Antibodies directed against ICOS and uses thereof

Abstract

The present invention provides antibodies directed against ICOS or a derivative thereof which neutralize ICOS engagement on Treg by inhibiting the fixation between ICOS and ICOS-L and abrogate proliferation of Treg induced by plasmacytoid dendritic cells. The present invention further provides antibodies directed against ICOS or a derivative thereof which induce IL-10 and IFNγ production, induce CD4− T cells proliferation, reduce Tconv proliferation, and increase the immunosuppressive function of Treg.

Claims

1. An antibody directed against ICOS which increases the immunosuppressive function of regulatory T cells, wherein said antibody comprises a CDRH3 comprising the sequence set forth in SEQ ID NO:25, or a CDRH3 comprising no more than three amino acid substitutions in the sequence set forth in SEQ ID NO:25.

2. An antibody directed against ICOS, wherein said antibody is selected from the group consisting of Icos 53-3, Icos 88-2 and Icos 92-17, respectively obtainable from the hybridoma deposited at the CNCM on Jul. 2, 2009 under the accession numbers CNCM I-4176, CNCM I-4177, and CNCM I-4178.

3. An antibody directed against ICOS, wherein said antibody has the following 6 CDRs: H-CDR1 of amino acid sequence GYSFTSYWIN (SEQ ID NO:23); H-CDR2 of amino acid sequence NIYPSDSYTNYNQMFKD (SEQ ID NO:24); H-CDR3 of amino acid sequence WNLSYYFDNNYYLDY (SEQ ID NO:25); L-CDR1 of amino acid sequence RSSKSLLHSNGNTYLY (SEQ ID NO:26); L-CDR2 of amino acid sequence RMSNLAS (SEQ ID NO:27); and L-CDR3 of amino acid sequence MQHLEYPWT (SEQ ID NO:28).

4. An antibody according to claim 3, wherein the nucleotidic sequences encoding the 6 CDRs are the following: H-CDR1: GGCTACAGTTTCACCAGCTACTGGATAAAC (SEQ ID NO:17); H-CDR2: AATATTTATCCTTCTGATAGTTATACTAACTACAATCAAA TGTTCAAGGAC (SEQ ID NO:18); H-CDR3: TGGAATCTTTCTTATTACTTCGATAATAACTACTACTTGG ACTAC (SEQ ID NO:19); L-CDR1: AGGTCTAGTAAGAGTCTCCTGCATAGTAATGGCAACACT TACTTGTAT (SEQ ID NO:20); L-CDR2: CGGATGTCCAACCTTGCCTCA (SEQ ID NO:21); L-CDR3: ATGCAACATCTAGAATATCCGTGGACG (SEQ ID NO:22).

5. A method of treatment comprising administering an effective amount of the antibody according to claim 1 to a patient suffering from a disease selected from the group consisting of autoimmune diseases, transplantation rejection or a graft versus host disease.

6. A method of treatment comprising administering an effective amount of the antibody according to claim 1 to a patient suffering from an inflammatory disorder.

7. The method according to claim 6, wherein said patient suffers from an inflammatory disorder selected in the group consisting of inflammatory disorder of the nervous system, mucosal inflammatory disease, inflammatory skin disease and autoimmune arthritis.

8. The method according to claim 6, wherein said patient suffers from an inflammatory disorder selected from multiple sclerosis, inflammatory bowel disease, asthma or tonsillitis, dermatitis, psoriasis, contact hypersensitivity and rheumatoid arthritis.

Description

FIGURE LEGEND

(1) FIG. 1: Ta-Treg strongly express ICOS, co-localized with Ta-pDC and proliferate in situ but do not proliferate in vitro.

(2) A—Tumor frozen sections were stained with anti ICOS Ab (green) and Ki67 Ab (brown) and secondary anti murine Ab conjugated to HRP and revealed respectively with histogreen and DAB (magnification 10× and 40× for the insert box)).

(3) B—Ki67 expression) was analyzed by multi color flow cylometry on Treg (CD4.sup.+CD127.sup.−CD25.sup.high) and Tconv (CD4.sup.+CD127.sup.+CD25.sup.low/−) within primary tumor (Ta-Treg, Ta-Tconv) or paired blood (Treg, Tconv).

(4) C—Purified Treg and Tconv from either primary tumor or healthy blood were cultured in a 96 well U-bottomed-plate in presence of 500 UI/ml of IL-2. Cell number was quantified every 4 days by enumeration.

(5) D-F Tumor frozen sections were stained with anti CD3 Ab (brown) and counterstained with hematoxylin (blue) (10× and 40× in the insert box) (D); CD3 Ab (green) and BDCA2 (brown) (20× and 40× in the insert box) (E); FoxP3 Ab (brown) and BDCA2 (green) (20× and 40× n the insert box) (F).

(6) FIG. 2: ICOS and ICOS-L blockade abrogates IL-10 secretion during pDC mediated T cell activation without interfering strongly on MoDC/T co-culture. Purified and R848-activated pDC or MoDC were co-cultured for 5 days with allogeneic memory CD4+ T cells in presence of Ctrl Ab, anti ICOS (314.8) or anti ICOS-L (MIH12). At day 5, IL-10 and IFNγ were quantified by ELISA in supernatants from pDC/T co-culture (A) and MoDC/T co-culture (B).

(7) FIG. 3: ICOS and CD3 co-stimulation favor Treg and Tconv proliferation as well as IL-10 but not IFNγ secretion in the presence of exogenous IL-2.

(8) A/B—FACS-sorted Treg or Tconv issued from tonsil were cultured for 5 days alone or with beads coated with CD3/IgG, CD3/88.2, CD3/CD28 agonist mAb in the presence of IL-2 (n=3). The proliferation was assessed by [.sup.3H]-Thymidine incorporation (A). IL-10 and IFNγ levels were measured by ELISA in the culture supernatant (B).

(9) C—CD4.sup.+ TaT cells sorted from tumor were cultured for 5 days with beads coated with antiCD3/IgG; antiCD3/88.2 or antiCD3/antiCD28 in the presence of exogenous IL-2 (100 UI/ml). The concentrations of IL-10 and IFNγ in the supernatant were quantified by ELISA.

(10) FIG. 4: ICOS engagement blocks CD28-induced IL-2 and consequently reduces proliferation and IFNγ secretion

(11) A—CFSE labeled CD4.sup.+ memory T cells were cultured for 5 days with the different beads alone or in presence of graded concentration of exogenous rhIL-2 (20 UI/ml and 100 UI/ml) and proliferation was assessed by CFSE dilution by flow cytometry.

(12) B—IL-2 detected by ELISA after 5 days culture with the different beads without exogenous IL-2.

(13) C—Blood CD4.sup.+ memory T lymphocytes from healthy donors were cultured for 5 days with the different beads alone or in presence of exogenous IL-2 (100 UI/ml). The IL-10 and IFNγ secretions were quantified by ELISA.

