HERV-K-DERIVED ANTIGENS AS SHARED TUMOR ANTIGENS FOR ANTI-CANCER VACCINE

Abstract

A composition or vaccine comprising at least one peptide, or an expression vector that induces expression of said at least one peptide in vivo, the peptide consisting of, or comprising, shared HERV-K derived antigens, and a pharmaceutically acceptable vehicle or excipient. Composition comprising Cytotoxic T Lymphocytes (CTLs) of a patient treated with such a peptide, or comprising T-cell Receptor (TCR) engineered T cells recognizing such a peptide.

Claims

1. A composition comprising 2, 3, 4, 5, 6 or 7 peptides, or one or more expression vector(s) that induce(s) expression of said 2, 3, 4, 5, 6 or 7 peptides in vivo, the peptides having from 9 to 100 amino acid residues, each one comprising at least one of the epitopes of sequences SEQ ID NO: 1-7, and each peptide comprising at least one different epitope with respect to the others; or at least one peptide, or an expression vector that induces expression of said at least one peptide in vivo, said peptide having from 9 to 100 amino acids residues and comprising 2, 3, 4, 5, 6 or 7 of the epitopes of sequences SEQ ID NO: 1-7; and a pharmaceutically acceptable vehicle or excipient.

2. The composition of claim 1, comprising or expressing one or more peptides comprising the epitopes of sequence SEQ ID NO: 1 and/or 6.

3. The composition of claim 1, comprising or expressing 3, 4, 5, 6, or the 7 peptides comprising the epitopes of sequences SEQ ID NO: 1 to 7.

4. The composition of claim 1 wherein the peptides comprise 9 to 100, 70, 50, 40, 30, 25, or 20 amino acid residues and at least one of said epitopes of sequence SEQ ID NO: 1 to 7.

5. The composition of claim 4, wherein each peptide of 9 to 100, 70, 50, 40, 30, 25, or 20 amino acid residues comprises one specific epitope of SEQ ID NO: 1, 2, 3, 4, 5, 6 or 7.

6. The composition of claim 1, wherein the peptide(s) comprise(s) 9 to 100, 70, 50, 40, 30, 25, or 20 consecutive amino acid residues of an HERV gag or pol including at least one of said epitopes of sequence SEQ ID NO: 1 to 7.

7. The composition of claim 1, wherein the peptide(s) comprise(s) 13, 14, 15, 16, 17, or 18, consecutive amino acid residues of an HERV gag or pol including at least one of said epitopes of sequence SEQ ID NO: 1 to 7.

8. The composition of claim 1, wherein the peptides are selected from the group consisting of SEQ ID NO: 8-14 and the composition comprises 2, 3, 4, 5, 6 or 7 of them.

9. The composition of claim 1, wherein the peptides consist of 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28 consecutive amino acid residues of sequences SEQ ID NO: 8-14, including the 9-mer epitope, and the composition comprises 2, 3, 4, 5, 6 or 7 of them.

10. Vaccine or immunogenic composition comprising the composition of claim 1 and a pharmaceutically acceptable vehicle or excipient and preferably an adjuvant.

11. Composition comprising Cytotoxic T Lymphocytes (CTLs) of a patient treated with a peptide as described in claim 1, or comprising T-cell Receptor (TCR) engineered T cells recognizing a peptide as described in any one of claims 1-7, and a pharmaceutical vehicle.

12. Composition according to claim 1, for use in treating cancer, in particular breast cancer, including triple negative breast cancer, ovarian cancer, melanoma, sarcoma, teratocarcinoma, bladder cancer, lung cancer (non small cell lung carcinoma and small cell lung carcinoma), head and neck cancer, colo-rectal cancer, glioblastoma and leukemias.

13. Isolated peptide selected from the group consisting of the peptides of sequence SEQ ID NO: 2-14, and the peptides having 10 to 100, 70, 50, 40, 30, 25, or 20 amino acids and comprising at least one of said peptides of SEQ ID NO: 2-7.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0136] FIG. 1: A. Correlation of the expression of the 19 HERV subtypes overexpressed in TNBC with immune signatures, showing a correlation with antigen presentation, cytotoxic and immunoregulatory pathway; positive correlation is indicated by each dot (light or dark grey); B. As an example, correlation of HERV-K 10 expression with the T cell signature.

