Novel tumor-specific antigens for acute myeloid leukemia (AML) and their uses

Abstract

Acute myeloid leukemia (AML) has not benefited from innovative immunotherapies, primarily due to the lack of actionable immune targets. Novel tumor-specific antigens (TSAs) shared by the majority of AML cells are described herein. Most of the TSAs described herein are derived from aberrantly expressed, non-mutated genomic sequences that are not expressed in normal tissues, such as intronic and intergenic sequences. Nucleic acids, compositions, cells, and vaccines derived from these TSAs are described. Use of the TSAs, nucleic acids, compositions, cells, and vaccines for the treatment of leukemias, such as AML, is also described.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 63 / 009,853, filed April 14, 2020, which is incorporated herein by reference in its entirety.

[0002] Sequence Listing Not applicable.

[0003] The present invention relates generally to cancer, and more specifically to acute myeloid leukemia-specific tumor antigens useful in T cell-based cancer immunotherapy. [Background technology]

[0004] Acute myeloid leukemia (AML), the most aggressive hematologic malignancy, is a heterogeneous disease characterized by aberrant epigenetic patterning, impaired mitochondrial proteostasis, and a relatively small number of mutations (Li et al., 2016, Ntziachristos et al., 2016, Ishizawa et al., 2019, Fennell et al., 2019). Notably, genetic and epigenetic changes in AML can occur years before diagnosis (Abelson et al., 2018, Desai et al., 2018). Furthermore, cure requires not only the elimination of bulk tumor cells but also the elimination of leukemic stem cells (Shlush et al., 2017, Boyd et al., 2018). Currently, most patients relapse after chemotherapy, and the 5-year overall survival rate is 40% for patients under 60 years of age and only 10–20% for patients over 60 years of age (who represent the majority of AML cases) ( Vasu et al., 2018 ).

[0005] In recent years, enthusiasm for cancer immunotherapy has been fueled primarily by two major breakthroughs: i) immune checkpoint therapy for the treatment of melanoma and selected types of solid tumors, and ii) chimeric antigen receptors for the treatment of lymphoid malignancies. However, AML has not benefited from such innovations, primarily due to the lack of actionable immune targets. In line with the notion that major histocompatibility complex (MHC)-associated peptides (MAPs) recognized by T cells are at the core of anticancer responses (Coulie et al., 2014), evidence suggests that AML cells must present immunogenic MAPs to CD8 T cells: i) AML cells express high densities of MHC class I molecules (Berlin et al., 2015), and ii) the bone marrow of AML patients contains CD8 T cells and possesses phenotypic and transcriptional characteristics of exhaustion (and therefore antigen recognition) (Knaus et al., 2018). However, the nature of AML antigens capable of eliciting a protective immune response remains unclear.

[0006] The first class of MAPs to attract the attention of cancer immunologists are tumor-associated antigens (TAAs), which are overexpressed on tumor cells compared with normal cells. Because high-affinity T cells that recognize self-antigens are eliminated by the central immune tolerance process of thymic selection, TAAs are inherently recognized by low-affinity T cells. Therefore, TAA-based vaccines have not convincingly impacted AML progression. The most studied AML TAA, Wilms tumor 1 (WT1), has yielded particularly disappointing results (Di Stasi et al., 2015; Maslak et al., 2018; Rashidi and Walter, 2016). Importantly, a recent report demonstrated that TCR gene therapy targeting a WT1-derived peptide, in which T cells are engineered to express a high-affinity TCR against a selected antigen, can durably prevent relapse in recipients of allogeneic hematopoietic stem cell transplants (Chapuis et al., 2019). Overall, these studies suggest that WT1-derived peptides are poorly immunogenic and will need to be targeted with engineered T cells to achieve their full therapeutic potential.

[0007] In contrast to TAAs, tumor-specific antigens (TSAs) are MAPs presented exclusively by tumor cells. Mutant TSAs (mTSAs), also known as neoantigens, have recently attracted considerable attention in the search for vaccines against solid tumors. Indeed, mTSAs can be highly immunogenic because they are not found in thymic medullary cells (mTECs), which induce central immune tolerance. However, mTSAs present two caveats. First, they are generally unique to each patient's tumor (individual neoantigens). Second, they are less common than initially predicted (Knaus et al., 2018). Consistent with the low mutational burden of AML cells, only one mTSA has been validated by mass spectrometry (MS) analysis of primary AML cells (van der Lee et al., 2019). The therapeutic potential of this mTSA, derived from a frameshift in the NPM1 gene, has yet to be evaluated, but available evidence suggests that it does not induce spontaneous immune responses in AML patients (van der Lee et al., 2019).

[0008] Given this, there is an urgent need to identify antigens that can elicit therapeutic immune responses against AML, which could be used as vaccines (± immune checkpoint inhibitors) or as targets for T cell receptor-based approaches (cell therapy, bispecific biologics).

[0009] This description makes reference to several documents, the contents of which are incorporated herein by reference in their entirety. Summary of the Invention

[0010] The present disclosure provides the following items 1 to 67. 1. A leukemia tumor antigen peptide (TAP) comprising one of the following amino acid sequences: [Table 1-1] [Table 1-2]

[0011] 2. The leukemia TAP according to item 1, comprising one of the amino acid sequences shown in SEQ ID NOs: 97 to 154.

[0012] 3. The leukemia TAP of item 1 or 2, wherein the leukemia TAP binds to an HLA-A*01:01 molecule and comprises the amino acid sequence NTSHLPLIY (SEQ ID NO: 48), HTDDIENAKY (SEQ ID NO: 67), YSHHSGLEY (SEQ ID NO: 89), ILDLESRY (SEQ ID NO: 134), VTDLLALTV (SEQ ID NO: 151), or LSDRQLSL (SEQ ID NO: 164), preferably ILDLESRY (SEQ ID NO: 134) or VTDLLALTV (SEQ ID NO: 151).

[0013] 4. The leukemia TAP binds to an HLA-A*02:01 molecule and has the amino acid sequence FLLEFKPVS (SEQ ID NO: 7), LLSRGLLFRI (SEQ ID NO: 11), LLDNILQSI (SEQ ID NO: 27), FLASFVEKTVL (SEQ ID NO: 32), ILASHNLTV (SEQ ID NO: 33), IQLTSVHLL (SEQ ID NO: 34), LELISFLPVL (SEQ ID NO: 35), LLLPESPSI (SEQ ID NO: 43), ALASHLIEA (SEQ ID NO: 51), AL DDITIQL (SEQ ID NO: 52), ALGNTVPAV (SEQ ID NO: 53), ALLPAVPSL (SEQ ID NO: 54), GLYYKLHNV (SEQ ID NO: 61), HLLSETPQL (SEQ ID NO: 65), KLLEKAFSI (SEQ ID NO: 72), SLWGQPAEA (SEQ ID NO: 77), SVFAGVVGV (SEQ ID NO: 82), VLVPYEPPQV (SEQ ID NO: 86), VLFGGKVSGA (SEQ ID NO: 104), KLQDKEIGL (SEQ ID NO: 108), TLNQ GINVYI (SEQ ID NO: 119), ALPVALPSL (SEQ ID NO: 123), ALDPLLLRI (SEQ ID NO: 130), KILDVNLRI (SEQ ID NO: 132), SLLSGLLRA (SEQ ID NO: 146), SLDLLPLSI (SEQ ID NO: 150), ILLEEQSLI (SEQ ID NO: 167), LTSISIRPV (SEQ ID NO: 168), TISECPLLI (SEQ ID NO: 169), ILLSNFSSL (SEQ ID NO: 171), RMVAYLQQL (SEQ ID NO: 183 ), or KLNQAFLVL (SEQ ID NO: 188), preferably VLFGGKVSGA (SEQ ID NO: 104), KLQDKEIGL (SEQ ID NO: 108), TLNQGINVYI (SEQ ID NO: 119), ALPVALPSL (SEQ ID NO: 123), ALDPLLLRI (SEQ ID NO: 130), KILDVNLRI (SEQ ID NO: 132), SLLSGLLRA (SEQ ID NO: 146), or SLDLLPLSI (SEQ ID NO: 150).

[0014] 5. The leukemia TAP binds to an HLA-A*03:01 molecule and has the amino acid sequence RSASSATQVHK (SEQ ID NO: 5), IVATGSLLK (SEQ ID NO: 18), KIKNKTKNK (SEQ ID NO: 19), KLLSLTIYK (SEQ ID NO: 20), ITSSAVTTALK (SEQ ID NO: 42), VILIPLPPK (SEQ ID NO: 44), NVNRPLTMK (SEQ ID NO: 74), SVYKYLKAK (SEQ ID NO: 91), VVFPFPVNK (SEQ ID NO: 105), ILFQNSALK (SEQ ID NO: 113), T 3. The leukemia TAP of item 1 or 2, comprising VIRIAIVNK (SEQ ID NO: 126), ISLIVTGLVK (SEQ ID NO: 131), HVSDGSTALK (SEQ ID NO: 159), IAYSVRALR (SEQ ID NO: 160), LSSRLPLGK (SEQ ID NO: 180), or RLVSSTLLQK (SEQ ID NO: 189), preferably VVFPFPVNK (SEQ ID NO: 105), ILFQNSALK (SEQ ID NO: 113), TVIRIAIVNK (SEQ ID NO: 126), or ISLIVTGLVK (SEQ ID NO: 131).

[0015] 6. The leukemia TAP of item 1 or 2, wherein the leukemia TAP binds to an HLA-A*11:01 molecule and comprises the amino acid sequence SASSATQVHK (SEQ ID NO: 6), AVLLPKPPK (SEQ ID NO: 45), ATQNTIIGK (SEQ ID NO: 96), SLLIIPKKK (SEQ ID NO: 106), SVQLLEQAIHK (SEQ ID NO: 121), STFSLYLKK (SEQ ID NO: 149), or RTQITKVSLKK (SEQ ID NO: 152), preferably SLLIIPKKK (SEQ ID NO: 106), SVQLLEQAIHK (SEQ ID NO: 121), STFSLYLKK (SEQ ID NO: 149), or RTQITKVSLKK (SEQ ID NO: 152).

[0016] 7. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-A*24:02 molecule and comprises the amino acid sequence LYFLGHGSI (SEQ ID NO: 13), NFCMLHQSI (SEQ ID NO: 36), KFSNVTMLF (SEQ ID NO: 71), IYQFIMDRF (SEQ ID NO: 92), LYPSKLTHF (SEQ ID NO: 95), or RYLANKIHI (SEQ ID NO: 145), preferably RYLANKIHI (SEQ ID NO: 145).

[0017] 8. The leukemia TAP of item 1 or 2, wherein the leukemia TAP binds to an HLA-A*26:01 molecule and comprises the amino acid sequence ETTSQVRKY (SEQ ID NO: 59) or TVPGIQRY (SEQ ID NO: 185).

[0018] 9. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-A*29:02 molecule and comprises one of the amino acid sequences VVFDKSDLAKY (SEQ ID NO: 88), FNVALNARY (SEQ ID NO: 99), or LGISLTLKY (SEQ ID NO: 138), preferably FNVALNARY (SEQ ID NO: 99) or LGISLTLKY (SEQ ID NO: 138).

[0019] 10. The leukemia TAP of item 1 or 2, wherein the leukemia TAP binds to an HLA-A*30:01 molecule and comprises the amino acid sequence TSRLPKIQK (SEQ ID NO: 26), LSWGYFLFK (SEQ ID NO: 29), or LSHPAPSSL (SEQ ID NO: 165).

[0020] 11. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-A*68:02 molecule and comprises the amino acid sequence NVSSHVHTV (SEQ ID NO: 50) or SSSPVRGPSV (SEQ ID NO: 148), preferably SSSPVRGPSV (SEQ ID NO: 148).

[0021] 12. The leukemia TAP binds to an HLA-B*07:02 molecule and has the amino acid sequence GPQVRGSI (SEQ ID NO: 8), SPQSGPAL (SEQ ID NO: 25), VPAPAQAI (SEQ ID NO: 40), APAPPPVAV (SEQ ID NO: 55), APDKKITL (SEQ ID NO: 56), KPMPTKVVF (SEQ ID NO: 73), SPADHRGYASL (SEQ ID NO: 78), SPQSAAAEL (SEQ ID NO: 79), SPVVHQSL (SEQ ID NO: 80), SPYRTPVL (SEQ ID NO: 81), PPRPLGAQV (SEQ ID NO: 98), GPGSRESTL (SEQ ID NO: 100), APGAAGQRL (SEQ ID NO: 107), TPGRSTQAI (SEQ ID NO: 110), APRGTAAL (SEQ ID NO: 111), SPVVRVGL (SEQ ID NO: 118), RPRGPRTAP (SEQ ID NO: 119), 120), TLRSPGSSL (SEQ ID NO: 128), TVRGDVSSL (SEQ ID NO: 129), LPSFSHFLLL (SEQ ID NO: 157), PRGFLSAL (SEQ ID NO: 161), IPLNPFSSL (SEQ ID NO: 163), LPSFSRPSGII (SEQ ID NO: 179), or SPARALPSL (SEQ ID NO: 184), preferably PPRPLGAQV (SEQ ID NO: 98), GPGSRESTL (SEQ ID NO: 100), APGAAGQRL (SEQ ID NO: 107), TPGRSTQAI (SEQ ID NO: 110), APRGTAAL (SEQ ID NO: 111), SPVVRVGL (SEQ ID NO: 118), RPRGPRTAP (SEQ ID NO: 120), TLRSPGSSL (SEQ ID NO: 128), or TVRGDVSSL (SEQ ID NO: 129).

[0022] 13. The leukemia TAP binds to an HLA-B*08:01 molecule and has the amino acid sequence SGKLRVAL (SEQ ID NO: 4), NPLQLSLSI (SEQ ID NO: 14), DLMLRESL (SEQ ID NO: 15), IALYKQVL (SEQ ID NO: 17), NILKKTVL (SEQ ID NO: 21), NPKLKDIL (SEQ ID NO: 22), NQKKVRIL (SEQ ID NO: 23), RLEVRKVIL (SEQ ID NO: 28), EGKIKRNI (SEQ ID NO: 31), LNHLRTSI (SEQ ID NO: 47), SIQRNLSL (SEQ ID NO: 49), IPHQRSSL (SEQ ID NO: 101), NLKEKKALF (SEQ ID NO: 103), ILKKNISI (SEQ ID NO: 104), ILKKNISI (SEQ ID NO: 105), ILKKNISI (SEQ ID NO: 106), ILKKNISI (SEQ ID NO: 107), ILKKNISI (SEQ ID NO: 109), ILKKNISI (SEQ ID NO: 110), ILKKNISI (SEQ ID NO: 111), ILKKNISI (SEQ ID NO: 112), ILKKNISI (SEQ ID NO: 113), ILKKNISI (SEQ ID NO: 114), ILKKNISI (SEQ ID NO: 115), ILKKNISI (SEQ ID NO: 116), ILKKNISI (SEQ ID NO: 117), ILKKNISI (SEQ ID NO: 118), ILKKNISI (SEQ ID NO: 119), ILKKNISI (SEQ ID NO: 120), ILKKNISI (SEQ ID NO: 121), ILKKNISI ( 114), VLKEKNASL (SEQ ID NO: 137), DLLPKKLL (SEQ ID NO: 139), SRIHLVVL (SEQ ID NO: 147), QIKTKLLGSL (SEQ ID NO: 156), TLKLKKIFF (SEQ ID NO: 170), MIGIKRLL (SEQ ID NO: 181), or NLKKREIL (SEQ ID NO: 182), preferably IPHQRSSL (SEQ ID NO: 101), NLKEKKALF (SEQ ID NO: 103), ILKKNISI (SEQ ID NO: 114), VLKEKNASL (SEQ ID NO: 137), DLLPKKLL (SEQ ID NO: 139), or SRIHLVVL (SEQ ID NO: 147).

[0023] 14. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-B*14:01 molecule and comprises the amino acid sequence DRELRNLEL (SEQ ID NO: 2), SNLIRTGSH (SEQ ID NO: 39), DQVIRLAGL (SEQ ID NO: 58), HQLYRASAL (SEQ ID NO: 66), SLQILVSSL (SEQ ID NO: 124), ERVYIRASL (SEQ ID NO: 133), LYIKSLPAL (SEQ ID NO: 136), IAGALRSVL (SEQ ID NO: 141), ISSWLISSL (SEQ ID NO: 162), DRGILRNLL (SEQ ID NO: 175), GLRLIHVSL (SEQ ID NO: 176), or GLRLLHVSL (SEQ ID NO: 177), preferably SLQILVSSL (SEQ ID NO: 124), ERVYIRASL (SEQ ID NO: 133), LYIKSLPAL (SEQ ID NO: 136), or IAGALRSVL (SEQ ID NO: 141).

[0024] 15. The leukemia TAP of item 1 or 2, wherein the leukemia TAP binds to an HLA-B*15:01 molecule and comprises the amino acid sequence KIKVFSKVY (SEQ ID NO: 10), AQMNLLQKY (SEQ ID NO: 57), GQKPVILTY (SEQ ID NO: 62), or AQKVSVGQAA (SEQ ID NO: 94).

[0025] 16. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-B*27:05 molecule and comprises the amino acid sequence RQISVQASL (SEQ ID NO: 1) or LRSQILSY (SEQ ID NO: 144), preferably LRSQILSY (SEQ ID NO: 144).

[0026] 17. The leukemia TAP of item 1 or 2, wherein the leukemia TAP binds to an HLA-B*38:01 molecule and comprises the amino acid sequence TQVSMAESI (SEQ ID NO: 46), HHLVETLKF (SEQ ID NO: 64), or THGSEQLHL (SEQ ID NO: 84).

[0027] 18. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-B*40:01 molecule and comprises the amino acid sequence REPYELTVPAL (SEQ ID NO: 75) or SEAEAAKNAL (SEQ ID NO: 76).

[0028] 19. The leukemic TAP of item 1 or 2, wherein the leukemic TAP binds to an HLA-B*44:03 molecule and comprises the amino acid sequence KEIFLELRL (SEQ ID NO: 127).

[0029] 20. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-B*51:01 molecule and comprises the amino acid sequence LPIASASLL (SEQ ID NO: 12), PFPLVQVEPV (SEQ ID NO: 24), PLPIVPAL (SEQ ID NO: 38), IAAPILHV (SEQ ID NO: 68), IPLAVRTI (SEQ ID NO: 115), LPRNKPLL (SEQ ID NO: 116), or LPSHSLLI (SEQ ID NO: 190), preferably IPLAVRTI (SEQ ID NO: 115) or LPRNKPLL (SEQ ID NO: 116).

[0030] 21. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-B*57:01 molecule and comprises the amino acid sequence GARQQIHSW (SEQ ID NO: 3), VTFKLSLF (SEQ ID NO: 16), KGHGGPRSW (SEQ ID NO: 41), GSLDFQRGW (SEQ ID NO: 63), KAFPFHIIF (SEQ ID NO: 69), GTLQGIRAW (SEQ ID NO: 93), RTPKNYQHW (SEQ ID NO: 122), ISNKVPKLF (SEQ ID NO: 125), KTFVQQKTL (SEQ ID NO: 135), ILRSPLKW (SEQ ID NO: 153), or LTVPLSVFW (SEQ ID NO: 183), preferably RTPKNYQHW (SEQ ID NO: 122), ISNKVPKLF (SEQ ID NO: 125), KTFVQQKTL (SEQ ID NO: 135), or ILRSPLKW (SEQ ID NO: 153).

[0031] 22. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-B*57:03 molecule and comprises the amino acid sequence GGSLIHPQW (SEQ ID NO: 60) or LGGAWKAVF (SEQ ID NO: 172).

[0032] 23. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-C*03:03 molecule and comprises the amino acid sequence PARPAGPL (SEQ ID NO: 37), IASPIALL (SEQ ID NO: 112), or HSLISIVYL (SEQ ID NO: 140), preferably IASPIALL (SEQ ID NO: 112) or HSLISIVYL (SEQ ID NO: 140).

[0033] 24. The leukemic TAP according to item 1 or 2, wherein the leukemic TAP binds to an HLA-C*05:01 molecule and comprises the amino acid sequence SLDLLPLSI (SEQ ID NO: 150).

[0034] 25. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-C*06:02 molecule and comprises the amino acid sequence IRMKAQAL (SEQ ID NO: 9), KATEYVHSL (SEQ ID NO: 70), VSFPDVRKV (SEQ ID NO: 87), IGNPILRVL (SEQ ID NO: 142), LSTGHLSTV (SEQ ID NO: 154), or LRKAVDPIL (SEQ ID NO: 166), preferably IGNPILRVL (SEQ ID NO: 142) or LSTGHLSTV (SEQ ID NO: 154).

