anti-CD45RC monoclonal antibody (chimeric human IgG1) used to generate the CAR was used as a positive control and a human non-reactive IgG1 as an isotype control, Ctrl-CAR Jurkat expressing a CAR recognizing a target non-expressed in this assay and non-transduced Jurkat cells were used as a negative control.

[1714] FIG. 15 shows that Jurkat cells transduced with the CD45RC-CAR lentiviral vector can be activated after a contact with human T cells. Human T cells were cultured in different cell ratio conditions with Jurkat cells transduced with CD45RC-CAR (sorted or not based on the expression of GFP) or Ctrl-CAR. The mean fluorescence intensity (MFI) of CD69, a marker of T cell activation, was evaluated by flow cytometry after 18h of culture. Ctrl-CAR Jurkat sorted or not and non-transduced Jurkat were used as a negative control.

[1715] FIG. 16 depicts CD45RC-CAR binding to the CD45RC target. HEK 293T were non-transduced or transduced with CD45RC-CAR lentiviral vector and cultured for 5 days. Cells were then incubated with biotinylated CD45R-ABC protein and then stained with streptavidin-APC-Cy7 and analyzed by flow cytometry.

[1716] FIG. 17 A-B shows the expansion of CD4.sup.+Tregs transduced with CD45RC-CAR lentiviral vectors. CD4.sup.+Tregs were activated between day 0 and day 1 and then transduced twice with CD45RC-CAR on day 1 and 2. GFP+ cells were sorted on day 7. CD45RC-CAR CD4.sup.+Tregs were analyzed for GFP expression after 7 or 14 days of expansion: (A) Representative histograms and dot plot of GFP expression in CD45RC-CAR transduced CD4.sup.+Tregs at day 7 and 14; (B) Percentage of cells expressing the CD45RC-CAR at day 7 and 14.

[1717] FIG. 18 shows that CD45RC-CAR CD4.sup.+Tregs are specifically activated through the CAR. CD45RC-CAR CD4.sup.+Tregs and Ctrl-CAR CD4.sup.+Tregs expanded for 14 days were cultured for 24h with coated CD45R-ABC protein, in the presence of brefeldin A for the 4 last hours, and then analyzed by flow cytometry for activation markers. MFI of each marker is expressed as a ratio to unstimulated cells.

[1718] FIG. 19 A-C shows that CD45RC-CAR CD4.sup.+Tregs induce cell death of CD45RC.sup.high T cells through apoptosis. Apoptosis was analyzed on PBMC incubated 4h with different ratios of allogenic CD4.sup.+ Tregs transduced with CD45RC-CAR (average of 60% GFP+ transduced cells), Ctrl-CAR (>90% LNGFR+ transduced cells) or non-transduced CD4.sup.+Tregs, as well as with anti-CD45RC mAb (ABIS-45RC at 10 μg/mL) followed by staining with annexin V and DAPI. Results are expressed as relative proportion of annexin V.sup.+ cells among (A) T cells, (B) CD45RC.sup.low/neg T cells, (C) CD45RC.sup.high T cells. The grey diamond-shaped isolated point represents apoptosis in the presence of the anti-CD45RC mAb.

EXAMPLES

[1719] The present invention is further illustrated by the following examples.

[1720] Throughout the examples, the following nomenclature applies:

[1721] “ABIS-45RC”: the murine anti-hCD45RC antibody of the invention, comprising: [1722] a heavy chain variable region with SEQ ID NO: 61; [1723] a heavy chain constant region with SEQ ID NO: 93; [1724] a light chain variable region with SEQ ID NO: 81; and [1725] a light chain constant region with SEQ ID NO: 94.

[1726] “Anti-45RC Variant A”: a humanized variant of ABIS-45RC, comprising: [1727] a heavy chain variable region with SEQ ID NO: 62; [1728] a heavy chain constant region with SEQ ID NO: 91; [1729] a light chain variable region with SEQ ID NO: 82; and [1730] a light chain constant region with SEQ ID NO: 92.

[1731] “Anti-45RC Variant B”: a humanized variant of ABIS-45RC, comprising: [1732] a heavy chain variable region with SEQ ID NO: 62; [1733] a heavy chain constant region with SEQ ID NO: 91; [1734] a light chain variable region with SEQ ID NO: 83; and [1735] a light chain constant region with SEQ ID NO: 92.

[1736] “Anti-45RC Variant C”: a humanized variant of ABIS-45RC, comprising: [1737] a heavy chain variable region with SEQ ID NO: 62; [1738] a heavy chain constant region with SEQ ID NO: 91; [1739] a light chain variable region with SEQ ID NO: 84; and [1740] a light chain constant region with SEQ ID NO: 92.

[1741] “Anti-45RC Variant D”: a humanized variant of ABIS-45RC, comprising: [1742] a heavy chain variable region with SEQ ID NO: 63; [1743] a heavy chain constant region with SEQ ID NO: 91; [1744] a light chain variable region with SEQ ID NO: 82; and [1745] a light chain constant region with SEQ ID NO: 92.

[1746] “Anti-45RC Variant E”: a humanized variant of ABIS-45RC, comprising: [1747] a heavy chain variable region with SEQ ID NO: 63; [1748] a heavy chain constant region with SEQ ID NO: 91; [1749] a light chain variable region with SEQ ID NO: 83; and [1750] a light chain constant region with SEQ ID NO: 92.

[1751] “Anti-45RC Variant F”: a humanized variant of ABIS-45RC, comprising: [1752] a heavy chain variable region with SEQ ID NO: 63; [1753] a heavy chain constant region with SEQ ID NO: 91; [1754] a light chain variable region with SEQ ID NO: 84; and [1755] a light chain constant region with SEQ ID NO: 92.

[1756] “Anti-45RC Variant G”: a humanized variant of ABIS-45RC, comprising: [1757] a heavy chain variable region with SEQ ID NO: 64; [1758] a heavy chain constant region with SEQ ID NO: 91; [1759] a light chain variable region with SEQ ID NO: 83; and [1760] a light chain constant region with SEQ ID NO: 92.

[1761] “Anti-45RC Variant H”: a humanized variant of ABIS-45RC, comprising: [1762] a heavy chain variable region with SEQ ID NO: 64; [1763] a heavy chain constant region with SEQ ID NO: 91; [1764] a light chain variable region with SEQ ID NO: 84; and [1765] a light chain constant region with SEQ ID NO: 92.

[1766] “Anti-45RC Variant I”: a humanized variant of ABIS-45RC, comprising: [1767] a heavy chain variable region with SEQ ID NO: 64; [1768] a heavy chain constant region with SEQ ID NO: 91; [1769] a light chain variable region with SEQ ID NO: 82; and [1770] a light chain constant region with SEQ ID NO: 92.

[1771] “Anti-45RC Variant A1”: a humanized variant of ABIS-45RC, comprising: [1772] a heavy chain variable region with SEQ ID NO: 101; [1773] a heavy chain constant region with SEQ ID NO: 91; [1774] a light chain variable region with SEQ ID NO: 85; and [1775] a light chain constant region with SEQ ID NO: 92.

[1776] “Anti-45RC Variant A2”: a humanized variant of ABIS-45RC, comprising: [1777] a heavy chain variable region with SEQ ID NO: 101; [1778] a heavy chain constant region with SEQ ID NO: 91; [1779] a light chain variable region with SEQ ID NO: 103; and [1780] a light chain constant region with SEQ ID NO: 92.

[1781] “Anti-45RC Variant A3”: a humanized variant of ABIS-45RC, comprising: [1782] a heavy chain variable region with SEQ ID NO: 65; [1783] a heavy chain constant region with SEQ ID NO: 91; [1784] a light chain variable region with SEQ ID NO: 85; and [1785] a light chain constant region with SEQ ID NO: 92.

[1786] “Anti-45RC Variant A4”: a humanized variant of ABIS-45RC, comprising: [1787] a heavy chain variable region with SEQ ID NO: 65; [1788] a heavy chain constant region with SEQ ID NO: 91; [1789] a light chain variable region with SEQ ID NO: 103; and [1790] a light chain constant region with SEQ ID NO: 92.

[1791] “Anti-45RC Variant A5”: a humanized variant of ABIS-45RC, comprising: [1792] a heavy chain variable region with SEQ ID NO: 62; [1793] a heavy chain constant region with SEQ ID NO: 91; [1794] a light chain variable region with SEQ ID NO: 85; and [1795] a light chain constant region with SEQ ID NO: 92.

[1796] “Anti-45RC Variant A6”: a humanized variant of ABIS-45RC, comprising: [1797] a heavy chain variable region with SEQ ID NO: 101; [1798] a heavy chain constant region with SEQ ID NO: 91; [1799] a light chain variable region with SEQ ID NO: 82; and [1800] a light chain constant region with SEQ ID NO: 92.

[1801] “Anti-45RC Variant A7”: a humanized variant of ABIS-45RC, comprising: [1802] a heavy chain variable region with SEQ ID NO: 121; [1803] a heavy chain constant region with SEQ ID NO: 91; [1804] a light chain variable region with SEQ ID NO: 85; and [1805] a light chain constant region with SEQ ID NO: 92.