(14) FIG. 5: Absence of expression of ICOS-L on breast tumor cell lines and primary breast tumor dilacerations

(15) A—ICOS-L expression was assessed by flow cytometry on breast tumor epithelial cell lines suspensions harvested in PBS-EDTA in absence of trypsin to avoid Ag deterioration.

(16) B—ICOS-L expression was assessed on tumor cells (CD45− cells) after 48 h culture in presence of control Ab (dashed line) or anti ICOS Ab (314.8) (continuous line).

(17) FIG. 6: Treatment of primary Neu15 mammary tumors with a surrogate rat anti-mouse anti ICOS Ab (17G9, IgG2b) slow down the tumor growth.

(18) FIG. 7:

(19) A: Treg cells numbers are increased within primary cervix cancer.

(20) B: Treg cells ICOS+ are increased within primary cervix cancer.

(21) FIG. 8: Increase of ICOS expressing Treg in non Hodgkin lymphoma (NHL)

(22) HD Hodgkin Disease

(23) FL Follicular Lymphoma

(24) DLBCL Diffuse Large B Cell Lymphoma

(25) MCL Mantle Cell Lymphoma

(26) MZL Marginal Zone Lymphoma

(27) FIG. 9: Presence of ICOS+ Treg cells within primary breast tumors has a negative impact on survival

(28) 120 paraffin embedded primary tumor samples with 10 years clinical follow up were tested for their expression of ICOS using a commercial anti ICOS rabbit polyclonal Ab. Mean of ICOS+ cells were assessed on six different spots. To perform the statistical analysis the median was used as cut-off to have equilibrated groups.

(29) Impact of ICOS expression according to the presence of ICOS in the primary tumor on Overall Survival (A) or Progression Free Survival (B) is shown.

EXAMPLE

Example 1: Characterisation of the Antibodies According to the Invention

(30) Material and Methods

(31) I. Cellular Biology

(32) 1—Selection/Cell Purification

(33) Isolation of Peripheral Blood Mononuclear Cells

(34) PBMCs (Peripheral Blood Mononuclear Cells) were isolated from peripheral stem cells transplantation of healthy volunteers (Etablissement Francais du sang, Marseille, France) by Lymphoprep gradient (Abcys). In tubes: ⅔ of blood are deposited dropwise over ⅓ of Lymphoprep and centrifuged for 20 minutes at 2000 rpm at 20° C. with no acceleration so as not to disturb the gradient. After centrifugation, the mononuclear cells are recovered and washed twice in PBS 1% FCS (Fetal Calf Serum)+heparin for 20 min at 1000 rpm at 20° C.

(35) The cells were then used immediately or frozen at −80° C. to 50.Math.10.sup.6 cells/ml in RPMI 1640 50% FCS 10% DMSO (Dimethyl Sulfoxide). After 24 h, the cells are transferred to the nitrogen for preservation

(36) Negative Selection of CD4

(37) CD4+ T lymphocytes were purified from PBMCs. After thawing, the cells were washed and diluted in 40 μL of sorting buffer (PBS 0.5% BSA 2 mM EDTA) for 10.10.sup.6 cells. The kit MACS human CD4+ T Cell Isolation Kit II were used (Miltenyi Biotec): 10 μL of a solution of monoclonal antibodies conjugated to biotin (primary labeling) are added and the mixture was incubated for 10 min at 4° C. with stirring.

(38) The cells are then put in contact with 20 μL of magnetic beads coupled with anti-biotin (secondary labeling) for 15 min at 4° C. with stirring. After washing with buffer sorting, cells were sorted to Automacs (Miltenyi). The negative fraction depleted in CD4+ T labeled cells is then isolated. This gives a population of CD4+ pure of about 95%.

(39) 2—Activation and Cell Culture

(40) Pre-Activation with Beads CD3/CD28 and then Stimulation with mAbs

(41) The CD4+ T cells are put to the concentration of 10.sup.6 cells/mL RPMI 10% FCS in the presence of beads CD3/CD28 (Dynabeads, Invitrogen) (1 cell/1 bead) and incubated for 48 h at 37° C. The cells are then separated from the beads with a magnet and Dynal Biotech rest overnight in RPMI 10% FCS at a concentration of 10.sup.6 cells/ml.

(42) On the other hand, mAbs anti-CD3 (OKT3), anti-ICOS (ICOS 88-2), and control IgG1 mAb (Sigma) were coated on a 96 flat well overnight at 4° C. The wells are coated with 50 ng/ml anti-CD3 supplemented with 20 pg/ml in other mAbs PBS 1×100 μL/wells. The next clay, the plate is washed with PBS, saturated two hours with PBS 5% FCS (200 μl per well). The CD4+ T cells with the previously incorporated CFSE (see below) are distributed on the coated plate at a rate of 2105 cells/200 μL of medium/well and incubated for 72 h at 37° C. At 48 h, supernatants were collected and at 72 h, cells were collected to analyze proliferation by flow cytometry (FIG. 4).

(43) Activation by Artificial APC

(44) Magnetic beads (Dynabeads M-450 Epoxy, Invitrogen) were washed in sodium phosphate buffer 0.1 M and then incubated with mAbs anti-CD3 (OKT3) at a concentration suboptimal of 1 mg/1.10.sup.7 beads representing 5% of the mAbs coupled with the beads, with the mAb anti-CD28 or ICOS (ICOS 88-2 or CD28.2), (2 μg/1.10.sup.7 beads, 10%). These artificial APC were incubated with mAbs in slow rotation overnight at 4° C. The next day, two washes are performed in PBS 0.1% BSA. Artificial APC are distributed on a one bead for a cell in a 96 round plate wells on which 2.10.sup.5 lymphocytes T CD4+/200 μl per well were deposited and then incubated for 72 h at 37° C. The CD4+ T cells have previously incorporated the CFSE. At 48 h, supernatants were collected and 72 h, cells were collected to analyze proliferation by flow cytometry.

(45) 3—Cell Proliferation

(46) The proliferation of lymphocytes is followed by the CFSE (carboxyfluorescein diacetate succinimidyl ester) (Molecular Probes, Invitrogen). The CFSE is cell permeable and nonfluorescent. When entering the cell, esterases cleave the acetate groups which become fluorescent whereas the cell become impermeable.

(47) The characteristic of CFSE is to be shared equitably in each newly formed cell at each division. It emit in the green radiation allows the simultaneously analyze of the number, the position and the stage of differentiation of the cells, the fluorescence intensity per cell being proportional to the concentration of CFSE. To label the cells with CFSE, the cell suspension is diluted in cold 1×PBS. Adding the CFSE: 5 μM to 10.Math.10.sup.6 cells. The cells are then placed in a water-bath at 37° C.

(48) After 8 to 10 minutes stirring, the cells were quickly placed on ice to stop the reaction. The cells are then centrifuged twice with 2 ml of PBS IX. Finally they are collected in the desired volume of RPMI 10% FCS for culture. The proliferation is determined thanks to flow cytometry.

