MEANS AND METHODS FOR DETERMINING T CELL RECOGNITION

Abstract

The invention provides improved screening methods for testing T cell recognition of T cell epitopes.

Claims

1. Method for determining whether a T cell recognizes a T cell epitope, comprising: inducing, enhancing and/or maintaining expression of Bcl-6 and/or STAT5 in at least one B cell; inducing, enhancing and/or maintaining expression of an anti-apoptotic nucleic acid in said at least one B cell; allowing expansion of said at least one B cell into a B cell culture; incubating B cells of said B cell culture with at least one compound; incubating the resulting B cells with T cells; and determining whether at least one T cell recognizes at least one T cell epitope of said at least one compound.

2. Method according to claim 1, comprising incubating said B cells with at least one peptide.

3. Method according to claim 1, wherein said at least one peptide has a length of between 5 and 40 amino acids, preferably between 8 and 35 amino acids or between 9 and 31 amino acids.

4. Method according to claim 1, wherein said B cells are incubated with at least 2, preferably with at least 5, preferably with at least 10 different peptides.

5. Method according to claim 1, wherein said T cells comprise CD8+ cytotoxic T cells and/or CD4+ helper T cells.

6. Method according to claim 1, wherein said at least one B cell and said T cells are from the same human individual.

7. Method according to claim 1, wherein said T cell epitope is from a modified self-antigen or from a non-self antigen or from an autoantigen.

8. Method according to claim 1, wherein said T cell epitope is a tumor-specific T cell epitope.

9. Method according to claim 8, wherein said tumor is a melanoma or epithelial cancer.

10. Method according to claim 1, wherein said T cell epitope is from a pathogen.

11. Method according to claim 10, wherein said pathogen is a virus, a bacterium or a parasite.

12. Method according to claim 1, wherein said T cell epitope is from an autoantigen.

13. Method according to claim 1, wherein it is determined whether at least one T cell has recognized at least one of said T cell epitopes by determining whether said at least one T cell is activated.

14. Method according to claim 13, wherein said T cell activation is determined by measuring cytokine release.

15. Method according to claim 1, further comprising preparing a medicament comprising T cells that recognize at least one T cell epitope of said compounds.

16. Method according to claim 1, further comprising identifying at least one T cell epitope recognized by a T cell.

17. Method according to claim 16, further comprising preparing an immunogenic composition, or a prophylactic agent or vaccine, comprising said at least one T cell epitope.

18. Method according to claim 16, further comprising preparing an immunogenic composition, or a prophylactic agent or vaccine, comprising an antigen presenting cell, preferably a B cell, which displays on its surface said at least one T cell epitope.

19. Method according to claim 1, wherein the B cells are incubated with T cells from an individual suffering from, or having suffered from, a disorder.

20. Method according to claim 19, wherein said individual is suffering from, or has suffered from, cancer, preferably melanoma or epithelial cancer.

21. Method according to claim 19, wherein said individual is suffering from, or has suffered from, a pathogen, preferably a virus, a bacterium or a parasite.

22. Method according to claim 19, wherein said individual is suffering from, or has suffered from, an autoimmune disease, preferably multiple sclerosis, diabetes or coeliac disease.

23. Method according to claim 1, wherein said T cells are from a sample from said individual, characterized in that the proportion of T cells specific for a T cell epitope that is associated with said disease, relative to the total amount of T cells, is lower than 24%, more preferably lower than 23%, more preferably lower than 20%, more preferably lower than 15%, more preferably lower than 10%, more preferably lower than 9%, more preferably lower than 8%, more preferably lower than 7%, more preferably lower than 6%, more preferably lower than 5% in said sample or in a T cell culture after in vitro expansion of said sample.

24. Method according to claim 1, wherein said T cells are from a sample from said individual, characterized in that the proportion of T cells specific for a T cell epitope that is associated with a disorder of interest, relative to the total amount of T cells, is lower than 1.9%, more preferably lower than 1.8%, more preferably lower than 1.7%, more preferably lower than 1.6%, more preferably lower than 1.5% in said sample or in a T cell culture after in vitro expansion of said sample.

25. Method according to claim 1, wherein said T cells are from a sample from said individual, characterized in that the proportion of T cells specific for a T cell epitope that is associated with a disorder of interest, relative to the total amount of T cells, is lower than 1.0%, more preferably lower than 0.9%, more preferably lower than 0.8%, more preferably lower than 0.7%, more preferably lower than 0.6%, more preferably lower than 0.5%, more preferably lower than 0.4%, more preferably lower than 0.3% in said sample or in a T cell culture after in vitro expansion of said sample.

26. Method according to claim 1, wherein a sample from said individual is used, or wherein T cells of a resulting T cell culture after in vitro expansion of said sample are used, wherein the percentage of T cells specific for a T cell epitope that is associated with said disease, relative to the total number of T cells in said sample or in said resulting in vitro T cell culture, is lower than 24%, more preferably lower than 23%, more preferably lower than 20%, more preferably lower than 15%, more preferably lower than 10%, more preferably lower than 9%, more preferably lower than 8%, more preferably lower than 7%, more preferably lower than 6%, more preferably lower than 5%.

27. Method according to claim 1, wherein a sample from said individual is used, or wherein T cells of a resulting T cell culture after in vitro expansion of said sample are used, wherein the percentage of T cells specific for a T cell epitope that is associated with a disorder of interest, relative to the total number of T cells in said sample or in said resulting in vitro T cell culture, is lower than 1.9%, more preferably lower than 1.8%, more preferably lower than 1.7%, more preferably lower than 1.6%, more preferably lower than 1.5%.

28. Method according to claim 1, wherein a sample from said individual is used, or wherein T cells of a resulting T cell culture after in vitro expansion of said sample are used, wherein the percentage of T cells specific for a T cell epitope that is associated with a disorder of interest, relative to the total number of T cells in said sample or in said resulting in vitro T cell culture, is lower than 1.0%, more preferably lower than 0.9%, more preferably lower than 0.8%, more preferably lower than 0.7%, more preferably lower than 0.6%, more preferably lower than 0.5%, more preferably lower than 0.4%, more preferably lower than 0.3%.