[1806] “Anti-45RC Variant A8”: a humanized variant of ABIS-45RC, comprising: [1807] a heavy chain variable region with SEQ ID NO: 122; [1808] a heavy chain constant region with SEQ ID NO: 91; [1809] a light chain variable region with SEQ ID NO: 85; and [1810] a light chain constant region with SEQ ID NO: 92.

[1811] “Anti-45RC Variant A9”: a humanized variant of ABIS-45RC, comprising: [1812] a heavy chain variable region with SEQ ID NO: 123; [1813] a heavy chain constant region with SEQ ID NO: 91; [1814] a light chain variable region with SEQ ID NO: 85; and [1815] a light chain constant region with SEQ ID NO: 92.

[1816] “Anti-45RC Variant A10”: a humanized variant of ABIS-45RC, comprising: [1817] a heavy chain variable region with SEQ ID NO: 124; [1818] a heavy chain constant region with SEQ ID NO: 91; [1819] a light chain variable region with SEQ ID NO: 85; and [1820] a light chain constant region with SEQ ID NO: 92.

[1821] “Anti-45RC Variant D1”: a humanized variant of ABIS-45RC, comprising: [1822] a heavy chain variable region with SEQ ID NO: 63; [1823] a heavy chain constant region with SEQ ID NO: 91; [1824] a light chain variable region with SEQ ID NO: 85; and [1825] a light chain constant region with SEQ ID NO: 92.

[1826] “Anti-45RC Variant I1”: a humanized variant of ABIS-45RC, comprising: [1827] a heavy chain variable region with SEQ ID NO: 67; [1828] a heavy chain constant region with SEQ ID NO: 91; [1829] a light chain variable region with SEQ ID NO: 85; and [1830] a light chain constant region with SEQ ID NO: 92.

[1831] “Anti-45RC Variant I2”: a humanized variant of ABIS-45RC, comprising: [1832] a heavy chain variable region with SEQ ID NO: 67; [1833] a heavy chain constant region with SEQ ID NO: 91; [1834] a light chain variable region with SEQ ID NO: 103; and [1835] a light chain constant region with SEQ ID NO: 92.

[1836] “MT2”: a murine anti-hCD45RC antibody commercially available at OriGene, under Ref. AM39022PU-N.

[1837] “Engineered Asn/Phe ABIS-45RC”: the murine ABIS-45RC, chimerized by engineering its LCVR by insertion of one residue in the CDR1 and substitution of one residue in the FR3 (as described in Example 8). Engineered Asn/Phe ABIS-45RC comprises: [1838] a heavy chain variable region with SEQ ID NO: 61; [1839] a heavy chain constant region with SEQ ID NO: 93; [1840] a light chain variable region with SEQ ID NO: 71 with X.sub.12 being Asn (N); and [1841] a light chain constant region with SEQ ID NO: 94.

[1842] “Engineered Ser/Phe ABIS-45RC”: the murine ABIS-45RC, chimerized by engineering its LCVR by insertion of one residue in the CDR1 and substitution of one residue in the FR3 (as described in Example 8). Engineered Ser/Phe ABIS-45RC comprises: [1843] a heavy chain variable region with SEQ ID NO: 61; [1844] a heavy chain constant region with SEQ ID NO: 93; [1845] a light chain variable region with SEQ ID NO: 71 with X.sub.12 being Ser (S); and [1846] a light chain constant region with SEQ ID NO: 94.

[1847] “Engineered Gly/Phe ABIS-45RC”: the murine ABIS-45RC, chimerized by engineering its LCVR by insertion of one residue in the CDR1 and substitution of one residue in the FR3 (as described in Example 8). Engineered Gly/Phe ABIS-45RC comprises: [1848] a heavy chain variable region with SEQ ID NO: 61; [1849] a heavy chain constant region with SEQ ID NO: 93; [1850] a light chain variable region with SEQ ID NO: 71 with X.sub.12 being Gly (G); and [1851] a light chain constant region with SEQ ID NO: 94.

[1852] “Chimeric N50A ABIS-45RC”: a chimeric variant of the murine ABIS-45RC, comprising the murine ABIS-45RC heavy chain and a humanized light chain. Chimeric N50A ABIS-45RC comprises: [1853] a heavy chain variable region with SEQ ID NO: 61; [1854] a heavy chain constant region with SEQ ID NO: 93; [1855] a light chain variable region with SEQ ID NO: 113; and [1856] a light chain constant region with SEQ ID NO: 92.

[1857] “Chimeric S52A ABIS-45RC”: a chimeric variant of the murine ABIS-45RC, comprising the murine ABIS-45RC heavy chain and a humanized light chain. Chimeric S52A ABIS-45RC comprises: [1858] a heavy chain variable region with SEQ ID NO: 61; [1859] a heavy chain constant region with SEQ ID NO: 93; [1860] a light chain variable region with SEQ ID NO: 126; and [1861] a light chain constant region with SEQ ID NO: 92.

[1862] “Chimeric N50X ABIS-45RC”: a chimeric variant of the murine ABIS-45RC, comprising the murine ABIS-45RC heavy chain and a humanized light chain. Chimeric N50X ABIS-45RC comprises: [1863] a heavy chain variable region with SEQ ID NO: 61; [1864] a heavy chain constant region with SEQ ID NO: 93; [1865] a light chain variable region with SEQ ID NO: 129, with X.sub.13 being any amino acid but Ala (A) or Asn (N); and [1866] a light chain constant region with SEQ ID NO: 92.

Example 1

[1867] Reactivity of ABIS-45RC

[1868] Material and Methods

[1869] PBMC Staining and Data Acquisition

[1870] 50 μL or 100 μL of fresh EDTA whole blood were stained with combinations of appropriate monoclonal antibodies (Abs) followed by erythrocyte lysis (versalyse, Beckman Coulter). After washing, cells were analyzed on a Navios flow cytometer and data analyzed using Kaluza software (Beckman Coulter, Marseille, France) and FlowJo Software (Tree Star Inc).

[1871] Antibodies and Flow Cytometry

TABLE-US-00207 TABLE 6 Antibodies used for flow cytometry analysis. For each antibody in the left column, the clone used is given in the right column. Antibody (specificity) Clone CD117 104D2D1 CD11c BU15 CD123 9F5 CD127 R34.34 CD14 RMO52 CD16 3G8 CD161 191B8 CD19 SJ25C1 CD25 2A3 CD3 SK7 CD3 UCHT1 CD336 Z231 CD4 13B8.2 CD45 J.33 CD56 N901 CD56 NCAM162 CD8 B9.11 CD8 SK1 Cocktail ILCs — CRTH2 BM16 HLADR L243 Lineage1 DCs — TCRab IP26A TCRgd IMMU510 Va24 6B11 Va7, 2 REA179 CD45RC MT2 CD45RC ABIS Streptavidine Alexa fluor 405

[1872] Results

[1873] As a first screening, ABIS-45RC did not react against CD45RC− cells sorted using the commercial anti-CD45RC mAb MT2 clone suggesting that ABIS-45RC could recognize CD45RC (data not shown).

[1874] To further characterize ABIS-45RC, the inventors analyzed its reactivity with human PBMCs.

[1875] As shown in FIGS. 1A and B, a fraction of T CD4.sup.+ and T CD8.sup.+ cells is ABIS-45RC.sup.high whereas the rest is ABIS-45RC.sup.low or ABIS-45RC.sup.−. Most B and NK cells as well as pDCs were ABIS-45RC.sup.high Most of NKT, iNKT MAIT, ILC2, ILC3 and of CD14.sup.int,CD16.sup.+ monocytes were ABIS-45RC.sup.high or ABIS-45RC.sup.low. CD14.sup.highCD16.sup.− monocytes, mDC, basophils and neutrophils were predominantly ABIS-45RC.sup.−. CD4.sup.+ Tregs and CD8.sup.+ Foxp3.sup.+ Tregs were largely ABIS-45RC.sup.low/-.

[1876] The analysis of the major PBMCs populations showed that ABIS-45RC had a pattern of reactivity comparable to the commercial anti-CD45RC mouse MT2 antibody (FIG. 1 and Picarda et al., 2017. JCI Insight. 2(3):e90088).

Example 2

[1877] Comparison of ABIS-45RC and the Commercial Anti-CD45RC Antibody “MT2”

[1878] Material and Methods

[1879] PBMC Isolation

[1880] Blood healthy volunteers is collected and peripheral blood mononuclear cells (PBMC) were isolated by Ficoll gradient centrifugation, which enables removal of unwanted fractions of blood products such as granulocytes, platelets and reaming red blood cell contaminants.

[1881] Antibodies and Flow Cytometry

[1882] Human PBMC were labeled with the ABIS-45RC antibody (at the indicated concentrations), an anti-CD3 antibody and an anti-CD45RC (mouse clone MT2, Biolegend)-FITC labeled at 1.33 mg/mL. The ABIS-45RC reactivity was revealed using a biotin donkey anti-human IgG.sup.+ Streptavidin PerCP-Cy 5.5 secondary antibody.