(49) II—Flow Cytometry

(50) CD4+ cells are diluted with 30% BSA PBS (50 μL/well) in a 96-well plate for 10 minutes at 4° C. to saturate nonspecific sites. They are then incubated for 30 minutes at 4° C. in the dark, with the desired antibodies coupled to a fluorochrome.

(51) After two washes in PBS 1×1% BSA 0.02% azide (centrifugation 2100 rpm, 3 min at 4° C.), cells were fixed in 200 μL of PBS 2% formaldehyde and placed in a flow cytometer (FACS Canto, BD Biosciences). The results are analyzed thanks to the FlowJo software.

(52) III—ELISA (Enzyme Linked Immunosorbent Assay)

(53) Culture supernatants of CD4+ T cells are collected at 48 h and stored at −20° C. for an assay on IL-10, IFNγ and TNFα.

(54) Results

(55) 1—Characterization of Anti-ICOS mAbs

(56) The inventors developed 5 anti-ICOS Abs. Their isotype was assayed by ELISA. For obtaining an indirect analysis of their affinity for their receptor, mAbs were tested using stable transfectants expressing ICOS. JICOS.1 cells were in the presence of an increasing range of anti-ICOS mAbs labeled with a probe coupled to a fluorochrome (PE-GAM: Goat anti mouse-PE) and the analysis was made thanks to cytometry flow.

(57) It was thus possible to determine the ED 50, i.e. the concentration of mAbs which 50% of sites are saturated. mAbs with the lowest ED 50 are those with the highest apparent affinity.

(58) Then the inventors tested the ability of anti-ICOS mAbs to inhibit the binding of ICOS L (a recombinant form of human IgG1 Fc domain) carried by the JICOS.1 cell.

(59) They used a gradient of concentration of anti-ICOS mAbs and they reveled the fixation to ICOS L Fc thanks to a probe coupled to a fluorochrom (GAH-PE: Goat anti-human-PE). The analysis was made by flow cytometry. The inventors thus determined the ID 50 i.e. the dose which inhibits 50% of the binding of ICOS L-Fc on ICOS.

(60) The more the ID 50 is little, the more mAb easily compete with recombinant ICOS Fc.

(61) The inventors thus evidenced that ICOS R 314-8 and ICOS R 53-3 have a high affinity for their binding site (ED 50<0.5 ug/ml) and a significant blocking potential (ID 50<1 mg/ml).

(62) The antibody ICOS R 314-8 was therefore chosen for being coupled to the fluorochrome Alexa Fluor 647 and used in flow cytometry analysis.

(63) 2—Anti-ICOS mAbs Differ in their Ability to Induce the Production of IL-10 by Activated CD4+ T Cell

(64) The inventors tested the ability of the mAbs to act as agonist antibodies, i.e. being able to have the same action as the natural ligand of ICOS, using functional tests. The studied parameter studied was the secretion of IL-10 since ICOS induces the production of IL-10 by LT.

(65) The agonistic potential of anti-ICOS mAbs were tested on CD4+ T cells, which were pre-activated with CD3/CD28 beads for 48 h and distributed on a plate where anti-CD3 mAb were coated for continuing the stimulation along with the various anti-ICOS mAbs.

(66) The culture supernatants were then assayed for 48 h for IL-10 and the secretion of IL-0 induced by the different anti-ICOS mAbs was compared based to the secretion of IL-110 induced by a commercially available anti-ICOS mAb (ICOS c)

(67) The anti-ICOS mAbs 53-3, 88-2 and 92-17 significantly increased IL-10 secretion of CD4+ and thus are agonist antibodies. Regarding, anti-ICOS mAbs 145-1 and 314-8, no significant increase in the production of IL-10 was detected.

(68) The inventors finally showed that anti-ICOS mAbs 53-3, 88-2 and 92-17 are better agonists than the commercially available anti-ICOS. Indeed, if one considers the commercially available anti-ICOS mAb as reference, the anti-ICOS mAb 88-2 causes increased secretion of IL-10 of +61%, anti-ICOS mAb 92-17 of +20% and anti-ICOS mAb 53-3 of +14%.

(69) The results are summarized in the following table:

(70) TABLE-US-00003 [ED] 50 [ID] 50 Agonist mAb Isotype (μg/ml) (μg/ml) effect ICOS 88-2 IgG1- L 1.60 17 +++ ICOS 314-8 IgG1- K 0.06 0.29 −

Example 2: Use of an Antagonist Antibody of the Invention and an Agonist Antibody of the Invention

(71) Material and Methods

(72) Immunohistochemistry

(73) Frozen primary breast tumor sections were stained with mouse anti-FOXP3, or anti-Ki67 and revealed using the ImmPRESS anti-mouse Ig peroxidase kit (Abcys, Paris, France) according to the supplier's instructions and DAB. Then, the second primary antibody (mouse anti-ICOS (53.3), anti-CD3, anti-BDCA2 was added and revealed with ImmPRESS kit and Histogreen (Abcys). The specificity of the staining was assessed using mouse isotype controls in place of the first or the second primary antibody.

(74) Purification of Mononuclear Cells from Breast Tumors, Tonsils and Healthy Blood

(75) Mononuclear cells (MNC) were purified, from healthy peripheral blood obtained from EFS or from enzymatic dilaceration of primary breast tumors or tonsils samples, by Ficoll density gradient centrifugation.

(76) Phenotypic Analysis of pDC and T Cells Subsets

(77) For extensive phenotypic analysis, pDC were identified among total MNC as CD4.sup.+CD123.sup.+ cells using FITC or PE anti-CD123 and PE-Cy5 anti-CD4 and PE-coupled antibodies against CD40, CD86 or ICOSL. T cells were identified as CD3.sup.+CD4.sup.+ cells. Treg were identified either by the multi color phenotype CD4.sup.+ CD127.sup.− CD25.sup.high or for their FoxP3 expression after gating on CD3.sup.+CD4.sup.+ T cells.

(78) Proliferation of Ta-Treg and Ta-Tconv or their blood counterpart was assessed by multicolor analysis allowing Treg CD4.sup.+ (CD127.sup.− CD25.sup.high) and Tconv (CD4.sup.+CD127.sup.+CD25.sup.Low/−) characterization associated with KI67 Ab staining.

(79) Flow cytometric analysis was performed on a FACScan (BD Biosciences) or an ADP Cyan (Beckman Coulter) and data were analyzed with Cell Quest Pro software (BD Biosciences) or FlowJo (Treestar).

(80) Purification of pDC

(81) pDC were purified from lineage (Lin)-negative enriched MNC by either magnetically activated cell sorting using CD304/BDCA-4 microbeads kit or negative depletion using pDC isolation kit (Miltenyi Biotec)) or FACS®-sorting (FACSVantage SE™ DiVa flow cytometer, BD Biosciences) as Lin.sup.−CD4.sup.+CD11c.sup.− cells. Purity was routinely>98%.