[0137] FIG. 2: A. Representative plots of dextramer staining of CD8+ T cells without or with stimulation by CMV pp65 peptide (respectively, upper quadrants) and without or with HERV peptide P5 (SEQ ID NO: 5) (respectively, lower quadrants). B. Percentage of dextramer positive CD8+ T cells at day 12 without or with stimulation with specific HERV peptides (P1 to P7—SEQ ID NO: 1 to 7) or controls on several donors' PBMCs.

[0138] FIG. 3: A. Representative plots of IFN-γ production by CD8+ T cells after contact with T2 cells pulsed with negative and positive controls (upper quadrants) and HERV peptides P1, P2, P3 (SEQ ID NO: 1, 2 and 3) (lower quadrants) B. Percentage of IFN-γ positive CD8+ T cells at day 12 without or with stimulation with specific HERV peptides (P1, P2 and P3 (SEQ ID NO: 1, 2 and 3) or controls on several donors' PBMCs.

[0139] FIG. 4: A. Representative plots of dextramer staining of CD8+ T cells without or with stimulation by CMV pp65 peptide (respectively, upper quadrants) and without or with HERV peptide P1 (SEQ ID NO: 1) (respectively, lower quadrants).

[0140] B. Fold change ratio between percentage of dextramer positive specific CD8+ T cells in the peptide stimulated condition versus non stimulated (P1 to P7—SEQ ID NO: 1 to 7) at day 12 on several donors' PBMCs.

[0141] C. Representative histograms of the number of IFN-γ+ and Granzyme-I3+ spots after 12 days stimulation and following 24h of contact with T2 cells pulsed with the cognate peptide.

[0142] FIG. 5: A. Representative plots of IFN-γ production by CD8+ T cells after contact with T2 cells pulsed with negative and positive controls (upper quadrants) and HERV peptides P1, P2, P3 (SEQ ID NO: 1 to 3) (lower quadrants), on donor a;

[0143] A (continued). representative plots of IFN-γ production by CD8+ T cells after contact with T2 cells pulsed with negative and positive controls (upper quadrants) and HERV peptides P4, P5, P6 and P7 (SEQ ID NO: 4 to 7) (lower quadrants) for donor b;

[0144] B. Percentage of IFN-γ positive CD8+ T cells at day 12 without or with stimulation with specific HERV peptides (P1 to P7—SEQ ID NO: 1 to 7) or controls on several donors' PBMCs.

[0145] FIG. 6: A. Representative plots of dextramer staining for P1 (SEQ ID NO: 1) after sorting and expansion of specific P1 CD8+ T cells (right quadrant) and their negative counterpart (left quadrant).

[0146] B. Representative histograms of the number of IFN-γ+ and Granzyme-I3+ spots, after 24h of culture of the P1 (SEQ ID NO: 1) specific CD8+ T cells with T2 cells pulsed with P1 or negative control (without peptide charged).

[0147] C. Representative curves of real-time cell death quantification in co-cultures of MDA-MB-231 cell line (pulsed or not with the peptide of interest), with P1 (SEQ ID NO: 1) specific CD8+ T cells or their negative counterpart.

[0148] D. Representative histograms of the percentage of intracellular staining of IFN-γ (PE) of P1 (SEQ ID NO: 1) specific CD8+ T cells (in black) versus their negative counterpart (non specific CD8+ T cells, in white) after 6h of co-culture with MDA-MB-231 cell line pulsed or not with the peptide of interest. The addition of an HLA-A2 blocking antibody was used as a control.

[0149] FIG. 7: A. Representative schema of dextramer positive tumor infiltrated lymphocytes (TILs) (Black quadrants) found in CD8 T cells expanded from TNBC dilacerate. P1 to P7 (SEQ ID NO: 1 to 7) represent dextramer for peptides 1 to 7, TNBC 1 to 4 represent 4 different patients B. Representative schema of dextramer positive TILs (Black quadrants) found in CD8 T cells expanded from ovarian tumor dilacerate. P1 to P7 (SEQ ID NO: 1 to 7) represent dextramer for peptides 1 to 7, Ovary 1 to 3 represent 3 different patients.