[1883] A Canto II cytometer (BD Biosciences) was used to measure fluorescence intensity and data were analyzed using the FLOWJO software (Tree Star Inc.). Cells were first gated by their morphology and dead cells were excluded by selecting DAPI-negative cells.

[1884] Cytotoxicity Analysis

[1885] Human PBMCs were incubated with medium at 37° C., isotype control antibody (Ms IgG1, clone 107.3, 10 μg/ml), ABIS-45RC or anti-CD45RC (mouse clone MT2) at 2.5 or 10 μg/ml for 10 minutes to 18 hours. Then, cells were stained with anti-CD3 (clone SK7, BD Biosciences), annexin-V, and DAPI. Percentage of apoptosis was obtained by gating on Annexin V.sup.+ and DAPI.sup.+ cells among T or non-T cells by flow cytometry.

[1886] Results

[1887] Both ABIS-45RC and Commercial Anti-CD45RC MT2 Antibodies Compete for the Same Epitope

[1888] A shown in FIG. 2, co-labelling with the commercial anti-CD45RC MT2 clone showed that both antibodies competed and thus recognized the same or close epitope of human CD45RC.

[1889] Cytotoxicity Induced by ABIS-45RC is Higher Compared to Commercial Anti-CD45RC

[1890] As shown in FIG. 3, ABIS-45RC was cytotoxic to T cells but not non-T cells.

[1891] Moreover, the T cells cytotoxicity was directly correlated to the level of CD45RC expression and importantly, ABIS-45RC performed better at 2.5 μg/mL as compared to the MT2 clone at 10 μg/mL.

Example 3

[1892] Affinity of ABIS-45RC

[1893] Material and Methods

[1894] Briefly, 1×10.sup.7 CD45RC.sup.high PBMCs or CHO cells expressing CD45RC after plasmid transfection were solubilized using the Mem-PER membrane isolation kit (Thermo-fisher). ABIS-45RC was immobilized on a biochip CM5 and cell membranes were incubated at 25° C. to measure affinity constants using single cycle kinetics and calibration free concentration analysis on a BIAcore 3000 and a BIAcore T200.

[1895] Results

[1896] Measurement of the affinity of CD45RC antibody was assessed by surface Plasmon Resonance (SPR), a technology for characterizing antibody-antigen interactions, and revealed an affinity (K.sub.D) of 5×10.sup.−8 M, with a K.sub.on of 2.91×10.sup.5 M.sup.−1.Math.sec.sup.−1 and a K.sub.off of 1.44×10.sup.−2 sec.sup.−1.

Example 4

[1897] Treatment of Graft-Versus-Host-Disease (GVHD) with ABIS-45RC

[1898] Material and Methods

[1899] PBMC Isolation

[1900] Blood was collected at the Établissement Français du Sang (Nantes, France) from healthy individuals. Written informed consent was provided according to institutional guidelines. PBMC were isolated by Ficoll-Paque density-gradient centrifugation (Eurobio, Courtaboeuf, France). Remaining red cells and platelets were eliminated with a hypotonic solution and centrifugation.

[1901] Animals

[1902] 8- to 12-week-old NOD/SCID/IL2Rγ.sup./- (NSG) mice were bred in our own animal facilities in SPF conditions (accreditation number C44-278).

[1903] GVHD Model

[1904] Adult NSG immunodeficient mice were whole-body sublethaly irradiated (irradiation dose of 2 Gy at day −1) to induce lesions in tissues that will favor the development of GVHD. The following day (day 0), 1.5×10.sup.7 PBMCs (including CD45RC.sup.high and CD45RC.sup.low/- T cells) from healthy volunteers were injected intravenously in these mice. Human PBMCs, and in particular T cells, react against and attack mouse tissues inducing lesions. These T cells and the lesions observed in liver, intestine, lungs and skin mimic the GVHD observed following bone marrow transplantation in humans or other GVHD experimental systems using rodents as donors and recipients. In particular, these tissue lesions typically induce a body weight loss that begins—depending on the number of PBMCs injected and in our experimental system—around day 13 after injection of the PBMCs. Body weight loss is monitored daily and animals are sacrificed when it drops to 20% of the original body weight to avoid unnecessary suffering.

[1905] Treatment

[1906] NSG mice were treated intraperitoneally with purified ABIS-45RC, with MT2 anti-CD45RC antibody or with an irrelevant control (an IVIg preparation used clinically comprising human purified IgG, and containing predominantly IgG1 antibodies) at 0.8 mg/kg from day 0 and every 2.5 days during 20 days.

[1907] NSG mice treated with ABIS-45RC or control antibodies also received intraperitoneally rapamycin from day 0 to day 10 at a suboptimal dose of 0.4 mg/day.

[1908] The experimental procedure is summarized in FIG. 4A.

[1909] Results

[1910] Treatment with PBMCs only induced weight loss, initiated around day 14, and, as shown in FIG. 4, death of all mice by day 33 (median survival: 11 days (FIG. 4B) to 15 days (FIG. 4C) days).

[1911] Treatment with control antibody and rapamycin only prolonged survival without reaching statistical significance (median survival: 21 days (FIG. 4C)).

[1912] While treatment with MT2 significantly prolonged survival of the mice (median survival: 19 days (FIG. 4B)), treatment with ABIS-45RC significantly increased this prolongation of survival of the mice up to 72 days (median survival: 35 days (FIG. 4B)).

[1913] Finally, combinatorial administration of ABIS-45RC with rapamycin completely prevented death as a consequence of GVHD (100% survival, FIG. 4C).

Example 5

[1914] Humanization of ABIS-45RC

[1915] The design for the humanization of ABIS-45RC by grafting of the CDRs into human germline antibody sequences was undertaken. ABIS-45RC was humanized by grafting the three CDRs from the LCVR (with SEQ ID NOs: 15, 16 and 17) into a human germline LCVR that was as homologous as possible to ABIS-45RC's LCVR. Similarly, the three CDRs from the HCVR (with SEQ ID NOs: 1, 4 and 3) were grafted into a human germline HCVR that was as homologous as possible to ABIS-45RC's HCVR.

[1916] In addition, a few amino acid residues in the framework regions (FR) of the selected human germline variable regions were changed to the amino acid residues that were present in the murine variable regions (so called back-mutations). Based upon information on the structure of immunoglobulin variable regions, and with the guidance of an homology molecular model of the Fv of ABIS-45RC, these few residues in the FRs were identified as having key roles in either maintaining the CDRs in the right conformation or in HCVR/LCVR packing, and thus they were retained in a first humanized version (version A) or substituted with their human germline counterparts, if possible, in subsequent humanized version (versions B and C). Under guidance of the homology molecular model, in versions B and C, when judged possible, the CDR residues were also substituted for their human germline counterparts.

[1917] Homology Model Building

[1918] A model of ABIS-45RC was constructed according to established protocols (Ramos, 2012. Methods Mol Biol. 907:39-55).

[1919] Light Chain

[1920] In this section, unless specified otherwise, amino acid numbering is based on SEQ ID NO: 81.

[1921] The LCVR's framework residues were used to search the sequences of solved antibody structures via protein BLAST. The top hits were Protein Data Bank (PDB) ID: 4NCC (2.50 Å resolution) having 83 of 89 FR residues identical, and 85 out of 89 FR residues similar, to those of ABIS-45RC's LCVR, and PDB ID: 1QOK (2.40 Å resolution) having 83 of 87 framework residues identical, and 84 out of 87 similar, to those of ABIS-45RC's LCVR.

[1922] The sequences of these two structures both differed from that of ABIS-45RC's LCVR (with SEQ ID NO: 81) with the substitutions T9A, T39P, R44K, N49S and P54A. In addition, PDB ID: 4NCC differed in sequence from ABIS-45RC's LCVR with the substitutions L95F, A99G and L105I.

[1923] A comparison of the two structures showed high homology. However, the carbon chains adopted slightly different conformations in the regions A13-E17 and E104-K106.

[1924] Based upon the results of the subsequent CDR searches, and the presence of the two N-terminal residues, the LCVR of the structure of PDB ID: 4NCC was selected as the LCVR framework template, and the rotameric conformation of L105I was selected (in PyMol) with reference to PDB ID: 1QOK.

[1925] Subsequently, the sequences of ABIS-45RC LCVR's CDR1, CDR2 and CDR3, with the addition of two residues on each end, were used to search the sequences of solved antibody structures via protein BLAST.

[1926] For CDR1, the top hits consisted of a cluster of sequences having 9 out of 9 identical residues. Amongst these were PDB ID: 4NCC and PDB ID: 1QOK. Thus, the PDB ID: 4NCC structure was adopted as the template for CDR1.

[1927] For CDR2, the top hits consisted of a cluster of sequences having 6 out of 7 identical residues. Amongst these were again PDB ID: 4NCC and PDB ID: 1QOK. The PDB ID: 4NCC structure was therefore also adopted as the template for CDR2.