[0035] 26. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-C*07:01 molecule and comprises the amino acid sequence IGNPILRVL (SEQ ID NO: 142), IYAPHIRLS (SEQ ID NO: 143), TVEEYLVNI (SEQ ID NO: 155), LHNEKGLSL (SEQ ID NO: 178), or VSRNYVLLI (SEQ ID NO: 186), preferably IGNPILRVL (SEQ ID NO: 142) or IYAPHIRLS (SEQ ID NO: 143).

[0036] 27. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-C*07:02 molecule and comprises the amino acid sequence TILPRILTL (SEQ ID NO: 30), SYSPAHARL (SEQ ID NO: 83), TQAPPNVVL (SEQ ID NO: 85), YYLDWIHHY (SEQ ID NO: 90), SLREPQPAL (SEQ ID NO: 109), PAPPHPAAL (SEQ ID NO: 117), or CLRIGPVTL (SEQ ID NO: 158), preferably SLREPQPAL (SEQ ID NO: 109) or PAPPHPAAL (SEQ ID NO: 117).

[0037] 28. The leukemia TAP according to item 1 or 2, wherein the leukemia TAP binds to an HLA-C*08:02 molecule and comprises the amino acid sequence AQDIILQAV (SEQ ID NO: 97), LTDRIYLTL (SEQ ID NO: 102), or AGDIIARLI (SEQ ID NO: 174), preferably AQDIILQAV (SEQ ID NO: 97) or LTDRIYLTL (SEQ ID NO: 102).

[0038] 29. The leukemia TAP of item 1 or 2, wherein the leukemia TAP binds to an HLA-C*12:03 molecule and comprises the amino acid sequence LSAHLSSL (SEQ ID NO: 173).

[0039] 30. The leukemic TAP of any one of items 1 to 29, wherein the TAP is encoded by a sequence located in a non-protein-coding region of the genome.

[0040] 31. The leukemia TAP according to item 30, wherein the non-protein-coding region of the genome is an untranslated transcribed region (UTR).

[0041] 32. The leukemia TAP according to item 30, wherein the non-protein-coding region of the genome is an intron.

[0042] 33. The leukemia TAP according to item 30, wherein the non-protein-coding region of the genome is an intergenic region.

[0043] 34. A combination comprising at least two of the leukemia TAPs defined in any one of items 1 to 33.

[0044] 35. A nucleic acid encoding a leukemia TAP according to any one of items 1 to 33, or a combination according to item 34.

[0045] 36. The nucleic acid according to item 35, which is an mRNA or a viral vector.

[0046] 37. A liposome comprising the leukemia TAP according to any one of items 1 to 33, the combination according to item 34, or the nucleic acid according to item 35 or 36.

[0047] 38. A composition comprising the leukemia TAP according to any one of items 1 to 33, the combination according to item 34, the nucleic acid according to item 35 or 36, or the liposome according to item 37, and a pharmaceutically acceptable carrier.

[0048] 39. A vaccine comprising the leukemia TAP according to any one of items 1 to 33, the combination according to item 34, the nucleic acid according to item 35 or 36, the liposome according to item 37, or the composition according to item 38, and an adjuvant.

[0049] 40. An isolated major histocompatibility complex (MHC) class I molecule, comprising within its peptide-binding groove a leukemic TAP according to any one of items 1 to 33.

[0050] 41. The isolated MHC class I molecule according to item 40, in the form of a multimer.

[0051] 42. The isolated MHC class I molecule according to item 41, wherein the multimer is a tetramer.

[0052] 43. An isolated cell comprising (i) a leukemic TAP according to any one of items 1 to 33, (ii) a combination according to item 34, or (iii) a vector comprising a nucleotide sequence encoding a TAP according to any one of items 1 to 33 or a combination according to item 34.

[0053] 44. An isolated cell expressing on its surface a major histocompatibility complex (MHC) class I molecule, wherein the MHC class I molecule contains, in their peptide-binding groove, a leukemia TAP according to any one of items 1 to 33 or a combination according to item 34.

[0054] 45. The cell according to item 44, which is an antigen-presenting cell (APC).

[0055] 46. ​​The cell according to item 45, wherein the APC is a dendritic cell.

[0056] 47. A T cell receptor (TCR) that specifically recognizes the isolated MHC class I molecule according to any one of items 40 to 42 and / or the MHC class I molecule expressed on the surface of the cell according to any one of items 44 to 46.

[0057] 48. The TCR according to item 47, wherein the TCR comprises a TCR beta (TCRβ) chain comprising a complementarity-determining region 3 (CDR3) comprising one of the amino acid sequences set forth in SEQ ID NOs: 191 to 219.

[0058] 49. An isolated cell, which expresses on its cell surface a TCR according to item 47 or 48.

[0059] 50.CD8 + 50. The isolated cell of item 49, which is a T lymphocyte.

[0060] 51. A cell population comprising at least 0.5% isolated cells as defined in item 49 or 50.

[0061] 52. A method for treating leukemia in a subject, comprising administering to the subject an effective amount of (i) a leukemia TAP according to any one of items 1 to 33, (ii) a combination according to item 34, (iii) a nucleic acid according to item 35 or 36, (iv) a liposome according to item 37, (v) a composition according to item 38, (vi) a vaccine according to item 39, (vii) a cell according to any one of items 43 to 46, 49, and 50, or (viii) a cell population according to item 51.

[0062] 53. The method according to item 52, wherein the leukemia is myeloid leukemia.

[0063] 54. The method according to item 53, wherein the myeloid leukemia is acute myeloid leukemia (AML).

[0064] 55. The method of any one of items 52 to 54, further comprising administering to the subject at least one additional anti-tumor agent or therapy.

[0065] 56. The method of item 55, wherein the at least one additional anti-tumor agent or therapy is a chemotherapeutic agent, immunotherapy, immune checkpoint inhibitor, radiation therapy, or surgery.

[0066] 57. Use of (i) a leukemia TAP according to any one of items 1 to 33, (ii) a combination according to item 34, (iii) a nucleic acid according to item 35 or 36, (iv) a liposome according to item 37, (v) a composition according to item 38, (vi) a vaccine according to item 39, (vii) a cell according to any one of items 43 to 46, 49 and 50, or (viii) a cell population according to item 51 for treating leukemia in a subject.

[0067] 58. Use of (i) a leukemia TAP according to any one of items 1 to 33, (ii) a combination according to item 34, (iii) a nucleic acid according to item 35 or 36, (iv) a liposome according to item 37, (v) a composition according to item 38, (vi) a vaccine according to item 39, (vii) a cell according to any one of items 43 to 46, 49 and 50, or (viii) a cell population according to item 51 for the manufacture of a medicament for treating leukemia in a subject.

[0068] 59. The use according to item 57 or 58, wherein the leukemia is myeloid leukemia.

[0069] 60. The use according to item 59, wherein the myeloid leukemia is acute myeloid leukemia (AML).

[0070] 61. The use according to any one of items 57 to 60, further comprising the use of at least one additional anti-tumor agent or therapy.

[0071] 62. The use according to item 61, wherein the at least one additional anti-tumor agent or therapy is a chemotherapeutic agent, immunotherapy, immune checkpoint inhibitor, radiation therapy, or surgery.

[0072] 63. (i) a leukemia TAP according to any one of items 1 to 33, (ii) a combination according to item 34, (iii) a nucleic acid according to item 35 or 36, (iv) a liposome according to item 37, (v) a composition according to item 38, (vi) a vaccine according to item 39, (vii) a cell according to any one of items 43 to 46, 49, and 50, or (viii) a cell population according to item 51, for treating leukemia in a subject.

[0073] 64. The leukemia TAP, combination, nucleic acid, liposome, composition, vaccine, cell, or cell population for use according to item 63, wherein the leukemia is myeloid leukemia.

[0074] 65. The leukemia TAP, combination, nucleic acid, liposome, composition, vaccine, cell, or cell population for use according to item 64, wherein the myeloid leukemia is acute myeloid leukemia (AML).

[0075] 66. The leukemia TAP, combination, nucleic acid, liposome, composition, vaccine, cell, or cell population for use according to any one of items 63 to 65, which is for use in combination with at least one additional anti-tumor agent or therapy.

[0076] 67. The leukemia TAP, combination, nucleic acid, liposome, composition, vaccine, cell, or cell population for use according to item 66, wherein the at least one additional anti-tumor agent or therapy is a chemotherapeutic agent, immunotherapy, immune checkpoint inhibitor, radiation therapy, or surgery.

[0077] Other objects, advantages and features of the present invention will become more apparent upon reading of the following non-limiting description of specific embodiments, given by way of example only with reference to the accompanying drawings. [Brief explanation of the drawings]

[0078] In the accompanying drawings: [Figure 1]These graphs demonstrate that hematopoietic progenitor cells are a better control than mTECs for detecting TSAs in AML. Figure 1A: Comparison of the effectiveness of k-mer depletion from each k-mer set of 19 AML specimens with either the combined k-mers from six mTEC samples or six MPC samples. For this comparison, the jellyfish database was generated in canonical mode, ignoring the occurrence of k-mers less than 2. Figure 1B: Overlap between the combined k-mers of all AML specimens and the k-mers from the six mTEC and six MPC samples used in Figure 1A. The parameters for database construction used in Figure 1A were reapplied here. Figure 1C: t-distributed stochastic neighbor embedding (t-SNE) analysis of expressed protein-coding genes (TPM ≥ 1) in purified cell populations from the indicated tissues. Figure 1D: Comparison of the total number of expressed protein-coding genes (TPM ≥ 1) in the indicated tissues and cell populations used to plot panel C. Pluri_stem: pluripotent stem cells, Ery: erythrocytes, Precu: progenitor cells, Lympho: lymphocytes, Granulo: granulocytes, Mono: monocytes. mTECs were compared with other tissues using the Mann-Whitney U test (****p<0.0001). Bars indicate mean and standard deviation. [Figure 2] Schematic diagrams of the MPC-based TSA discovery approach are shown. (A) Schematic diagram of the workflow for TSA discovery based on mTEC k-mer depletion. (B) Schematic diagram of the workflow for ERE-derived MAP discovery. (C) Schematic diagram of the workflow for the mTEC+MPC k-mer depletion TSA discovery approach. (D) Schematic diagram of the workflow for the DKE approach. The workflow for the AML#1 sample is shown here. A fold change of 10 was used as the minimum to consider k-mers as overexpressed (other filters were also applied; see Methods). For the other three approaches, the resulting database of full-frame translated contigs was concatenated with the individualized canonical proteome, followed by MS identification of MAPs eluted from the same AML sample used for RNA sequencing. [Figure 3]Figure 3A shows that the MPC-based approach identifies the majority of TSA-high genes in AML. The σ of the distribution (black plot) is given. Figure 3A: For each AML specimen (n = 19), the proportion of MAPs derived from transcripts separated into 10 distinct groups (deciles) based on their TPM expression. Decile 10 has the most highly expressed transcripts, and decile 1 has the least expressed transcripts. The boxes indicate the median, 25th, and 75th percentiles of the distribution, with whiskers extending to the minimum and maximum values. Figure 3B: Normal distribution of the cumulative frequency of MAPs (dots) as a function of the logarithm of the total number of RNA-seq reads (rphm) that can encode them in AML specimens identified by MS. The mean (μ) and standard deviation (σ) of the distribution (black plot) are given. Figure 3C: Probabilities were calculated based on the normal distribution parameters in Figure 3B for RNA-seq to generate MAPs after the different indicated fold changes (FC, original rphm × FC). Figure 3D: A decision tree was used to separate the MAPs of interest (MOI) into TAA, HSA, and TSA high. "Normal tissue" refers to all tissues (GTEx, purified hematopoietic cells, and mTEC), while blood / BM refers to purified hematopoietic cells only. Figure 3E: Comparison of MOI counts obtained by each indicated proteogenomic approach. Figure 3F: Venn diagram comparing TSA high identities between the indicated approaches. Figure 3G: Pearson correlation between observed retention time and predicted retention time (left) or hydrophobicity index (right). Figure 3H: Median and interquartile range frequencies of successful re-identification of the indicated MAPs by Comet. [Figure 4]This indicates that TSA-highs are primarily derived from intron translation and are shared among many patients. Figure 4A: Heatmap showing the mean RNA expression (logarithm of rphm+1) of each identified TSA-high in either total normal tissue from GTEx (n = 12–50 depending on available samples), normal sorted hematopoietic cell populations (n ​​= 3–16 depending on available samples), or mTECs (n = 11). TAAs evaluated as safe in clinical trials have also been reported. Prec: Progenitor cells. Figure 4B: Comparison of the fold change in TSA-high between mean rphm expression in 19 AML specimens and MPCs (n = 16). Dots indicate each MOI, boxes indicate the median, 25th, and 75th percentiles of the distribution, and whiskers extend to the minimum and maximum values. Figure 4C: Distribution of biotypes (genomic regions or events) that generated the indicated MOIs. Exon-intron: peptides overlapping exon-intron junctions (retention introns), ncRNA: non-coding RNA, OoF translation: out-of-frame translation. Figure 4D: TSA-high RNA expression in 19 AML samples and 437 Leucegene patients. Figure 4E: Population coverage by HLA allotypes capable of presenting TSA-high (19 AML sample alleles presenting TSA-high + promiscuous binders calculated by MHC cluster). This was calculated using the IEDB Population Coverage Tool (www.iedb.org). Bars indicate the frequency of individuals within the global population carrying up to six allotypes (x-axis), and the cumulative percentage of population coverage is shown as dots. Figure 4F: Distribution of HLA-TSA-high complexes in the Leucegene cohort based on TSA-high RNA expression (considered expressed if rphm ≥ 2), patient HLA alleles (OptiType), and promiscuous binders. Figure 4G: Number of predHLA-TSA high complexes in Leucegene patients at diagnosis and relapse. Figure 4H: RNA expression of TSA high that can be presented by HLA alleles in paired purified AML blasts from 15 patients at diagnosis and relapse (data from (Toffalori et al., 2019)).Comparisons were performed using the Wilcoxon paired signed-rank test. Figure 4I: Comparison of shared genes (considered expressed if rphm>0) among TSA-high samples in sorted blast cells (n=12) or leukemia stem cells (LSC, n=8) as reported elsewhere (Corces et al., 2016). Figure 4J: RNA expression of HLA-ABC molecules in the samples shown in Figure 4I. Mean + SD is shown. Figure 4K: GSEA analysis comparing Leucegene patients expressing above the median (rphm>0) TSA-high (n=207) with other patients (n=230) for the indicated LSC signature gene set (Eppert et al., 2011). NES: Normalized Enrichment Score. [Figure 5] The presence of multiple TSA-high mutations correlates with better survival. Figure 5A: Kaplan-Meier survival analysis between Leucegene patients expressing high (n = 98, upper quartile of Figure 5B) versus low (n = 275, all other patients) levels of the HLA-TSA-high complex. Statistical significance was determined by the log-rank test. Figure 5B: Forest plot for multivariate analysis of 5-year overall survival. HR: adjusted hazard ratio, CI: confidence interval, adv: adverse, fav: favorable, int: moderate. NPM1 / FLT3 interaction = presence of both NPM1 variants and FLT3-ITD. Figure 5C: Log-rank p-values ​​calculated after removing the indicated number of TSA-high mutations from the analysis performed in (A). 1000 permutations were performed for each number, and the mean + SD is reported. Figure 5D: Percentage of significant p-values ​​obtained in Figure 5C. Figure 5E: Comparison of log-rank p-values ​​recalculated after alternative removal of each TSA-high mutation from the analysis in Figure 5A. Figure 5F: Comparison of log-rank p-values ​​recalculated after alternative removal of each HLA allele from the analysis in Figure 5A. [Figure 6]Figure 6A: Comparison of immunogenicity scores (Repitope) among MOIs, thymic stromal cell-derived MAPs, and HIV MAPs. Figure 6B: Median and interquartile range of mean RNA expression across 11 available mTEC samples for MOIs, 5112 non-immunogenic MAPs, and 1411 immunogenic MAPs (curated from the IEDB (Ogishi and Yotsuyanagi, 2019)). Figure 6C: IFN-γ ELISpot assay of healthy PBMCs after stimulation of DCs pulsed with the indicated peptides. Results from two independent experiments are combined. Figure 6D: ELISpot assay of the indicated TSA-high (single donor). Figure 6E: Flow cytometry analysis of cytokine secretion of T cells expanded in the presence of the indicated peptides. Figure 6F: Representative flow cytometry plot of the indicated dextramer frequency among T cells expanded in the presence of the indicated peptides. Figure 6G: FEST assay: Significant T cell clonotype expansion after 10 days of stimulation with three different TSA-high pools (5 peptides / pool). Figure 6H: TCR CDR3s per 1,000 TCR reads (CPK, as a measure of clonotype diversity) in Leucegene patients with high versus low counts of the indicated predHLA-MOI (related to Figures 4F and 11D). Figure 6I: Frequency of TSA-high-responsive clonotypes (n = 66-164 / group) in Leucegene, as predicted by ERGO. Figure 6J: Frequency of TAA-responsive clonotypes (n = 74-207 / group) in Leucegene, as predicted by ERGO. Figure 6K: Frequency of clonotypes recognizing the pred MOI presented in the examined samples among all anti-MOI clonotypes (normalized by the number of pred-presented MOIs) (related to I and J). Patients with an anti-presMOI clonotype count of 0 were ignored. Figure 6L: Correlation between RNA expression of CD8A and CD8B genes and the number of TSA highs expressed above 2 rphm in Leucegene. Figure 6M: Correlation between RNA expression of CD8A and CD8B genes and the number of predHLA-TSA highs in Leucegene.Figure 6N: Volcano plot of differential gene expression analysis comparing patients with normalized TSA high predilection above the median vs. below the median. Dots indicate genes upregulated in patients above the median. Figure 6O: GO term analysis of the upregulated genes in Figure 6N. [Figure 7] High TSA expression is associated with immune editing, AML driver mutations, and epigenetic abnormalities. Figure 7A: Pearson correlation between the number of HE-TSA highs and the expression of the indicated genes across the complete Leucegene cohort (n=437). In the first panel, HLA-A, -B, and -C expression values ​​were summed. Figure 7B: Comparison of PD-L1 (CD274) gene expression between Leucegene patients expressing above the median HE-TSA high and other patients (stratified as a function of NPM1 mutation status). Figure 7C: Network analysis of GO term enrichment between genes was inversely correlated with the number of HE-TSA highs. Node size is proportional to the size of the gene set. Figure 7D: Network analysis of GO term enrichment between genes was positively correlated with the number of HE-TSA highs. Figure 7E: Comparison of the number of patients expressing above the median HE-TSA high with the number of other patients between wild-type and mutant patients for the indicated genes. Statistical significance established by Fisher's exact test (**p<0.01, ***p<0.0001). Figure 7F: Comparison of HE-TSA counts among patients with 0 to 3 mutations in either NPM1, FLT3, or DNMT3A. Figure 7G: Unsupervised consensus clustering of intron retention ratios for Leucegene patients (n=437, columns) as determined by IRFinder. Rows represent the 1211 top introns with the highest variability and significance for the consensus clustering, clustered hierarchically. Patient FAB types are indicated below the heatmap, along with p-values ​​indicating significant association with the indicated consensus cluster (Fisher's exact test, *p<0.05, **p<0.01, ***p<0.001, ***p<0.0001). [Figure 8A] FIG. 1 is a diagram of the concept of k-mer occurrence. [Figure 8B]10 is a graph showing an example of k-mer frequency distribution in the occurrence function for sample 05H143. [Figure 8C] FIG. 1 is a graph illustrating a comparison of occurrence thresholds used between mTEC-only and mTEC+MPC k-mer depletion approaches (each dot is a different AML sample). [Figure 8D] Graph showing the overlap of k-mer identities between unique k-mer combinations obtained from all 19 AML specimens obtained after depletion of either mTEC or mTEC+MPC. [Figure 9A] FIG. 1 is a schematic diagram providing details of differential k-mer expression analysis and MS database construction. As an example, the construction of the MS database for sample AML#1 is presented. FC: fold change. FIG. 2 is a diagram providing details of differential k-mer expression analysis and MS database construction. As an example, the construction of the MS database for sample AML#1 is presented. FC: fold change. [Figure 9B] Graph showing the average database size (line) versus the cumulative number of personalized canonical peptide identifications (peptides derived from personalized canonical proteomes, either alone (Canon.) or concatenated with contig sequences in the four indicated approaches). [Figure 9C] Venn diagrams comparing the identity overlap of canonical peptides identified based on each approach with peptides identified based on the individualized canonical proteome alone are shown. [Figure 10A] Graph showing a comparison of the percentage of MHC-I associated peptides (MAPs) of interest (MOI) identified by each TSA identification approach. [Figure 10B] Graph showing a comparison of the total number of AML samples (out of 19 samples used to identify TSA in this study) expressing TSA high (rphm>0) identified by either mTEC+MPC k-mer depletion or differential k-mer expression approaches. [Figure 11A] Graph showing distribution of number of TSA highs with RNA expression ≧2 rphm in the Leucegene cohort (n=437). [Figure 11B] Graph showing survival comparison between patients in the Leucegene cohort (n=372, patients sequenced at diagnosis and for whom survival data was available) presenting with a high number of TSA expressed at levels of 2 rphm or greater (top quartile of distribution in left panel) versus patients presenting with low levels (remaining cohort). [Figure 11C] Graphs showing the distribution of HLA-MOI complexes across the entire Leucegene cohort based on RNA expression (considered expressed if rphm≧2), each patient's HLA allele, and promiscuous binder predictions (optitypes and MHC clusters) for HSA (FIG. 11C), TAA (FIG. 11D), and TSA low (FIG. 11E). [Figure 11D] Graphs showing the distribution of HLA-MOI complexes across the entire Leucegene cohort based on RNA expression (considered expressed if rphm≧2), each patient's HLA allele, and promiscuous binder predictions (optitypes and MHC clusters) for HSA (FIG. 11C), TAA (FIG. 11D), and TSA low (FIG. 11E). [Figure 11E] Graphs showing the distribution of HLA-MOI complexes across the entire Leucegene cohort based on RNA expression (considered expressed if rphm≧2), each patient's HLA allele, and promiscuous binder predictions (optitypes and MHC clusters) for HSA (FIG. 11C), TAA (FIG. 11D), and TSA low (FIG. 11E). [Figure 11F] Graphs showing survival comparisons between patients (n=372, patients sequenced at diagnosis and for whom survival data was available) from the Leucegene cohort presenting high levels of HLA-MOI complexes (top quartile of distribution in top panel) versus patients presenting low levels (remainder of cohort) for HSA (FIG. 11F), TAA (FIG. 11G), and TSA low (FIG. 11H). [Figure 11G]Graphs showing survival comparisons between patients (n=372, patients sequenced at diagnosis and for whom survival data was available) from the Leucegene cohort presenting high levels of HLA-MOI complexes (top quartile of distribution in top panel) versus patients presenting low levels (remainder of cohort) for HSA (FIG. 11F), TAA (FIG. 11G), and TSA low (FIG. 11H). [Figure 11H] Graphs showing survival comparisons between patients (n=372, patients sequenced at diagnosis and for whom survival data was available) from the Leucegene cohort presenting high levels of HLA-MOI complexes (top quartile of distribution in top panel) versus patients presenting low levels (remainder of cohort) for HSA (FIG. 11F), TAA (FIG. 11G), and TSA low (FIG. 11H). [Figure 12A] Pearson correlation between the number of HE-TSA highs and the expression of the indicated genes across the complete Leucegene cohort (n=437) is shown. [Figure 12B] A graph showing a comparison of the expression of the indicated genes in patients with high pred presentation levels of TSA high versus the remaining patients is shown (related to Figure 4F). [Figure 12C] Pearson correlation between ZNF445 expression and the number of retained introns in the Leucegene cohort (analysis by IRFinder, defined as retained if >10% of transcripts were retained) is shown. [Figure 12D] Graph showing comparison of patients expressing above median HE-TSA high with other patients between wild-type and mutant patients for the indicated genes. Statistical significance established by Fisher's exact test. [Figure 12E] Graph showing a comparison of patients expressing above median HE-TSA elevation with other patients, among patients who did or did not undergo allogeneic HSCT. Statistical significance established by Fisher's exact test. [Figure 12F]Graph showing the distribution of FAB types in patients with high HE-TSA counts above or below the median high HE-TSA count across the entire Leucegene cohort. [Figure 12G] Graph showing the distribution of 2008 WHO classification of patients with high HE-TSA counts above or below the median high HE-TSA count across the entire Leucegene cohort. [Figure 12H] Graph showing the distribution of cytogenetic profiles of patients with high HE-TSA counts above or below the median high HE-TSA count across the Leucegene cohort. DETAILED DESCRIPTION OF THE INVENTION