29. Method according to claim 23, wherein said disease or said disorder is selected from the group consisting of cancer, preferably melanoma or epithelial cancer, an infectious disease, preferably a viral infection, a bacterial infection or a parasite infection, and an autoimmune disease, preferably multiple sclerosis, diabetes or coeliac disease.

30. Method according to claim 1, wherein said anti-apoptotic nucleic acid comprises a gene encoding an anti-apoptotic molecule, preferably of the BCL2 family, preferably Bcl-xL, Mcl-1, Bcl-2, A1, Bcl-w, Bcl2L10, or a functional part or homologue thereof, most preferably Bcl-xL or Mcl 1.

31. A medicament comprising T cells that recognize a disorder-associated T cell epitope, preferably a tumor-specific T cell epitope or a T cell epitope from a pathogen or a T cell epitope from an autoantigen, when obtained with a method according to claim 1.

32. An immunogenic composition, or a prophylactic agent or vaccine, comprising a disorder-associated T cell epitope, preferably a tumor-specific T cell epitope or a T cell epitope from a pathogen or a T cell epitope from an autoantigen, when obtained with a method according to claim 1.

33. Use of a B cell for presenting an epitope of interest, characterized in that the in vitro replicative life span of said B cell is prolonged by inducing, enhancing and/or maintaining expression of Bcl-6 and/or STAT5 in said B cell and by inducing, enhancing and/or maintaining expression of an anti-apoptotic nucleic acid in said B cell.

34. Use according to claim 33, wherein said epitope is a T cell epitope.

35. Method or use according to claim 1, wherein said T cells are CD4+ T cells or CD8+ T cells.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0132] FIG. 1. Experimental setup for the identification of neo-antigen specific CD4.sup.+ T cells in tumor lesions. Whole exome-sequencing of tumor and healthy material in combination with RNA-sequencing identifies tumor-specific, non-synonymous mutations within genes with confirmed RNA-expression. This information was used to create a library of putative neo-epitopes comprised of 31 amino-acid long peptides extending each identified mutation by 15 amino-acids on either side.

Autologous B cells immortalized by retroviral gene transfer of Bcl-6 and Bcl-xL enable the profiling of CD4.sup.+ T cell reactivity against all MHC-class 11 haplotypes of the subject. Stimulation of CD4.sup.+ T cell cells obtained from melanoma lesion of the same subject by neo-antigen peptide-loaded Bcl-6/Bcl-xL immortalized B cells enables the detection of pre-existing CD4.sup.+ T cell reactivity with high sensitivity.

[0133] FIG. 2. Characteristics of the mutational landscape in the analyzed melanoma lesions.

[0134] FIG. 3. Detection of neo-epitope specific CD4.sup.+ T cells in human melanoma lesions. (a) Mean IFN- concentration in culture supernatant after 48 h co-culture of peptide loaded, autologous B-cells with in vitro expanded intratumoral CD4, T cells (n=2-3). Dotted line indicates mean IFN- production of CD4.sup.+ T cells after co-culture with unloaded B-cells. Error bars depict s.d. CIRH1.sub.P>L P=0.0026. GART.sub.V>A P=0.0645. ASAP1.sub.P>L P=0.0063, RND3.sub.P>S P=0.0061. (b,c) Detection of intracellular IFN- levels 24 h after co-culture of peptide loaded autologous B-cells with in vitro expanded, intratumoral CD4.sup.+ T cells for (b) NKIRTIL018 and (c) NKIRTIL034. Flow cytometry plots depicting single, live, CD4.sup.+, T cells from a representative experiment. Controls indicate frequency of single, live, CD4.sup.+, IFN-.sup.+ T cells after co-culture with unloaded B-cells. Bar graphs depict mean IFN- concentration over multiple experiments (n=3). Error bars depict s.d. CIRH1.sub.P>L P<0.0001, GART.sub.V>A P=0.0028. ASAP1.sub.P>L P=0.0012, RND3.sub.P>S P=0.037.

[0135] FIG. 4. Detection of neo-epitope specific CD4.sup.+ T cells in a melanoma lesion using BCL-6/BCL-XL or Epstein-Barr virus (EBV) immortalized, autologous B cells. IFN- concentration in culture supernatant after 48 h co-culture of in vitro expanded intratumoral CD4.sup.+ T cells with peptide loaded, autologous B-cells immortalized by stable transfection with BCL-6/BCL-XL (triangles) or EBV infection (circles) derived from NKIRTIL018.

[0136] FIG. 5. Detection of T.sub.H1, T.sub.H2 and T.sub.H17 cytokines by intratumoral CD4.sup.+ T cells in response to putative neo-epitopes. Cytokine concentration in culture supernatant after 48 h co-culture of peptide loaded, autologous B-cells with in vitro expanded intratumoral CD4.sup.+ T cells from (a) NKIRTIL018 and (b) NKIRTIL034. Dotted line indicates cytokine concentrations after co-culture of CD4.sup.+ T cells with unloaded B-cells.

[0137] FIG. 6. Analysis of cross-reactivity of intratumoral CD4.sup.+ T cells against a non-autologous, mutated peptide library. IFN- concentration in culture supernatant after 48 h co-culture of in vitro expanded intratumoral CD4.sup.+ T cells obtained from (a) NKIRTIL018 and (b) NKIRTIL034 with autologous B-cells loaded with peptide libraries of the respective other subject. Dotted line indicates IFN- production of CD4.sup.+ T cells after co-culture with unloaded B-cells. Identified neo-epitopes of NKIRTIL018 and NKIRTIL034 were used as positive control samples.

[0138] FIG. 7. Isolation and characterization of neo-epitope specific CD4.sup.+ T cells in human melanoma lesions. (a) IFN- concentration in culture supernatant after 48 h co-culture of CIRH1.sub.P>L reactive CD4.sup.+ T cell clones derived from NKIRTIL018 with unloaded (white bars) or peptide loaded (black bars) autologous B cells. (b) T cell receptor (TCR) repertoire diversity among neo-antigen reactive CD4.sup.+ T cell clones of NKIRTIL018 and NKIRTIL034. Numbers indicate different TCR clonotypes; grey segments indicate TCR clonotypes identified in at least two analyzed T cell clones. (c) IFN- concentration in culture supernatant after 48 h co-culture of neo-antigen reactive CD4.sup.+ T cell clones derived from NKIRTIL018 with autologous B-cells loaded with indicated concentrations of mutated peptide (open squares) or wildtype peptide (black circles). TCR clonotypes are indicated as determined in (b). (d) Mean IFN- concentration in culture supernatant after 48 h co-culture of neo-antigen reactive CD4-T cell clones derived from NKIRTIL018 with autologous B-cells loaded with truncated variants of the mutated peptide (n=2). Error bars depict s.d. CIRH1.sub.P>L P=0.1012 and P=0.1514, ASAP1.sub.P>L P=0.0005 and P=0.0474.