[1928] For CDR3, the top hit, containing no gaps, was PDB ID: 1QOK, having 13 out of 13 identical residues. PDB ID: 4NCC was however a close second, having 12 out of 13 identical residues. A comparison of the two structures showed essentially identical conformations, excepting for the L95F substitution. Thus, the PDB ID: 4NCC structure was adopted as the template for CDR3, and the rotameric conformation of L95F was selected (in PyMol) with reference to PDB ID: 1QOK.

[1929] It was thus not necessary to fit any CDR templates to the LCVR framework template because PDB ID: 4NCC was selected as the primary template for all of the LCVR's CDRs.

[1930] Finally, the LCVR partial model was manually subjected to mutagenesis at 8 positions (in PyMol), with selection of optimal rotamers, in order to match the ABIS-45RC LCVR sequence.

[1931] Heavy Chain

[1932] In this section, unless specified otherwise, amino acid numbering is based on SEQ ID NO: 61.

[1933] Next, the HCVR's framework residues were used to search the sequences of solved antibody structures via protein BLAST. The top hit was PDB ID: 30PZ (3.40 Å resolution), having 84 out of 90 framework residues identical, and 85 out of 90 similar, to those of ABIS-45RC's HCVR.

[1934] Since PDB ID: 30PZ was missing the N terminal residue, and was resolved with fairly poor resolution, additional hits with the highest identity/similarity scores were also surveyed. The top amongst these were PDB ID: 4CAD (2.50 Å resolution), having 78 out of 91 framework residues identical, and 87 out of 91 similar, to those of ABIS-45RC's HCVR; and PDB ID: 1RUR (1.50 Å resolution), having 75 out of 91 framework residues identical, and 87 out of 91 similar, to those of ABIS-45RC's HCVR.

[1935] A comparison of the PDB ID: 30PZ and PDB ID: 4CAD structures showed high homology with alternative residue rotamers being the principal differences.

[1936] A comparison of the PDB ID: 30PZ and PDB ID: 1RUR structures similarly showed high homology; however, there was a significant conformational change in the V.sub.H-FR2 loop L45-G.sub.49 relative to the PDB ID: 30PZ and PDB ID: 4CAD structures.

[1937] Further, based upon sequence, ABIS-45RC's HCVR and PDB ID: 4CAD were predicted to exhibit the Honegger Type III (Honegger & Pluckthun, 2001. J Mol Biol. 309(3):687-99) conformation of the N-terminal strand 5-12 because of the presence of a glutamine in position 6. However, PDB ID: 1RUR was predicted to exhibit the Honegger Type I conformation, due to the presence of a glutamic acid in position 6. Nevertheless, the three structures exhibited the identical conformation of the 5-12 strand. Also, the sequences of ABIS-45RC's HCVR, PDB ID: 4CAD and PDB ID: 1RUR were predicted to adopt the K-form (kinked base conformation) defined by the revised Shirai's rules for HCVR's CDR3 (Kuroda et al., 2008. Proteins. 73(3):608-20).

[1938] Based upon the results of the subsequent CDR searches, higher overall sequence similarity, structural concordance with PDB ID: 30PZ, and higher experimental resolution, the HCVR of the structure PDB ID: 1RUR was selected as the HCVR framework template; however, the 45-49 loop of PDB ID: 4CAD (having the same conformation as that of PDB ID: 30PZ) was substituted for that of PDB ID: 1RUR in the HCVR template using two residues N and C-terminal overhangs on the 45-49 ends to anchor the loop template fragment to the framework template.

[1939] Subsequently, the sequences of HCVR's CDR1, CDR2 and CDR3, with the addition of two residues on each end, were used to search the sequences of solved antibody structures via protein BLAST.

[1940] For CDR1, there was a cluster of sequence hits having 9 out of 12 identical residues. Amongst these was PDB ID: 1RUR. Thus, the PDB ID: 1RUR structure was selected as the template for CDR1.

[1941] For CDR2, the top hit was PDB ID: 3NTC (1.55 Å resolution) having 8 out of 12 residues identical, and 9 out of 12 similar, to those of ABIS-45RC's HCVR. However, PDB ID: 1RUR was a close second having 7 of 12 residues identical, and 9 out of 12 similar, to those of ABIS-45RC's HCVR. Comparison of the two structures showed essentially identical conformations, and the higher identity of PDB ID: 3NTC was due to 2 C terminal residues added to the CDR2 sequence for purposes of BLAST search. Thus, the PDB ID: 1RUR structure was preferred as the template for CDR2.

[1942] For CDR3, the top two hits, containing no gaps, were PDB ID: 1NGY (2.20 Å resolution) and PDB ID: 1NGZ (1.60 Å resolution), both having 8 out of 11 residues identical, and 9 out of 11 similar, to those of ABIS-45RC's HCVR. A comparison of the two structures showed a significantly different mainchain conformation. Without willing to be bound to a theory, the Inventors hypothesized that this difference might be due to the residue in position 101. In PDB ID: 1NGY, a larger methionine cannot adopt the orientation of the smaller serine of PDB ID: 1NGZ, which directs its sidechain into the core of the protein. Since the desired substitution to match the ABIS-45RC's HCVR sequence is F101, the PDB ID: 1NGY structure was adopted as the template for CDR3. Next, in order to complete the HCVR partial model, the CDR3 template was grafted onto the modified PDB ID: 1RUR HCVR template using the two residues overhang on its ends to anchor the CDR template fragment to the framework template (in PyMol).

[1943] Finally, the HCVR partial model was manually subjected to mutagenesis at 23 positions (in PyMol), with selection of optimal rotamers, in order to match the ABIS-45RC's HCVR sequence.

[1944] Final Model Assembly

[1945] Subsequently, the best tertiary arrangement of the HCVR and LCVR partial models were selected to assemble the final model. The HCVR and LCVR template sequences were submitted to the Packing Angle Prediction Server (PAPS) (Abhinandan & Martin, 2010. Protein Eng Des Sel. 23(9):689-97) to find a predicted best-fit tertiary arrangement. The PAPS server predicted that the solved antibody structure PDB ID: 1MNU, with a relative packing angle of −45.6°, would provide the best tertiary arrangement of HCVR and LCVR. Thus, the final model was assembled by fitting the backbone coordinates of the conserved anchor segments of the HCVR and LCVR partial models to PDB ID: 1MNU (in PyMol).

[1946] Lastly, the coordinates of this final model were subjected to a round of energy minimization employing GROMACS (Van der Spoel et al., 2005. J Comput Chem. 26(16):1701-18) with the GROMOS96 force-field (Scott et al., 1999. J Phys Chem A. 103(19):3596-3607).

[1947] Human Germlines

[1948] For the design of CDR-grafted versions of ABIS-45RC's HCVR and LCVR, two times three human germlines were selected: [1949] IGHV1-2*01, IGHV5-51*01 and IGHV3-11*05 for the HCVR; and [1950] IGKV1-9*01, IGKV6-21*02 and IGKV3-11*01 for the LCVR.

[1951] Design of humanized HCVR and LCVR The humanized versions A for both HCVR and LCVR are conservative versions that explicitly minimize and/or avoid alteration of CDR residues. These versions are thus expected to give a similar or better binding and/or potency activity as a chimeric antibody (ABIS-45RC's HCVR [SEQ ID NO: 61] and LCVR [SEQ ID NO: 81] fused to human constant regions [SEQ ID NOs: 91 and 92]).

[1952] The humanized versions B for both HCVR and LCVR are designed to reach a percentage of sequence identity with the closest human germline of at least 85%. This can be achieved by germlining (i.e., substituting the mouse residue with the corresponding human germline residue) FR and/or CDR amino acid residues. 85% is the cut-off percentage identity necessary to get the substem -zu- for “humanized”, denomination according to the 2014 World Health Organization (WHO) guidance on antibody International Nonproprietary Names (INN).

[1953] The humanized versions C for both HCVR and LCVR are designed to reach the highest degree of humanness (i.e., the highest degree of sequence identity with the corresponding human germline). Following inspection of the homology molecular model, a number of residues have been identified as candidates for germlining. Therefore, all the residues that could reasonably be germlined have been taken into consideration.

[1954] HCVR, Using IGHV1-2*01

[1955] To design humanized HCVR version A from IGHV1-2*01, the murine CDRs (with SEQ ID NOs: 1, 4 and 3) were grafted into IGHV1-2*01 and 4 residues in FR2 and FR3 were back-mutated to the parental murine residues, to maintain the full activity of the antibody. These residues are 148, L70, A72 and V97 in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 62 and shares 81.6% sequence identity with IGHV1-2*01 human germline.

[1956] To design humanized HCVR version B from IGHV1-2*01, in addition to version A, amino acid residues in the CDR2 were further germlined (i.e., substituted by the corresponding IGHV1-2*01 human germline residues). These residues are D56G, A58T, S60Y, N61A and K65Q in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 65 and shares 86.7% sequence identity with IGHV1-2*01 human germline.

[1957] To design humanized HCVR version C from IGHV1-2*01, in addition to version B, 2 amino acid residues in the CDR2 were further germlined. These residues are D50R and E62Q in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 68 and shares 88.8% sequence identity with IGHV1-2*01 human germline.