(82) In Vitro Generation of Monocytes Derived DC (MoDC)

(83) MoDC were obtained from blood purified monocytes after 7 days differentiation in GM-CSF (100 ng/ml)+IL-4 (50 UI/ml) (Schering Plough, Kenilworth USA).

(84) CD4.sup.+ Memory T Cells and Treg Purification

(85) Memory CD4.sup.+ T cells (>95% purity) were obtained from MNC after magnetic depletion including anti-CD45RA Ab, as described (Gobert et al, 2009). CD4.sup.+CD25.sup.highCD127.sup.− Treg and CD4.sup.+CD25.sup.−CD127.sup.low/+ conventional CD4.sup.+ T cells were FACS®-sorted from purified memory CD4.sup.+ T cells (Purity>98%).

(86) To follow their in vitro proliferation, purified memory CD4.sup.+ T cells were stained at day 0 with CFSE. Viable cells were selected by DAPI exclusion or Live and Dead reagent in case of cell permeabilisation (200,000 and 5,000 minimum events were gated on the total cell population and on purified cells respectively).

(87) DC-T Cell Co-Cultures

(88) Allogeneic memory CD4.sup.+ T cells, Treg or CD4.sup.+ Tconv cells were cultured at 3×10.sup.4 to 5×10.sup.4 cells in complete medium with IL-2 (100 IU/ml) and highly purified TApDC, healthy pDC or MoDC that were pre-activated for 24 h with IL-3, GM-CSF (10 ng/ml) in the presence of R848. Addition of T lymphocytes on pre-activated DC subsets was done in triplicate in 96-well round-bottomed plates at a ratio of 1:5 (DC/T cells) and co-cultured for 5 days. Proliferation was assessed either by CFSE dilution in experiments analysing FoxP3 expression or by DNA synthesis analyzed by .sup.3H-TdR uptake.

(89) The impact of ICOS/ICOSL engagement was assessed by addition of ctrl Ab, commercial (ISA-3) or proprietary (314.8) anti ICOS Ab or anti ICOSL (MIH12) in the cultures. To assess T cell cytokines secretion by ELISA, cells were co-cultured with pDC or TApDC, and supernatants harvested at day 5 were centrifuged and stored at −20° C.

(90) Stimulation of Tconv and Treg with Artificial Beads

(91) Artificial APC were produced as described in example 1. Treg (3×10.sup.4) or Tconv (1×10.sup.5) sorted from tonsil or Ta-CD4.sup.+ T cells (1×10.sup.5) purified from tumors were cultured for 5 days with artificial beads at a ratio 1:1 (artificial APC:T cell) in the presence of IL-2 (100 UI/ml) in 96 U-bottomed wells under 200 μl. Proliferation was assessed either by CFSE dilution or by DNA synthesis analyzed by .sup.3H-TdR uptake.

(92) Cytokines Detection in T Cell Cultures Supernatants by ELISA

(93) IL-10, IFNγ and IL-2 in 5 days culture supernatants were quantified by ELISA using commercial kits from Bender Medsystems according to manufacturer's instructions.

(94) Result

(95) The data presented below are intended to analyze the impact of two antibodies against ICOS (i.e. blocking MAb 314.8; agonist MAb 88.2) developed by the inventors in order to validate

(96) i) the blockade of ICOS by antagonistic 314.8 MAb as a new promising drug candidate to abrogate the immunosuppressive response observed in breast cancer; and

(97) ii) the engagement of ICOS by the agonist 88.2 MAb on CD4.sup.+ T cells to favour the amplification of Treg that would be of interest in the field of autoimmunity.

(98) Ta-Treg that Highly Express ICOS are Present within Lymphoid Aggregates in Primary Breast Tumors and Proliferate In Situ

(99) The inventors have previously demonstrated the presence of Tumor associated regulatory T cells (Ta-Treg) expressing CD25.sup.high and FoxP3 in primary breast tumors within lymphoid aggregates correlating with a poor prognosis and increased risk of metastasis (Gobert et al., 2009). These Ta-Treg represent 15% to 25% of total CD4.sup.+ TaT cells, are highly activated as they express ICOS, CD39, GITR and HLA-DR and suppress TaTconv proliferation and cytokines secretion (IL-2, IFNγ).

(100) These Ta-Treg proliferate within the primary breast tumor environment in situ (Gobert et al., 2009) as demonstrated by either the presence of ICOS.sup.+ Treg co-expressing Ki67 on tumor frozen sections, (FIG. 1A) or the higher proportion of KI67.sup.+ cells within purified Ta-Treg and Treg from blood (respectively 8% and 4%) compared to Ta-Tconv and Tconv (3% and 0.3% respectively) (FIG. 1B).

(101) In contrast with these in vivo results, the inventors demonstrated that in vitro stimulation of purified Ta-Treg with expand beads (beads coated with agonist anti CD3 and anti CD28) is not capable to favor Ta-Treg amplification in contrast to that observed with purified TaTconv or purified Treg or Tconv from blood of healthy donors (FIG. 1C).

(102) The inventors hypothesized that the ICOS engagement is essential for Ta-Treg proliferation and functions.

(103) A) Use of an Antagonist Antibody of the Invention

(104) Blockade of ICOS/ICOS-L Interaction Through Antagonist ICOS mAb (314.8) Ta-Treg Interact In Situ with Ta-pDC within Lymphoid Aggregates in Primary Breast Carcinoma

(105) Several studies reported the expression of ICOS-L, the specific ligand of ICOS, on pDC (Janke et al., 2006). Using immuno-histochemistry on tumor frozen sections, the inventors observed that Ta-CD3.sup.+ T cell present within the lymphoid aggregates surrounding the tumor are in interaction with Ta-pDC BDCA2.sup.+ (FIGS. 1D and 1E). A double staining with FoxP3 and BDCA2 Ab revealed that Ta-Treg are in close contact with Ta-pDC in these lymphoid aggregates, suggesting that this interaction will favor the ICOS engagement by ICOS-L in tumors (FIG. 1F).

(106) Ta-pDC are Activated but Did not Express ICOS-L as a Potential Consequence of In Situ ICOS/ICOS-L Interaction

(107) After purification from tumor disaggregation, Ta-pDC show an activated phenotype as they express up regulated levels of CD86 and CD40 compared to healthy blood and matched patient's blood pDC. As reported by several groups (Ito et al., 2007; Janke et al., 2006), freshly isolated healthy blood pDC express low levels of ICOS-L that is strongly unregulated after IL-3 exposure or TLR7/8 ligation, which is not observed on other DC subsets (mDC, MoDC). Interestingly, contrasting with their activated status (CD86.sup.+CD40.sup.+), Ta-pDC are devoid of membrane ICOS-L expression. In contrast freshly isolated paired patients' blood pDC or healthy blood pDC express ICOS-L. After a 24 h culture period in IL-3 or upon TLR7/8 ligation, sorted Ta-pDC reacquire a strong ICOS-L expression demonstrating their ability to modulate this ICOS-L expression (data not shown). Among CD3.sup.+ TaT cells, ICOS is strongly expressed on Ta-Treg (69.9% MFI: 361) in contrast to TaTconv (23% MFI: 83) or TaCD8.sup.+ (2% MFI: 50). These results indicate that in situ ICOS/ICOS-L interaction leads to ICOS-L down regulation on Ta-pDC membrane.