[0139] FIG. 8. Isolation of neo-epitope reactive CD4.sup.+ T cells from human melanoma lesions. IFN- concentration in culture supernatant after 48 h co-culture of (a) GART.sub.V>A and ASAP1.sub.P>L reactive CD4.sup.+ T cell clones derived from NKIRTIL018 and (b) RND3.sub.P>S reactive CD4.sup.+ T cell clones derived from NKIRTIL034 with unloaded (white bars) or peptide loaded (black bars) autologous B cells.

[0140] FIG. 9. Recognition of mutated and wildtype peptide by neo-antigen reactive T cell clones of NKIRTIL018. IFN- concentration in culture supernatant after 48 h co-culture of GART1.sub.V>A reactive CD4.sup.+ T cell clones derived from NKIRTIL018 with autologous B-cells loaded with indicated concentrations of mutated peptide (open squares) or wildtype peptide (black circles).

[0141] FIG. 10. Identification and enumeration of neo-antigen reactive CD4+ T cell products used for adoptive T cell therapy. (a) Mean IFN concentration in culture supernatant after a 48 h co-culture of peptide loaded, autologous B cells with in vitro expanded intratumoral CD4+ T cells. Dotted line indicates mean IFN production of CD4+ T cells after co-culture with unloaded B cells. (b) Detection of intracellular IFN levels after a 24 h co-culture of peptide loaded autologous B cells with either in vitro expanded intratumoral CD4+ T cells or with the TIL infusion product of NKIRTIL027. Flow cytometry plots depict single live CD4+ IFN+ T cells from a representative experiment. Bar graphs depict mean IFN concentrations over multiple experiments (n=3). P=0.0061 (in vitro expanded intratumoral CD4+ T cells) and 0.0011 (TIL infusion product). Error bars depict s.d. (c) Expression of CD 137 on CD4.sup.+ T cells within the T cell infusion product of subject BO after 16 h co-culture with autologous tumor cells or with RPS12.sub.V>I peptide loaded EBV immortalized B cells. Flow cytometry plots depict single, live. CD4.sup.+ T cells.

[0142] FIG. 11. Detection of neo-epitope specific CD8+ T cells in human melanoma lesion. (a) IFN-g concentration in culture supernatant after 48 h co-culture of 582 different 31-AA peptides loaded on autologous B-cells with in vitro expanded intratumoral CD8.sup.+ T cells (n=1). (b) IFN-g concentration in culture supernatant after 48 h co-culture of unloaded B cells, B cells loaded with an irrelevant epitope and B cells loaded with the TTC37.sub.A>V 31-AA peptide. (c) Percentage of CD8.sup.+ multimer.sup.+ T cells, after staining with HLA-A*01:01 multimers loaded with the TTC37 undecamer epitope.

EXAMPLES

Example 1

Intratumoral CD4.SUP.+ T Cell Reactivity Against Mutated Antigens is Commonly Observed in Human Melanoma

Methods

[0143] Generation of TIL material, tumor cell lines and Bcl-6/Bcl-xL transduced B cells. PBMC and TIL material was obtained from individuals with stage IV melanoma in accordance with Dutch guidelines, when applicable following signed informed consent and after approval of the medical ethical committees at the NKI-AVL (Medisch Ethische Toetsingscommissie).

[0144] PBMC material was prepared by Ficoll-Isopaque density centrifugation. TIL material and short-term tumor lines were obtained from resected melanoma lesions. Fresh tumor material was minced and digested overnight in RPMI 1640 (Life Technologies) supplemented with penicillin-streptomycin (Roche), 0.01 mg ml.sup.1 pulmozyme (Roche) and 1 mg ml.sup.1 collagenase type IV (BD Biosciences). A tumor line was obtained by culture of the resulting cell suspension in RPMI 1640 supplemented with penicillin-streptomycin (Roche) and 10% (v/v) heat-inactivated Fetal Bovine Serum (Sigma-Aldrich). TIL were obtained by culturing the suspension cells in RPMI 1640 supplemented with penicillin-streptomycin, 10% (v/v) AB serum (Sanquin Blood Supply and Life Technologies), L-glutamine (Life Technologies) and 6000 IU ml.sup.1 rhIL-2 (Novartis).

[0145] Autologous B cells from PBMC material were immortalized as part of collaboration agreement by AIMM Therapeutics by Bcl-6/Bcl-xL gene transfer as previously described.sup.15,16. For a detailed description, see also WO 2007/067046. Bcl-6/Bcl-xL transduced B cells were cultured in IMDM (Life Technologies) supplemented by 10% (v/v) Fetal bovine serum (Hyclone), penicillin-streptomycin (Roche) and 50 ng ml.sup.1 rm-IL21 (AIMM Therapeutics) and stimulated every 3-5 days by irradiated (50 Gy) mouse L cell fibroblasts expressing CD40L (2:1 B cell-to/L cell ratio).

[0146] Exome sequencing. Genomic DNA was extracted from cell pellets using a DNeasy purification kit (Qiagen), fragmented using the (ovaris S220 Focused-ultrasonicator (Woburn). DNA libraries were created using the Illumina TruSeq DNA library preparation kit. Exonic sequences were enriched capturing DNA fragments with the Sure Select Human All Exon 50 Mb Target Enrichment system (Agilent).sup.28, according to Agilent protocols with modifications. 1:2 of standard bait reaction was used and Block #3 in the hybridization mixture was replaced with a custom NKI-Block #3 to support the TruSeq DNA libraries in which the indexes require additional blocking. NKI-Block #3 consists of equal amounts of two DNA oligos (IDT-DNA) at 16.6 ug/ul:

TABLE-US-00001 NKI3.1 5AGATCGGAAGAGCACACGTCTGAACTCCAGTCACNNNNNNATCTCGTA TGCCGTCTTCTGCTTG/3ddC/3 NKI3.2 5CAAGCAGAAGACGGCATACGAGATNNNNNNGTGACTGGAGTTCAGACG TGTGCTCTTCCGATCT/3ddC/3

[0147] Captured library fragments were split into two fractions and both were PCR enriched (13 cycles) using the Illumina P5 and P7 oligonucleotides (IDT-DNA).