[0079] The terms and symbols used in this specification for genetics, molecular biology, biochemistry, and nucleic acid are based on standard treatises and texts in the field, such as Kornberg and Baker, DNA Replication, Second Edition (WH Freeman, New York, 1992), Lehninger, Biochemistry, Second Edition (Worth Publishers, New York, 1975), Strachan and Read, Human Molecular Genetics, Second Edition (Wiley-Liss, New York, 1999), Eckstein, editor, Oligonucleotides and Analogs: A Practical Approach (Oxford University Press, New York, 1991); Gait, editor, Oligonucleotide Synthesis: A Practical Approach (IRL Press, Oxford, 1984), etc. All terms should be understood in the typical sense established in the relevant art.

[0080] The articles "a" and "an" are used herein to refer to one or to more than one (i.e., to at least one) of the grammatical object of the article. By way of example, "an element" means one element or more than one element. Throughout this specification, unless the context requires otherwise, the words "comprise", "comprises", and "comprising" will be understood to imply the inclusion of the stated step or element or group of steps or elements but not the exclusion of any other step or element or group of steps or elements.

[0081] The recitation of ranges of values ​​herein, unless otherwise stated herein, is merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, and each separate value is incorporated herein as if it were individually recited herein. Every subset of values ​​within a range is also incorporated herein as if it were individually recited herein.

[0082] All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context.

[0083] Any and all examples provided herein, or the use of exemplary language (e.g., "etc."), are intended only to better illustrate the invention and do not pose a limitation on the scope of the invention unless otherwise stated.

[0084] No language in the specification should be construed as indicating any element not claimed as essential to the practice of the invention.

[0085] As used herein, the term "about" has its ordinary meaning. The term "about" is used to indicate that a value includes the inherent variation of error for the device or method being employed to determine the value, or encompasses values ​​that are near the recited value, e.g., within 10% or 5% of the recited value (or range of values).

[0086] In the study described herein, the inventors used a proteogenomics-based approach to identify TSA candidates from 19 AML specimens. The majority of these TSAs are derived from aberrantly expressed, non-mutated genomic sequences, such as non-exonic sequences (e.g., intronic and intergenic sequences), that are not expressed in normal tissues. The expression of these AML TSA candidates was shown to correlate with mutations in epigenetic modifiers (e.g., DNMT3A) and the expression of ZNF445, a regulator of genomic imprinting. It has also been shown that AML TSA candidates are highly shared between patients, expressed in both blasts and leukemia stem cells, and their HLA presentation is associated with markers of immunoediting and better overall survival. Therefore, the novel AML TSA candidates identified herein may be useful for leukemia T cell-based immunotherapy.

[0087] Thus, in one aspect, the present disclosure relates to a leukemia TAP (or leukemia tumor-specific peptide) comprising or consisting of one of the following amino acid sequences: [Table 2-1] [Table 2-2]

[0088] Generally, peptides (e.g., TAPs) presented in the context of HLA class I vary in length from about 7 or 8 to about 15, or preferably 8 to 14, amino acid residues. In some embodiments of the methods of the present disclosure, longer peptides comprising a TAP sequence as defined herein are artificially loaded into cells, such as antigen-presenting cells (APCs), where they are processed by the cells and the TAP is presented by MHC class I molecules on the surface of the APCs. In this method, peptides / polypeptides longer than 15 amino acid residues can be loaded into APCs and processed by proteases in the APC cytoplasm to provide the corresponding TAP as defined herein for presentation. In some embodiments, the precursor peptides / polypeptides used to generate the TAPs defined herein are, for example, 1000, 500, 400, 300, 200, 150, 100, 75, 50, 45, 40, 35, 30, 25, 20, or 15 amino acids or less. Thus, all methods and processes using TAPs described herein include the use of longer peptides or polypeptides (including naturally occurring proteins), i.e., tumor antigen precursor peptides / polypeptides, to induce the presentation of the "final" 8-14 TAPs after processing by cells (APCs). In some embodiments, the leukemia TAPs described herein are approximately 8-14, 8-13, or 8-12 amino acids in length (e.g., 8, 9, 10, 11, 12, or 13 amino acids in length), small enough to fit directly onto an HLA class I molecule. In embodiments, the TAP contains 20 or fewer amino acids, preferably 15 or fewer amino acids, and more preferably 14 or fewer amino acids. In embodiments, the TAP contains at least 7 amino acids, preferably at least 8 or fewer amino acids, and more preferably at least 9 amino acids.

[0089] As used herein, the term "amino acid" includes both L- and D-forms of naturally occurring amino acids as well as other amino acids (e.g., naturally occurring amino acids, non-naturally occurring amino acids, amino acids not encoded by nucleic acid sequences, etc.) used in peptide chemistry to prepare synthetic analogs of TAP. Examples of naturally occurring amino acids are glycine, alanine, valine, leucine, isoleucine, serine, threonine, etc. Other amino acids include, for example, non-genetically encoded forms of amino acids, as well as conservative substitutions for L-amino acids. Naturally occurring non-genetically encoded amino acids include, for example, β-alanine, 3-aminopropionic acid, 2,3-diaminopropionic acid, α-aminoisobutyric acid (Aib), 4-amino-butyric acid, N-methylglycine (sarcosine), hydroxyproline, ornithine (e.g., L-ornithine), citrulline, t-butylalanine, t-butylglycine, N-methylisoleucine, phenylglycine, cyclohexylalanine, norleucine (Nle), norvaline, 2-naphthylalanine, pyridylalanine, 3-benzothienylalanine, 4-chlorophenylalanine, 2-fluoro ... Examples of amino acids include phenylalanine, 3-fluorophenylalanine, 4-fluorophenylalanine, penicillamine, 1,2,3,4-tetrahydro-isoquinoline-3-carboxylic acid, β-2-thienylalanine, methionine sulfoxide, L-homoarginine (Hoarg), N-acetyllysine, 2-aminobutyric acid, 2-aminobutyric acid, 2,4-diaminobutyric acid (D- or L-), p-aminophenylalanine, N-methylvaline, homocysteine, homoserine (HoSer), cysteic acid, ε-aminohexanoic acid, δ-aminovaleric acid, and 2,3-diaminobutyric acid (D- or L-). These amino acids are well known in the fields of biochemistry / peptide chemistry. In embodiments, TAP contains only naturally occurring amino acids.

[0090] In embodiments, the TAPs described herein include peptides with altered sequences containing substitutions of functionally equivalent amino acid residues compared to the sequences described herein. For example, one or more amino acid residues within a sequence can be substituted with another amino acid of similar polarity (having similar physicochemical properties) that acts as a functional equivalent, resulting in a silent alteration. Substitutes for amino acids within a sequence can be selected from other members of the class to which the amino acid belongs. For example, positively charged (basic) amino acids include arginine, lysine, and histidine (as well as homoarginine and ornithine). Nonpolar (hydrophobic) amino acids include leucine, isoleucine, alanine, phenylalanine, valine, proline, tryptophan, and methionine. Uncharged polar amino acids include serine, threonine, cysteine, tyrosine, asparagine, and glutamine. Negatively charged (acidic) amino acids include glutamic acid and aspartic acid. The amino acid glycine can be included in either the nonpolar amino acid family or the uncharged (neutral) polar amino acid family. Substitutions made within a family of amino acids are generally understood to be conservative substitutions. The TAPs described herein can include any L-amino acid, any D-amino acid, or a mixture of L- and D-amino acids. In embodiments, the TAPs described herein include all L-amino acids.

[0091] In an embodiment, in the sequence of TAP comprising or consisting of one of the sequences of SEQ ID NOs: 1 to 190, preferably SEQ ID NOs: 97 to 154, amino acid residues that do not substantially contribute to the interaction with the T cell receptor can be modified by substituting them with other amino acids that do not substantially affect T cell responsiveness and do not eliminate binding to the relevant MHC.

[0092] TAP may also be N- and / or C-terminally capped or modified to prevent degradation and improve stability, affinity, and / or uptake. Thus, in another aspect, the present disclosure provides a compound of formula Z 1 -XZ 2wherein X is a TAP comprising or consisting of one of the amino acid sequences of SEQ ID NOs: 1 to 190, preferably SEQ ID NOs: 97 to 154.

[0093] In embodiments, the amino terminal residue of TAP (i.e., the N-terminal free amino group) is, for example, a moiety / chemical group (Z 1 ) is modified (e.g., for protection against degradation). 1 may be a straight or branched chain alkyl group of 1 to 8 carbons, or an acyl group (R—CO—), where R is a hydrophobic moiety (e.g., acetyl, propionyl, butanyl, isopropionyl, or iso-butanyl), or an aroyl group (Ar—CO—), where Ar is an aryl group. In embodiments, the acyl group is a C1-C 16 or C3-C 16 In a further embodiment, Z is a saturated C1-C6 acyl group (linear or branched, saturated or unsaturated), or an unsaturated C3-C6 acyl group (linear or branched), such as an acetyl group (CH3-CO-, Ac). 1 The carboxy-terminal residue of TAP (i.e., the free carboxy group at the C-terminus of TAP) may be modified (e.g., for protection against degradation) by, for example, amidation (replacement of an OH group with an NH2 group), and thus, in such cases, Z 2 is an NH group. In embodiments, Z 2 may be a hydroxamate group, a nitrile group, an amide (primary, secondary, or tertiary) group, an aliphatic amine of 1 to 10 carbons such as methylamine, iso-butylamine, iso-valerylamine, or cyclohexylamine, an aromatic or arylalkyl amine such as aniline, naphthylamine, benzylamine, cinnamylamine, or phenylethylamine, an alcohol, or CHOH. 2In an embodiment, TAP comprises one of the amino acid sequences of SEQ ID NOs: 1 to 190, preferably SEQ ID NOs: 97 to 154. In an embodiment, TAP consists of one of the amino acid sequences of SEQ ID NOs: 1 to 190, preferably SEQ ID NOs: 97 to 154, i.e., Z 1 and Z 2 does not exist.

[0094] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-A*01:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 48, 67, 89, 134, 151, or 164, SEQ ID NO: 134 or 151.

[0095] In another aspect, the disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-A*02:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 7, 11, 27, 32, 33, 34, 35, 4351, 52, 53, 54, 61, 65, 72, 77, 82, 86, 104, 108, 119, 123, 130, 132, 146, 150, 167, 168, 169, 171, 183, or 188, preferably SEQ ID NO: 104, 108, 119, 123, 130, 132, 146, or 150. Due to HLA allele promiscuity (certain HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-A*02:05, HLA-A*02:06, and / or HLA-A*02:07 molecules.

[0096] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-A*03:01 molecules, comprising or consisting of the sequence of SEQ ID NO: 5, 18, 19, 20, 42, 44, 74, 91, 105, 113, 126, 131, 159, 160, 180, or 189, preferably SEQ ID NO: 105, 113, 126, or 131. Due to HLA allele promiscuity (certain HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-A*11:01 molecules.

[0097] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-A*11:01 molecules, comprising or consisting of the sequence of SEQ ID NO: 6, 45, 96, 106, 121, 149, or 152, preferably SEQ ID NO: 106, 121, 149, or 152. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-A*03:01, HLA-A*31:01, and / or HLA-A*68:01 molecules.

[0098] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-A*24:02 molecules, comprising or consisting of the sequence of SEQ ID NO: 13, 36, 71, 92, 95, or 145, preferably SEQ ID NO: 145. Due to HLA allele promiscuity (certain HLA alleles present similar epitopes, see Table 4), the above identified TAPs may additionally bind to HLA-A*23:01 molecules.

[0099] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-A*26:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 59 or 185. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-A*25:01 and / or HLA-A*66:01 molecules.

[0100] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-A*29:02 molecules, comprising or consisting of the sequence of SEQ ID NO: 88, 99, or 138, preferably SEQ ID NO: 99 or 138. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-A*30:02 and / or HLA-B*15:02 molecules.

[0101] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-A*30:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 26, 29, or 165.

[0102] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-A*68:02 molecule, comprising or consisting of the sequence of SEQ ID NO: 50 or SEQ ID NO: 148, preferably SEQ ID NO: 148.

[0103] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-B*07:02 molecule, comprising or consisting of the sequence of SEQ ID NO: 8, 25, 40, 55, 56, 73, 78, 79, 80, 81, 98, 100, 107, 110, 111, 118, 120, 128, 129, 157, 161, 163, 179, or 184, preferably SEQ ID NO: 98, 100, 107, 110, 111, 118, 120, 128, or 129. Due to HLA allele promiscuity (certain HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-B*35:02, HLA-B*35:03, HLA-B*55:01, and / or HLA-B*56:01 molecules.

[0104] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-B*08:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 4, 14, 15, 17, SEQ ID NO: 21, 22, 23, 28, 31, 47, 49, 101, 103, 114, 137, 139, 147, 156, 170, 181, or 182, preferably SEQ ID NO: 101, 103, 114, 137, 139, or 147.

[0105] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-B*14:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 2, 39, 58, 66, 124, 133, 136, 141, 162, 175, 176, or 177, preferably SEQ ID NO: 124, 133, 136, or 141.

[0106] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-B*15:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 10, 57, 62, or 94. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-B*15:02, HLA-B*15:03, and / or HLA-B*46:01 molecules.

[0107] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-B*27:05 molecules, comprising or consisting of the sequence of SEQ ID NO: 1 or 144, preferably SEQ ID NO: 144. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may additionally bind to HLA-B*27:02 molecules.

[0108] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-B*38:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 4, 64, or 84. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to an HLA-B*39:01 molecule.

[0109] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-B*40:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 75 or 76. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAP may further bind to HLA-B*18:01, HLA-B*40:02, HLA-B*41:02, HLA-B*44:02, HLA-B*44:03, and / or HLA-B*45:01 molecules.

[0110] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-B*44:03 molecules, comprising or consisting of the sequence of SEQ ID NO: 127. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAP may further bind to HLA-B*18:01, HLA-B*40:01, HLA-B*40:02, HLA-B*41:02, HLA-B*44:02, and / or HLA-B*45:01 molecules.

[0111] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-B*51:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 12, 24, 38, 68, 115, 116, or 190, preferably SEQ ID NO: 115 or 116. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-B*35:02, HLA-B*35:03, HLA-B*52:01, HLA-B*53:01, HLA-B*55:01, and / or HLA-B*56:01 molecules.

[0112] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-B*57:01 molecules, comprising or consisting of the sequence of SEQ ID NO: 3, 16, 41, 63, 69, 93, 122, 125, 135, 153, or 183, preferably 122, 125, 135, or 153. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-A*32:01 and / or HLA-B*58:01 molecules.

[0113] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-B*57:03 molecule, comprising or consisting of the sequence of SEQ ID NO: 60 or 172.

[0114] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-C*03:03 molecule, comprising or consisting of the sequence of SEQ ID NO: 37, 112, or 140, preferably SEQ ID NO: 112 or 140. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-B*46:01, HLA-C*03:02, HLA-C*03:04, HLA-C*08:01, HLA-C*08:02, HLA-C*12:02, HLA-C*12:03, HLA-C*15:02, and / or HLA-C*16:01 molecules.

[0115] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-C*05:01 molecules, comprising or consisting of the sequence of SEQ ID NO: 150. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may additionally bind to HLA-C*08:01 and / or HLA-C*08:02 molecules.

[0116] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-C*06:02 molecules, comprising or consisting of the sequence of SEQ ID NO: 9, 70, 87, 142, 154, or 166, preferably SEQ ID NO: 142 or 154. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-B*27:02, HLA-C*07:01, and / or HLA-C*07:02 molecules.

[0117] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-C*07:01 molecule, comprising or consisting of the sequence of SEQ ID NO: 142, 143, 155, 178, or 186, preferably SEQ ID NO: 142 or 143. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-B*27:02, HLA-C*07:01, HLA-C*07:02, and / or HLA-C*14:02 molecules.

[0118] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-C*07:02 molecules, comprising or consisting of the sequence of SEQ ID NO: 30, 83, 85, 90, 109, 117, or 158, preferably SEQ ID NO: 109 or 117. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-B*27:02, HLA-C*07:01, HLA-C*07:02, and / or HLA-C*14:02 molecules.

[0119] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to HLA-C*08:02 molecules, comprising or consisting of the sequence of SEQ ID NO: 97, 102, or 174, preferably SEQ ID NO: 97 or 102. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAPs may further bind to HLA-C*03:03, HLA-C*03:04, HLA-C*05:01, HLA-C*08:01, and / or HLA-C*15:02 molecules.

[0120] In another aspect, the present disclosure provides a leukemia TAP (or tumor-specific peptide), preferably an AML TAP, that binds to an HLA-C*12:03 molecule, comprising or consisting of the sequence of SEQ ID NO: 173. Due to HLA allele promiscuity (particular HLA alleles present similar epitopes, see Table 4), the above identified TAP may further bind to HLA-B*46:01, HLA-C*03:02, HLA-C*03:03, HLA-C*03:04, HLA-C*08:01, HLA-C*12:03, HLA-C*15:02, and / or HLA-C*16:01 molecules.

[0121] In one embodiment, TAP is encoded by a sequence located in an untranslated transcribed region (UTR), i.e., the 3'-UTR or 5'-UTR region. In another embodiment, TAP is encoded by a sequence located in an intron. In another embodiment, TAP is encoded by a sequence located in an intergenic region. In another embodiment, TAP is encoded by a sequence located in an exon and results from a frameshift.