[0150] The invention will now be described using non-limiting examples, referring to the drawings.

Example

[0151] Identification of HERV Sequences

[0152] HERVs DNA sequences from different families (including HERV-K, HERV-H, HERV-W, HERV-E and ERV3) were extracted from the Genbank database. BLAST was used to localize these sequences on the human genome (GRCH37), keeping the position with at least 98% of similarity on at least 85% of the queried sequences and no gap. 66 functional HERVs sequences have thus been identified.

[0153] Analysis of HERV Sequences in TNBC

[0154] HERVs expression was analyzed in a pre-existing database of 84 breast cancer samples containing 42 TNBC. Comparison was made with 56 normal beast sample (51 peritumoral and 5 mammal reduction samples). RNA was extracted from fresh tumor biopsies performing DNAse treatment and poly A selection. If presenting a sufficient quality (RNA integrity number>6.5), functional HERVs sequences are aligning with the RNAseq data.

[0155] Multiple component analysis of the 66 Human endogenous retroviruses (HERV) subtypes was performed. 42 HERVs are expressed. 19 HERVs specifically characterize triple negative breast cancer (TNBC) upon normal tissue and ER+ subtypes: HERVK 10, 17, 22, 7, 6, 21, 25, 11, 20, 16, 23, 1, 5; HERVH_4, 7; and HERV3_1.

[0156] Peptide Selection and Synthesis

[0157] Common regions in the Gag and Pol shared between the 19 overexpressed HERVs were identified after alignement of the reads on a reference genome. Using different epitope prediction tools (NetMHC I & II), potential strong epitope binders for the most frequent MHC I and II alleles were identified. Among them, 7 predicted 9-mer strong binders for HLA-A*0201 were selected and synthetized: 4 Gag and 3 Pol Peptides (JPT peptide technology, Berlin, Germany). Peptides identity was confirmed by mass spectrometry by the seller. Purity>95% was expected and determined by high-performance liquid chromatography. Lyophilised peptides were dissolved in deionized water <5% DMSO, aliquoted and conserved at −20° C. until use.

[0158] 29mer GMX peptides (SEQ ID NO: 8-14) containing the 9-mer peptides strong binder for class I MHC (SEQ ID NO: 1-7), plus lateral sequences of class II motifs (10-mer on each side, except for peptide SEQ ID NO: 12, where the sequence SEQ ID NO: 5 is on the C-terminal) were identified and analyzed for synthesis.

[0159] Bioinformatics Analysis of the Correlations Between HERVs Expression and T Cell Signatures

[0160] Different signatures were used to evaluate the immune characteristics of the tumor: MCP counter signatures (ref http://cit.ligue-cancernet/?p=1338&lang=en) for T cells, CD8 T cells, NK cells, Cytotoxic lymphocytes, as well as Fibroblasts, Neutrophils, and Endothelial cells for a control analysis; Estimate package (http://bioinformatics.mdanderson.org/estimate/index.html) for ImmuneScore, StromalScore; Immunophenogram 15 for Effector cells, immunomodulators (immune checkpoints), suppressor cells (regulatory T cells and MDSCs). Specific genes will be also evaluated (like OCT4, TRIM28, SETDB1) as well as EMT signature (SSGSEA and Jean-Paul Thierry signature) and signatures associated specifically with the tumor subtype. Correlation between these signatures and HERVs will be analyzed using classical statistical methods. FIG. 1A shows a significant correlation between 19 HERVs and specific immune signatures as antigen presentation-related, cytotoxic or immunomodulatory signatures. As an example, HERV-K 10 expression strongly correlates with the T cell signature in breast cancer (FIG. 1B).