[1958] Various other humanized versions of the HCVR from IGHV1-2*01 were further designed, starting from version B. Indeed, to get a well humanized monoclonal antibody, 85% is supposed to be sufficient (version B from IGHV1-2*01 shares 86.7% sequence identity with IGHV1-2*01 human germline). In order to reduce the risk of introducing mutations, versions D, E, F, G and H were thus designed to reach 85% and no more.

[1959] The resulting HCVR versions D, E, F, G and H from IGHV1-2*01 are as set forth in SEQ ID NOs: 101, 121, 122, 123 and 124, respectively, and all share 85.7% sequence identity with IGHV1-2*01 human germline.

[1960] HCVR, Using IGHV5-51*01

[1961] To design humanized HCVR version A from IGHV5-51*01, the murine CDRs (with SEQ ID NOs: 1, 4 and 3) were grafted into IGHV5-51*01 and 6 residues in FR1, FR2 and FR3 were back-mutated to the parental murine residues, to maintain the full activity of the antibody. These residues are A24, T28, 148, L70, L83 and V97 in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 63 and shares 79.6% sequence identity with IGHV5-51*01 human germline.

[1962] To design humanized HCVR version B from IGHV5-51*01, in addition to version A, 1 amino acid residue in the FR1 and 6 amino acid residues in the CDR2 were further germlined. These residues are A24G, D56S, A58T, S60Y, N61S, K63S and K65Q in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 66 and shares 86.7% sequence identity with IGHV5-51*01 human germline.

[1963] To design humanized HCVR version C from IGHV5-51*01, in addition to version B, 1 amino acid residue in the FR1 and 2 amino acid residues in the CDR2 were further germlined. These residues are T28S, D50I and E62P in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 69 and shares 89.8% sequence identity with IGHV5-51*01 human germline.

[1964] HCVR, Using IGHV3-11*05

[1965] To design humanized HCVR version A from IGHV3-11*05, the murine CDRs (with SEQ ID NOs: 1, 4 and 3) were grafted into IGHV3-11*05 and 9 residues in FR1, FR2 and FR3 were back-mutated to the parental murine residues, to maintain the full activity of the antibody. These residues are Y27, T30, 148, G49, L70, A72, T74, A79 and V97 in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 64 and shares 76.5% sequence identity with IGHV3-11*05 human germline.

[1966] To design humanized HCVR version B from IGHV3-11*05, in addition to version A, 2 amino acid residues in the FR1, 6 amino acid residues in the CDR2 and 1 amino acid residue in the FR3 were further germlined. These residues are Y27F, T30S, D56S, A58T, S60Y, N61A, E62D, K63S and A79L in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 67 and shares 89.8% sequence identity with IGHV3-11*05 human germline.

[1967] To design humanized HCVR version C from IGHV3-11*05, in addition to version B, 1 amino acid residue in the FR2 and 1 amino acid residue in the CDR2 were further germlined. These residues are I48V and P53S in SEQ ID NO: 61. The resulting HCVR is as set forth in SEQ ID NO: 70 and shares 87.8% sequence identity with IGHV3-11*05 human germline.

[1968] LCVR, Using IGKV1-9*01

[1969] To design humanized LCVR version A from IGKV1-9*01, the murine CDRs (with SEQ ID NOs: 15, 16 and 17 with X.sub.12 being absent) were grafted into IGKV1-9*01 and 3 residues in FR2 and FR3 were back-mutated to the parental murine residues, to maintain the full activity of the antibody. These residues are F35, W46 and Y70 in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 82 and shares 83.2% sequence identity with IGKV1-9*01 human germline.

[1970] To design humanized LCVR version B from IGKV1-9*01, in addition to version A, 1 amino acid residue in the CDR1 and 1 amino acid residue in the FR2 were further germlined. These residues are S24R and F35Y in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 85 and shares 85.3% sequence identity with IGKV1-9*01 human germline.

[1971] To design humanized LCVR version C from IGKV1-9*01, in addition to version B, 2 amino acid residues in the CDR2 and 1 amino acid residue in the FR3 were further germlined. These residues are N49A, P54Q and Y70F in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 88 and shares 88.4% sequence identity with IGKV1-9*01 human germline.

[1972] A LCVR version D from IGKV1-9*01 was further designed, in order to introduce an extra residue (Ser; S) in the CDR1 as found in the human germline IGKV1-9*01. The introduction of the extra residue in the CDR1 loop has shown that the binding activity was conserved (data not shown). The introduction of a serine residue in the CDR1 of version B brings the sequence identity to 86.3%, so in order to reduce the risk of introducing mutations, version D was designed where Kabat residue L36 was reverted to the original mouse residue Phe (F). The resulting LCVR is as set forth in SEQ ID NO: 103.

[1973] LCVR, Using IGKV6-21*02

[1974] To design humanized LCVR version A from IGKV6-21*02, the murine CDRs (with SEQ ID NOs: 15, 16 and 17 with X.sub.12 being absent) were grafted into IGKV6-21*02 and 4 residues in FR2 and FR3 were back-mutated to the parental murine residues, to maintain the full activity of the antibody. These residues are F35, W46, Y48 and Y70 in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 83 and shares 81.1% sequence identity with IGKV6-21*02 human germline.

[1975] To design humanized LCVR version B from IGKV6-21*02, in addition to version A, 1 amino acid residue in the CDR1, 1 amino acid residue in the FR2, 1 amino acid residue in the CDR2 and 1 amino acid residue in the FR3 were further germlined. These residues are S24R, F35Y, L53S and Y70F in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 86 and shares 85.3% sequence identity with IGKV6-21*02 human germline.

[1976] To design humanized LCVR version C from IGKV6-21*02, in addition to version B, 1 amino acid residue in the CDR3 was further germlined. This residue is Q88H in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 89 and shares 86.3% sequence identity with IGKV6-21*02 human germline.

[1977] LCVR, Using IGKV3-11*01

[1978] To design humanized LCVR version A from IGKV3-11*01, the murine CDRs (with SEQ ID NOs: 15,16 and 17 with X.sub.12 being absent) were grafted into IGKV3-11*01 and 3 residues in FR2 and FR3 were back-mutated to the parental murine residues, to maintain the full activity of the antibody. These residues are F35, W46 and Y70 in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 84 and shares 84.2% sequence identity with IGKV3-11*01 human germline.

[1979] To design humanized LCVR version B from IGKV3-11*01, in addition to version A, 1 amino acid residue in the CDR1 was further germlined. This residue is S24R in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 87 and shares 85.3% sequence identity with IGKV3-11*01 human germline.

[1980] To design humanized LCVR version C from IGKV3-11*01, in addition to version B, 1 amino acid residue in the FR2, 3 amino acid residues in the CDR2 and 1 amino acid residue in the FR3 were further germlined. These residues are F35Y, L53R, P54A, S55 and Y70F in SEQ ID NO: 81. The resulting LCVR is as set forth in SEQ ID NO: 90 and shares 90.5% sequence identity with IGKV3-11*01 human germline.

Example 6

[1981] Production, Purification and Characterization of Humanized Anti-45RC Antibodies

[1982] Analytical size exclusion chromatography (SEC-HPLC) and differential scanning calorimetry (DSC) were used to compare the profile and the thermal stability, respectively, of 9 humanized anti-45RC variants A to I. These variants correspond to antibody comprising the “versions A” HCVR and LCVR described in Example 5. Analytical size exclusion chromatography (SEC-HPLC) and differential scanning calorimetry (DSC) were also used to compare the profile and the thermal stability, respectively, of 4 other humanized anti-45RC variants A1, A2, I1 and I2.

[1983] Material and Methods

[1984] SEC-HPLC

[1985] A Shimadzu Prominence HPLC system was used, with a Superdex 200 Increase 5/150 GL column (GE Healthcare). The column was previously calibrated in the same buffer and conditions used during sample analysis (using the Molecular Weight SEC Calibration kits from GE Healthcare, in PBS 1×, at 0.25 mL/min, with the column oven set to 30° C.).

[1986] All samples were centrifuged (20.000 g, 5 minutes, 4° C.) and had their protein content quantitated by Nanodrop ND-1000 spectrophotometer with IgG analysis program, prior to SEC analysis.

[1987] The isocratic program was set to inject about 15 μg of each sample, at 0.25 m/min during 18 minutes. After SEC analysis, 280 nm chromatogram was extracted from the raw data, and analyzed by peak integration.

[1988] DSC

[1989] A MicrocalTM VP-Capillary DSC system was used to perform differential scanning calorimetry experiments.

[1990] Samples in 1×PBS buffer were centrifuged (20.000 g, 5 minutes, 4° C.), and had their protein content quantitated Nanodrop ND-1000 spectrophotometer with IgG analysis program, prior to DSC analysis. Samples were then diluted in PBS to a final concentration of 1 mg/mL.

[1991] The pre-equilibration time was 3 minutes and the thermograms that followed were acquired between 20 and 110° C. with a scanning rate of 60° C./hour, a filtering period of seconds and medium feedback.

[1992] Prior to sample analysis, 5 buffer/buffer scans were measured to stabilize the instrument, and a buffer/buffer scan was performed between each protein/buffer scan.