(108) Blockade of ICOS/ICOS-L Interaction Through Antagonistic Anti ICOS MAb (314.8) Abrogate ICOS-L Downregulation on pDC

(109) To test this hypothesis, healthy blood T cells were cultured with TLR7-activated pDC purified from tonsil. The inventors observed after 24 h culture period with increased T:pDC ratio a dose dependent ICOS-L downregulation on pDC. Interestingly the addition of our antagonist MAb against ICOS (314.8) abrogates totally this ICOS-L downregulation on pDC, result that is not reproduced using commercial anti ICOS antibody (ISA-3) (data not shown).

(110) The results demonstrate Ta-pDC and Ta-Treg interactions through ICOS/ICOS-L and indicate that ICOS ligation could be involved in Ta-Treg activation and proliferation.

(111) Coculture of CD4.sup.+ T Cells as Well as Purified Treg with Activated pDC but not MoDC Induced Treg Proliferation that is Blocked with 314.8

(112) To test the ability of ICOS/ICOS-L interactions to induce Treg amplification, the inventors cultured total memory CD4.sup.+ T cell with healthy blood purified allogeneic TLR7/8 (R848)-activated pDC or mDC. Among purified CD4.sup.+ T cells, 3.5% expressed FoxP3 (data not shown). After 5 days of co-culture with pDC the proportion of FoxP3.sup.high expressing cells, corresponding to Treg, rises to 12.3% and the addition of 314.8 Ab blocks by 80% this enrichment in FoxP3.sup.high cells. In contrast, coculture of CD4.sup.+ T cell with activated mDC was not able to favor a distinct FoxP3.sup.high subpopulation among CD4.sup.+ T cells, and the addition of 314.8 has no significant effect (6.3% to 8%).

(113) Similar results were obtained with CD4.sup.+ T cells purified from tumor. Ta-Treg Foxp3.sup.+ represent 9% of freshly purified CD4.sup.+ TaT cells (data not shown). Their co-culture with R848 activated pDC increases the proportion of Ta-Treg to 14.5% whereas the addition of 314.8 leads to a decrease of Ta-Treg proportion to 4.5%, below the starting level.

(114) FACS sorted purified Treg or Tconv populations stained with CFSE were cultured with R848-activated pDC or LPS-activated MoDC to analyze their proliferation capacity by flow cytometry (dilution of CFSE expression). First, the inventors observed that in absence of exogenous IL-2, that activated MoDC do not induce purified Treg proliferation whereas Tconv strongly proliferate. In contrast coculture with activated pDC is able to induce a strong proliferation of both purified Treg and Tconv.

(115) The addition of anti-ICOS 314.8 MAb strongly reduces Treg and Tconv proliferation when pDC are used as APC whereas Tconv proliferation is unchanged in cocultures with MoDC. In this experiment, ICOS or ICOS-L blockade with commercial antibodies (ISA-3 mAb or MIH-12 MAb) do not affect neither Treg nor Tconv proliferation in pDC/T co-cultures.

(116) These data demonstrate that the anti-ICOS 314.8 MAb neutralizes ICOS engagement on Treg and abrogates their expansion induced by pDC.

(117) ICOS and ICOS-L Blockade Abrogate IL-10 Secretion During pDC Mediated T Cell Activation without Interfering Strongly on MoDC/T Co-Culture.

(118) 314.8 MAb also reduces Tconv proliferation in response to activated pDC stimulation. The inventors determined the impact of 314.8 on IFNγ and IL-10 secretion by Elisa during Tconv and allogeneic R848 activated pDC (FIG. 2A) or and LPS activated MoDC (FIG. 2B) co-cultures. In these settings IL-10 secretion is totally abrogated by 314.8 mAb (217+/−31 pg/ml in control and 13+/−6 pg/ml with 314.8). Whereas, the IFNγ secretion is slightly reduced upon the addition of 314.8 MAb on co-cultures with pDC (32% reduction 507+/−53 pg/ml in control condition and 341+/−73 pg/ml with 314.8) (FIG. 2A). In Tconv/MoDC co-cultures ICOS inhibition leads to slightly increased secretions of IL-10 and IFNγ (FIG. 2B).

(119) B) Use of an Agonist Antibody According to the Invention

(120) Use of Agonist Anti ICOS MAb (88.2) to Mimick ICOS Engagement

(121) To perfect their understanding on ICOS functions on Treg and Tconv, the inventors generated a model of artificial APC using beads coated with agonist MAbs leading to CD3 (OKT3); CD28 (CD28.2) and/or ICOS (88.2, Table 1) signaling on purified T cells.

(122) ICOS Engagement with an Agonist MAb (88.2) on Treg Induced their Proliferation and their Capacity to Secrete High Amounts of IL-10

(123) First the inventors observed that Treg from healthy donors proliferate in response to anti CD3/88.2 beads in presence of exogenous IL-2 (FIG. 3A). As previously reported (Simpson et al., 2010; Ito et al., 2008) upon activation through TCR and ICOS engagement in presence of IL-2, both purified Tconv and Treg subpopulations secrete high amounts of IL-10 (311+/−22 pg/ml and 426+/−48 pg/ml respectively) and low levels of IFNγ (205+/−8 pg/ml and 381+/−12 pg/ml). This result contrasts with data obtained using the anti CD3/anti CD28 beads. In this model Tconv secrete large amounts of IFNγ (1213+/−72 pg/ml) and low levels of IL-10 (69+/−58 pg/ml) whereas Treg secrete IL-10 and low levels of IFNγ (422+/−36 pg/ml and 305+/−31 pg/ml respectively) (FIG. 3B).

(124) Similar experiments with T cells purified from tumor demonstrated that CD4.sup.+ TaT cells produce similar levels of IL-10 in response to ICOS and CD28 while the levels of IFNγ are weaker in response to ICOS compared to CD28 engagement (FIG. 3C).