[0122] The TAP of the present disclosure can be produced by expression in a host cell containing a nucleic acid encoding TAP (recombinant expression) or by chemical synthesis (e.g., solid-phase peptide synthesis). Peptides can be readily synthesized by manual and / or automated solid-phase procedures well known in the art. Suitable synthesis can be carried out, for example, by utilizing the "T-boc" or "Fmoc" procedure. Techniques and procedures for solid-phase synthesis are described, for example, in "Solid Phase Peptide Synthesis: A Practical Approach," by E. Atherton and R.C. Sheppard, published by IRL, Oxford University Press, 1989. Alternatively, MiHA peptides can be prepared by segment condensation, as described, for example, in (Liu et al., Tetrahedron Lett. 37:933-936, 1996; Baca et al., J. Am. Chem. Soc. 117:1881-1887, 1995; Tam et al., Int. J. Peptide Protein Res. 45:209-216, 1995; Schnolzer and Kent, Science 256:221-225, 1992; Liu and Tam, J. Am. Chem. Soc. 116:4149-4153, 1994; Liu and Tam, Proc. Natl. Acad. Sci. USA 91:6584-6588, 1994; and Yamashiro and Li, Int. J. Peptide Protein Res. Res. 31:322-334, 1988). Other methods useful for synthesizing TAP are described in Nakagawa et al., J. Am. Chem. Soc. 107:7087-7092, 1985. In embodiments, TAP is chemically synthesized (synthetic peptide). Another embodiment of the present disclosure relates to non-naturally occurring peptides, which consist of or consist essentially of the amino acid sequences defined herein and are synthetically produced (e.g., synthesized) as pharmaceutically acceptable salts.The salts of TAP according to the present disclosure are substantially different from the peptides in their in vivo state, since the peptides produced in vivo do not have salts. Non-natural salt forms of the peptides can modulate the solubility of the peptides, particularly in the context of pharmaceutical compositions containing the peptides, such as the peptide vaccines disclosed herein. Preferably, the salts are pharmaceutically acceptable salts of the peptides.

[0123] In embodiments, the TAP described herein is substantially pure. A compound is "substantially pure" when it is separated from compounds that naturally accompany it. Typically, a compound is substantially pure when it is at least 60%, more usually 75%, 80%, or 85%, preferably greater than 90%, and more preferably greater than 95%, by weight of the total material in a sample. Thus, for example, a polypeptide chemically synthesized or produced by recombinant techniques will generally be substantially free of its naturally associated components, e.g., components of the macromolecule from which it is derived. A nucleic acid molecule is substantially pure when it is not immediately contiguous (i.e., not covalently linked) with coding sequences with which it is normally contiguous in the naturally occurring genome of the organism from which the nucleic acid is derived. A substantially pure compound can be obtained, for example, by extraction from a natural source, by expression of a recombinant nucleic acid molecule encoding the peptide compound, or by chemical synthesis. Purity can be measured using any suitable method, such as column chromatography, gel electrophoresis, HPLC, etc. In embodiments, the TAP is in solution. In another embodiment, the TAP is in a solid form, e.g., lyophilized.

[0124] In another aspect, the present disclosure further provides (isolated) nucleic acids encoding the TAP or tumor antigen precursor peptides described herein. In embodiments, the nucleic acids comprise from about 21 nucleotides to about 45 nucleotides, from about 24 to about 45 nucleotides, e.g., 24, 27, 30, 33, 36, 39, 42, or 45 nucleotides. "Isolated," as used herein, refers to a peptide or nucleic acid molecule that is separated from other components present in the molecule's natural environment or from macromolecules of the naturally occurring source (e.g., other nucleic acids, proteins, lipids, sugars, etc.). "Synthetic," as used herein, refers to a peptide or nucleic acid molecule that is not isolated from its natural source, e.g., produced through recombinant technology or using chemical synthesis. Nucleic acids of the present disclosure can be used for recombinant expression of a TAP of the present disclosure and can be contained within a vector or plasmid, such as a cloning vector or expression vector, that can be transfected into a host cell. In embodiments, the present disclosure provides a cloning vector, expression vector, or viral vector, or plasmid, that includes a nucleic acid sequence encoding a TAP of the present disclosure. Alternatively, a nucleic acid encoding a TAP of the present disclosure can be integrated into the genome of a host cell. In either case, the host cell expresses the TAP or protein encoded by the nucleic acid. As used herein, the term "host cell" refers not only to the particular subject cell but also to the progeny or potential progeny of such a cell. A host cell can be any prokaryotic cell (e.g., E. coli) or eukaryotic cell (e.g., insect cells, yeast, or mammalian cells) capable of expressing a TAP described herein. A vector or plasmid contains elements necessary for the transcription and translation of an inserted coding sequence and may contain other components, such as resistance genes, cloning sites, etc. Methods well known to those skilled in the art can be used to construct expression vectors containing a peptide or polypeptide coding sequence and appropriate transcriptional and translational control / regulatory elements operably linked thereto. These methods include in vitro recombinant DNA techniques, synthetic techniques, and in vivo genetic recombination.Such techniques are described in Sambrook et al. (1989) Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Press, Plainview, NY, and Ausubel, FM et al. (1989) Current Protocols in Molecular Biology, John Wiley & Sons, New York, NY. "Operably linked" refers to the juxtaposition of components, particularly components that enable the normal function of the nucleotide sequence to be carried out. Thus, a coding sequence operably linked to a regulatory sequence refers to a nucleotide sequence configuration in which the coding sequence can be expressed under the regulatory control, i.e., transcriptional and / or translational control, of the regulatory sequence. As used herein, "regulatory / control region" or "regulatory / control sequence" refers to a non-coding nucleotide sequence involved in regulating the expression of an encoding nucleic acid. Thus, the term regulatory region includes promoter sequences, regulatory protein binding sites, upstream activator sequences, and the like. Vectors (e.g., expression vectors) may contain necessary 5' upstream and 3' downstream regulatory elements for efficient gene transcription and translation in the respective host cells, such as promoter sequences (e.g., CMV, PGK, and EFla promoters), ribosome recognition and binding TATA boxes, and 3' UTR AAUAAA transcription termination sequences. Other suitable promoters include the constitutive promoters of the simian vims 40 (SV40) early promoter, mouse mammary tumor virus (MMTV) promoter, HIV LTR promoter, MoMuLV promoter, avian leukosis virus promoter, EBV immediate early promoter, and Rous sarcoma vims promoter. Human gene promoters may also be used, including, but not limited to, actin promoter, myosin promoter, hemoglobin promoter, and creatine kinase promoter. In certain embodiments, an inducible promoter is also contemplated as part of a vector expressing TAP. This provides a molecular switch that can turn on or off expression of a polynucleotide sequence of interest.Examples of inducible promoters include, but are not limited to, metallothionine promoters, glucocorticoid promoters, progesterone promoters, or tetracycline promoters. Examples of vectors include plasmids, autonomously replicating sequences, and transposable elements. Additional exemplary vectors include, but are not limited to, plasmids, phagemids, cosmids, artificial chromosomes (e.g., yeast artificial chromosomes (YACs)), bacterial artificial chromosomes (BACs), or P1-derived artificial chromosomes (PACs), bacteriophages (e.g., lambda phage or M13 phage), and animal viruses. Examples of animal virus categories useful as vectors include, but are not limited to, retroviruses (including lentiviruses), adenoviruses, adeno-associated viruses, herpesviruses (e.g., herpes simplex viruses), poxviruses, baculoviruses, papillomaviruses, and papovaviruses (e.g., SV40). Examples of expression vectors are the Lenti-X™ Bicistronic Expression System (Neo) vector (Clontrch), pClneo vector (Promega) for expression in mammalian cells, and pLenti4 / V5-DEST™, pLenti6 / V5-DEST™, and pLenti6.2N5-GW / lacZ (Invitrogen) for lentivirus-mediated gene transfer and expression in mammalian cells. The coding sequence of TAP disclosed herein can be ligated into such expression vectors for expression of TAP in mammalian cells.

[0125] In certain embodiments, the nucleic acid encoding the TAP of the present disclosure is provided in a viral vector. The viral vector may be derived from a retrovirus, lentivirus, or foamy virus. As used herein, the term "viral vector" refers to a nucleic acid vector construct that contains at least one element of viral origin and has the ability to be packaged into a viral vector particle. The viral vector may contain coding sequences for various proteins described herein in place of non-essential viral genes. The vectors and / or particles can be utilized to introduce DNA, RNA, or other nucleic acids into cells either in vitro or in vivo. Many forms of viral vectors are known in the art.

[0126] In embodiments, nucleic acids (DNA, RNA) encoding the TAP of the present disclosure are contained within a liposome or any other suitable vehicle.

[0127] In another aspect, the disclosure provides MHC class I molecules comprising (i.e., presenting or bound to) one or more TAPs of SEQ ID NOs: 1-190, preferably SEQ ID NOs: 97-154. In one embodiment, the MHC class I molecule is an HLA-A1 molecule, and in a further embodiment, an HLA-A*01:01 molecule. In another embodiment, the MHC class I molecule is an HLA-A2 molecule, and in a further embodiment, an HLA-A*02:01 molecule. In another embodiment, the MHC class I molecule is an HLA-A3 molecule, and in a further embodiment, an HLA-A*03:01 molecule. In another embodiment, the MHC class I molecule is an HLA-A11 molecule, and in a further embodiment, an HLA-A*11:01 molecule. In another embodiment, the MHC class I molecule is an HLA-A24 molecule, and in a further embodiment, an HLA-A*24:02 molecule. In another embodiment, the MHC class I molecule is an HLA-A26 molecule, and in a further embodiment, an HLA-A*26:01 molecule. In another embodiment, the MHC class I molecule is an HLA-A29 molecule, and in a further embodiment, an HLA-A*29:02 molecule. In another embodiment, the MHC class I molecule is an HLA-A30 molecule, and in a further embodiment, an HLA-A*30:01 molecule. In another embodiment, the MHC class I molecule is an HLA-A68 molecule, and in a further embodiment, an HLA-A*68:02 molecule. In another embodiment, the MHC class I molecule is an HLA-B07 molecule, and in a further embodiment, an HLA-B*07:02 molecule. In another embodiment, the MHC class I molecule is an HLA-B08 molecule, and in a further embodiment, an HLA-B*08:01 molecule. In another embodiment, the MHC class I molecule is an HLA-B14 molecule, and in a further embodiment, an HLA-B*14:01 molecule. In another embodiment, the MHC class I molecule is an HLA-B15 molecule, and in a further embodiment, an HLA-B*15:01 molecule. In another embodiment, the MHC class I molecule is an HLA-B27 molecule, and in a further embodiment, an HLA-B*27:05 molecule. In another embodiment, the MHC class I molecule is an HLA-B38 molecule, and in a further embodiment, an HLA-B*38:01 molecule.In another embodiment, the MHC class I molecule is an HLA-B40 molecule, and in a further embodiment, an HLA-B*40:01 molecule. In another embodiment, the MHC class I molecule is an HLA-B44 molecule, and in a further embodiment, an HLA-B*44:02 molecule or an HLA-B*44:03 molecule. In another embodiment, the MHC class I molecule is an HLA-B57 molecule, and in a further embodiment, an HLA-B*57:01 or an HLA-B*57:03 molecule. In another embodiment, the MHC class I molecule is an HLA-C03 molecule, and in a further embodiment, an HLA-C*03:03 molecule. In another embodiment, the MHC class I molecule is an HLA-C04 molecule, and in a further embodiment, an HLA-C*04:01 molecule. In another embodiment, the MHC class I molecule is an HLA-C05 molecule, and in a further embodiment, an HLA-C*05:01 molecule. In another embodiment, the MHC class I molecule is an HLA-C06 molecule, and in a further embodiment, an HLA-C*06:02 molecule. In another embodiment, the MHC class I molecule is an HLA-C07 molecule, and in a further embodiment, an HLA-C*07:01 or HLA-C*07:02 molecule. In another embodiment, the MHC class I molecule is an HLA-C08 molecule, and in a further embodiment, an HLA-C*08:02 molecule. In another embodiment, the MHC class I molecule is an HLA-C12 molecule, and in a further embodiment, an HLA-C*12:03 molecule.

[0128] In embodiments, TAP is non-covalently linked to an MHC class I molecule (i.e., TAP is loaded or non-covalently bound into the peptide-binding groove / pocket of the MHC class I molecule). In another embodiment, TAP is covalently linked / bound to an MHC class I molecule (alpha chain). In such constructs, TAP and an MHC class I molecule (alpha chain) are produced as a synthetic fusion protein, typically with a short (e.g., 5-20 residues, preferably about 8-12, e.g., 10) flexible linker or spacer (e.g., a polyglycine linker). In another aspect, the present disclosure provides nucleic acids encoding fusion proteins comprising a TAP as defined herein fused to an MHC class I molecule (alpha chain). In embodiments, the MHC class I molecule (alpha chain)-peptide complex is multimerized. Thus, in another aspect, the present disclosure provides multimers of MHC class I molecules loaded (covalently or non-covalently) with a TAP as described herein. Such multimers can be attached to a tag, e.g., a fluorescent tag, that allows for detection of the multimer. Numerous strategies have been developed for the production of MHC multimers, including MHC dimers, tetramers, pentamers, octamers, etc. (reviewed in Bakker and Schumacher, Current Opinion in Immunology 2005, 17:428-433). MHC multimers are useful, for example, for the detection and purification of antigen-specific T cells. Thus, in another aspect, the present disclosure provides a method for the production of CD8 MHC multimers specific for TAP as defined herein. + A method for detecting or purifying (isolating, enriching) T lymphocytes is provided, comprising contacting a cell population with a multimer of MHC class I molecules loaded (covalently or non-covalently) with TAP, and detecting CD8 bound by the MHC class I multimer. + and detecting or isolating CD8 T lymphocytes bound by MHC class I multimers. + T lymphocytes may be isolated using known methods, for example, fluorescence activated cell sorting (FACS) or magnetic activated cell sorting (MACS).

[0129] In yet another aspect, the disclosure provides cells (e.g., host cells), and in embodiments, isolated cells comprising a nucleic acid, vector, or plasmid of the disclosure (i.e., a nucleic acid or vector encoding one or more TAPs) described herein. In another aspect, the disclosure provides cells expressing an MHC class I molecule (e.g., an MHC class I molecule of one of the alleles disclosed above) bound to or presenting a TAP according to the disclosure. In one embodiment, the host cell is a eukaryotic cell, e.g., a mammalian cell, preferably a human cell, cell line, or immortalized cell. In another embodiment, the cell is an antigen-presenting cell (APC). In one embodiment, the host cell is a primary cell, cell line, or immortalized cell. In another embodiment, the cell is an antigen-presenting cell (APC). Nucleic acids and vectors can be introduced into cells via conventional transformation or transfection techniques. The terms "transformation" and "transfection" refer to techniques for introducing foreign nucleic acid into host cells, including calcium phosphate or calcium chloride co-precipitation, DEAE-dextran-mediated transfection, lipofection, electroporation, microinjection, and viral-mediated transfection. Suitable methods for transforming or transfecting host cells can be found, for example, in Sambrook et al. (supra), and other laboratory manuals. Methods for introducing nucleic acids into mammalian cells in vivo are also known and can be used to deliver the vectors or plasmids of the present disclosure to a subject for gene therapy.

[0130] Cells, such as APCs, can be loaded with one or more TAPs using various methods known in the art. As used herein, "loading cells" with TAP means that RNA or DNA encoding TAP or TAP is transfected into cells, or alternatively, APCs are transformed with a nucleic acid encoding TAP. Cells can also be loaded by contacting them with exogenous TAP, which can directly bind to MHC class I molecules present on the cell surface (e.g., cells pulsed with peptide). TAP can also be fused to a domain or motif (e.g., an endoplasmic reticulum (ER) retrieval signal, a C-terminal Lys-Asp-Glu-Leu sequence (see Wang et al., Eur J Immunol. 2004 Dec;34(12):3582-94)) that promotes its presentation by MHC class I molecules.

[0131] In another aspect, the present disclosure provides compositions or combinations / pools of peptides comprising any one or any combination of the TAPs (or nucleic acids encoding such peptides) defined herein. In embodiments, the compositions comprise any combination of the TAPs defined herein (any combination of 2, 3, 4, 5, 6, 7, 8, 9, 10, or more TAPs) or any combination of nucleic acids encoding such TAPs. Compositions comprising any combination / subcombination of the TAPs defined herein are encompassed by the present disclosure. In another embodiment, the combinations or pools may include one or more known tumor antigens.

[0132] Thus, in another aspect, the present disclosure provides compositions comprising any one or any combination of TAPs defined herein and cells expressing an MHC class I molecule (e.g., an MHC class I molecule of one of the alleles disclosed above). APCs for use in the present disclosure are not limited to a particular type of cell, and may be any cell type, including CD8 +This includes professional APCs such as dendritic cells (DCs), Langerhans cells, macrophages, and B cells, which are known to present proteinaceous antigens on their cell surface for recognition by T lymphocytes. For example, APCs can be obtained by inducing DCs from peripheral blood monocytes and then contacting (stimulating) them with TAPs either in vitro, ex vivo, or in vivo. APCs can also be activated to present TAPs in vivo (one or more of the TAPs disclosed herein are administered to a subject), and APCs that present TAPs are induced within a subject's body. The phrases "inducing APCs" or "stimulating APCs" include contacting or loading cells with one or more TAPs or nucleic acids encoding TAPs, resulting in the presentation of TAPs on the cell surface by MHC class I molecules. As described herein, according to the present disclosure, TAP can be indirectly loaded using, for example, a longer peptide / polypeptide (including a natural protein) containing the sequence of TAP, which is then processed (e.g., by a protease) inside the APC to generate a TAP / MHC class I complex on the surface of the cell. After loading the APC with TAP and allowing the APC to present the TAP, the APC can be administered to a subject as a vaccine. For example, ex vivo administration can include: (a) collecting APCs from a first subject; (b) contacting / loading the APCs of step (a) with TAP to form an MHC class I / TAP complex on the surface of the APCs; and (c) administering the peptide-loaded APCs to a second subject in need of treatment.

[0133] The first subject and the second subject may be the same subject (e.g., an autologous vaccine) or different subjects (e.g., an allogeneic vaccine). Alternatively, the present disclosure provides use of a TAP (or a combination thereof) described herein for manufacturing a composition (e.g., a pharmaceutical composition) for inducing antigen-presenting cells. In addition, the present disclosure provides a method or process for manufacturing a pharmaceutical composition for inducing antigen-presenting cells, the method or process comprising mixing or combining a TAP, or a combination thereof, with a pharmaceutically acceptable carrier. Cells such as APCs expressing MHC class I molecules (e.g., HLA-A1, HLA-A2, HLA-A3, HLA-A11, HLA-A24, HLA-A25, HLA-A29, HLA-A32, HLA-B07, HLA-B08, HLA-B14, HLA-B15, HLA-B18, HLA-B39, HLA-B40, HLA-B44, HLA-C03, HLA-C04, HLA-C05, HLA-C06, HLA-C07, HLA-C12, or HLA-C14 molecules) loaded with any one, or any combination, of the TAPs defined herein, are expressed as CD8 + T lymphocytes (e.g., autologous CD8 + Thus, in another aspect, the present disclosure relates to a method for stimulating / expanding T lymphocytes, more specifically CD8 T lymphocytes, by using a TAP (or a nucleic acid or vector encoding the same) as defined herein, a cell expressing an MHC class I molecule, and a T lymphocyte, more specifically CD8 T lymphocytes. + T lymphocytes (e.g., CD8 + a cell population comprising T lymphocytes), or any combination thereof.

[0134] In embodiments, the composition further comprises a buffer, excipient, carrier, diluent, and / or medium (e.g., culture medium). In further embodiments, the buffer, excipient, carrier, diluent, and / or medium is a pharmaceutically acceptable buffer, excipient, carrier, diluent, and / or medium. As used herein, "pharmaceutically acceptable buffer, excipient, carrier, diluent, and / or medium" includes any and all solvents, buffers, binders, lubricants, fillers, thickeners, disintegrants, plasticizers, coatings, barrier layer formulations, lubricants, stabilizers, release retardants, dispersion media, coatings, antibacterial and antifungal agents, isotonicity agents, and the like that are physiologically compatible, do not interfere with the effectiveness of the biological activity of the active ingredient, and are not toxic to the subject. The use of such media and agents for pharmaceutically active substances is well known in the art (Rowe et al., Handbook of Pharmaceutical Excipients, 2003, 4). th edition, Pharmaceutical Press, London UK). Use of any conventional media or agent in the compositions of the present disclosure is contemplated, except to the extent that such media or agents are incompatible with the active compounds (peptides, cells). In embodiments, the buffer, excipient, carrier, and / or medium is a non-naturally occurring buffer, excipient, carrier, and / or medium. In embodiments, one or more of the TAPs defined herein, or nucleic acids (e.g., mRNAs) encoding the one or more TAPs, are contained within or complexed with a liposome (e.g., a cationic liposome) or other suitable carrier (see, e.g., Vitor MT et al., Recent Pat Drug Deliv Formul. 2013 Aug;7(2):99-110).