TABLE-US-00002 P5primer:5AATGATAGGGCGACCACCGAGATCT3, P7primer:5CAAGCAGAAGACGGCATACGAG3.

[0148] Both PCR reactions quantified on a BioAnalyzer DNA7500 Chip (Agilent), equally combined and diluted to 10 nM concentrations for paired-end 75 bp sequencing on a Illumina HiSeq2000 sequencer. Reads were aligned to the human reference genome GRCh37 using BWA version 0.5.10.sup.20. PCR duplicates were filtered using Picard (http://picard.sourceforge.net) and realignment around insertions and deletions (indels) was performed using GATK toolkit.sup.21.

[0149] Somatic single nucleotide variants (SNV) were called using Somatic-sniper.sup.22 and filtered using a somatic score cut-off>34 and a minimum of 4 reads in both tumor and control. Somatic indels were called using the GATK somatic indel detector and filtered using a minimum variation frequency of 25% and having at least 5 reads showing the indel. Germ-line variants in the vicinity of detected somatic variants were identified using Samtools.sup.23 and filtered using minimum coverage and minimum number of alternate reads of 10 and 6 reads, respectively. SNPeff.sup.24 was then used to predict the effect of all variants on the Ensembl gene build version 65. Using a custom Perl script and the Ensembl API, coxling variants were edited into the cDNA sequence and subsequently translated into protein sequence. These protein sequences were separately generated for normal (only germline variants) and tumor samples (germline variants and tumor specific mutations).

[0150] RNA-sequencing. RNA was isolated using TriZol reagent (Life Technologies). Poly-A selected RNA libraries were prepared using the TruSeq RNA library protocol (Illumina) and Paired-end 50 bp sequencing was performed on an Illumina HiSeq2000. Reads were aligned to human reference genome GRCh37 using Tophat 1.4.sup.25. Expression values were calculated as FPKM using Cufflinks.sup.26.

[0151] Peptide synthesis. Peptides were synthesized at the NKI Peptide synthesis facility (Amsterdam) using preloaded Wang resin with a SYRO II robot using standard Fmoc Solid Phase Peptide Chemistry, with PyBop and Dipea as activator and base.

[0152] Generation of CD4.sup.+ T cell material from TIL and detection of neo-antigen reactivity. Cell-sorting was performed on a FACSAria I (BD Biosciences) or MoFlo Astrios (Beckman Coulter). Bulk CD4.sup.+ T cell populations were generated from cryo-preserved TIL material by sorting of live single CD4.sup.+ T cells stained antibody against CD8 (BD Biosciences; SK1; 1:50) and CD4 (BD Biosciences; SK3; 1:50). Isolated live single CD4.sup.+ CD8.sup. T cells were expanded using 30 ng ml.sup.1 CD3-specific antibody (OKT-3; Janssen-Cilag) and 3,000 IU ml.sup.1 rh-IL-2 (Novartis) in 1:1 (v/v) medium mixture of RPMI 1640 and AIM-V (Life Technologies) supplemented with 10% AB serum (Life Technologies). Glutamax (Life Technologies) at a 1:200 T cell/feeder cell ratio to obtain pure CD4.sup.+ T cell populations (routinely >97% CD4.sup.+) which were used for determination of T cell reactivity against neo-epitopes.

[0153] Detection of neo-epitope reactive CD4+ T cells. 110.sup.5 Bcl-6/Bcl-xL transduced B cells per well were loaded with peptide (20 g ml.sup.1 unless otherwise indicated) in 96-round-bottom plates for 18-24 hours in 200 l IMDM medium (Life Technologies) supplemented by 10% (v/v) Fetal bovine serum (Hyclone), penicillin-streptomycin (Roche) and 50 ng ml.sup.1 rm-IL21 (AIMM Therapeutics). Afterwards, medium was removed and 110.sup.5 CD4.sup.+ T cells were added per well in 200 l RPMI 1640 (Life Technologies) supplemented by 10% (v/v) AB serum (Life Technologies), penicillin-streptomycin (Roche) and 50 ng ml.sup.1 rm-IL21 (AIMM Therapeutics). 48 hours later, culture supernatant was harvested and analyzed using Human T.sub.H1/T.sub.H2/T.sub.H17 cytometric bead array (BD Biosciences) or IFN- Flex bead E7 cytometric bead array (BD Biosciences) according to manufacturer's guidelines. For detection of intracellular levels of IFN- by flow cytometry. CD4.sup.+ T cells were stimulated with peptide-loaded B cells for 24 hours. Subsequently, cells were stained using IR-Dye (Life Technologies) for exclusion of dead cells and the Cytofix/Cytoperm kit (BD Biosciences) and an antibody against IFN- (BD Biosciences; 25723; 1:50) according to the manufacturer's guidelines. For isolation of live, IFN- producing CD4.sup.+ T cells. T cells were stimulated with peptide-loaded B cells for 6 hours. Subsequently, cells were stained using the IFN- secretion capture kit (Miltenyi Biotec) and an antibody for CD4 (BD Biosciences; SK3; 1:50). Single, live IFN- producing CD4.sup.+ T cells were sorted by flow cytometry and collected in 96-well round-bottom culture plates containing 210.sup.5 irradiated PBMCs, 30 ng ml.sup.1 CD3-specific antibody (OKT-3; Janssen-Cilag) and 3.000 IU ml.sup.1 recombinant human IL-2 (Novartis) in 200 l 1:1 (v/v) medium mixture of RPMI 1640 and AIM-V (Life Technologies) supplemented with 10% AB serum (Life Technologies), penicillin-streptomycin (Roche) and Glutamax (Life Technologies). After 7 days, 100 l was replaced with fresh medium supplemented with rh-IL-2 (3,000 IU ml.sup.1 final) and T cell specificity was confirmed by assessing IFN- in response to neo-epitope after 14 days.