[0161] PBMCs Culture for Specific CD8+ HERV+ Stimulation

[0162] PBMCs from HLA-A2 donors where cultured for 12 days in AIM-V medium (Thermo Fisher Scientific) supplemented with 5% AB human serum (pool of 5 donors from EFS Lyon, filtered) and 20 UI/mL IL-2 (PROLEUKIN aldesleukine, Prometheus, Vevey, Switzerland) enriched with 5 μg/mL of R848 (InvivoGen, San Diego, USA) and 10 μg/mL of Poly-IC (InvivoGen) and the peptide of interest (10 μM). Cultures were performed in a 96 U-bottom wells plate, 1.5×10.sup.5 cells per well, and 20 wells were performed for each peptide condition. 100 μL of medium were changed and enriched with R848, Poly-IC, IL-2 and the peptide of interest (peptides of sequences SEQ ID NO: 1-7) to achieve the same final concentration on day 3. IL-2 and the peptide of interest where added on day 6 and on day 10 IL-2 only. Positive control was cultured with 0.1 μg/mL of PP65 (JPT peptide technology) in the dextramer experiment, a CMV peptide that is presented by the MHC class I and specifically stimulate CD8+ T cells. For IFN-γ experiment, 0.4 μg/mL CEF peptide (Mabtech) were used, consisting in a pool of 23 MHC class I restricted viral peptides from human CMV, EBV and influenza virus which stimulate CD8+ T cells to preferentially synthetize IFN-γ.

[0163] Dextramer Assay and Sorting

[0164] On day 12, cells from same conditions were pooled together in polypropylene tubes, washed with 2 mL of FACS buffer and resuspended in FACS buffer. Conditions were stained with 10 μL of the corresponding dextramer (Immudex, Copenhague, Danemark) for 15 minutes at room temperature in the dark. Zombie Near Infra-Red (NIR) fixable viability kit (Zombie NIR, biolegend, Paris, France) was used at 1/400 to assess viability. Anti CD3 (BV421, Biolegend) and anti CD8 (FITC, Beckman coulter, Brea, USA) antibodies were then added to each condition ( 1/10 in the assay of FIG. 2, 1/25 for FIG. 4) and left 20 minutes in the dark at 4° C. Cells were then washed two times with 2 ml of FACS buffer and resuspended in 350 μL of FACS buffer until analysis. Analysis was performed on FACS Fortessa (BD) to discriminate multimer specific HERV CD8+ T cells.

[0165] Results in FIG. 2A show the population stained by dextramer in non-specific (left panels) and specific (right panels) peptide-pulsed PBMCs. In upper panels, up to 72% of CD8+ T cells resulted positive after stimulation with the positive control pp65 peptide from CMV vs. 2.02% in non-stimulated PBMCs (indicating a possible presence of memory T cells against CMV due to previous infections). Interestingly PBMCs stimulated with a HERV peptide (e.g. SEQ ID NO: 5 in the lower panel) resulted in a 34.2% of dextramer positive CD8+ T cells vs. 0.04% in the non-stimulated condition. Results obtained on four different donors are summarized in FIG. 2B.

[0166] Results in FIG. 4A show the population stained by dextramer in non-specific (left panels) and specific (right panels) peptide-pulsed PBMCs. In upper panels, up to 23.40% of CD8+ T cells resulted positive after stimulation with the positive control pp65 peptide from CMV vs. 3.63% in non-stimulated PBMCs (indicating a possible presence of memory T cells against CMV due to previous infections). Interestingly PBMCs stimulated with a HERV peptide (e.g. SEQ ID NO: 1 in the lower panel) resulted in a 0.63% of dextramer positive CD8+ T cells vs. 0.091% in the non-stimulated condition. Results obtained on 12 different donors are summarized in FIG. 4B, showing a significant increase in dextramer positive CD8+ T cells for P1, P4 and P6 (e.g. SEQ ID NO: 1, 4, 6) and slightly for P2 and P3 (e.g. SEQ ID NO: 2, 3)

[0167] Peptide stimulated PBMCs after the 12-day culture were used to perform an ELISPOT assay for IFN-γ and Granzyme-β (FIG. 2C). 2.10.sup.5 cells per well were put in co-culture with T2 cell specific for the peptide of interest in a ratio 10:1. ELISPOT was revealed after 24h showing specific IFN-γ and Granzyme-β spots for P1 and P6 (e.g. SEQ ID NO: 1, 6).