[0135] In another aspect, the present disclosure provides a composition comprising any one or any combination of TAPs (or nucleic acids encoding such peptides) defined herein and one or more of a buffer, excipient, carrier, diluent, and / or medium. For compositions comprising cells (e.g., APCs, T lymphocytes), the composition comprises a suitable medium that allows for the maintenance of viable cells. Representative examples of such media include saline, Earl's Buffered Salt Solution (Life Technologies®), or PlasmaLyte® (Baxter International®). In embodiments, the composition (e.g., pharmaceutical composition) is an "immunogenic composition," "vaccine composition," or "vaccine." As used herein, the terms "immunogenic composition," "vaccine composition," or "vaccine" refer to a composition or formulation that comprises one or more TAPs or vaccine vectors and that, when administered to a subject, is capable of eliciting an immune response against one or more TAPs present therein. The use of vaccines or vaccine vectors to induce an immune response in a mammal includes vaccines or vaccine vectors administered by any conventional route known in the vaccine art, for example, via a mucosal (e.g., ocular, intranasal, pulmonary, oral, gastric, intestinal, rectal, vaginal, or urinary) surface, via a parenteral (e.g., subcutaneous, intradermal, intramuscular, intravenous, or intraperitoneal) route, or by topical administration (e.g., via a transdermal delivery system such as a patch). In embodiments, TAP (or a combination thereof) is conjugated to a carrier protein (conjugate vaccine) to increase the immunogenicity of TAP. Accordingly, the present disclosure provides compositions (conjugates) comprising TAP (or a combination thereof) or a nucleic acid encoding TAP (or a combination thereof) and a carrier protein. For example, TAP can be conjugated or complexed with Toll-like receptor (TLR) ligands (see, e.g., Zom et al., Adv Immunol. 2012, 114:177-201) or polymers / dendrimers (see, e.g., Liu et al., Biomacromolecules. 2013 Aug12;14(8):2798-806).In embodiments, the immunogenic composition or vaccine further comprises an adjuvant. An "adjuvant" refers to a substance that, when added to an immunogenic agent, such as an antigen (TAP, nucleic acid, and / or cell according to the present disclosure), non-specifically enhances or potentiates the immune response in a host to the agent upon exposure to the mixture. Examples of adjuvants currently used in the field of vaccines include: (1) mineral salts (aluminum salts such as aluminum phosphate and aluminum hydroxide, calcium phosphate gel), squalene; (2) oil-based adjuvants (oil emulsions and surfactant-based formulations), such as MF59 (microfluidized detergent-stabilized oil-in-water emulsion), QS21 (purified saponin), and AS02 [SBAS2] (oil-in-water emulsion + MPL + QS-21); (3) particulate adjuvants, such as virosomes (unilamellar liposomal vehicles incorporating influenza hemagglutinin), AS04 (aluminum salt of [SBAS4] containing MPL), ISCOMS (structural complexes of saponin and lipids), and polylactide-co-glycolide (PLG); and (4) microbial derivatives (tenta). (4) endogenous human immunomodulators, such as human GM-CSF or human IL-12 (cytokines that can be administered either as proteins or as encoded plasmids), Immudaptin (C3d tandem arrays), and / or (5) inert vehicles, such as monophosphoryl lipid A (MPL), Detox (MPL plus M. Phlei cell wall skeleton), AGP [RC-529] (synthetic acylated monosaccharide), DC_Chol (lipidic immunostimulator that can self-assemble into liposomes), OM-174 (a lipid A derivative), CpG motifs (synthetic oligonucleotides containing immunostimulatory CpG motifs), modified LT and CT (bacterial toxin immunoglobulins genetically engineered to provide a non-toxic adjuvant effect), (6) endogenous human immunomodulators, such as human GM-CSF or human IL-12 (cytokines that can be administered either as proteins or as encoded plasmids), Immudaptin (C3d tandem arrays), and / or (7) inert vehicles, such as gold particles.

[0136] In one embodiment, TAP or a composition comprising it is in lyophilized form. In another embodiment, TAP or a composition comprising it is a liquid composition. In a further embodiment, TAP is present in the composition at a concentration of about 0.01 μg / mL to about 100 μg / mL. In yet another embodiment, TAP is present in the composition at a concentration of about 0.2 μg / mL to about 50 μg / mL, about 0.5 μg / mL to about 10, 20, 30, 40, or 50 μg / mL, about 1 μg / mL to about 10 μg / mL, or about 2 μg / mL.

[0137] As described herein, cells such as APCs expressing MHC class I molecules loaded with or bound to any one of the TAPs defined herein, or any combination thereof, can be used to induce CD8 + These molecules can be used to stimulate / proliferate T lymphocytes. Thus, in another aspect, the present disclosure provides T cell receptor (TCR) molecules capable of interacting with or binding to the MHC class I molecule / TAP complexes described herein, as well as nucleic acid molecules encoding such TCR molecules, and vectors comprising such nucleic acid molecules. TCRs according to the present disclosure can specifically interact with or bind to TAP loaded on or presented by MHC class I molecules, preferably on the surface of living cells in vitro or in vivo.

[0138] In embodiments, an anti-leukemia (e.g., anti-AML) TCR according to the present disclosure comprises a TCR beta (β) chain comprising a complementarity determining region 3 (CDR3) comprising one of the amino acid sequences set forth in SEQ ID NOs: 191-219.

[0139] In embodiments, the TCR is specific for one or more of the following TAPs: SLLSGLLRA, ALPVALPSL, ALDPLLLRI, IASPIALL, and / or SLDLLPLSI, and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 191-199. In embodiments, the TCR is specific for the TAP SLLSGLLRA and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 191-199. In embodiments, the TCR is specific for the TAP ALPVALPSL and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 191-199. In embodiments, the TCR is specific for the TAP ALDPLLLRI and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 191-199. In embodiments, the TCR is specific for TAP IASPIALL and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 191 to 199. In embodiments, the TCR is specific for TAP SLDLLPLSI and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 191 to 199.

[0140] In another embodiment, the TCR is specific for one or more of the following TAPs: LTDRIYLTL, VLFGGKVSGA, LGISLTLKY, FNVALNARY, and / or TLNQGINVYI and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 200-209. In embodiments, the TCR is specific for TAP LTDRIYLTL and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 200-209. In embodiments, the TCR is specific for TAP VLFGGKVSGA and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 200-209. In embodiments, the TCR is specific for TAP LGISLTLKY and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 200-209. In embodiments, the TCR is specific for TAP FNVALNARY and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 200 to 209. In embodiments, the TCR is specific for TAP TLNQGINVYI and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 200 to 209.

[0141] In another embodiment, the TCR is specific for one or more of the following TAPs: LRSQILSY, KILDVNLRI, HSLISIVYL, KLQDKEIGL, and / or AQDIILQAV and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 210-219. In embodiments, the TCR is specific for TAP LRSQILSY and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 210-219. In embodiments, the TCR is specific for TAP KILDVNLRI and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 210-219. In embodiments, the TCR is specific for TAP HSLISIVYL and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 210-219. In embodiments, the TCR is specific for TAP KLQDKEIGL and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 210 to 219. In embodiments, the TCR is specific for TAP AQDIILQAV and comprises a TCR β chain comprising a CDR3 comprising one of the amino acid sequences set forth in SEQ ID NOs: 210 to 219.

[0142] In embodiments, a TCR according to the present disclosure recognizes one or more of the above-mentioned TAPs bound to HLA-A*02:01, HLA-A*29:02, HLA-B*15:01, HLA-B27:05, HLA-C*01:02, and / or HLA-C*03:04 molecules. In embodiments, a TCR according to the present disclosure recognizes one or more of the above-mentioned TAPs bound to HLA-A*02:01 molecules. In embodiments, a TCR according to the present disclosure recognizes one or more of the above-mentioned TAPs bound to HLA-A*29:02 molecules. In embodiments, a TCR according to the present disclosure recognizes one or more of the above-mentioned TAPs bound to HLA-B*15:01 molecules. In embodiments, a TCR according to the present disclosure recognizes one or more of the above-mentioned TAPs bound to HLA-B27:05 molecules. In embodiments, a TCR according to the present disclosure recognizes one or more of the above-mentioned TAPs bound to an HLA-C*01:02 molecule. In embodiments, a TCR according to the present disclosure recognizes one or more of the above-mentioned TAPs bound to an HLA-C*03:04 molecule.

[0143] The term TCR, as used herein, refers to a member of the immunoglobulin superfamily that has a variable binding domain, a constant domain, a transmembrane region, and a short cytoplasmic tail (see, e.g., Janeway et al., Immunobiology: The Immune System in Health and Disease, 3rd Ed., Current Biology Publications, p. 4:33, 1997), and is capable of specifically binding to an antigenic peptide bound to an MHC receptor. TCRs can be found on the surface of cells and generally consist of a heterodimer having an α chain and a β chain (also known as TCRα and TCRβ, respectively). Similar to immunoglobulins, the extracellular portion of a TCR chain (e.g., α chain, β chain) comprises two immunoglobulin regions: a variable region (e.g., a TCR variable α region or Vα, and a TCR variable β region or Vβ; typically, amino acids 1-116 according to Rabat numbering at the N-terminus) and one constant region adjacent to the cell membrane (e.g., a TCR constant domain α or Cα, typically, amino acids 117-259 according to Rabat; a TCR constant domain β or Cβ, typically, amino acids 117-295 according to Rabat). Also similar to immunoglobulins, the variable domains comprise complementarity-determining regions (CDR3 in each chain) separated by framework regions (FR). In certain embodiments, TCRs are found on the surface of T cells (or T lymphocytes) and associate with the CD3 complex.

[0144] TCRs, and specifically nucleic acids encoding TCRs of the present disclosure, can be applied to, for example, T lymphocytes (e.g., CD8 + T lymphocytes) or other types of lymphocytes can be genetically transformed / modified to generate novel T lymphocyte clones that specifically recognize the MHC class I / TAP complex. In certain embodiments, T lymphocytes (e.g., CD8 + T lymphocytes (e.g., CD8 T lymphocytes) are transformed to express one or more TCRs that recognize TAP, and the transformed cells are administered to the patient (autologous cell transfusion). In certain embodiments, T lymphocytes (e.g., CD8 T lymphocytes) obtained from a donor are transfected with TCRs that recognize TAP. +In another embodiment, the present disclosure provides a method for administering T lymphocytes (e.g., CD8 T lymphocytes) transformed / transfected with a vector or plasmid encoding a TAP-specific TCR to a recipient (allogeneic cell transfusion). + In a further embodiment, the present disclosure provides a method of treating a patient with autologous or allogeneic cells transformed with a TAP-specific TCR. In certain embodiments, the TCR is expressed in primary T cells (e.g., cytotoxic T cells) by replacing the endogenous locus (e.g., the endogenous TRAC and / or TRBC locus) using, for example, CRISPR, TALEN, zinc finger, or other targeted disruption systems.

[0145] In another embodiment, the present disclosure provides a nucleic acid encoding the TCR described above. In a further embodiment, the nucleic acid is present in a vector, such as the vector described above.

[0146] In a still further embodiment, there is provided the use of a tumor antigen-specific TCR in the production of autologous or allogeneic cells for the treatment of cancer (leukemia, e.g., AML).

[0147] In some embodiments, patients treated with the compositions (e.g., pharmaceutical compositions) of the present disclosure are treated before or after treatment with allogeneic stem cell transplantation (ASCL), allogeneic lymphocyte infusion, or autologous lymphocyte infusion. The compositions of the present disclosure include allogeneic T lymphocytes (e.g., CD8) activated ex vivo against TAP. + T lymphocytes), TAP-loaded allogeneic or autologous APC vaccines, TAP vaccines, and allogeneic or autologous T lymphocytes (e.g., CD8 +These include CD8 T lymphocytes (e.g., T lymphocytes), or lymphocytes transformed with a tumor antigen-specific TCR. Methods for providing T lymphocyte clones capable of recognizing TAP according to the present disclosure can be generated for and specifically target tumor cells expressing TAP in a subject (e.g., a transplant recipient), e.g., an ASCT and / or donor lymphocyte infusion (DLI) recipient. Thus, the present disclosure provides CD8 T lymphocytes that encode and express a T cell receptor that can specifically recognize or bind to a TAP / MHC class I molecule complex. + Providing T lymphocytes. The T lymphocytes (e.g., CD8 + The CD8 T lymphocytes may be recombinant (engineered) or naturally selected T lymphocytes. + At least two methods for producing T lymphocytes are provided, which involve contacting undifferentiated lymphocytes with TAP / MHC class I molecule complexes (typically expressed on the surface of cells such as APCs) under conditions conducive to T cell activation and induction of T cell proliferation, which can be done in vitro or in vivo (i.e., in patients administered an APC vaccine in which APCs are loaded with TAP, or in patients treated with a TAP vaccine). Combinations or pools of TAPs bound to MHC class I molecules can be used to generate CD8 T lymphocytes that can recognize multiple TAPs. + Alternatively, tumor antigen-specific or target T lymphocytes can be generated by integrating MHC class I molecule / TAP complexes (i.e., engineered or recombinant CD8 +TAP-specific TCRs can be produced / generated in vitro or ex vivo by cloning one or more nucleic acids (genes) encoding TCRs (more specifically, alpha and beta chains) that specifically bind to TAP (e.g., T lymphocytes). Nucleic acids encoding TAP-specific TCRs of the present disclosure can be obtained from T lymphocytes activated against TAP ex vivo (e.g., by APCs loaded with TAP) or from individuals that exhibit an immune response to peptide / MHC molecule complexes, using methods known in the art. TAP-specific TCRs of the present disclosure can be recombinantly expressed in host cells and / or host lymphocytes obtained from the graft recipient or graft donor and, optionally, differentiated in vitro to provide cytotoxic T lymphocytes (CTLs). Nucleic acids (transgenes) encoding the TCR alpha and beta chains can be introduced into T cells (e.g., from the subject to be treated or another individual) using any suitable method, such as transfection (e.g., electroporation) or transduction (e.g., use of viral vectors) (e.g., calcium phosphate-DNA co-precipitation, DEAE-dextran mediated transfection, polybrene mediated transfection, electroporation, microinjection, liposome fusion, lipofection, protoplast fusion, retroviral infection, biolistics). Engineered CD8 T cells expressing a TCR specific for TAP can be used. + T lymphocytes can be expanded in vitro using well-known culture methods.

[0148] The present disclosure provides methods for generating immune effector cells that express a TCR described herein. In one embodiment, the method comprises transfecting or transducing immune effector cells (e.g., immune effector cells isolated from a subject, such as a subject with leukemia (e.g., AML)) such that the immune effector cells express one or more TCRs described herein. In certain embodiments, immune effector cells are isolated from an individual and genetically modified without further in vitro manipulation. Such cells can then be directly readministered to the individual. In further embodiments, immune effector cells are first activated and stimulated to proliferate in vitro before being genetically modified to express a TCR. In this regard, the immune effector cells can be cultured before or after being genetically modified (i.e., transduced or transfected to express a TCR described herein).

[0149] Prior to in vitro manipulation or genetic modification of immune effector cells described herein, a cell source can be obtained from a subject. In particular, immune effector cells for use with the TCRs described herein include T cells. T cells can be obtained from several sources, including peripheral blood mononuclear cells (PBMCs), bone marrow, lymph node tissue, umbilical cord blood, thymus tissue, tissue from an infection site, ascites, pleural effusion, spleen tissue, and tumors. In certain embodiments, T cells can be obtained from a unit of blood drawn from a subject using any number of techniques known to those skilled in the art, such as FICOLL™ separation. In one embodiment, cells from an individual's circulating blood can be obtained by apheresis. The apheresis product typically contains lymphocytes, including T cells, monocytes, granulocytes, B cells, other nucleated white blood cells, red blood cells, and platelets. In one embodiment, cells collected by apheresis can be washed to remove the plasma fraction and place the cells in an appropriate buffer or medium for subsequent processing. In one embodiment, the cells are washed with PBS. In alternative embodiments, the wash solution may lack calcium, magnesium, or many, but not all, divalent cations. As will be appreciated by those skilled in the art, the wash step can be accomplished by methods known to those skilled in the art, for example, by using a semi-automated flow-through centrifuge. After washing, the cells can be resuspended in various biocompatible buffers or other saline solutions, with or without buffers. In certain embodiments, undesirable components of the apheresis sample can be removed by resuspending the cells directly in the medium. In certain embodiments, T cells are isolated from peripheral blood mononuclear cells (PBMCs) by lysing red blood cells and depleting monocytes (e.g., by centrifugation through a PERCOLL™ gradient). Specific subpopulations of T cells, such as CD28+, CD4+, CD8+, CD45RA+, and CD45RO+ T cells, can be further isolated by positive or negative selection techniques. For example, enrichment of T cell populations by negative selection can be achieved by a combination of antibodies directed against surface markers unique to the negatively selected cells.One method for use herein is cell sorting and / or selection by negative magnetic immunoadhesion or flow cytometry, which uses a cocktail of monoclonal antibodies directed against cell surface markers present on the negatively selected cells. For example, to enrich for CD8+ cells by negative selection, the monoclonal antibody cocktail typically includes antibodies against CD14, CD20, CD11b, CD16, HLA-DR, and CD4. Flow cytometry and cell sorting can also be used to isolate cell populations of interest for use in the present disclosure. PBMCs can be used directly for TCR genetic modification using the methods described herein. In certain embodiments, after isolation of PBMCs, T lymphocytes are further isolated, and in certain embodiments, both cytotoxic and helper T lymphocytes can be sorted into naive, memory, and effector T cell subpopulations, either before or after genetic modification and / or expansion.

[0150] The present disclosure provides isolated immune cells (e.g., CD8 + The present disclosure also provides a method for detecting a TAP or a combination thereof (i.e., one or more TAPs bound to an MHC class I molecule) according to the present disclosure and a CD8 T lymphocyte capable of recognizing the TAP. + In another aspect, the present disclosure provides a composition comprising a CD8 T lymphocyte that specifically recognizes one or more MHC class I molecule / TAP complexes described herein. + Cell populations or cell cultures enriched in T lymphocytes (e.g., CD8 + Such enriched populations can be obtained by ex vivo expansion of specific T lymphocytes using cells such as APCs expressing MHC class I molecules that are loaded with (e.g., present) one or more of the TAPs disclosed herein. As used herein, "enriched" refers to the proliferation of tumor antigen-specific CD8 T lymphocytes in a population.+ In a further embodiment, the percentage of TAP-specific CD8 T lymphocytes in the cell population is significantly higher than in the native population of cells, i.e., a population that has not been subjected to the step of ex vivo expansion of the specific T lymphocytes. + In some embodiments, the proportion of T lymphocytes is at least about 0.5%, e.g., at least about 1%, 1.5%, 2%, or 3%. In some embodiments, the proportion of TAP-specific CD8 + The proportion of T lymphocytes is about 0.5 to about 10%, about 0.5 to about 8%, about 0.5 to about 5%, about 0.5 to about 4%, about 0.5 to about 3%, about 1% to about 5%, about 1% to about 4%, about 1% to about 3%, about 2% to about 5%, about 2% to about 4%, about 2% to about 3%, about 3% to about 5%, or about 3% to about 4%. CD8 T lymphocytes that specifically recognize one or more MHC class I molecule / peptide (TAP) complexes of interest are also included. + Such cell populations or cultures enriched for T lymphocytes (e.g., CD8 + T lymphocyte populations) can be used in tumor antigen-based cancer immunotherapy, as described in detail below. In some embodiments, TAP-specific CD8 + The population of T lymphocytes can be further enriched using affinity-based systems, such as multimers of MHC class I molecules loaded (covalently or non-covalently) with TAP as defined herein. Thus, the present disclosure provides methods for the detection of TAP-specific CD8 + A purified or isolated population of T lymphocytes is provided, e.g., TAP-specific CD8 + The proportion of T lymphocytes is at least about 50%, 60%, 70%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, or 100%.

[0151] The present disclosure further provides a method for producing the above-mentioned immune cells (CD8 + T lymphocytes) or TAP-specific CD8 + The present invention relates to a pharmaceutical composition or vaccine comprising a population of T lymphocytes. Such a pharmaceutical composition or vaccine may comprise one or more pharmaceutically acceptable excipients and / or adjuvants, as described above.

[0152] The present disclosure further relates to the use of any TAP, nucleic acid, expression vector, T cell receptor, cell (e.g., T lymphocyte, APC), and / or composition according to the present disclosure, or any combination thereof, as a medicament or in the manufacture of a medicament. In embodiments, the medicament is for the treatment of cancer, e.g., a cancer vaccine. The present disclosure relates to any TAP, nucleic acid, expression vector, T cell receptor, cell (e.g., T lymphocyte, APC), and / or composition according to the present disclosure (e.g., vaccine composition), or any combination thereof, for use in the treatment of cancer, e.g., as a cancer vaccine. The TAP sequences identified herein can be used for the production of synthetic peptides: i) used for in vitro stimulation and expansion of tumor antigen-specific T cells to be injected into tumor patients, and / or ii) used as vaccines to induce or enhance anti-tumor T cell responses in cancer patients.