[1993] The data was fitted to a non-2-state unfolding model, with the pre- and post-transition adjusted baseline subtracted. The calorimetric enthalpy (ΔH) is determined as the area under the peak of the transition, whereas the van′t Hoff enthalpy (ΔHv) is determined from the model used.

[1994] Results

[1995] SEC-HPLC

[1996] A summary of the SEC parameters is given in Table 7 below.

TABLE-US-00208 TABLE 7 SEC parameters of the humanized ABIS-45RC variants A-I, A1, A2; I1 and I2. ABIS-45RC Area Calculated variant Peak # RT Area % MW A 1 4.939 373336 10.62 510 2 5.212 790998 22.50 404 3 6.099 2351531 66.88 190 B 1 4.671 1305266 26.47 640 2 5.013 1451918 29.44 478 3 5.899 2174527 44.10 225 C 1 4.896 405089 14.04 529 2 5.148 527782 18.29 426 3 5.980 1788411 61.97 210 4 7.708 164621 5.70 48 D 1 4.939 126757 5.38 510 2 5.268 430061 18.25 385 3 6.137 1799214 76.37 184 E 1 4.738 607732 18.87 605 2 5.039 686150 21.31 468 3 5.925 1925975 59.82 220 F 1 4.811 791680 13.93 568 2 5.124 1306598 22.99 435 3 5.977 3585434 63.08 211 G 1 4.768 185266 3.54 589 2 5.092 889883 16.99 447 3 5.893 4159958 79.46 226 H 1 4.896 105313 1.90 529 2 5.158 506367 9.11 423 3 5.962 4941812 88.99 213 I 1 5.142 123472 3.16 429 2 6.081 3783288 96.84 193 A1 1 5.074 129770 2.65 478 2 6.098 4775239 97.35 189 A2 1 5.074 129770 2.65 478 2 6.098 4775239 97.35 189 I1 1 4.885 131671 2.5 567 2 5.146 355754 6.78 447 3 6.142 4760685 90.7 181 I2 1 4.864 32276 0.73 578 2 5.150 168687 3.79 446 3 6.181 4246352 95.48 175 RT: retention time (in minutes) MW: molecular weight (in kDa)

[1997] Table 7 above shows in bold the peaks corresponding to the anti-CD45RC antibodies (peak 3 for each of variants A to H, I1 and I2; and peak 2 for each of variants I, A1 and A2), with RT and calculated MW expected for a monomeric, non-precipitated and non-dissociated antibody.

[1998] DSC

[1999] A summary of the DSC parameters is given in Table 8 below.

TABLE-US-00209 TABLE 8 DSC parameters of the humanized ABIS-45RC variants A-I, A1, A2; I1 and I2. Denaturation of the antibody happens in two steps, hence two melting temperatures are given, one of each step. ABIS-45RC variant Conc. T.sub.1/2 ΔH T.sub.onset T.sub.m1 T.sub.m2 A 0.0035 7.09 763 55.59 66.44 80.64 B 0.0056 7.09 776 55.96 64.73 81.01 C 0.0031 7.93 596 55.95 63.05 79.74 D 0.0023 5.83 863 60.09 70.91 80.10 E 0.0042 5.83 762 57.09 71.28 80.05 F 0.0099 5.41 867 58.63 71.15 80.34 G 0.0087 5.01 824 60.76 69.93 80.79 H 0.0093 4.59 877 59.79 68.14 80.24 I 0.0055 5.42 809 62.07 70.83 81.27 A1 0.0049 3.32 969 61.09 68.17 81.91 A2 0.0070 5.83 1150 60.21 66.92 81.94 I1 0.0067 3.75 1080 63.08 73.08 81.84 I2 0.0055 4.16 1030 63.68 72.43 82.02 Conc.: concentration, in mM. T.sub.1/2: width of transition at half height of the peak, in ° C. ΔH: calorimetric enthalpy of unfolding, in cal/M. T.sub.onset: temperature at which the unfolding transition begins, in ° C. T.sub.m1: denaturing/melting temperature of the first step, in ° C. T.sub.m2: denaturing/melting temperature of the second step, in ° C.

Example 7

[2000] Reactivity of Humanized ABIS-45RC Variants A-I

[2001] Material and Methods

[2002] PBMC Isolation

[2003] Blood healthy volunteers is collected and peripheral blood mononuclear cells (PBMC) were isolated by Ficoll gradient centrifugation, which enables removal of unwanted fractions of blood products such as granulocytes, platelets and reaming red blood cell contaminants.

[2004] Antibodies and Flow Cytometry

[2005] Human PBMC were labeled with the murine ABIS-45RC antibody or each of the humanized ABIS-45RC antibodies variants A-I (at 2 μg/mL and 1 μg/mL); and an anti-CD3 antibody. The murine and humanized ABIS-45RC antibodies reactivity was revealed using a biotin donkey anti-human IgG.sup.+ Streptavidin PerCP-Cy 5.5 secondary antibody.

[2006] A Canto II cytometer (BD Biosciences) was used to measure fluorescence intensity and data were analyzed using the FLOWJO software (Tree Star Inc.). Cells were first gated by their morphology and dead cells were excluded by selecting DAPI-negative cells.

[2007] Results

[2008] Both ABIS-45RC and Commercial Anti-CD45RC MT2 Antibodies Compete for the Same Epitope

[2009] Labelling with either of the humanized ABIS-45RC antibodies (variant A, FIG. 5A; variant B, FIG. 5B; variant C, FIG. 5C; variant D, FIG. 5D; variant E, FIG. 5E; variant F, FIG. 5F; variant G, FIG. 5G; variant H, FIG. 5H; variant I, FIG. 5I) or the murine ABIS-45RC (FIG. 5J) shows that the antibodies recognized human CD45RC in a similar manner.

Example 8

[2010] Engineered Antibodies

[2011] The CDR1 of ABIS-45RC's LCVR has a canonical structure unique to mouse antibodies, with a length of 10 amino acid residues (SEQ ID NO: 15 with X.sub.12 being absent, i.e., SASSSVSYMH).

[2012] For the design of humanized versions A, B and C of the LCVR described in Example 5, this 10-amino-acid-residue CDR1 was grafted into the human germlines, with backmutation and/or germlining but no addition or deletion of any residue. However, in human germlines, the LCVR's CDR1 has a minimum length of 11 amino acid residues.

[2013] Therefore, to increase the humanness of the humanized antibodies, the Inventors have sought to engineer ABIS-45RC V.sub.L-CDR1 “SASSSVSYMH” to extend it by one extra residue. One candidate position is position 8 in SEQ ID NO: 15 (designated as X.sub.12), i.e., between S30 and Y31 in SEQ ID NO: 81. In all of the candidate germlines for the humanization design, this position is occupied with Asn (N), Ser (S) or Gly (G), while in the murine germline, this position is empty.

[2014] In order to investigate the structural relevance and stability of such insertion, ABIS-45RC V.sub.L-CDR1 was expanded by insertion of an asparagine, i.e., SEQ ID NO: 15 with X.sub.12 being Asn (N), i.e., SASSSVSNYMH. A search of the sequences of solved antibody structures via protein BLAST was then conducted. The top hit was the LCVR CDR1 of the structure PDB ID: 5CMA. Subsequently, this structural segment was grafted onto the ABIS-45RC model using the two-residue overhang on its ends to anchor the CDR template fragment to the model. It was observed that, in order to accommodate the additional residue, there was a conformational change that shifted the neighboring residue, thereby presenting a slight steric clash with Y70. However, in all of the human germlines, this residue is a more accommodating phenylalanine.

[2015] Engineered Mouse Antibody

[2016] Based on the above, the Inventors have engineered the ABIS-RC45 antibody, by inserting an asparagine residue (Asn, N) in the VL-CDR1, and further mutating Y70 of SEQ ID NO: 81 into a phenylalanine (Phe, F). The resulting “engineered Asn/Phe ABIS-RC45” LCVR is set forth in SEQ ID NO: 71, with X.sub.12 being Asn (N).

[2017] Two other mouse antibodies have also been produced on the same basis, by inserting a serine residue (Ser, S) or a glycine residue (Gly, G) in the V.sub.L-CDR1, and further mutating Y70 of SEQ ID NO: 81 into a phenylalanine (Phe, F). The two resulting “engineered Ser/Phe ABIS-RC45” and “engineered Gly/Phe ABIS-RC45” LCVR are set forth in SEQ ID NO: 71, with X.sub.12 being Ser (S) or Gly (G), respectively.

[2018] Engineered Humanized Antibodies

[2019] Based on the above, engineered humanized LCVR versions A, B and C (as described in Example 5) can be further designed, as set forth in SEQ ID NOs: 72-80 where X.sub.12 is Asn (N), Ser (S) or Gly (G) and the residue in position 70 is a Phe (F).

Example 9

[2020] Reactivity of Engineered Asn/Phe ABIS-45RC

[2021] Material and Methods

[2022] Blood from healthy volunteers was collected and peripheral blood mononuclear cells (PBMC) were isolated by Ficoll gradient centrifugation, which enables removal of unwanted fractions of blood products such as granulocytes, platelets and reaming red blood cell contaminants.