(125) ICOS Engagement Blocks CD28-Induced IL-2 and Consequently Reduces Proliferation and IFNγ Secretion

(126) Whereas, CD4.sup.+ memory T cells proliferate in response to anti CD3/anti CD28 stimulation independently of exogenous IL-2, no proliferation is observed in response to anti CD3/88.2 stimulation (FIG. 4A). The addition of hIL-2 rescues this proliferation in a dose dependent manner. Interestingly, the co-engagement of ICOS and CD28 in the absence of IL-2 reduces significantly the proliferation of CD4.sup.+ memory T cells compared to only CD28 engagement, and this is completely rescued in the presence of 100 UI/ml IL-2. Interestingly, ICOS ligation through 88.2 mAb abrogates IL-2 secretion detected with anti CD3/anti CD28 stimulation (FIG. 4B). Taken together, this argues in favor of a reduced spontaneous IL-2 secretion when ICOS and CD28 are co-engaged compared to the only CD28 engagement suggesting an ICOS inhibitory function on CD28-induced IL-2 secretion.

(127) Moreover, even in the presence of exogenous IL-2, the inventors observed a 50% reduction of IFNγ produced by Tconv, when ICOS and CD28 are triggered compared with anti CD3/anti CD28 beads (FIG. 4C).

(128) In contrast, whereas IL-10 secretion is strictly IL-2 dependent when cells are activated under ICOS triggering, as previously described (Ito 2008, Paulos 2010), the addition of ICOS signal does not affect the IL-10 secretion induced by anti CD3/anti CD28 (FIG. 4C).

(129) All together these results demonstrate that ICOS ligation reduced the ability of anti CD3/anti CD28 to favor Th1 polarisation (through the reduced IFNγ production) but sustains the IL-10 production favoring the development of an immunosuppressive environment.

(130) ICOS Engagement Through 88.2 MAb Increased the Treg Suppressive Function

(131) To assess that ICOS engagement can be associated with an immunosuppressive T cell response, the inventors setup suppressions assays in the absence of exogenous IL-2 to compare anti CD3/anti CD28/IgG and anti CD3/anti CD28/88.2 beads efficiency. The addition of the ICOS signaling (88.2) strongly increases the suppressive function of Treg compared with the anti CD3/anti CD28/IgG1 (51% inhibition in condition one Treg for 4 Tconv anti CD3/anti CD28/88.2 compared to 21% inhibition with anti CD3/anti CD28/IgG). All together these results demonstrate that ICOS engagement favors an immunosuppressive T cell response that could result either from an increased Tconv sensitivity to suppression or a stronger Treg suppressive ability.

Example 3: Analysis of Prognostic Impact of Detection of ICOS.SUP.+ Treg Cells within Primary Breast Tumors

(132) 120 paraffin embedded primary tumor samples with 10 years clinical follow up were tested for their expression of ICOS using a commercial anti ICOS rabbit polyclonal Ab (Spring Biosciences). ICOS.sup.+ cells were quantified in double blind on 6 different replicates for each tumor and mean of the results were compiled (data not shown). To perform the statistical analysis the inventors used the median as cut-off to have equilibrated groups.

(133) In univariate analysis the inventors demonstrate that the presence of ICOS.sup.+ cells (>1.66 ICOS.sup.+ cells/spot) correlated with high tumor grade (p=0.007), expression of Estrogen receptor by tumor cells (p=0.018), luminal A/B molecular subtypes (p<0.001) and absence of Her2/neu overexpression (p=0.035).

(134) The impact of ICOS.sup.+ cells detection within primary breast tumors on overall survival (OS) or progression free survival (PFS) was investigated.

(135) Whereas 6/59 deaths were observed in the ICOS− group, 14/61 patients deceased in the ICOS.sup.+ demonstrating the significant prognostic value of ICOS.sup.+ detection on OS (Log Rank test p value=0.0465) (FIG. 9A). Similar analysis performed on PFS demonstrated ICOS.sup.+ cells was associated with a poorer overall survival, with progression in 11/59 in ICOS.sup.− group, whereas 20/61 patients progressed in the ICOS.sup.+ group (p=0.0285) (FIG. 9B).

Example 4: Confirmation of the Existence of In Situ Interaction of pDC with ICOS.SUP.+ Treg in the Tumor Environment

(136) Ex vivo co-culture of tumor cell dilacerations in the presence of anti ICOS 314.8 MAb or Ctrl Ab for 48h in presence of IL-3 (20 ng/ml). At the end of the culture period, the expression of ICOS-L on pDC is observed only in the presence of the anti ICOS 314.8 MAb and not with control Ab, demonstrating that the down regulation of ICOS-L on pDC is mediated through an interaction with ICOS.sup.+ cells (data not shown).

Example 5: Epithelial Breast Tumor Cells from Either Established Cell Lines or Fresh Tumor Samples do not Express ICOS-L in Contrast to Melanoma or Glioma Tumor Cells Even after Ex Vivo Culture with Anti ICOS Antibody 314.8

(137) Breast epithelial tumor cells lines were harvested in PBS EDTA to avoid trypsin-associated degradation of the Ag and cells were stained with anti ICOS-L antibody to evaluate the expression at cell surface by flow cytometry. None of the cell lines tested was found positive for ICOS-L (FIG. 5A). Similar analyses were performed on 48 h cultured primary tumor disagregations in presence of anti ICOS Ab (314.8) or control Ab plus IL-3 (20 ng/ml) (FIG. 5B).

Example 6: Impact of a Surrogate Rat Anti-Mouse Anti ICOS Ab (17G9, IgG2b) on Mammary Tumor Growth in a Syngenic Mammary Tumor Model

(138) Mouse mammary tumor model was obtained in female FVB mice in 28-35 days after orthotopic injection of the Neu 15 cell line. The generated tumors appear significantly infiltrated by activated Ta-pDC. ICOS.sup.high TATreg and resting TATconv.

(139) Injection of 17G9 antibody (50 μg/ml) intra-peritoneally three times a week from day 11 after tumor implantation results in a reduced Neu15 tumor size at late time points compared to the injection of control Ab (LTF2, IgG2b) (p=0.053) (FIG. 6).

Example 7: A: Treg Cells Numbers are Increased within Primary Cervix Cancer

(140) Cervical samples were obtained from patients either with dysplasia (CIN2/3, n=18) or cancer (n=14). Normal cervical tissue was used as control (n=11). Samples were obtained by both enzymatic and physical dissociations. After washing, mononuclear cells were incubated with labeled mAbs and Tregs enumerated as CD127.sup.lowCD25.sup.brightCD4.sup.+ T cells. The percentage of Treg within the CD4.sup.+ subset is depicted. Treg were increased within cervical cancer samples in comparison to normal tissue and dysplasia. Hence, this increase is associated to the cancer development (FIG. 7A).

(141) B: Treg Cells ICOS.sup.+ are Increased within Primary Cervix Cancer.

(142) Cervical samples were obtained from patients either dysplasia (CIN2/3, n=5) or cancer (n=12). Normal cervical tissue was used as control (n=5). Samples were obtained by both enzymatic and physical dissociation. After washing mononuclear cells were incubated with labeled using ICOS mAbs and Tregs enumerated. The percentage of Treg ICOS within the CD4.sup.+ subset is depicted ICOS.sup.+ Treg are present within tissues with only a trend to their increase in cervical cancer due the limited numbers of samples analyzed (FIG. 7B).