[0154] Identification of TCR sequences. cDNA of T cell clones was generated and used to prepare DNA libraries with the Illumina TruSeq DNA library preparation kit. The resulting DNA libraries were sequenced on a Illumina MiSeq sequenzer using Paired-end 150 bp chemistry.

[0155] Sequencing reads in FASTQ files were mapped to the human genome, build NCBI36/hg18, using BWA.sup.20 and SAMtools.sup.23. PCR duplicates in resulting BAM files were filtered using Picard (http://picard.sourceforge.net). CDR3 TCR sequences were identified as previously reported.sup.27. TCR and - sequences were conferred with an in-house developed python script.

[0156] Statistical analysis. Differences in cytokine concentrations and frequencies of cytokine-producing T were compared using a two-tailed Student's t test. P values <0.05 were considered significant; Significance was indicated as P<0.05 (*), P<0.01 (**) and P<0.001 (***).

Results

[0157] Tumor-specific neo-antigens arising as a consequence of mutations in human cancers.sup.1,2 are thought to be important for the efficacy of clinically used cancer immunotherapies.sup.3-5. While tumor-specific CD4.sup.+ T cell responses are known and growing evidence suggests that neo-antigens may be commonly recognized by intratumoral CD8.sup.+ T cells.sup.3,4,6, it is unknown whether neo-antigen specific CD4.sup.+ T cells commonly reside within human tumors. Here, we use immortalized Bcl-6/Bcl-xL transduced B cells to measure the occurrence of CD4.sup.+ T cell responses against putative neo-epitopes that are identified by tumor exome sequencing. Using this approach, we show the presence of neo-antigen reactive CD4.sup.+ T cells in 4 out of 5 melanoma patients analyzed, including melanoma patients who demonstrate a clinical response after adoptive T cell therapy.

[0158] Based on I) the evidence that supports a role for CD4.sup.+ T cells in the efficacy of cancer immunotherapies.sup.7-12, II) the proposed correlation between mutational load and clinical response to immunotherapy.sup.13, and III) the recent observation that neo-antigen specific CD4.sup.+ T cells can mediate tumor-regression in a metastatic cholangiocarcinoma.sup.10, we wish to understand whether neo-antigen specific CD4.sup.+ T cell reactivity is a rare or common phenomenon in human cancers.

[0159] The average mutational load of melanoma is high.sup.1. Furthermore, the tumor-infiltrating lymphocyte (TIL) products that are generated for cellular therapy of melanoma.sup.14 regularly contain substantial fractions of CD4+ T cells, potentially mediating clinical effects.sup.7,10. Because of these data, we chose to examine the occurrence of neo-antigen specific CD4.sup.+ T cell reactivity in a set of melanoma specimens with varying mutational loads.

[0160] To assess the occurrence of intratumoral CD4.sup.+ T cell responses against non-synonymous somatic mutations within these tumors, we used whole exome-sequencing and RNA-sequencing data to first identify the entire set of tumor-specific, non-synonymous mutations within expressed genes. Subsequently CD4.sup.+ T cell reactivity against any of these mutated peptides was analyzed by the use of retrovirally Bcl-6 and Bcl-xL immortalized autologous B cells.sup.15,16 (see also WO 2007/067046) which were loaded with mutated peptides (FIG. 1).

[0161] This screening platform was validated by the analysis of three melanoma lesionsNKIRTIL018. NKIRTIL034 and NKIRTIL045from patients who underwent palliative metastasectomy. While all three tumors showed the expected UV induced mutational signature, total mutational load varied considerably (range: 180-464 somatic mutations. FIG. 2).sup.1,2. On average 153 mutations (Range: 99-187) were identified as candidate neo-epitopes (defined as a tumor-specific, non-synonymous mutation in a gene with confirmed RNA-expression). Peptides (31 amino-acids) that covered the individual mutations were then loaded onto the Bcl-6/Bcl-xL immortalized, autologous B cells, and the resulting targets were incubated with in vitro expanded, intratumoral CD4.sup.+ T cells (routinely 97% CD4+). Subsequently, we assessed culture supernatants for the presence of the T.sub.H1. T.sub.H2 and T.sub.H17 cytokines IFN-, TNF-, IL-10, IL-2, IL-4, IL-6 and IL-17a. For subject NKIRTIL018, IFN- production of tumor-derived CD4.sup.+ T cells was observed in response to three mutated gene products. CIRH1A P333L (CIRH1A.sub.P>L). GART V551A (GART.sub.V>A), and ASAP1 P941L (ASAP1.sub.P>L) (FIG. 3a).

[0162] Importantly, detection of these neo-antigen specific CD4.sup.+ T cells was readily feasible as a result of the constant low background observed when using autologous Bcl-6/Bcl-xL immortalized B cells. Contrary, detection of these neo-antigen specific CD4.sup.+ T cells was impossible with autologous Epstein-Barr virus (EBV) immortalized B cells (FIG. 4).

[0163] This demonstrates that the use of Bcl-6/Bcl-xL immortalized B cells is superior over the use of EBV immortalized B cells.

[0164] Melanoma-derived CD4.sup.+ T cells of subject NKIRTIL034 also showed neo-antigen reactivity, in this case to a mutation within the Rho family GTPase 3 RND P49S (RND3.sub.P>S). Only in the subject with the lowest mutational load, NKIRTIL027, reactivity against neo-antigens as measured by production of TH-cytokines was not observed within the intratumoral CD4.sup.+ T cell compartment (FIG. 3a and data not shown). The presence of neo-antigen reactive CD4.sup.+ T cells within the melanoma lesions of NKIRTIL018 and NKIRTIL034 was confirmed by analysis of intracellular IFN- levels upon antigen stimulation (FIG. 3b,c). Of note, in this experiment wherein Bcl-6/Bcl-xL immortalized B cells were used, the control frequency of single, live, CD4+, IFN-.sup.+ T cells after co-culture with unloaded B cells was 0.078% (FIG. 3b) or 0.277% (FIG. 3c). This means that only 0.078% or 0.277% of the CD4+ T cells that were co-cultured with unloaded Bcl-6/Bcl-xL immortalized B cells displayed IFN- production (indicative for T cell activation). Importantly, these control frequencies are much lower than the control frequency of 1.98% that was obtained when EBV-immortalized B cells were used (see Example 3 and FIG. 10c).