[2023] Human PBMCs were labeled with ABIS-45RC or with engineered Asn/Phe ABIS-45RC and an anti-CD3 antibody. The reactivity was revealed using a biotin donkey anti-human IgG.sup.+ Streptavidin PerCP-Cy 5.5 secondary antibody.

[2024] A Canto II cytometer (BD Biosciences) was used to measure fluorescence intensity and data were analyzed using the FLOWJO software (Tree Star Inc.). Cells were first gated by their morphology and dead cells were excluded by selecting DAPI-negative cells.

[2025] Results

[2026] As shown in FIG. 6, labelling with either ABIS-45RC (left panel) or the engineered Asn/Phe ABIS-45RC (right panel) shows that both antibodies recognized human CD45RC in a similar manner.

Example 10

[2027] Reactivity of Humanized ABIS-45RC

[2028] Material and Methods

[2029] Blood from healthy volunteers was collected and peripheral blood mononuclear cells (PBMC) were isolated by Ficoll gradient centrifugation, which enables removal of unwanted fractions of blood products such as granulocytes, platelets and reaming red blood cell contaminants.

[2030] Human PBMCs were labeled with ABIS-45RC or with humanized ABIS-45RC at 20, 5, 1.25 or 0.3 μg/mL and an anti-CD3 antibody. The reactivity was revealed using a biotin donkey anti-human IgG.sup.+ Streptavidin PercpCy 5.5 secondary antibody.

[2031] A Canto II cytometer (BD Biosciences) was used to measure fluorescence intensity and data were analyzed using the FLOWJO software (Tree Star Inc.). Cells were first gated by their morphology and dead cells were excluded by selecting DAPI-negative cells.

[2032] Results

[2033] Labelling with either murine ABIS-45RC (FIG. 7A) or the humanized ABIS-45RC variant A1 (FIG. 7B) or variant A3 (FIG. 7C), at various concentrations, shows that both antibodies recognized human CD45RC in a similar manner.

Example 11

[2034] Cell Death Induction by Humanized ABIS-45RC Variants

[2035] Material and Methods

[2036] Human PBMCs were incubated with medium, isotype control Ab or anti-CD45RC variants (10 μg/mL) for 6 hours. Then, cells were stained with anti-CD3 and anti-CD45RA, annexin V and DAPI. Percentage of total apoptosis was obtained by gating on DAPI.sup.+ annexin V.sup.++DAPI.sup.− Annexin V.sup.+ cells among T or non-T cells by flow cytometry.

[2037] Results

[2038] ABIS-45RC or the humanized variants A1 or A3 efficiently induced cell death of CD3.sup.+ cells (FIG. 8A) but not CD3.sup.− cells (FIG. 8B).

Example 12

[2039] Treatment of Human Skin Rejection with ABIS-45RC and Humanized Variant A1

[2040] Material and Methods

[2041] PBMC Isolation

[2042] Blood was collected at the Établissement Français du Sang (Nantes, France) from healthy individuals. Written informed consent was provided according to institutional guidelines. PBMC were isolated by Ficoll-Paque density-gradient centrifugation (Eurobio, Courtaboeuf, France). Remaining red cells and platelets were eliminated with a hypotonic solution and centrifugation.

[2043] Animals

[2044] 8- to 12-week-old NOD/SCID/IL2Rγ.sup./- (NSG) mice were bred in our own animal facilities in SPF conditions (accreditation number C44-278).

[2045] Human Skin Transplantation Model

[2046] Human skins were obtained from healthy volunteers from abdominoplasty surgery and transplantation was performed as previously described (Bézie et al., 2018. Front Immunol. 8:2014). One month later, 5×10.sup.6 PBMCs from allogeneic healthy volunteers were intravenously injected with or without antibodies.

[2047] Graft rejection was scored from 0 to 5 based on dryness (score 1), rigidity (score 2), scab (score 3), partial loss (score 4) and complete loss of the skin (score 5) by macroscopic observation.

[2048] Human PBMCs engraftment was monitored in blood by flow cytometry.

[2049] Treatment

[2050] NSG mice were treated intraperitoneally with purified ABIS-45RC or humanized variant A1 antibodies at 0.8 mg/kg from day 0 and every 2.5 days during 20 days, together with intraperitoneal administration of rapamycin from day 0 to day 10 at a suboptimal dose of 0.4 mg/day.

[2051] Results

[2052] Treatment with PBMCs only induced weight loss, initiated around day 14, and, as shown in FIG. 9, death of all mice by day 33.

[2053] The inventors previously showed that treatment with rapamycin only did not prolonged survival (median survival: 21 days (Bézie et al., 2018. Front Immunol. 8:2014)). Here, treatment with ABIS-45RC or the humanized variant A1 completely abrogated the skin graft rejection.

Example 13

[2054] Cell Engineering with the CD45RC-CAR

[2055] Material and Methods

[2056] Lentiviral Vectors

[2057] A self-inactivating second-generation lentiviral vector encoding for CAR-CD45R was generated. In this vector, the EF1alpha promoter controls the following sequences in the indicated order: a signal peptide, the variable regions of the heavy and light chains of the anti-CD45RC mAb (ABIS-45RC) fused by a linker, the transmembrane region of CD8-CD28 costimulatory signals region, the CD3zeta transducing signals region, P2 Å self-splicing sequences, GFP. The lentiviral vector was pseudotyped with VSV-G. Ctrl-CAR, used as a control, encodes the variable heavy and light chains of a monoclonal antibody having an antigenic specificity other than CD45RC (as well as the transmembrane region of CD8.sup.− CD28 co-stimulatory signals region, and the CD3zeta transducing signals) under the control of the EF1alpha promoter and LNGFR under the control of a CMV promoter. This Ctrl-CAR lentiviral vector was also VSV-G pseudotyped.

[2058] Cell Transduction

[2059] HEK (293) or Jurkat cells were harvested and counted at day 1, then plated in 6 wells plate at 500 000 cells/well in 2 mL of DMEM 10% FBS, 10 μg/mL penicillin-streptomycin, 2 nM L-Glutamin. At day 2, the medium was removed by pipetting and 2 mL of fresh medium (pre-warmed at 37° C.) is added. In parallel, 500p L of the following preparation was added per well: [2060] 2.5 μg of DNA is diluted in 250 μL of Optimem (Gibco, life technology) during 5 min at room temperature (RT) [2061] 10 μL of lipofectamine TM 2000 (Life technology, Invitrogen) is dilute in 250 μL of Optimem during 5 min at RT [2062] DNA and lipofectamine are mixed 20 min at RT.

[2063] GFP Detection

[2064] A Canto II cytometer (BD Biosciences) was used to measure fluorescence intensity and data were analyzed using the FLOWJO software (Tree Star Inc.). Cells were gated by their morphology and then dead cells were excluded by selecting DAPI.sup.− negative cells. GFP was analyzed by flow cytometry after staining of the cells with Protein L (Genscript).

[2065] Results

[2066] CAR were designed to have at least one extracellular domain, optionally a hinge domain, a transmembrane domain, at least one co-stimulatory domain and at least one primary signaling domain (FIG. 10). One CAR construct of the invention is a CAR harboring an scFv CD45RC, a CD8α hinge domain and transmembrane domain, a CD28 co-stimulatory domain and a CD3ζ primary signaling domain. Optionally, GFP coding sequences may be present immediately 3′ of the CD45RC-CAR sequences separated from them by a T2 Å self-splicing sequence (not drawn). In this case, GFP may be used as a surrogate marker of CAR-CD45RC expression.

[2067] The inventors showed that HEK cells are able to be transfected with the CD45RC-CAR as compared to a GFP mock plasmid as shown by the GFP staining (FIG. 11). In addition, this CAR construction can also be transduced with lentiviral vectors in a human immortalized T-cell cell line (Jurkat). Indeed, the level of GFP, showing the presence of the CAR is of 58.7% in Jurkat cells transduced with the CD45RC CAR and only of 1.25% in Jurkat cells non transduced (FIG. 12). In particular, the surface expression of the CAR in Jurkat cells was confirmed by Protein L staining and demonstrated a good correlation with GFP staining, demonstrating altogether that the lentivirus was functional (FIG. 13).

Example 14

[2068] CD45RC-CAR Induction of Target T Cell Apoptosis

[2069] Material and Methods

[2070] Total T cells were negatively sorted using magnetic sorting (Miltenyi Biotech) from human PBMC. Cells were labelled with CPD 670 and plated in 96 wells bottom V plates (20,000 cells per well in complete medium RPMI (10% SVF, amino acid, penicillin/streptomycin, glutamine, sodium pyruvate, HEPES). 100,000 Jurkat cells were plated in 96-well bottom plate in complete DMEM medium and transduced with 10 uL of the CD45RC-CAR or Ctrl-CAR control lentiviral vectors described in Example 13, incubated 2 days at 370° C. Jurkat cells were then counted and added to T cells at different ratios of T:Jurkat cells in complete RPMI medium (ratio 1:0-1:1-1:5-1:10). The cells were incubated 18h at 37° C. After incubation, the cells were marked with annexin V in annexin buffer for 20 min. The DAPI was added in the annexin buffer and cells were directly analyzed in a Canto II cytometer (BD Biosciences).