[0153] In another aspect, the present disclosure provides for the use of a TAP (SEQ ID NOS: 1-190, preferably SEQ ID NOS: 97-154), or combinations thereof (e.g., peptide pools), as described herein, as a vaccine for treating cancer in a subject. The present disclosure also provides a TAP, or combinations thereof (e.g., peptide pools), as described herein, for use as a vaccine for treating cancer in a subject. In embodiments, the subject is administered TAP-specific CD8 + Thus, in another aspect, the present disclosure provides a method of treating cancer (e.g., reducing the number of tumor cells, killing tumor cells) by administering to a subject in need thereof an effective amount of CD8 T lymphocytes that recognize (i.e., express a TCR that binds to) one or more MHC class I molecule / TAP complexes (expressed on the surface of cells such as APCs). + In an embodiment, the method comprises administering (injecting) the CD8 T lymphocytes. +After administration / infusion of T lymphocytes, the method further comprises administering to the subject an effective amount of a TAP or a combination thereof, and / or cells (e.g., APCs such as dendritic cells) expressing MHC class I molecules loaded with TAPs. In yet a further embodiment, the method comprises administering to a subject in need thereof a therapeutically effective amount of dendritic cells loaded with one or more TAPs. In yet a further embodiment, the method comprises administering to a patient in need thereof a therapeutically effective amount of allogeneic or autologous cells expressing a recombinant TCR that binds to a TAP presented by an MHC class I molecule.

[0154] In another aspect, the present disclosure provides a method for treating cancer in a subject (e.g., reducing the number of tumor cells, killing tumor cells) by administering to a subject a CD8+ antibody that recognizes one or more MHC class I molecules and that is loaded with (presents) TAP or a combination thereof. + In another aspect, the present disclosure provides a use of CD8 T lymphocytes that recognize one or more MHC class I molecules and are loaded with (presenting) TAP or a combination thereof for the preparation / manufacture of a medicament for treating cancer in a subject (e.g., for reducing the number of tumor cells and killing tumor cells). + In another aspect, the present disclosure provides a method for treating cancer in a subject (e.g., reducing the number of tumor cells, killing tumor cells), comprising administering to a subject a CD8 T lymphocyte that recognizes one or more MHC class I molecules and that is loaded with (presents) TAP or a combination thereof. + In a further embodiment, the use provides TAP-specific CD8 + After the use of T lymphocytes, the method further includes the use of an effective amount of TAP (or a combination thereof) and / or cells (e.g., APCs) expressing one or more MHC class I molecules that are loaded with (present) TAP.

[0155] The present disclosure also provides a method of generating an immune response in a subject against tumor cells (leukemia cells, AML cells) expressing human class I MHC molecules loaded with any of the TAPs disclosed herein or a combination thereof, comprising administering cytotoxic T lymphocytes that specifically recognize class I MHC molecules loaded with a TAP or a combination of TAPs. The present disclosure also provides the use of cytotoxic T lymphocytes that specifically recognize class I MHC molecules loaded with any of the TAPs or a combination of TAPs disclosed herein to generate an immune response against tumor cells expressing human class I MHC molecules loaded with a TAP or a combination thereof.

[0156] In embodiments, the methods or uses described herein further comprise, prior to the treatment / use, determining the HLA class I alleles expressed by the patient and administering or using a TAP that binds to one or more of the HLA class I alleles expressed by the patient. For example, if the patient is determined to express HLA-A1*01 and HLA-C05*01, then any combination of TAPs of (i) SEQ ID NOs: 48, 67, 89, 134, 151, and / or 164 (which bind to HLA-A1*01), and / or SEQ ID NO: 150 (which binds to HLA-C05*01) may be administered or used to the patient.

[0157] In embodiments, the cancer is a blood cancer, preferably a leukemia, such as acute lymphocytic leukemia (ALL), acute myeloid leukemia (AML), chronic myeloid leukemia (CML), hairy cell leukemia (HCL), and myelodysplastic syndrome (MDS). In embodiments, the leukemia is AML. AML treated by the methods and uses described herein can be any type or subtype of AML (e.g., low-risk, intermediate-risk, or high-risk AML), including AML with genetic abnormalities such as AML with a translocation between chromosome 8 and chromosome 21 [t(8;21)], AML with a translocation or inversion in chromosome 16 [t(16;16) or inv(16)], AML with the PML-RARA fusion gene, AML with a translocation between chromosome 9 and chromosome 11 [t(9;11)], AML with a translocation between chromosome 6 and chromosome 9 [t(6:9)], AML with a translocation in chromosome 3 [t(16;16) or inv(16)], or AML with a translocation in chromosome 4 [t(16;16) or inv(16)]. or AML with an inversion [t(3;3) or inv(3)], AML (megakaryocyte) with a translocation [t(1:22)] between chromosome 1 and chromosome 22, AML with the BCR-ABL1 (BCR-ABL) fusion gene, AML with a mutated NPM1 gene, AML with biallelic mutations in the CEBPA gene, AML with a mutated RUNX1 gene, AML with a mutated ASX1 gene, AML with mutated IDH1 and / or IDH2 genes, AML with a mutated FLT3 gene, AML with myelodysplasia-related changes, and AML associated with previous chemotherapy or radiation.

[0158] In embodiments, the TAP, nucleic acid, expression vector, T cell receptor, cell (e.g., T lymphocyte, APC), and / or composition according to the present disclosure, or any combination thereof, may be administered in combination with one or more additional active agents or therapies, for example, chemotherapy (e.g., vinca alkaloids, agents that interfere with microtubule formation (e.g., colchicine and its derivatives), anti-angiogenic agents, therapeutic antibodies, EGFR targeting agents, tyrosine kinase targeting agents (e.g., tyrosine kinase inhibitors), transition metal complexes, proteasome inhibitors, antimetabolites (e.g., nucleoside analogs), alkylating agents, platinum-based agents, anthracycline antibiotics, topoisomerase inhibitors, macrophages, and the like, to treat cancer. These include: steroids, retinoids (e.g., all-trans retinoic acid or its derivatives), geldanamycin or its derivatives (17-AAG), surgery, radiation therapy, immune checkpoint inhibitors (immunotherapeutics), immune checkpoint inhibitors (immunotherapeutics (e.g., PD-1 / PD-L1 inhibitors such as anti-PD-1 / PD-L1 antibodies, CTLA-4 inhibitors such as anti-CTLA-4 antibodies, B7-1 / B7-2 inhibitors such as anti-B7-1 / B7-2 antibodies, TIM3 inhibitors such as anti-TIM3 antibodies, BTLA inhibitors such as anti-BTLA antibodies, CD47 inhibitors such as anti-CD47 antibodies, GITR inhibitors such as anti-GITR antibodies), antibodies against tumor antigens, cell-based therapies (e.g., CAR In embodiments, the TAP, nucleic acid, expression vector, T cell receptor, cell (e.g., T lymphocyte, APC), and / or composition according to the present disclosure is administered / used in combination with one or more chemotherapeutic agents used to treat AML, or in combination with other AML therapies (e.g., stem cell / bone marrow transplant).

[0159] The additional therapy may be administered before, simultaneously with, or after administration of the TAP, nucleic acid, expression vector, T cell receptor, cell (e.g., T lymphocyte, APC), and / or composition according to the present disclosure.

[0160] Modes for Carrying Out the Invention The present invention is illustrated in further detail by the following non-limiting examples.

[0161] Example 1: Materials and Methods AML specimens Diagnostic AML samples (cryovials of DMSO-frozen leukemic blasts) were obtained from the Banque de cellules leucemiques du Quebec program (BCLQ, bclq.org). Table 1 provides technical and clinical characteristics of the samples. 100 million cells of each AML sample (except 14H124, see section below) were thawed (1 minute in a 37°C water bath) and resuspended in 48 ml of 4°C PBS. Two million cells (1 ml) were pelleted and resuspended in 1 ml of Trizol for RNA sequencing, and the remaining 98 million were pelleted and flash-frozen in liquid nitrogen for mass spectrometry analysis. [Table 3-1] [Table 3-2] [Table 3-3] [Table 3-4] NC: Unclassifiable by FAB criteria. HLA was determined by Optitype based on RNA-Seq data for each sample. Clinical data were provided by the Banque de cellules leucemiques du Quebec program (BCLQ, bclq.org).

[0162] Other Data Sources Human mTEC samples have been prepared and sequenced for our team's previous research needs (#GSE127825 and #GSE127826) (Larouche et al., 2020; Laumont et al., 2018) or published by other teams (E-MTAB-7383) (Fergusson et al., 2018). Only six mTEC samples previously used by our group for TSA discovery have been used in the k-mer depletion approach (Laumont et al., 2018). Eleven MPC samples used as primary normal controls were sequenced by the IRIC genome platform and previously published by the Leucegene group (#GSE98310, #GSE51984). All other normal samples used in this study were downloaded from dbGap (www.ncbi.nlm.nih.gov / gap / ), Arrayexpress (www.ebi.ac.uk / arrayexpress / ), or GEO (www.ncbi.nlm.nih.gov / geo / ). Using Leucegene's entire cohort of 437 RNA-sequenced AML samples, we found that TSA 高 The clinical significance of this data was studied. RNA sequencing data have been previously published and are available separately (#GSE49642, #GSE52656, #GSE62190, #GSE66917, #GSE67039) (Lavallee et al., 2015; Macrae et al., 2013; Pabst et al., 2016). Curated LSC and BLAST RNA-Seq data have been published elsewhere and were obtained from #GSE74246 (Corces et al., 2016). RNA-Seq data for pre- and post-relapse AML blasts (matched samples) have been published elsewhere (Toffalori et al., 2019), and the HLA typing of these samples was kindly provided by Dr. Luca Vago. All data obtained from external sources were aligned to the GRCh38 genome using STAR v2.5.1b.

[0163] Growth of 14H124 AML cells in NSG mice Because only 20 million cells were available for this patient, blast cells from patient 14H124 were thawed, washed with PBS, and then transferred 24 hours after sublethal total body irradiation (2.5 Gy, 137Cs-γ source) to 10 NOD-scid IL-2Rγ null (NSG) mice (2 × 10 6 The cells were intravenously injected into 122 mice (100 μl / mouse). Human AML cell engraftment was assessed in peripheral blood on day 122 by flow cytometry. Briefly, 100 μl of blood was collected by tail vein bleeding, depleted of red blood cells using RBC lysis buffer (eBioscience), washed with staining buffer (PBS + 3% FBS), stained with anti-human CD45-Pacific Blue (HI30, Biolegend) and anti-mouse CD45-PECy5 (30-F11, BD) for 20 minutes at 4°C, and washed with PBS. Data were acquired on a FACS Canto II flow cytometer (Becton Dickinson) and analyzed with Flowjo® software 7.0 (Tree Star Inc., Ashland, OR).

[0164] Chimerism of greater than 1% of human cells was found in 8 / 10 mice. Mice were sacrificed within 188-264 days after implantation for signs of disease (anemia, >20% weight loss, or palpable tumors), and bone marrow, spleen, and solid tumors (found in the interscapular region, neck and hip region, or kidney, liver, and lymph nodes) were harvested. Tumors were snap-frozen in liquid nitrogen for future processing by mass spectrometry. AML cells were harvested by crushing spleens and flushing femurs and tibias (bone marrow harvest) with 4°C PBS. Cells were depleted of red blood cells, filtered (100 μm) to remove debris, counted, and analyzed by flow cytometry (5 × 10 ) as detailed above. 5 Cells >1 cm in size that were available for RNA sequencing were either processed for RNA sequencing (cells) to assess their purity or (all remaining cells) were lysed in Trizol® (Invitrogen) and cryopreserved for future RNA sequencing. 3One tumor with 6 million bone marrow-derived blast cells at >99% purity (Figure 14A-B) was selected and processed for MAP identification by mass spectrometry. Two mice with no humanization in the peripheral blood (graft failure) at day 122 did not develop any signs of disease and were sacrificed at the end of the experiment (day 264). All sacrifices were performed humanely by CO2 asphyxiation followed by cervical dislocation. Mice were evaluated three times weekly for signs of disease and monitored daily throughout the experiment.

[0165] RNA extraction, library preparation, and sequencing RNA extraction was performed using RNeasy® Mini extraction columns (Qiagen) with Trizol® / chloroform extraction and purification. 400 ng of total RNA was used for library preparation. Total RNA quality was assessed using a BioAnalyzer Nano (Agilent), and all samples had an RIN greater than 8. Library preparation was performed using the KAPA mRNAseq Hyperprep Kit (KAPA, catalog number KK8581). Ligation was performed at a final concentration of 51 nM Illumina Truseq index. 12 PCR cycles were required to amplify the cDNA library. Sample 14H124 was run separately using 4 million cells and 1 μg of total RNA. Library preparation was performed similarly to the previous samples, except that amplification was performed with 10 PCR cycles instead of 12. Libraries were quantified using QuBit and a BioAnalyzer DNA1000. All libraries were diluted to 10 nM and normalized by qPCR using the KAPA Library Quantification Kit (KAPA; catalog number KK4973). Libraries were pooled to equimolar concentrations. Sequencing was performed on an Illumina Nextseq500 using 150 cycles (2 x 80 bp) of the Nextseq High Output Kit with 2.8 pM of pooled library. Approximately 120-200 M paired-end PF reads were generated per sample. Library preparation and sequencing were performed at the Immunology and Cancer's Genomics Platform (IRIC).

[0166] Database generation for shotgun mass spectrometry identification 1) Generation of individualized canonical proteomes. This was performed as previously detailed (Laumont et al., 2018). Briefly, RNA-Seq reads were trimmed using Trimmomatic v0.35 and aligned to GRCh38.88 using STAR v2.5.1b (Dobin et al., 2013), run with default parameters except for the --alignSJoverhangMin, --alignMatesGapMax, --alignIntronMax, and --alignSJstitchMismatchNmax parameters, which were replaced with default values ​​of 10, 200,000, 200,000, and "5-155", respectively, to generate bam files. Single-base variants with a minimum alternative count setting of 5 were identified using freeBayes 1.0.2-16-gd466dde (arXiv:1207.3907). Transcript expression was quantified in transcripts per million (tpm) using kallisto v0.43.0 (Bray et al., 2016) with default parameters. Finally, pyGeno was used to generate fasta files of individualized canonical proteomes by inserting high-quality sample-specific single-nucleotide variants (freeBayes quality > 20) into the reference exome and exporting sample-specific sequences of known proteins produced by expressed transcripts (tpm > 0).

[0167] 2) Generation of AML-specific proteomes by mTEC k-mer depletion (Figure 2A). This was performed as previously detailed (Laumont et al., 2018). Briefly, the R1 and R2 fastq files for each sample were trimmed as reported above, and R1 reads were reverse-complemented using the fastx_reverse_complement function in FASTX-Toolkit v0.0.14. k-mer databases (24 or 33 lengths) were generated with Jellyfish v2.2.3 (Marcais and Kingsford, 2011). A single database was generated for each AML sample, while the six mTEC samples were combined into their own database by concatenating their fastq files. Because the duration of k-mer assembly (see below) exponentially increases beyond 30 million k-mers, the 33-nucleotide-long k-mer database for each AML was filtered based on a sample-specific threshold of occurrence (the number of times a given k-mer appears in the database) to reach a maximum of 30 million k-mers for the assembly step (Table 1). After this filtering, k-mers that existed at least once in the mTEC k-mer database were removed from each sample database, and the remaining k-mers were assembled into contigs using NEKTAR, a software developed in our facility. Briefly, one of the presented 33-nucleotide-long k-mers was randomly selected as a seed, extending from both ends with consecutive k-mers overlapping by 32 nucleotides on the same strand (the -r option was disabled and used as the linear set of k-mers). The assembly process stopped either when a k-mer could not be assembled or when multiple k-mers fit (the -a1 option for linear assembly). In such cases, a new seed is selected and the assembly process is restarted until all k-mers from the proposed list have been used once. Finally, the contigs were three-frame translated using a Python script in our facility, the amino acid sequences were split at internal stop codons, and the resulting subsequences were concatenated with each sample in each individualized canonical proteome.

[0168] 3) Generation of ERE-specific proteomes (Figure 2B). For each sample, RNA-Seq reads were aligned to the human reference genome (GRCh38.88) using STAR (Dobin et al., 2013) with default parameters. Using the intersect function in BEDtools (PMID 20110278), reads were separated into two datasets: those mapping perfectly to either the ERE sequence or the canonical gene. Reads from the ERE read dataset were discarded if their sequences were also present in the canonical read dataset. Unmapped reads, secondary alignments, and low-quality reads were then discarded from the ERE read dataset in samtools view (PMID 19505943). The remaining ERE reads were then in silico translated into ERE polypeptides in all possible reading frames. The ERE polypeptides were spliced ​​at the stop codon, downstream sequences were discarded, and only upstream sequences of 8 or more amino acids (i.e., the minimum length of the MAP) were retained. The resulting ERE proteome was then concatenated with the individualized canonical proteome of each sample.

[0169] 4) Generation of an AML-specific proteome by mTEC + MPC k-mer depletion (Figure 2C). To implement this approach, we used the same method described for mTEC k-mer depletion with the following modifications: (i) An additional normal k-mer database was generated in Jellyfish by combining the fastq files of 11 MPC samples used as k-mer controls. Because these samples were not sequenced in stranded mode, the k-mer database was generated with the -C option and the R1 fastq files were not reverse-complemented. (ii) AML k-mers present in either the mTEC or MPC k-mer database were removed. As a result, the number of k-mers filtered by this step was higher than in the mTEC k-mer depletion approach. (iii) Due to the higher efficacy of k-mer depletion by normal samples, it was possible to pre-filter AML k-mers using a lower occurrence threshold (Table 1 and Figure 7C), which dramatically changed the identity of k-mers present in these databases compared to mTEC k-mer depletion alone (Figure 7D). Importantly, to exclude possible sequencing errors, we did not use an occurrence threshold below 3. All other procedures were performed as reported in the "Generation of AML-specific proteomes by mTEC k-mer depletion" section.

[0170] 5) Generating AML-specific proteomes by differential k-mer expression (Figure 2D). Differential k-mer analysis was performed using a customized DE-kupl, a computational pipeline that generates a k-mer database from fastq files, normalizes k-mer abundance, filters k-mers based on their occurrence and their inter-sample sharing, compares k-mer abundance between samples in two different conditions using statistical tests, assembles differentially expressed k-mers into contigs, aligns the contigs to the genome, and annotates the contigs based on their genome alignment (Figure 8) (Audoux et al., 2017). Specifically, AML samples were first compared to 11 MPC controls using a DE-kupl run with the following parameters: diff_method Ttest, kmer_length 33, gene_diff_method limma-voom, data_type WGS, lib_type unstranded, min_recurrence 6, min_recurrence_abundance 3, pvalue_threshold 0.05, and log2fc_threshold 0.1. This returned a diff-counts.tsv file containing the sequences and normalized counts of 33-nucleotide-long k-mers that were significantly differentially expressed between AML and MPC samples (FDR < 0.05) and were present with a minimum of three occurrences in at least six samples (either MPC or AML).Because custom rules for k-mer filtering were desired, we did not apply any restrictions on the k-mer fold change (log2fc_threshold 0.1) in DE-kupl. Rather, we manually filtered the list of k-mers provided in the diff-counts.tsv file to retain all k-mers that were (i) completely absent (count = 0) in all MPC samples (and thus present in at least six AML samples), or (ii) present in at least six AML samples (>30% of samples) with a fold change of 10-fold or greater, or (iii) present in a single MPC sample and less abundant than the lowest abundance in an AML sample, or (iv) present in at least six AML samples with a fold change of 5-fold or greater and an FDR of 0.000001 or less. Based on these rules, we obtained ~41 × 10 k-mers. 6 A new diff-counts.tsv file containing k-mers was generated and used to perform k-mer assembly with DE-kupl, resulting in ~2.1 × 10 6 The resulting file, merged-diff-counts.tsv, contained the contigs. Finally, the generated contigs were mapped and annotated on the GRCh38 human genome using the annot function in DE-kupl.

[0171] To obtain individualized contig sequences for each AML sample, the DiffContigsInfos.tsv output of DE-kupl annot was used to construct a bed file of all contigs with a length of 34 nucleotides or greater (derived from the assembly of at least two k-mers), which were aligned without gaps, insertions, or deletions (CIGAR without N / D / I). Next, we used the bed files and the bedtools, samtools, and bcftools suites to extract personalized contig sequences (bedtools getfasta-fi consensus.fasta-bed contigs.bed-name>>output.fasta) from the consensus genome generated from the bam files (reads mapped to GRCh38 using STAR, see the "Generating personalized canonical proteomes" section) for each AML sample (samtools mpileup-C50-uf ref_genome.fasta sample.bam|bcftools call-c|vcfutils.pl vcf2fq-d 8-D 100|awk' / ^@chr.$|^chr..$|^@GL........$|^@KI........$ / , / ^+$ / '|sed' / ^+ / d'|tr”@”“>”>consensus.fasta) for each AML sample.