[0165] As shown in FIGS. 3b and 3c, the frequency of the CIRH1a.sub.P>L-recognizing T cells was 0.096% (i.e. 0.174% minus the control frequency of 0.078%). The frequency of the GART.sub.V>A-recognizing T cells was 0.053% (i.e. 0.131% minus the control frequency of 0.078%). The frequency of the ASAP1.sub.P>L-recognizing T cells was 0.264% (i.e. 0.342% minus the control frequency of 0.078%) and the frequency of the RND3.sub.P>S-recognizing T cells was 0.246% (i.e. 0.523% minus the control frequency of 0.277%). Hence, it is clear that the neo-antigen recognizing T cells were present in very low frequencies. Nevertheless, they could still be detected, thanks to the very sensitive screening methods according to the present invention wherein Bcl-6/Bcl-xL immortalized B cells are used for T cell neo-antigen presentation. These T cells could not have been detected using EBV-immortalized B cells as antigen presenting cells.

[0166] Neo-antigen reactive CD4.sup.+ T cells did not show production of any of the other cytokines tested (FIG. 5).

[0167] T cell receptors can trigger T cell function upon interaction with a large number of different epitopes, and in the above screens, a diverse T cell pool is tested for recognition of a sizable set of peptides. Thus, we assessed in a set of validation screens whether the observed T cell reactivity represented a genuine neo-antigen driven T cell response, rather than T cell cross reactivity. In these screens, the ability of CD4.sup.+ T cell pools from NKIRTIL018 and NKIRTIL034 to react against the set of potential neo-epitopes of the other subject was evaluated. This analysis showed that IFN- production by tumor-derived CD4.sup.+ T cells was patient mutanome specific as none of the other subject's neo-epitopes were recognized (FIG. 6), indicating that the presence of CD4.sup.+ T cells with neo-antigen reactivity identified through the use of Bcl-6/Bcl-xL immortalized B cells reflects a true neo-antigen specific CD4+ T cell response, driven by expression of the antigen in the autologous melanoma.

[0168] In order to further characterize the neo-antigen reactive CD4, T cell compartment in these melanoma patients, we generated a panel of neo-epitope reactive CD4.sup.+ T cell clones from TIL by isolation of antigen-specific IFN- producing CD4.sup.+ T cells (FIG. 7+8). To assess the capacity of neo-antigen reactive CD4.sup.+ T cells to discriminate between the mutated cognate peptide and its parental sequence, neo-antigen specific CD4+ T cells were stimulated with different concentrations of either peptide (FIG. 7c+FIG. 9). For all T cell clones, specific for either ASAP1.sub.P>L, CIRH1A.sub.P>L, or GART.sub.V>A, recognition of the mutant peptide was detectable at concentrations that were 100 to >1,000 lower than that required for recognition of the parental peptide (FIG. 7c+FIG. 9). Thus, both by their restriction towards the autologous mutanome set and by their preferential recognition of the mutant peptide over its wild-type counterpart, these tumor-resident CD4+ T cell responses are defined as true neo-antigen driven T cell reactivities.

[0169] Truncation of two of the identified neo-epitopes revealed that peptides of 13 amino acids (RKITFLIIRCLISC; CIRH1.sub.P>L) and 19 amino acids (KPPPGDLPLKPTELAPKPQ; ASAP1.sub.P>L) were still recognized with high efficiency (FIG. 7c). For both epitopes, the mutated residue was located at a central position within truncated epitope, consistent with an essential role of this amino acid in T cell activation.sup.17,18. For each of the four identified neo-epitopes, TCR alpha-beta sequences were obtained for a small set (8-11) of T cell clones. This analysis demonstrated that the TCR repertoire of neo-epitope specific T cell responses is generally oligoclonal to polyclonal, with 2-7 identified TCR clonotypes for these 4 epitopes (FIG. 7b). Hence, the neo-antigen specific T cell responses towards these 4 epitopes are not due to the outgrowth of rare neo-antigen reactive CD4+ T cells but are oligo- to polyclonal.

Example 2

[0170] Based on the observation that neo-antigen specific CD4.sup.+ T cell reactivity can readily be detected on the basis of cancer exome data (FIG. 3), and that a T cell product containing a high frequency of neo-antigen specific CD4.sup.+ T cells was recently shown to mediate partial regression of a cholangiocarcinoma.sup.10, we analyzed whether a neo-antigen reactive CD4.sup.+ T cell compartment is also prevalent in melanoma patients who experience a clinical response upon adoptive T cell therapy.

[0171] The first patient, NKIRTIL027, was a stage IV melanoma patient who exhibited a partial clinical response upon TIL therapy. Exome sequencing revealed a very high mutational burden in the tumor of this patient (total of 1393 somatic mutations) (FIG. 2). Bcl-6/Bcl-xL transduced B cells were loaded with the collection of 582 candidate neo-epitopes, identified after filtering for non-synonymous and RNA expressed mutations, and used as targets for CD4.sup.+ T cells expanded from an autologous melanoma lesion. Strikingly, this analysis identified CD4.sup.+ T cell reactivity against 7 different mutated gene products (CPT1A G212S (CPT1A.sub.G>S), HERC4 P768S (HERC4.sub.P>S). GYLTL1B D597E (GYLTL1B.sub.D>E), KRTAP4-11 P187S (KRTAP4-11.sub.P>S), LEMD2 P495L (LEMD2.sub.P>L), DTNBP1 D334H (DTNBP1.sub.D>H), MFSD9 P219L (MFSD9.sub.P>L) as detected by IFN- production in culture supernatants and independent confirmation by intracellular detection of IFN- production. In particular, this analysis identified strong CD4+ T cell reactivity against the mutated gene product LEM Domain Containing 2 P495, (LEMD2.sub.P>L), as detected by IFN secretion (FIG. 10a).