[0168] Cytotoxicity Assay with T2 Cells Contact

[0169] On day 12 of PBMCs culture, T2 cells were washed in RPMI and resuspended into AIM-V medium (Thermo Fisher Scientific). T2 (SD Cell line) are a lymphoblast cell line deficient in the transporter associated with antigen processing (TAP) protein, and therefore cannot present endogenous peptides on the class I MHC, but can be used to monitor the CTL response to an exogenous antigen of interest in a non-competitive environment. T2 cells were first pulsed with HERV peptide by adding 10 μg/mL of corresponding peptide to 2M T2 cell for 2 hours at 37° C. PBMCs were pooled and counted and put in co-culture with corresponding T2 cells at a respective concentration of 1:5 into a new 96 U-well plate.

[0170] After 4 hours of incubation at 37° C., cells from co-culture were washed and pooled in V-well plate according to their staining condition in FACS buffer. Zombie NIR (Biolegend) was used at 1/400 to assess viability. Anti CD3 (PercP, Biolegend) and anti CD8 (FITC, Beckman coulter) antibodies were added 1/10 per condition and left for 25 minutes at 4° C. Cells were washed again and then fixed with fixation/permeabilization solution kit (Invitrogen, Carlasbad, USA) according to the manufacturer's instructions for 15 minutes at room temperature. Cells were washed two times in FACS buffer and kept at 4° c.

[0171] On day 13, cells were permeabilized with the permeabilization solution kit (Invitrogen) for 5 minutes at room temperature and anti IFN-γ (PE, Biolegend) antibody was added 1/20 to the solution for additional 25 minutes at 4° C. Cells were washed two times and resuspended in 350 μL of FACS buffer before FACS analysis. Analysis was performed on FACS Fortessa (BD) to explore the specific cytotoxicity and degranulation against T2 cells expressing HERV sequences.

[0172] Results in FIG. 3A show IFN-γ production of stimulated CD8+ T cells against T2 cells not pulsed (P0: upper-left panel), pulsed with a different peptide (Pneg: upper-central panel), or with each cognate peptide (positive control Ppos: upper/right panel, HERV peptides of SEQ ID NO 1, 2 and 3 in lower panels); results show a specific IFN-γ production against T2 cells expressing HERV peptides. Results of 5 different donors are summarized in FIG. 3B, showing an IFN-γ production in ˜10% (in median) of CD8+ T cell stimulated with P1 (SEQ ID NO: 1) and ˜5% (in median) of CD8+ T cells stimulated with P2 or P3 (SEQ ID NO: 2 and 3).

[0173] Results in FIG. 5A show IFN-γ production of stimulated CD8+ T cells against T2 cells not pulsed (P0: upper-left panel in both donor a and donor b quadrants), pulsed with a different peptide (Pneg: upper-central panel in both donor a and donor b quadrants), or with each cognate peptide (positive control Ppos: upper/right panel, HERV peptides of SEQ ID NO 1 to 3 in lower panels for donor a and HERV peptides of SEQ ID NO 4 to 7 in lower panels for donor b); results show a specific IFN-γ production against T2 cells expressing HERV peptides. Results of 12 different donors are summarized in FIG. 5B, showing an IFN-γ production in a significant number of CD8+ T cells stimulated with P1 (SEQ ID NO: 1), P2 and P3 (SEQ ID NO: 2 and 3) and in a moderate number of CD8+ T cells stimulated with P6 (SEQ ID NO: 6).

[0174] Cytotoxicity of P1 Specific CD8+ T Cells

[0175] The dextramer-stained cells were sorted by FACS Aria (BD) to separate peptide-specific CD8+ T cells from the unspecific counterpart. Both fractions were collected and expanded separately on feeder cells in 96 round-well plates for 14 days. The purity of the specific versus unspecific fraction was evaluated at day 14 (FIG. 6A) resulting in >90% of dextramer positive CD8+ T cells in the positive fraction versus <5% in the negative fraction. These cells were used for the cytotoxicity experiments.