[0172] The portions of contigs not covered by the reads (N) were removed using sed (sed-E"s / NNN+ / Λn / g"), and all contigs were written to a fasta file. Sequences of contigs that aligned with gaps, insertions, or deletions (and those not retrievable from the consensus genome) and were reported as expressed by the relevant sample in DiffContigsInfos.tsv were added to this fasta file. Finally, using our in-house Python scripts (previously published or included in pyGeno (Daouda et al., 2016; Laumont et al., 2018)), the contigs were six-frame translated, converting ambiguous amino acid sequences into all possible sequences (since contigs overlapping with single-nucleotide variations can encode multiple different amino acid sequences). The amino acid sequences were split at internal stop codons, and the resulting subsequences were concatenated with each individualized canonical proteome for each sample.

[0173] 6) Database Size Validation - Figure 9B-C. Given that the MS databases used in the four proteogenomic approaches in this study presented variably expanded sizes compared to the canonical (individualized) proteome database, we investigated how these larger sizes affected MS identification. First, we compared the cumulative number of peptides identified by each approach across 19 AML samples (Figure 9B). This showed that despite significant differences in database size between the approaches, the number of identified peptides varied only slightly compared to the canonical proteome (up to approximately 9% for the ERE approach). Next, because each database was concatenated with each sample's respective individualized canonical proteome, we reasoned that an appropriately sized database should enable the identification of peptides derived from canonical proteins of similar identity to the canonical proteome alone. As shown in Figure 9C, across all AML samples, the majority (88.2%-96.2%) of peptides annotated as encoding proteins identified by each approach were common to those identified based on the canonical proteome alone. Based on these findings, it was concluded that different database sizes are suitable for reliable MS identification.

[0174] Isolation of MHC-associated peptides W6 / 32 antibody (BioXcell) was incubated with PureProteome Protein A magnetic beads (Millipore) at a ratio of 1 mg of antibody per mL of slurry in PBS for 60 minutes at room temperature. The antibody was covalently crosslinked to the magnetic beads using dimethylpimelidate as described. The beads were stored at 4°C in PBS (pH 7.2) and 0.02% NaN3. For frozen cell pellet samples (98 million cells / pellet), cells were thawed, resuspended in 1 mL of PBS (pH 7.2), and solubilized by adding 1 mL of detergent buffer containing PBS (pH 7.2), 1% (w / v) CHAPS (Sigma), supplemented with a protease inhibitor cocktail (Sigma). For tumor samples, the samples were cut into small pieces (cubes, approximately 3 mm in size), and 5 mL of ice-cold PBS containing a protein inhibitor cocktail was added. First, the tissue fragments were homogenized twice for 20 seconds using an Ultra Turrax T25 homogenizer (IKA-Labortechnik) set at 20,000 rpm, followed by 20 seconds using an Ultra Turrax T8 homogenizer (IKA-Labortechnik) set at 25,000 rpm. 550 μl of ice-cold 10x lysis buffer (5% w / v CHAPS) was then added to the sample. The cell pellet and tumor sample were incubated at 4°C with endoverture for 60 minutes, then spun down at 10,000 g for 20 minutes at 4°C. The supernatant was transferred to a new tube containing 1 mg of W6 / 32 antibody covalently cross-linked protein A magnetic beads per sample and incubated at 4°C with endoverture for 180 minutes. The samples were placed on a magnet to collect the MHC I complexes bound to the magnetic beads. The magnetic beads were washed first with 8 x 1 mL of PBS, then with 1 x 1 mL of 0.1X PBS, and finally with 1 x 1 mL of water. The MHC I complexes were eluted from the magnetic beads by acid treatment using 0.2% formic acid (FA). To remove any remaining magnetic beads, the eluate was transferred to a 2.0 mL Costar Spin-X centrifuge tube filter (0.45 μm, Corning) and spun down at 855 g for 2 minutes.The peptide-containing filtrate was separated from MHC I subunits (HLA molecules and β-2 macroglobulin) using a homemade stage tip packed with 21 mm diameter octadecyl (C-18) solid-phase extraction disks (EMPORE). The stage tip was pre-washed first with methanol, then with 80% acetonitrile (ACN) in 0.2% trifluoroacetic acid (TFA), and finally with 0.2% FA. The sample was loaded onto the stage tip and washed with 0.2% FA. Peptides were eluted with 30% ACN in 0.1% TFA, dried using a vacuum centrifuge, and then stored at -20 °C until MS analysis.

[0175] mass spectrometry The dried peptide extract was resuspended in 4% formic acid and loaded onto a homemade C18 analytical column (15 cm x 150 μm i.d. packed with C18 Jupiter Phenomenex) using a 56 min gradient (10H005) or a 106 min gradient (all other samples) of 0% to 30% acetonitrile (0.2% formic acid) and a flow rate of 600 nL / min on an EasyLC II system. Samples were analyzed on a Q-Exactive HF mass spectrometer (Thermo Fisher Scientific) in positive ion mode with a Nanospray 2 source at 1.6 kV. Each full MS spectrum acquired at a resolution of 60,000 was followed by 20 MS / MS spectra, and the most abundant multiply charged ions were analyzed at a resolution of 30,000 and 5 × 10 4 (10H005) or 2 x 10 4 An automatic gain control target of 0.05 (for all other samples), an injection time of 100 ms (10H005) or 500 ms (15H023, 15H063, 15H080, 05H149) or 800 ms (for all other samples), and a collision energy of 25% were selected for MS / MS sequencing.

[0176] Synthetic peptides If sufficient material is available, TSA 高The amino acid sequence of was further verified with synthetic peptides as previously described (Zhao et al., Cancer Immunol Res. 2020 Feb11 doi:10.1158 / 2326-6066. CIR-19-0541. [Epub ahead of print]).

[0177] Bioinformatics analysis All analyses were performed on trimmed data, and all alignments were performed using STAR as described in the previous section, and all alignments were performed on GRCh38.88 unless otherwise stated.

[0178] All liquid chromatography (LC)-MS / MS (LC-MS / MS) data were searched against a relational database using PEAKS X (Bioinformatics Solution Inc.). For peptide identification, the precursor and fragment ion tolerances were set to 10 ppm and 0.01 Da, respectively. The occurrence of oxidation (M) and deamidation (NQ) was set as variable modifications.

[0179] 1) MAP Identification. After peptide identification, a list of unique peptides was obtained for each sample, and a 5% false discovery rate (FDR) was applied to the peptide scores. The binding affinity of the sample to its HLA alleles was predicted using NetMHC4.0 (Andreatta and Nielsen, 2016). Only peptides 8–11 amino acids long with a percentile rank of 2% or less were used for further annotation.

[0180] 2) Identification and validation of MAPs of interest (MOIs). For both k-mer depletion approaches, this was performed using a similar approach as previously described (Laumont et al., 2018). Briefly, each MAP and its coding sequence was queried against the relevant AML and normal canonical proteomes (constructed for all mTECs and MPCs, as detailed above) or a cancer and normal 24-nucleotide k-mer database (constructed from either combined mTECs or combined MPCs, as detailed above), respectively. MAPs detected in the normal canonical proteome were excluded regardless of their coding sequence detection status. MAPs not detected in either the MAP, normal canonical proteome, or normal k-mers were flagged as MOI candidates. MAPs not present in both canonical proteomes but present in both k-mer databases were required to have their RNA coding sequences overexpressed by at least 10-fold in AML compared to normal samples and were flagged as MOIs. Finally, a MAP corresponding to several RNA sequences (derived from different proteins) can be flagged as an MOI only if their respective coding sequences are consistently flagged as MOIs.

[0181] For the ERE approach, an ERE status of "yes," "may," or "no" was assigned to each individual MAP based on the presence of the ERE and its amino acid sequence in the personalized canonical proteome. For "may" candidates, the expression level of the peptide's coding sequence in the ERE read and canonical read datasets (i.e., the minimum occurrence of a 24-nucleotide k-mer set of the peptide) was calculated. Only "may" candidates with at least 10-fold higher expression in the ERE read dataset were considered as ERE MAPs. The remaining ERE MAP candidates were then manually validated in IGV (Robinson et al., Nat Biotechnol. 2011 Jan;29(1):24-6) to determine whether the peptide's coding sequence contained germline polymorphisms and had the appropriate orientation compared to the ERE sequence and the canonical annotated sequence (if applicable).

[0182] For the differential k-mer approach, the complete list of MAPs was queried against the AML-specific proteome to be flagged as MOI candidates. Next, the RNA expression of each MAP (following the procedure described in the next section) was assessed in 19 AML specimens and in 11 MPCs used as controls in DE-kupl. All MAPs with a minimum fold change of 5 between normal and cancer samples were flagged as MOIs. Because the MAP RNA expression assessment procedure relies on the reference genome for quantification, candidate MOIs derived from mutations could not be properly quantified and were systematically flagged as MOI candidates. To unambiguously verify the presence of each MOI at the RNA level in each AML sample in which they were identified, the MOI coding sequences were extracted from the DiffContigsInfos.tsv output of DE-kupl and queried against the associated fastq files (sequence in the forward R2 fastq and reverse complement in the reverse R1 fastq). MOIs that failed this test were discarded.

[0183] For all lists of MOI candidates (four different approaches), leucine and isoleucine variants cannot be distinguished by standard MS approaches. MOIs for which existing variants were flagged as non-MOIs were discarded unless they exhibited higher RNA expression than the variants. MS / MS spectra for all MOIs were manually inspected to remove any false positives. Finally, genomic locations were assigned to all MOIs by mapping reads containing their coding sequences on the reference genome using BLAT (a tool from the UCSC Genome Browser). MOIs with reads that did not match a matched genomic location or matched hypervariable regions (e.g., MHC, Ig, or TCR genes) were excluded. For those with a matched genomic location, IGV was used to filter out MOIs with coding sequences that overlapped with known germline polymorphisms (dbSNP149).

[0184] 3) Quantification of MAP coding sequences in RNA-Seq data. To unambiguously assess the RNA expression of each MAP, all MAP amino acid sequences were reverse-translated into all possible nucleotide sequences. Next, all possible sequences were mapped onto the genome with GSNAP (Wu et al., 2016) using the -n1000000 option to localize all genomic regions that could encode a given MAP. To confidently capture MAPs encoded by overlapping splice site sequences, potential MAP coding sequences were mapped to the transcriptome (cDNA and non-coding RNA) to extract a large portion (80 nucleotides) of the reference transcriptome sequence (using the --length80 option in samtools faidx), which was then mapped to the reference genome (using GSNAP, the --use-splicing and --novelsplicing=1 options). Genome alignments of all reads containing their coding sequences were also performed for the MOIs generated by the different TSA discovery pipelines. The GSNAP output was filtered to retain only exact matches between the sequence and the reference, generating a 'bed' file containing all possible genomic regions susceptible to coding for a given MAP. Using samtools view (-F256 option), grep, and wc (-l option), the number of reads containing the MAP coding sequence at each genomic location in each desired RNA-Seq sample (e.g., AML, GTEX, or normal sample) was counted and aligned to the reference genome using STAR (bam file). Finally, all read counts (from different regions and coding sequences) for a given MAP were summed and normalized by the total number of reads sequenced in each evaluated sample to obtain reads per hundred million (RPHM) counts.

[0185] 4) Immunogenicity assessment. Immunogenicity prediction of MOI was performed using Repitope (Ogishi and Yotsuyanagi, 2019). Feature calculations were performed using the predefined MHCI_Human_MinimumFeatureSet variable, and the FeatureDF_MHCI and FragmentLibrary files provided in the package's Mendeley repository (https: / / data.mendeley.com / datasets / sydw5xnxpt / 1) were updated (July 12, 2019).

[0186] 5) MOI presentation and expression by AML patients. To identify all possible HLA alleles capable of presenting a given MAP (promiscuous binder), we used the MHCcluster online tool (http: / / www.cbs.dtu.dk / services / MHCcluster / ) (Thomsen et al., 2013). HLA alleles with a clustering value of 0.4 or less were considered to be capable of presenting the same MAP. To evaluate MOI presentation by AML patients in the Leucegene cohort, their HLA type was first determined using Optitype. A given MOI was considered presented if its expression at the RNA level was higher than 2 rphm (rather than 0 rphm to maximize the probability of presentation) and if the patient expressed an HLA allele capable of presenting the MOI (as predicted by NetMHC4.0 for the original identification of the presenting molecule for each MOI found, and MHC cluster for the identification of promiscuous binders). If a patient expressed two different HLA alleles capable of presenting the same MOI, the MOI was considered to be presented twice.

[0187] High TSA 高 To assess molecular features associated with expression, TSA 高 was considered to be expressed in a given patient if its expression in this patient was higher than the median expression across the entire cohort (calculated based only on non-null values).高 Total number of (#HE-TSA 高 ) were counted for each patient and used to perform correlation analyses with gene expression and association with mutations or other clinical features (see next section).

[0188] 6) Survival analysis. Survival data for 374 patients in the Leucegene cohort were kindly donated by the Leucegene team (https: / / leucegene.ca). High-count HLA-TSA, calculated as above, were used. 高 Complex (HLA-restricted TSA 高 A survival analysis was performed to assess the association between the number of TSAs (presentation) and clinical outcome (overall survival). 高 Depending on the total number of HLA-TSA, two groups were identified: high expressers (HLA-TSA) 高 Patients were divided into low expressers (highest quartile of expression counts) and low expressers (all other patients). Survival was compared between the two groups using Kaplan-Meier curves, and significance was assessed by the log-rank test in GraphPad Prism v7.0. Multivariate analyses were performed using the R package survivalAnalysis v0.1.1, incorporating age as a continuous variable, mutations coded as present / absent (1 / 0), and cytogenetic risk assessments treated as individual groups, with intermediate vs. favorable risk and adverse vs. favorable risk.

[0189] 7) Mutation analysis. Mutation data for NPM1, FLT3-ITD, FLT3-TKD, IDH1(R132), and biallelic CEBPA were obtained from previously published data on the Leucegene cohort (Audemard et al., 2019; Lavallee et al., 2016). Mutations in ASXL1, TP53, DNMT3A, IDH2 (R140 and R172 only), WT1, RUNX1 et TET2 were detected with Freebayes and filtered to remove the following mutations: (i) mutations with a variant allele frequency (VAF) of less than 20%, (ii) mutations flagged as SNPs in the COSMIC database (https: / / cancer.sanger.ac.uk / cosmic), (iii) mutations with low putative impact (5'UTR premature start codon gain variants, splice region variants, and synonymous variants, stop retention variants, synonymous variants), and (iv) missense SNPs with a mild impact on protein structure and function as predicted by FATHMM-XF (http: / / fathmm.biocompute.org.uk / fathmm-xf / ) (Rogers et al. al., 2018 ), (iv) insertions and deletions with AAAAA+ or TTTTT+, and (v) mutations that are only flagged as germline in the COSMIC database ( Tate et al., 2018 ).

[0190] 8) Gene Expression Analysis. Quantification of expression of all transcripts was performed using kallisto v0.43.0 with default parameters. Kallisto transcript-level count estimates were converted to gene-level counts using the R package tximport. Counts were normalized using the TMM algorithm and edgeR was used to output counts per million (cpm) values. For further analysis, only protein-coding genes were retained (as reported in Ensembl's BioMart tool (useast.ensembl.org / biomart)). Each gene expression and HE-TSA were compared. 高Exhaustive Pearson correlations between counts were performed using the cor.test function in R. Correlations were performed for all non-NPM1 / FLT3-ITD / DNMT3A-mutated and non-FAB-M1 patients. p-values ​​were corrected for multiple comparisons with the Benjamini-Hochberg method (p.adjust in R). Only genes with an FDR < 0.00001 in at least one of three correlation analyses, an FDR < 0.001 in three analyses, consistent correlation coefficients (positive or negative) in three analyses, and a correlation coefficient > 0.3 or < -0.3 in at least one analysis were retained for downstream processing.

[0191] t-Distributed Stochastic Neighbor Embedding (t-SNE) analysis was performed using the Rtsne package on the identities of expressed genes obtained from the aggregation of Kallisto transcript-level abundance estimates to gene abundance estimates from tximport (expression = 1 if tpm ≥ 1, expression = 0 if tpm < 1). Only protein-coding genes (which are most susceptible to MAP generation) were used for this analysis.

[0192] 9) GO Term and Enrichment Map Analysis. Biological Process Gene Ontology (GO) term overrepresentation was performed using BiNGO v3.0.3 (Maere et al., 2005) in Cytoscape v3.7.2, using a hypergeometric test and a significance cutoff of 0.005 or less FDR-adjusted p-value. The output from BiNGO was imported into EnrichmentMap v3.2.1 (Merico et al., 2010) in Cytoscape to cluster redundant GO terms and visualize the results. EnrichmentMap was generated using a Jaccard similarity coefficient cutoff of 0.25, a p-value cutoff of 0.001, and an FDR-adjusted cutoff of 0.005. Networks were visualized in Cytoscape using the default "Prefuse Force-Directed Layout" setting with default settings and 600 iterations. Groups of similar GO terms were manually circled.

[0193] 10) Intron Retention and NMF Clustering. Intron retention (IR) analysis of the complete Leucegene cohort and 11 primary MPC samples was performed using IRFinder v1.2.5 (Middleton et al., 2017). Introns with an IRatio ≥ 10% (introns with ≥ 10% transcript retention) and a minimum coverage of 3 reads were considered retained. Introns were filtered to retain only those retained in at least two AML samples and not in any MPC samples. The 10% most variable introns (by coefficient of variation of IRatio across the complete cohort) were selected for further analysis (6988 introns). Unsupervised consensus clustering results were generated using the NMF v0.21.0 package in R (Gaujoux and Seoighe, 2010) for the selected intron IRatios, using the default Brunet algorithm and 200 iterations for rank exploration and clustering runs. For clustering solutions with 3–15 clusters, cluster results were selected by considering the cophenetic score profile and average silhouette width of the consensus membership matrix.

[0194] Abundance heatmaps were generated by identifying the top 2% of introns in the NMF metagene (W matrix) output file. Removing duplicate names resulted in a list of 1,211 introns. For each Leucegene sample, a matrix of these intron IRatios was generated and reordered to match the NMF clustering output. Hierarchical clustering of introns with a central correlation distance metric and complete linkage was performed using the heatmap.3 package in R.

[0195] ELISPOT assay 1) Generation of Monocyte-Derived Dendritic Cells. Monocyte-derived dendritic cells were generated from frozen PBMCs as previously described (Vincent et al., Biology of Blood and Marrow Transplantation: Journal of the American Society for Blood and Marrow Transplantation, 22 Oct 2013, 20(1):37-45; Laumont et al., Nat Commun. 2016 Jan 5;7:10238). Briefly, DCs were prepared from adherent PBMC fractions by culturing them for 8 days in X-VIVO™ 15 medium (Lonza Bioscience) supplemented with 5% human serum (Sigma-Aldrich), sodium pyruvate (1 mM), IL-4 (100 ng / mL, Peprotech), and GM-CSF (100 ng / mL, Peprotech). After 7 days of culture, DCs were matured overnight with IFN-γ (1000 IU / mL, Gibco) and LPS (100 ng / mL, Sigma-Aldrich). Two hours into the maturation process, DCs were loaded with 2 μg / mL of peptide and then irradiated (40 Gy) before being used as APCs in T-DC cultures. In the control group, DCs were pulsed with a mixture containing Melan-A, NS3, and Gag-A2 peptides (all three of which bind to HLA-A*02:01).

[0196] 2) In vitro peptide-specific T cell proliferation. Thawed PBMCs were first transfected with human CD8 + CD8 T cell isolation kit (Miltenyi Biotech) was used. +T cells were enriched and co-incubated with DCs pulsed with autologous peptides at an APC:T cell ratio of 1:10. Expanding T cells were cultured in Advanced RPMI medium (Gibco) supplemented with 8% human serum (Sigma-Aldrich), L-glutamine (Gibco), and cytokines for 4 weeks (with restimulation with pulsed DCs every 7 days). During the first week of co-culture, IL-12 (10 ng / mL) and IL-21 (30 ng / mL) were added to the medium. Two days later, IL-2 (100 UI / mL) was also added to the cytokine mixture. During the second week, IL-2 (100 UI / mL), IL-7 (10 ng / mL), IL-15 (5 ng / mL), and IL-21 (30 ng / mL) were added to the medium. During the final two weeks of co-culture, IL-2 (100 UI / mL), IL-7 (10 ng / mL), and IL-15 (5 ng / mL) were used. Medium supplemented with the appropriate cytokine mixture was added to the co-cultures every two days. At the end of the fourth week of co-culture, cells were harvested for ELISPOT assays.