[0172] Of note, the T cell product infused into subject NKIRTIL027 contained a substantial fraction of CD4.sup.+ T cells (70% of all CD3.sup.+ T cells; data not shown). Thus, we assessed the frequency of neo-antigen reactive CD4+ T cells within the T cell product used for adoptive T cell therapy. This analysis revealed the presence of 7 different neo-epitope reactive CD4.sup.+ T cell populations within the infusion cell product with an average frequency of 2.9% (Range: 1.7-4.5%) of total CD4.sup.+ T cells. FIG. 10b depicts LEMD2.sub.P>L, reactive CD4+ T cells (3.8% of total CD4+ T cells) that were observed within the T cell product used for adoptive T cell therapy by intracellular cytokine staining.

[0173] Of note, also in this experiment wherein Bcl-6/Bcl-xL immortalized B cells were used, the control frequencies of single, live, CD4+, IFN-.sup.+ T cells after co-culture with unloaded B cells were 0.52% (In vitro expanded intratumoral CD4+ T cells from an autologous melanoma lesion) and 0.20% (TIL infusion product used for adoptive T cell therapy); see FIG. 10b. Importantly, these control frequencies are much lower than the control frequency of 1.98% that was obtained when EBV-immortalized B cells were used (see Example 3 and FIG. 10c).

[0174] The frequency of the LEMD2.sub.P>L-recognizing T cells in the in vitro expanded intratumoral CD4+ T cells was 4.74% (i.e. 5.26% minus the control frequency of 0.52%). The frequency of the LEMD2.sub.P>L-recognizing T cells in the TIL infusion product used for adoptive T cell therapy was 3.33% (i.e. 3.53% minus the control frequency of 0.20%).

Example 3

[0175] Next, we analyzed neo-antigen specific CD4.sup.+ T cell reactivity in a stage IV melanoma patient (BO) who received multiple infusions of in vitro expanded, autologous tumor-specific T cells obtained by stimulation of peripheral blood mononuclear cells (PBMCs) with autologous tumor cells (ref 19). Following treatment, this patient experienced a complete tumor remission, now ongoing for 7 years. Furthermore. CD4.sup.+ T cells present within the T cell product showed strong recognition of the autologous melanoma line (FIG. 10c). As Bcl-6/Bcl-xL immortalized B cells were not yet available for this subject, autologous EBV-immortalized B cells had to be used in spite of the larger background noise. The EBV-immortalized B cells were loaded with 31-mer peptides covering the 501 non-synonymous mutations within expressed genes that were detected in this patient. In spite of the higher background noise due to the use of EBV-transformed APCs, this screen identified one prominent CD4.sup.+ T cell response (24% of total CD4.sup.+ T cells within the infusion product; i.e. 26.0% minus the control frequency of 1.98%) that was directed against a mutant version of ribosomal protein S12 (RPS12 V104I) (FIG. 10c).

[0176] Of note, in this experiment wherein EBV-immortalized B cells were used, the control frequency of single, live, CD4+, IFN-.sup.+ T cells after co-culture with unloaded B cells, is 1.98%. This means that 1.98% of the CD4+ T cells that were co-cultured with unloaded EBV-immortalized B cells displayed IFN- production (indicative for T cell activation). Importantly, this control frequency is much higher than the frequencies of each of the four neo-epitope-specific T cells that were found in Example 1. As shown in FIGS. 3b and 3c, the frequency of the CIRH1a.sub.P>L-recognizing T cells was 0.096% (i.e. 0.174% minus the control frequency of 0.078%). The frequency of the GART.sub.V>A-recognizing T cells was 0.053% (i.e. 0.131% minus the control frequency of 0.078%). The frequency of the ASAP1.sub.P>L-recognizing T cells was 0.264% (i.e. 0.342% minus the control frequency of 0.078%) and the frequency of the RND3.sub.P>S-recognizing T cells was 0.246% (i.e. 0.523% minus the control frequency of 0.277%). It is clear that these low concentrations of neo-epitope-specific T cells would not have been detected using EBV-immortalized B cells in view of the much higher background noise that is present when EBV-immortalized B cells are used. Hence, if EBV-immortalized B cells were used as APCs, all four neo-epitope specific T cells identified in Example 1 would have been missed. This emphasizes the superiority of the use of Bcl-6/Bcl-xL immortalized B cells as T cell (neo)epitope presenting APCs.

Conclusion

[0177] The above data show that B cells that are immortalized with a method according to WO 2007/067046 are preferred APCs for testing T cell recognition of T cell epitopes.

[0178] As shown in Examples 1 and 2, the use of Bcl-6/Bcl-xL immortalized B cells is superior over the use of EBV immortalized B cells, because EBV immortalized B cells provide high levels of background noise. Contrary, when Bcl-6/Bcl-xL immortalized B cells are used as APCs, background noise is much lower, if present at all. This is for instance shown in FIG. 4.

[0179] This is also apparent from a comparison between the low control frequencies of FIGS. 3b and 3c, obtained with the use of Bcl-6/Bcl-xL immortalized B cells as APCs (0.078% and 0.277%, respectively) and the high control frequency of FIG. 10, obtained with the use of EBV-immortalized B cells as APCs (1.98%). In fact, the control frequency of FIG. 10 is much higher that the frequencies of the neo-epitope-specific T cells depicted in FIG. 3.

[0180] As a consequence, a more sensitive detection method is provided by the present invention, which enables detection of T cells which are present in low frequencies. For instance, 7 different neo-epitope reactive CD4+ T cell populations with an average frequency of 2.9% of total CD4+ T cells were identified in Example 1, using Bcl-6/Bcl-xL immortalized B cells as APCs. With EBV immortalized B cells, only one prominent CD4+ T cell response with a much higher frequency of 24% of total CD4+ T cells was detected (i.e. 26.0% minus the control frequency of 1.98%).

Example 4

Detection of Neo-Epitope Specific CD8.SUP.+ T Cells in a Human Melanoma Lesion.

[0181] This Example shows that the use of Bcl-6/Bcl-xL immortalized B cells as APCs, as described in the previous Examples for the identification of neo-epitope specific CD4.sup.+ T cell responses, is also suitable to identify neo-epitope specific CD8.sup.+ T cell responses.