[2071] Results

[2072] Human T cells were cultured during 18 hours in presence of Jurkat cells previously transduced with a CD45RC-CAR of a Ctrl-CAR and then apoptosis was evaluated by flow cytometry. In presence of Jurkat cells harboring the CD45RC-CAR, T cells were 15% apoptotic and only 7-8% in Ctrl-CAR or untransduced cells (FIG. 14). The level of apoptosis observed was equivalent to the level obtained with an anti-CD45RC antibody. Thus, Jurkat cells expressing the CD45RC-CAR can induce apoptosis in human T cells.

Example 15

[2073] CD45RC-CAR Induction of T Cell Activation

[2074] Material and Methods

[2075] Total T cells were negatively sorted using magnetic sorting (Miltenyi Biotech) from human PBMC. Cells were labelled with CPD 670 and plated in 96 wells bottom V plates (20,000 cells per well in complete medium RPMI (10% SVF, amino acid, penicillin/streptomycin, glutamine, sodium pyruvate, HEPES). 100,000 Jurkat cells were plated in 96-well bottom plate in complete DMEM medium and transduced with 10 uL of the CD45RC-CAR or Ctrl-CAR control lentiviral vectors described in Example 13, incubated 2 days at 37° C. Jurkat cells were then counted and added to T cells at different ratios of T: Jurkat cells in complete RPMI medium (ratio 1:0-1:1-1:5-1:10). The cells were incubated 18h at 37° C. After incubation, the cells were marked with an anti-CD69 mAb for 20 min. The DAPI was added and the cells were directly analyzed in a Canto II cytometer (BD Biosciences).

[2076] Results

[2077] Human T cells were cultured during 18 hours in presence of Jurkat cells transduced with a CD45RC-CAR or a Ctrl-CAR and then T cell activation was measured by flow cytometry using the CD69 activation marker.

[2078] The inventors observed that the CD45RC-CAR induce a T cells activation (as shown by CD69 expression) compared to Ctrl-CAR and untransduced Jurkat cells. Interestingly, when the cells are sorted, meaning that the CD45RC-CAR is expressed by all cells, the activation is even better compared to non-sorted CD45RC-CAR Jurkat cells (FIG. 15). Altogether, this shown that the CAR is functional.

Example 16

[2079] CD45RC-CAR Induction of CD4.sup.+ Treg Activation and of Target T Cell Apoptosis

[2080] Material and Methods

[2081] CD4.sup.+ Treg Lentiviral Transduction and Expansion Protocol

[2082] At day 0, CD4.sup.+ CD127.sup.1ow CD25.sup.+ CD45RC.sup.− Tregs were sorted on a FACSAria™ and seeded at 10.sup.5 cells in 100 μL medium per in 96-well flat bottom plates previously coated with 1 μg/mL anti-CD3 monoclonal antibodies (clone OKT3). Medium used at day 0 was RPMI 1640 supplemented with penicillin, streptomycin, sodium pyruvate, HEPES buffer, amino acids, glutamine, 5% CTS serum, anti-CD28 mAbs (clone CD28.2 at 1 μg/mL) and 1000 U/mL IL-2. At day 1 and day 2, cells were transduced with 10 μL of the lentiviral vectors described in Example 13, which code for anti-CD45RC chimeric antigen receptor (CD45RC-CAR) and GFP, or for a control CAR having a different antigenic specificity (Ctrl-CAR) and LNGFR. At day 3, medium was added to reach a 10% CTS serum final concentration. At day 7, cells were harvested and sorted on CAR expression based on GFP or LNGFR and then newly stimulated with anti-CD3 and anti-CD28 mAbs for a second round of 7 days of expansion. Cytokines were freshly added in culture medium every 2 days, and fresh medium was added when required.

[2083] CD45RC-CAR Detection

[2084] HEK 293T cells were also transduced with the lentivirus encoding for CD45RC-CAR once at day 1. CD4.sup.+ Tregs were transduced with lentivirus as explained above. HEK 293T and CD4.sup.+ Tregs were analyzed 5 days after the last transduction. Cells were treated with human Fc block (BD Biosciences) and then incubated with 2 μg/mL of CD45R-ABC protein (R&D) biotinylated in house diluted in PBS BSA 1% for 1 h at 4° C. CD45R-ABC-biotin protein was revealed with streptavidin-APC-Cy7 at 4 μg/mL. DAPI was added to exclude dead cells and the cells were analyzed in a FACSCanto™.

[2085] CAR-Mediated Activation Assay

[2086] A total of 10.sup.5 CD45RC-CAR CD4.sup.+ Tregs or Ctrl-CAR CD4.sup.+ Tregs were plated in 96-wells U-bottom plates previously coated with CD45R-ABC protein (1 μg/mL in PBS, 1h30 at 37° C.) in complete medium RPMI (10% CTS serum, amino acid, penicillin/streptomycin, glutamine, sodium pyruvate, HEPES). The cells were incubated 24h at 37° C. and brefeldin A was added the 4 last hours of culture. After incubation, the cells were marked with a viability dye and stained extracellularly with anti-CD69, anti-CD25, anti-CD71 mAbs for 30 min. Cells were fixed and permeabilized and stained intracellularly with anti-CTLA-4 mAb for 1 h. Cells were analyzed in the FACSCanto™.

[2087] Apoptosis Assay

[2088] Human PBMC were labelled with CPD670 and plated in 96 wells V-bottom plates (50,000 cells per well) in complete medium RPMI (10% human AB serum, amino acid, penicillin/streptomycin, glutamine, sodium pyruvate, HEPES). CD4.sup.+ Tregs non-transduced or transduced with CD45RC-CAR or Ctrl-CAR were counted after 15 days of expansion and added to allogenic PBMC at different ratios of PBMC:Tregs in complete RPMI medium (ratio 1:0-1:1-1:2-1:5). The cells were incubated 4h at 37° C. After incubation, cells were stained with anti-CD3, anti-CD19, anti-CD14, anti-CD56 and anti-CD45RA mAbs and then cells were marked with annexin V in annexin buffer for 20 min. DAPI was added in the annexin buffer and the cells were rapidly analyzed in the FACSCanto™. Anti-CD45RA mAb was used as surrogate markers for identification of CD45RC.sup.high cells in each cell subset. As a control of apoptosis, ABIS-45RC was incubated with the PBMCs at 10 μg/mL. The percentage of total apoptotic cells was calculated by the sum of annexin.sup.+ DAPI.sup.+ cells and annexin.sup.+ DAPI.sup.− cells.

[2089] Results

[2090] Human HEK 293 were transduced with a lentiviral vector encoding for CD45RC-CAR and GFP and analyzed by cytofluorimetry (FIG. 16). CD45RC-CAR detection using a CD45R-ABC protein labeled with biotin and revealed with streptavidin-APC-Cy7 showed a clear positive signal and co-expression with GFP as compared to non-transduced cells, demonstrating that the CD45RC-CAR is expressed and functional in its specificity to CD45RC.

[2091] Transduction of CD4.sup.+ Tregs with CD45RC-CAR and cell sorting selection using GFP at day 7 followed by in vitro expansion during 14 days allowed obtention of a population of 74.38%+/−17.3% of transduced cells (FIG. 17), demonstrating that CD4.sup.+ Tregs can be transduced and that CD45RC-CAR CD4.sup.+ Tregs can be expanded.

[2092] The inventors then analyzed the activation of CD45RC-CAR-transduced and expanded CD4.sup.+ Tregs by incubating the cells with CD45R-ABC protein and analyzing activation markers by cytofluorimetry (FIG. 18). They observed in CD4.sup.+ Tregs transduced by CD45RC-CAR, but not by Ctrl-CAR, increased protein expression of CD69, CD25, CD71 and CTLA-4 upon incubation with CD45R-ABC protein for 24h, demonstrating that the CAR is signaling and directly activating the Tregs.

[2093] Apoptosis on different cell populations of PBMC by allogenic CAR-transduced or not transduced CD4.sup.+ Tregs after 14 days of expansion was analyzed and compared to the apoptosis obtained with the anti-CD45RC mAb used to generate the CD45RC-CAR (FIG. 19). The inventors observed that CD45RC-CAR-transduced CD4.sup.+ Tregs induced apoptosis of CD45RC.sup.high T cells equivalent to the anti-CD45RC mAb, in contrast to Ctrl-CAR and non-transduced CD4.sup.+ Tregs that did not induce apoptosis. This apoptotic effect of CD45RC-CAR CD4.sup.+ Tregs was dose dependent, increasing with higher ratios of PBMCs:CD4.sup.+ Tregs.

[2094] Altogether, these results showed that the CD45RC-CAR is functional since it induced activation of CD45RC-CAR-transduced CD4.sup.+ Tregs. Moreover, CD45RC-CAR-transduced CD4.sup.+ Tregs induced apoptosis of CD45RC.sup.high T cells.