[0176] Sorted and expanded CD8+ T cells were evaluated for their cytotoxic potential by ELISPOT assay (Cellular technology limited, CTL). 4.10.sup.4 P1 specific CD8+ T cells were co-cultured with T2 cells previously pulsed with the peptide P1. The number of spots counted indicates a production of IFN-γ and Granzyme-β (˜800 spots for both cytokines) by the P1 specific CD8+ T cells against target cells pulsed with the cognate peptide in comparison to negative control (FIG. 6B). These experiments indicate that those cells are specific and functional.

[0177] The in silico analysis of HERV expression performed on the cell lines HMEC (HLA-A2 human mammary epithelial cells) and MDA-MB-231 (HLA-A2 Triple negative breast cancer cell line), showed an overexpression of HERVs in the MDA-MB-231 cell line in comparison to the HMEC.

[0178] The HLA-A2 TNBC cell line MDA-MB-231 was used as target for a real-time analysis of cell death induced by P1 specific CD8+ T cells. 5.10.sup.3 MDA-MB-231 cells were pulsed or not with the peptide P1 and were allowed to adhesion in 96-well plates. After adhesion P1 CD8+ T cells or their negative counterpart (control) were added into the wells. The co-culture was performed in the presence of the Cytotox green dye (Essenbioscience) which enters into the cells when the plasma membrane integrity diminishes, yielding a 100-1000-fold increase in fluorescence upon binding to deoxyribonucleic acid (DNA). The kinetics shows a very significant increase in cell death when MDA-MB-231 are co-cultured with the P1-specific T cells in comparison to their negative counterpart. As expected, a further increase of cell death was observed when target MDA-MB-231 cells were pulsed with P1 and co-cultured with P1-specific T cells (probably due to an increase in the number of HLA-peptide 1 complexes on target cells) (FIG. 6C).

[0179] Furthermore, after 6 hours of co-culture between P1 specific CD8+ T cells or their negative counterpart and MDA-MB-231 HLA-A2 TNBC cell line, an intracellular staining of IFN-γ was performed by FACS (FIG. 6D). Results shows an increase in the percentage of IFN-γ producing cells when the P1 specific cells are in the co-culture in comparison to the co-cultures with the negative counterpart. This percentage is slightly increased when MDA-MB-231 are previously pulsed with P1. The use of an anti HLA-A2 specific blocking antibody can reverse this effect, demonstrating its specificity.

[0180] Altogether, these experiments show that P1-specific CD8+ T cells specifically recognize and are functional against target cells presenting the cognate peptide (T2 cells in this experiment) and specifically recognize and kill tumor cells expressing endogeneously HERV-derived antigens (MDA-MB-231 TNBC cell line in this experiment).

[0181] Dextramer Staining of Tumor Infiltrated Lymphocytes (TILs) from TNBC and Ovarian Cancer

[0182] Tumors were dilacerated in small pieces and digested with collagenase IV and DNAse for 45 minutes. The cells obtained were resuspended in 5% human serum RPMI and distributed in 96 well (5.10.sup.4 cells per well). Anti-CD3/CD28 microbeads (Miltenyi biotech) were added to the well in a ratio of 1:4 with cells in the presence of IL-2 (F.C. 100 IU/ml). TILs were cultured for 14 days by changing medium at day 5, 7, 9 and 12 and the number of the cells was adapted to have 0.5×10.sup.6 cells/ml.

[0183] At day 14 a dextramer staining was performed (see dextramer assay paragraph) for the 7 peptides of interest (P1 to P7—SEQ ID NO: 1 to 7) and dextramer specific CD8+ T cells were identified by FACS analysis in TNBC (FIG. 7A) and ovarian cancer (FIG. 7B). Results show that specific CD8+ T cells for all the peptides of interest can be found in tumors, with some variations according to the sample. This confirms the in vivo immunogenicity of the peptides and shows that these peptides can be naturally processed and can elicit a detectable immune response during tumor development. CD8+ T cells specific for P1 and P6 are more frequently represented in the tumors and are peptides of choice as a basis of a vaccine or immunogenic composition. These results also militates in favor of a combination of several peptides in order to provide for a vaccine or an immunogenic composition that is usable in patients of unknown status with respect to this reactivity, and preferably comprising P1 and/or P6.