[0182] Based on exome- and RNA sequencing data of patient NKIRTIL027 (see Example 2), 582 tumor-specific single nucleotide variants were identified with RNA expression >0 FPKM. 31 amino acid peptides covering these mutations were synthesized and loaded on Bcl-6/Bcl-xL immortalized autologous B cells. A 48 co-culture with CD8 enriched tumor infiltrating T cells (TIL) was performed and IFNg concentration was measured in the culture supernatant. Incubation of CD8 enriched TIl with one peptide (TTC37.sub.A>V) resulted in a signal above background (i.e. above the control signals of unloaded B cells). This is shown in FIG. 11a. As a validation of this result, an independent co-culture in which IFNg concentration was measured in the culture supernatant was performed (FIG. 11b). In parallel, epitope predictions were performed in which the 31 AA epitope was fed into netMHCpan to test for affinity to the patient's HLA-A and HLA-B alleles. A putative candidate undecamer epitope was predicted to bind to HLA-A*01:01. HLA-A*01:01 TTC37.sub.A>V multimers were generated and conjugated to two different fluorochromes. Staining of the CD8.sup.+ enriched T cell product resulted in a double positive multimer staining of 0.737% of all CD8.sup.+ T cells (FIG. 11c).

[0183] The applicability of the use of Bcl-6/Bcl-xL immortalized B cells as APCs for both screening neo-epitope specific CD8 and CD4 restricted T cell responses greatly increases the value of this analytic platform in the field of cancer immunotherapy.

REFERENCES OF EXAMPLES 1-4

[0184] 1. Alexandrov, L. B., et al. Signatures of mutational processes in human cancer. Nature 500, 415-421 (2013). [0185] 2. Vogelstein. B., et al. Cancer genome landscapes. Science 339, 1546-1558 (2013). [0186] 3. Robbins, P. F., et al. Mining exomic sequencing data to identify mutated antigens recognized by adoptively transferred tumor-reactive T cells. Nature medicine 19, 747-752 (2013). [0187] 4. van Rooij, N., et al. Tumor exome analysis reveals neoantigen-specific T-cell reactivity in an ipilimumab-responsive melanoma. Journal of clinical oncology: official journal of the American. Society of Clinical Oncology 31, e439-442 (2013). [0188] 5. Lu, Y. C., et al. Mutated PPP1R3B is recognized by T cells used to treat a melanoma patient who experienced a durable complete tumor regression. J Immunol 190, 6034-6042 (2013). [0189] 6. Wick. D. A., et al. Surveillance of the tumor mutanome by T cells during progression from primary to recurrent ovarian cancer. Clinical cancer research: an official journal of the American Association for Cancer Research 20, 1125-1134 (2014). [0190] 7. Friedman. K. M., et al. Tumor-specific CD4+ melanoma tumor-infiltrating lymphocytes. J Immunother 35, 400-408 (2012). [0191] 8. Kitano. S., et al. Enhancement of tumor-reactive cytotoxic CD4(+) T cell responses after ipilimumab treatment in four advanced melanoma patients. Cancer immunology research 1(2013). [0192] 9. Quezada, S. A., et al. Tumor-reactive CD4(+) T cells develop cytotoxic activity and eradicate large established melanoma after transfer into lymphopenic hosts. The Journal of experimental medicine 207, 637-650 (2010). [0193] 10. Tran, E., et al. Cancer immunotherapy based on mutation-specific CD4+ T cells in a patient with epithelial cancer. Science 344, 641-645 (2014). [0194] 11. Kenter. G. G., et al. Vaccination against HPV-16 oncoproteins for vulvar intraepithelial neoplasia. The New England journal of medicine 361, 1838-1847 (2009). [0195] 12. Ossendorp, F., Mengede. E., Camps, M., Filius, R. & Melief, C. J. Specific T helper cell requirement for optimal induction of cytotoxic T lymphocytes against major histocompatibility complex class II negative tumors. The Journal of experimental medicine 187, 693-702 (1998). [0196] 13. Champiat, S., Ferte, C., Lebel-Binay. S., Eggermont, A. & Soria. J. C. Exomics and immunogenics: Bridging mutational load and immune checkpoints efficacy. Oncoimmunology 3, e27817 (2014). [0197] 14. Dudley, M. E., et al. Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science 298, 850-854 (2002). [0198] 15. Kwakkenbos, M. J., et al. Genetic manipulation of B cells for the isolation of rare therapeutic antibodies from the human repertoire. Methods 65, 38-43 (2014). [0199] 16. Kwakkenbos, M. J., et al. Generation of stable monoclonal antibody-producing B cell receptor-positive human memory B cells by genetic programming. Nature medicine 16, 123-128 (2010). [0200] 17. Kessels, H. W., et al. The impact of self-tolerance on the polyclonal CD8+ T cell repertoire. J Immunol 172, 2324-2331 (2004). [0201] 18. Calis. J. J., et al. Properties of MHC class I presented peptides that enhance immunogenicity. PLoS computational biology 9, e1003266 (2013). [0202] 19. Verdegaal, E. M., et al. Successful treatment of metastatic melanoma by adoptive transfer of blood-derived polyclonal tumor-specific CD4+ and CD8+ T cells in combination with low-dose interferon-alpha. Cancer immunology, immunotherapy: CII 60, 953-963 (2011). [0203] 20. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754-1760 (2009). [0204] 21. DePristo, M. A., et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nature genetics 43, 491-498 (2011). [0205] 22. Larson. D. E., et al. SomaticSniper: identification of somatic point mutations in whole genome sequencing data. Bioinformatics 28, 311-317 (2012). [0206] 23. Li, H., et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078-2079 (2009). [0207] 24. Cingolani, P., et al. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly 6, 80-92 (2012). [0208] 25. Trapnell. C., Pachter. L. & Salzberg, S. L. TopHat: discovering splice junctions with RNA-Seq. Bioinformatics 25, 1105-1111 (2009). [0209] 26. Trapnell. C., et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nature biotechnology 28, 511-515 (2010). [0210] 27. Linnemann, C., et al. High-throughput identification of antigen-specific TCRs by TCR gene capture. Nature medicine 19, 1534-1541 (2013). [0211] 28. Varela, I., et al. Exome sequencing identifies frequent mutation of the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature 469, 539-542 (2011).