Example 8: Increase of ICOS Expressing Treg in Non Hodgkin Lymphoma (NHL)

(143) The inventors have analyzed the Teg numbers and the expression of ICOS on Treg in LNH samples. Fresh lymphoma cells leased from lymph nodes were collected from 45 patients with informed consent. Lymphoma samples correspond to Hodgkin disease (HD, n=11), follicular lymphoma (FL, n=13), diffuse large B cell lymphoma (DLBCL, n=10), mantle cell lymphoma (MCL, n=5) and marginal zone lymphoma (MZL, n=6). Detection of Treg cells was performed by incubation for 20 min at 4° C. with anti-ICOS-PE (Becton Dickinson™), anti-CD3-ECD, anti-CD4-Pacific Blue (Beckman Coulter®), anti-CD127 FITC, anti-CD25 APC-Cy7 and LIVE/DEAD® Fixable Dead Cell Stain Kit (Invitrogen™). After staining, each cell preparation was washed twice in PBS, fixed with 2% paraformaldehyde and analyzed on a FACS LSR2 flow cytometer (Becton Dickinson™). Data were analyzed using FlowJo Software (TreeStar™). Treg were increased in all lymphoma samples except HD. Most of Treg displayed an increased expression of ICOS in comparison to control lymph nodes (FIG. 8).

Example 9: Sequencing of Icos 314.8 (CNCM I-4180)

(144) Total RNA was extracted from provided frozen hybridoma cells and cDNA was synthesized. Then, RT-PCR was performed to amplify the variable regions (heavy and light chains) of the MAb. The MAb variable regions of the heavy and light chains were cloned into a cloning vector separately, then the obtained sequences were analyzed to deduce the sequences of the MAb.

(145) Materials

(146) Hybridoma cells ICOS 314.8 (CNCM I-4180); TRIzol® Plus RNA Purification System (Invitrogen, Cat. No: 15596-026); SuperScript™ III First-Strand Synthesis System (Invitrogen, Cat. No: 18080-051).

(147) Methods

(148) Total RNA Extraction

(149) Total RNA was isolated from the hybridoma cells according to the technical manual of TRIzol® Plus RNA Purification System. The total RNA was checked by gel electrophoresis.

(150) RT-PCR

(151) Total RNA was reverse transcribed into cDNA using isotype specific anti-sense primer or universal primer and whole procedure was according to the technical manual of SuperScript™ III First-Strand Synthesis System. The antibody fragment will be amplified according to the standard operation protocol of RACE method of GenScript.

(152) Cloning of Antibody Genes

(153) Target PCR products of antibody genes were cloned into the cloning vector separately according to standard molecular cloning procedures.

(154) Screening and Sequencing

(155) Colony screening was employed to screen clones with inserts of correct sizes, and no less than ten independent positive colonies were sequenced for each antibody fragment.

(156) Results and Analysis

(157) Total RNA Extraction

(158) Total RNA of the sample was run alongside of DL3000 DNA marker on a 1.5% agarose/GelRed™ gel electrophoresis.

(159) PCR Product of Antibody Genes

(160) 4 μl PCR products of each sample were run alongside of DL3000 DNA marker on a 1.5% agarose/GelRed™ gel electrophoresis.

(161) Sequencing Results and Analysis

(162) The sequencing results are as follows. The consensus DNA sequences and corresponding amino acid sequences are listed below:

(163) TABLE-US-00004 Heavy chain: DNA sequence (426 bp): Leader sequence-FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (SEQ ID NO: 13) ATGGGATGGCGCTGTATCATCCTCTTCTTGGTATCAACAGCTACAGGTGT CCACTCCCAGGTCCAACTACACCAGCCTGGGACTGAACTTATGAAGCCTG CGCCTTCACTGAACCTGTCCTCCAAGGCTTCTGGCTACACCTTCACCACC TACTGGATGCACTGGGTGAAGCACAGGCCTCGACAAGGCCTTCAGTCCAT CGCAGAGATTGATCCTTCTGATACTTATCTTAACTACAATCAAAACTTTA AGGCCAAGGCCACATTGACTGTAGACAAATCCTCCAGCACAGCCTACATA CACCTCACCAGCCTGACATCTGAGGACTCTGCGGTCTATTTTTGTGCGAG ATCCCCTGATTACTACGGTACTAGTCTTGCCTGGTTTGATTACTGGGGCC AAGGGACTCTGGTCACTGTCTCTACA Heavy chain: Amino acids sequence (142 AA): Leader sequence-FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (SEQ ID NO: 14) MGWRCIILFLVSTATGVHSQVQLQQPGTELMKPGASVKLSCKASGYTFTT YWMHWVKQRPGQGLEWIGEIDPSDSYVNYNQNFKGKATLTVDKSSSTAYI QLSSLTSEDSAVYFCARSPDYYGTSLAWFDYWGQGTLVTVST Light chain: DNA sequence (396 bp): Leader sequence-FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (SEQ ID NO: 15) ATGAGGTGCCTAGCTGAGTTCCTGGGGCTGCTTGTGCTCTGGATCCCTGG AGTCATTGGGGATATTGTGATGACTCAGGCTGCACCCTCTGTACCTGTCA CTCCTGGAGAGTCAGTATCCATCTCCTGCAGGTCTAGTAAGAGTCCCCTG CATAGTAACGGCAACATTTACTTATATTGGTTCCTGCAGAGGCCAGGCCA GTCTCCTCAGTTCCTGATATATCGGATGTCCAACCTTGCCTCAGGAGTCC CAGACAGGTTCAGTGGCAGTGGGTCAGGAACTACTTTCACACTGAAAATC AGTAGAGTGGAGGCTGAGGATGTGGGTGTTTATTACTGTATGCAACATCT AGAATATCCGTACACGTTCGGAGGGGGGACCAAGCTCCTAAATAAAA Light chain: Amino acids sequence (132 AA): Leader sequence-FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (SEQ ID NO: 16) MRCLAEFLGLLVLWIPGVIGDIVMTQAAPSVPVTPGESVSISCRSSKSPL HSNGNIYLYWFLQRPGQSPQLLIYRMSNLASGVPDRFSGSGSGTTFTLKI SRVEAEDVGVYYCMQHLEYPYTFGGGTKLEIK

(164) Thus, sequences of ICOS 314.8 (CNCM I-4180) can be resumed as follows:

(165) TABLE-US-00005 Aminoacid  DNA sequence sequence H-CDR1 GGCTACACCTTCACCACCTACT GYTFTTYWMH GGATGCAC (SEQ ID NO: 7) (SEQ ID NO: 1) H-CDR7 GAGATTGATCCTTCTGATAGTT EIDPSDSYVNYNQNFKG ATGTTAACTACAATCAAAACTT (SEQ ID NO: 8) TAAGGGC (SEQ ID NO: 2) H-CDR3 TTTGATTAC FDY (SEQ ID NO: 3) (SEQ ID NO: 9) L-CDR1 AGGTCTAGTAAGAGTCCCCTGC RSSKSPLHSNGNIYLY ATAGTAACGGCAACATTTACTT (SEQ ID NO: 10) ATAT (SEQ ID NO: 4) L-CDR2 CGGATGTCCAACCTTGCCTCA RMSNLAS (SEQ ID NO: 5) (SEQ ID NO: 11) L-CDR3 ATGCAACATCTAGAATATCCGT MQHLEYPYT ACACG (SEQ ID NO: 6) (SEQ ID NO: 12)

Example 10: Sequencing of Icos 88.2 (CNCM I-4177)

(166) Total RNA was extracted from provided frozen hybridoma cells and cDNA was synthesized. Then, RT-PCR was performed to amplify the variable regions (heavy and light chains) of the MAb. The MAb variable regions of the heavy and light chains were cloned into a cloning vector separately, then the obtained sequences were analyzed to deduce the sequences of the MAb.

(167) Materials

(168) Hybridoma cells ICOS 88.2 (CNCM I-4177); TRIzol® Plus RNA Purification System (Invitrogen, Cat. No: 15596-026); SuperScript™ III First-Strand Synthesis System (Invitrogen, Cat. No: 18080-051).

(169) Methods

(170) Total RNA Extraction

(171) Total RNA was isolated from the hybridoma cells according to the technical manual of TRIzol® Plus RNA Purification System. The total RNA was checked by gel electrophoresis.

(172) RT-PCR

(173) Total RNA was reverse transcribed into cDNA using isotype specific anti-sense primer or universal primer and whole procedure was according to the technical manual of SuperScript™ III First-Strand Synthesis System. The antibody fragment will be amplified according to the standard operation protocol of RACE method of GenScript.

(174) Cloning of Antibody Genes

(175) Target PCR products of antibody genes were cloned into the cloning vector separately according to standard molecular cloning procedures.

(176) Screening and Sequencing

(177) Colony screening was employed to screen clones with inserts of correct sizes, and no less than ten independent positive colonies were sequenced for each antibody fragment.

(178) Results and Analysis

(179) Total RNA Extraction

(180) Total RNA of the sample was run alongside of DL3000 DNA marker on a 1.5% agarose/GelRed™ gel electrophoresis.

(181) PCR Product of Antibody Genes

(182) 4 μl PCR products of each sample were run alongside of DL3000 DNA marker on a 1.5% agarose/GelRed™ gel electrophoresis.

(183) Sequencing Results and Analysis

(184) The sequencing results are as follows. The consensus DNA sequences and corresponding amino acid sequences are listed below:

(185) TABLE-US-00006 Heavy chain: DNA sequence (429 bp): Leader sequence-FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (SEQ ID NO: 29) ATGGGATGGAGCTGTATCATCCTCTTCTTGGTAGCAACAGCTACAGGTGT CCACTCCCACCTCCAACTGCAGCAGCCTGGGGCTGAGCTGGTGAGGCCTG GGGCTTCAGTGAAGCTGTCCTGCAAGGCTTCTGGCTACAGTTTCACCACC TACTGGATAAACTGGGTGAAGCAGACGCCTGGACAAGGCCTTGAGTGGAT CGGAAATATTTATCCTTCTGATAGTTATACTAACTACAATCAAATGTTCA AGGAGAAGGCCACATTGACTGTAGACAAATCCTCCAACACACCCTACATG CACCTCACCAGCCCGACATCTGAGGACTCTGCGCTCTATTACTGTACAAG ATGGAATCTTTCTTATTACTTCGATAATAACTACTACTTGGACTACTGGG GCCAAGGCACCACTCTCACACTCTCCTCA Heavy chain: Amino acids sequence (143 AA): Leader sequence-FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (SEQ ID NO: 30) MGWSCIILFLVATATGVHSQVQLQQPGAELVRPGASVKLSCKASGYSFTS YWINWVKQRPGQGLEWIGNIYPSDSYTNYNQMFKDKATLTVDSKKNTAYM QLTSPTSEDSAVYYCTRWNLSYYFDNNYYLDYWGQGTTLTVSS Light chain: DNA sequence (396 bp): Leader sequence-FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (SEQ ID NO: 31) ATGAGGTGCCTAGCTGAGTTCCTGGGGCTGCTTGTGCTCTGGATCCCTGG AGCCATTGGGGATATTGTGATGACTCAGGCTGCACCCTCTGTACCTGTCA CTCCTGGAGAGTCAGTATCCATCTCCTGCAGGTCTAGTAAGAGTCTCCTG CATAGTAATGGCAACACTTACTTGTATTGGTTCCTGCAGAGGCCAGGCCA GTCTCCTCAACTCCTGATATATCGGATGTCCAACCTTGCCTCAGGAGTCC CAGACAGGTTCAGTGGCAGTGGGTCAGGAACTGCTTTCACACTGAGAATC AGTAGAGTGGAGGCTGAGGATGTGGGTGTTTATTACTGTATGCAACATCT AGAATATCCGTGGACGTTCGGTGGAGGCACCAAGCTGGAAATCAAA  Light chain: Amino acids sequence (132 AA): Leader sequence-FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (SEQ ID NO: 32) MRCLAEFLGLLVLWIPGAIGDIVMTQAAPSVPVTPGESVSISCRSSKSLL HSNGNTYLYWFLQRPGQSPQLLIYRMSNLASGVPDRFSGSGSGTAFTLRI SRVEAEDVGVYYCMQHLEYPWTFGGGTKLEIK 

(186) Thus, sequences of ICOS 88.2 (CNCM I-4177) can be resumed as follows:

(187) TABLE-US-00007 Aminoacid  DNA sequence sequence H-CDR1 GGCTACAGTTTCACCAGCTA GYSFTSYWIN CTGGATAAAC (SEQ ID NO: 23) (SEQ ID NO: 17) H-CDR2 AATATTTATCCTTCTGATAG NIYPSDSYTNYNQMFKD TTATACTAACTACAATCAAA (SEQ ID NO: 24) TGTTCAAGGAC (SEQ ID NO: 18) H-CDR3 TGGAATCTTTCTTATTACTT WNLSYYFDNNYYLDY CGATAATAACTACTACTTGG (SEQ ID NO: 25) ACTAC (SEQ ID NO: 19) L-CDR1 AGGTCTAGTAAGAGTCTCCT RSSKSLLHSNGNTYLY GCATAGTAATGGCAACACTT (SEQ ID NO: 26) ACTTGTAT (SEQ ID NO: 20) L-CDR2 CGGATGTCCAACCTTGCCTC RMSNLAS A (SEQ ID NO: 27) (SEQ ID NO: 21) L-CDR3 ATGCAACATCTAGAATATCC MQHLEYPWT GTGGACG (SEQ ID NO: 28) (SEQ ID NO: 22)