[0197] 3) IFNγ ELISPOT assay. To perform the experiment, an ELISpot human IFNγ kit (R&D Systems, USA) was used according to the manufacturer's recommendations. The collected CD8 + T cells were seeded and incubated for 24 hours at 37°C in the presence of irradiated, peptide-pulsed PBMCs (40 Gy) used as stimulators. As a negative control, selected CD8 + T cells were incubated with irradiated, unpulsed PBMCs. Spots were counted using an ImmunoSpot S5 UV Analyzer (Cellular Technology Ltd, Shaker Heights, OH) as described in the reagent set manufacturer's protocol. IFN-γ production was calculated using a 10 μg / mL ELISA after subtracting spot counts from negative control wells. 6 CD8 + It was expressed as the number of peptide-specific spot-forming cells (SFC) per T cell.

[0198] Immunogenicity prediction Immunogenicity prediction of MOI was performed using Repitope (Ogishi and Yotsuyanagi, 2019). Feature calculations were performed using the predefined MHCI_Human_MinimumFeatureSet variable, and the FeatureDF_MHCI and FragmentLibrary files provided in the package's Mendeley repository (https: / / data.mendeley.com / datasets / sydw5xnxpt / 1) were updated (July 12, 2019).

[0199] TCR and cytotoxic T cell signature analysis TCR repertoire analysis was performed on RNA-seq data from 437 Leucine patients using TRUST4 software (Li et al., 2017) with default parameters. T cell clonotype diversity was estimated by normalizing the number of TCR CDR3s (complete and partial) per kilo TCR read (CPK). ERGO (Springer et al., 2020) predictions of interactions between the amino acid sequence of the complete TCRβ CDR3 detected by TRUST4 and MOI were performed through a freely available web portal (http: / / tcr.cs.biu.ac.il / ) using an autoencoder-based model and VDJdb as the training database.

[0200] For cytotoxic T cell signature analysis, predicted HLA-TSA per patient 高 The number of pairs was calculated by dividing the number of TSAs with rphm expression ≥ 2. 高 Normalized TSA by dividing by the count of 高 The presentation level of HLA-TSA was obtained. 高 Patient samples with no count or not collected at the time of diagnosis were discarded from the analysis. The remaining 361 patients were analyzed using their normalized TSA 高Patients were grouped according to presentation level, and patients with presentations above the median of the distribution were compared with other patients (below the median) through differential gene expression analysis. Analysis was performed in R 3.6.1. Raw read counts were converted to counts per million (cpm) normalized to library size, and low-expressing genes were filtered out by retaining genes with cpm > 1 in at least two samples using edgeR 3.26.8 (Robinson et al., 2010) and limma 3.40.6 (Ritchie et al., 2015). This was followed by voom transformation and linear modeling using lmfit in LIMMA. Finally, adjusted t-statistics were calculated in eBayes. Genes with a p-value ≤ 0.01 and -0.3 ≥ log2(FC) ≥ 0.3 were considered significantly differentially expressed.

[0201] Cytokine secretion assay and dextramers After three rounds of stimulation using peptide-loaded monocyte-derived dendritic cells and cytokines according to (Janelle et al., 2015), 1.0 × 10 6Cells were incubated for 4 hours with dimethyl sulfoxide (DMSO), 5 μg / ml of the peptide of interest, 5 μg / ml of a control peptide (negative control), or 50 ng / ml of phorbol 12-myristate 13-acetate (PMA) and 500 ng / ml of ionomycin (positive control, Sigma-Aldrich) in the presence of 7.5 μg / ml Brefeldin A (Sigma-Aldrich, Oakville, ON). Cells were then stained with cell surface antibodies and fixed and permeabilized using Cytofix / Cytoperm buffer for intracellular staining according to the manufacturer's instructions (BD Biosciences, Mississauga, ON). Permeabilized cells were incubated with antibodies directed against IFNγ, IL-2, and TNFα (BD Biosciences) for 20 min at 4°C, resuspended in phosphate-buffered saline (PBS) supplemented with 2% fetal bovine serum (FBS; ThermoFisher, Waltham, MA, USA), and then acquired. Acquisition was performed using an LSRII flow cytometer (BD Biosciences), and data were analyzed using FlowJo™ V10 software (BD Biosciences). For multimer staining, 1.0 × 10 6 Cells were stained with a custom-made fluorescent dextramer (Immudex, Copenhagen, Denmark) for 45 minutes at 4°C, followed by CD8 monoclonal antibody (eBiosciences, San Diego, CA) for 30 minutes at 4°C. Cells were washed with PBS 2% FBS and then acquired on an LSRII cytometer (BD Biosciences). Data were analyzed using FlowJo™ V10 software (BD Biosciences).

[0202] FEST assay For the FEST assay, T cells were cultured as previously described (Danilova et al., 2018) with minor modifications. Briefly, on day 0, thawed PBMCs from a healthy donor (BioIVT) were enriched for T cells using a human pan T cell isolation kit (Miltenyi). T cells were cultured at 2 × 10 in AIM V medium supplemented with 50 μg / mL gentamicin (ThermoFisher Scientific) and 1% HEPES. 6 The T cell negative fraction was irradiated with 30 G γ, washed, and resuspended at 2.0 × 10 cells / mL in AIM V medium supplemented with 50 μg / mL gentamicin and 1% HEPES. 6 Both T cells and irradiated T cell depleted cells were resuspended at 1 mL per well in 3 TSA 高 Pool (5 TSA per pool) 高 , 1 μM final concentration for each TSA) or without peptide was added to a 12-well plate. Cells were cultured at 37°C, 5% CO for 10 days. On days 3 and 7, half of the culture medium was replaced with fresh culture medium containing 100 IU / mL IL-2, 50 ng / mL IL-7, and 50 ng / mL IL-15 (day 3) and 200 IU / mL IL-2, 50 ng / mL IL-7, and 50 ng / mL IL-15 (day 7). On day 10, human CD8 + Cells were harvested and CD8 T cells were isolated using a T cell isolation kit (Miltenyi). + The cells were further isolated. As a negative control, CD8 + T cells were also isolated from freshly thawed, uncultured PBMCs of the same healthy donor using a Qiagen DNA blood mini kit (Qiagen). +DNA was extracted from T cells. TCR Vβ CDR3 sequencing was performed using the ImmunoSEQ platform (Adaptive Biotechnologies) at survey resolution. Raw data exported from the immunoSEQ portal were processed with the FEST web tool (www.stat-apps.onc.jhmi.edu / FEST) using no minimum number of templates and the "Ignore baseline threshold" parameter.

[0203] Quantification and statistical analysis Unless explicitly stated in the figure legends, all statistical tests comparing two conditions were performed using the Mann-Whitney U test. All correlations were assessed using the Pearson correlation coefficient. Unless otherwise stated, all boxes in box plots represent the median, 25th, and 75th percentiles of the distribution, and whiskers span the 10th and 90th percentiles. Unless otherwise stated, all bar graphs represent the mean and standard deviation (SD). Plots and statistical tests were primarily performed using GraphPad Prism v7.00. For all statistical tests, **** indicates p<0.0001, *** indicates p<0.001, ** indicates p<0.01, and * indicates p<0.05.

[0204] Example 2: Purified hematopoietic progenitor cells are a useful control for detecting TSA in AML. MS is the only available technology capable of directly identifying MAPs (Ehx and Perreault, 2019; Shao et al., 2018). MS-based identification of MAPs is typically performed through the use of software tools that match acquired tandem MS spectra to a database of protein sequences provided by the user. However, reference protein databases contain only canonical protein sequences and therefore cannot identify MAPs derived from mutations and aberrantly expressed non-canonical genomic regions (which are the main source of aeTSAs) (Laumont et al., 2018). A proteogenomics strategy for constructing a tailored MS database for the identification of global TSAs has been previously described. The tailored database is constructed for each tumor sample and must meet two criteria: it must be comprehensive enough to include all potential TSAs, yet be limited in size, since an expanded reference database increases the risk of false discoveries (Nesvizhskii et al., 2014; Chong et al., 2020). Database construction begins with (i) RNA sequencing of tumor samples, which represents the core of the data; (ii) in silico slicing of RNA-seq reads into 33-nucleotide-long subsequences (k-mers); and (iii) subtraction of normal k-mers to create modules containing only cancer-specific k-mers. As with many aspects of cancer research, a challenging issue is the selection of negative controls (here, a source of normal k-mers). Previous studies have used k-mers from mTECs as normal controls. However, in the case of AML, we tested a different type of negative control: sorted myeloid progenitor cells (MPCs, which contain granulocytic / monocytic progenitors and various types of granulocytic progenitors).

[0205] To compare the values ​​of mTEC and MPC as negative controls, we first compared the similarity between 19 targeted AML specimens (see Table 1 for characteristics), six mTEC samples, and six MPC samples for which high-coverage RNA-seq had previously been performed (Maiga et al., 2016). Notably, MPC depleted, on average, 16.4% more k-mers from AML than mTEC, indicating greater transcriptome overlap between MPC and AML than between mTEC and AML (Figure 1A). Thus, mTEC and MPC (approximately 8.7 × 10 8 for approximately 9.9 × 10 8 ), whereas MPCs yielded similar k-mer counts compared with mTECs (approximately 1.9 × 10 8 , approximately 22%) than AML (approximately 3.3 × 10 8 , approximately 33%) (Figure 1B). To confirm that lineage differences are at the origin of this higher similarity, we performed t-SNE clustering of AML samples along with RNA-seq arrays of sorted epithelial and hematopoietic cells downloaded from various sources based on the identity of expressed protein-coding genes (see Methods). This showed that AML samples clustered with hematopoietic cells, while mTECs clustered with epithelial cells (Figure 1C). Importantly, mTECs expressed the highest diversity of genes, consistent with their biological function (Figure 1D). Overall, these results indicate that, despite the transcriptome diversity of mTECs, MPCs are better normal controls than mTECs for the discovery of TSAs in AML. As a corollary, the size of the database of AML-specific k-mers is smaller when MPC k-mers are subtracted instead of mTEC k-mers.

[0206] Example 3: Development of an MPC-based TSA discovery approach In addition to capturing the entire AML TSA landscape, we evaluated four strategies for constructing a reference database. The first two strategies have been previously reported (Figure 2A, B), while the other two are novel (Figure 2C, D). Importantly, MS analysis of AML specimens was performed only once; therefore, each of the four different TSA discovery approaches was performed on the same MS spectrum for each AML sample. The first strategy results in mTEC subtraction (Figure 2A) (Laumont et al., 2018). The second strategy focuses specifically on MAPs encoded by EREs, which may be a rich source of TSAs (Figure 2B) (Larouche et al., 2020).

[0207] The third strategy depleted k-mers from both mTECs and MPCs (Figure 2C). Notably, prior to the depletion step, we filtered k-mers based on their occurrence (the number of times a k-mer occurs in the same sample; Figure 8A, B) to limit the final number of k-mers for contig assembly to approximately 30 million (assembling more k-mers would be too demanding in terms of computational time). As a result, mTEC+MPC k-mer depletion removed more k-mers from AML samples than mTEC alone, reducing the occurrence threshold by approximately 2-3 fold, thereby enabling us to discover MAPs that would have been missed in the database with the mTEC k-mer depletion approach (Figure 8C, D).

[0208] The fourth strategy aimed to circumvent a major caveat of the k-mer depletion strategy: the lack of comparison between k-mer abundances in normal and cancer samples. Specifically, in the k-mer depletion strategy, the presence of a k-mer in normal controls results in the filtering of this k-mer in cancer samples, even if its frequency is 100-fold higher in cancer compared to normal controls. Briefly, differential k-mer expression (DKE) analysis was performed using the DE-kupl computational protocol (Audoux et al., 2017) with several in-house adjustments and can be summarized as follows (Figure 2D and Figure 9A): (i) pre-filtering of k-mers present in at least 30% of AML samples (occurrence ≥ 3 times), (ii) normalization of k-mer abundance, (iii) statistical comparison of k-mer abundances by a user-defined algorithm, (iv) assembly of significantly differentially overexpressed k-mers (minimum fold change 10) into contigs, and (v) alignment of the contigs to establish their origin. Because they were the most closely related normal samples, MPCs were selected for this study and used as normal controls. The 19 AML samples were compared with the 11 available high-coverage MPC samples. Next, personalized contig sequences were generated for each AML sample (based on read coverage and SNP calling at the genomic locations of differentially expressed contigs), translated into all possible read frames, and combined with personalized canonical proteomes to identify MAPs (Figure 2).

[0209] Example 4: MPC-based approach improves TSA in AML 高 Identify the majority of Each of the four TSA discovery approaches identified thousands of MAPs across 19 AML samples (Table 2). To be considered actionable TSAs, MAPs must be abundantly presented by AML cells and either not presented by normal cells or presented at a low enough level to not elicit T cell recognition, as epitope density plays a key role in target cell eradication by CD8 T cells (Cosma and Eisenlohr, 2019). Because MAPs are preferentially derived from highly abundant transcripts (Figure 3A and Pearson et al., 2016), we established two key thresholds: (i) the RNA expression level at which the probability of generating a MAP in normal tissue can be considered low, and (ii) the fold change in RNA expression (FC) required to significantly increase the probability of presenting a MAP. To achieve this, we assessed the RNA expression of all identified MAPs in each AML sample and found it followed a normal distribution plotted as a cumulative frequency distribution (Figure 3B). This demonstrated that expression below 8.55 reads per hundred million (RPHM) resulted in a less than 5% probability of generating MAP. Considering that AML cells express similar levels of MHC molecules compared to normal granulocytes, and that granulocytes express the highest levels of MHC-I among normal tissues (Berlin et al., 2015; Boegel et al., 2018), 8.55 RPHM was established as the first threshold for all tissues. Based on the same distribution, the influence of different FCs on the probability of generating MAP can also be evaluated (Figure 3C). This showed that FCs between 2 and 5 tend to have a greater impact on the probability than larger FCs. Therefore, 5 was adopted as the minimum FC threshold.

[0210] Based on these two thresholds, we established a decision tree to separate MAPs based on their RNA expression in a wide range of normal adult tissues, including AML, MPC, other normal hematopoietic cells, and mTEC (Figure 3D). In summary, all MAPs expressed at less than 8.55 RPHM in normal tissues and at higher levels in AML than MPC were flagged as TSAs because their detection is evidence of their presentation on the AML cell surface, while their probability of being presented by normal tissues is low. Furthermore, TSAs with an FC of at least 5 between AML and MPC were flagged as TSAs because they have the highest probability of being presented exclusively by AML cells. 高 Other MAPs that were overexpressed in hematopoietic cells compared to other tissues but did not meet these criteria were classified as TAAs or hematopoietic-specific antigens (HSAs) (Figure 3D).

[0211] After a pre-filtering step (see Methods) and decision tree-based classification for each pipeline, we obtained four lists of MAPs of interest (MOIs) (Table 2). The mTEC depletion approach resulted in the highest percentage of HSA, while TSA 高 The majority of TSAs were identified by the MPC-based approach (Figures 3E, F, and 10A). The overlap between both MPC-based approaches was low because the DKE approach pre-filters k-mers with minimal occurrence in minimal patients. Therefore, most TSAs identified by the depletion approach were 高 were less shared between patients compared to those identified by the DKE approach (Figure 9B). Overall, these results suggest that the DKE approach may be beneficial in identifying TSAs in AML. 高 It is most suitable for identifying additional, less shared TSAs, which can be complemented by an MPC-based k-mer depletion approach. 高 It is shown that it is possible to identify [Table 4-1] [Table 4-2] [Table 4-3] [Table 4-4] [Table 4-5] [Table 4-6] [Table 4-7] [Table 4-8] [Table 4-9] [Table 4-10] [Table 4-11] [Table 4-12] [Table 4-13] [Table 4-14] Biotypes were manually assigned upon inspection of the peptide-coding sequence at the indicated genomic location. Immunogenicity scores were calculated using Repitope. HLA alleles correspond to those most likely to present the peptide in a given sample, as predicted by netMHC4.0. Validation of synthetic peptides was performed using TSA. 高 This was carried out only.

[0212] To assess the robustness of MOI identification, the observed mean retention time (RT) of a given peptide was correlated with two best-in-class metrics for validating MAPs identified by high-throughput MS: RT calculated by the DeepLC algorithm (Bouwmeester et al., 2020) and hydrophobicity index assessed by SSRcalc (Krokhin, 2006), both predicted based on peptide sequence. This showed that the RT distribution of non-canonical MOIs correlated well with the prediction and was not significantly different (F-test) from the distribution of canonical proteome-derived peptides, supporting correct identification (Figure 3G). Finally, all MS database searches (initially performed with PEAKS software) were repeated with the Comet algorithm. The percentage of re-identifications showed no significant difference between non-canonical MOIs and canonical peptides (Figure 3H, left panel). Among the MOIs, 58 TSAs were identified. 高 We re-identified 52 of the MAPs (90%) (Figure 3H, right panel). This large overlap between MAPs identified by the two different search engines further supports the robustness of non-canonical MOI identification.

[0213] Example 5: TSA 高 is an immunogenic MAP that is primarily derived from the translation of an intron. The combined results of the four TSA discovery approaches resulted in a total of 47 HSAs, 49 TAAs, and 36 TSAs. 低 , and 58 T.S.A. 高 Table 2 lists the main characteristics of all MOIs. By definition, TSA was expressed below the threshold in all organs (from GTEx), as well as in mTECs and normal hematopoietic cells (Figure 4A). Importantly, TSA in normal tissues 高 Expression of coding RNAs was systematically inferior to that of TAAs previously used in clinical trials without off-target toxicity (Chapuis et al., 2019; He et al., 2020; Legat et al., 2016; Qazilbash et al., 2017). Consistent with this, an HLA-ligand atlas containing human MAPs identified in 29 non-malignant tissues did not show any significant association with TSAs.高 (https: / / www.biorxiv.org / content / 10.1101 / 778944v1). This is because TSA 低 and 58 TSA 高 The safety of targeting TAA (without redundancy) is supported. TAAs showed elevated expression in at least one normal tissue, whereas HSA expression was restricted to the hematopoietic compartment. Comparison of FC between AML specimens and MPCs revealed that TSA 高 showed the highest overexpression along with TAA (median 22-fold), while HSA was expressed at the highest level in healthy cells (median 0.6-fold) (Figure 4B). Overall, these results suggest that TSA 高 This demonstrates that this method combines the advantages of both TSA specificity / safety and TAA overexpression.

[0214] Although TSAs were mostly derived from what are considered non-coding regions of the genome, only 13% of them were derived from canonical protein exons, and 58% of them were identified as being derived from introns (Figure 4C). None were derived from mutations, consistent with the low mutation burden of AML (Lawrence et al., 2013). TAAs were primarily derived from protein-coding exons, but the origin of HSAs was also dominated by non-coding regions, consistent with previous studies reporting tissue-specific intron retention and ERE expression patterns (Middleton et al., 2017; Larouche et al., 2020). Eight TSAs were identified. 高 Although derived from canonical protein-coding genes, they can be considered safe targets given their low expression in normal tissues compared to safe TAAs (Figure 4A). Supporting their relevance as therapeutic targets, three of them are derived from known AML biomarkers (LTBP1, MYCN, and PLPPR3), while the other five have unknown functions or are involved in proliferation, differentiation, or drug resistance (Table 3). [Table 5]

[0215] The therapeutic value of TSA depends in part on the extent to which it is shared by the patient. 高 To assess sharing, we analyzed the Leucegene cohort (Lavallee et al., 2015; Macrae et al., 2013; Pabst et al., 2016), which contains RNA-seq data from purified AML blasts for 437 patients. Because most MAPs can be presented by different HLA allotypes, the identified TSAs 高 The presentation of TSAs was first assessed by taking into account promiscuous binders. Individual TSAs were then analyzed by using the MHC cluster tool (Thomsen et al., 2013), which clusters together HLA alleles that present similar epitopes. 高 Based on these data, it is possible to estimate the complete set of HLA allotypes that can present with TSA (Table 4). 高 It has been shown that individuals with one or more HLA-I allotypes capable of presenting TSA are then able to express an individual TSA only if the TSA-encoding transcript is expressed and the patient has an HLA allotype capable of presenting this TSA. 高 was considered to be present in a given AML sample. Based on these criteria, the TSA per patient in the Leucegene cohort 高 The median number of TSAs was 4, and 93.6% of patients had at least one TSA. 高 It can be predicted that the phenotype will be similar to that of the phenotype of the cerebrospinal fluid (C1F) (Figure 4F). [Table 6-1] [Table 6-2] [Table 6-3]

[0216] TSA in AML samples analyzed at initial diagnosis 高 When the number of TSAs that could be presented by patient HLA alleles was compared with that at relapse (mismatched samples), no difference was observed between the two groups (Figure 4G). In another study (Toffalori et al., 2019), the number of TSAs that could be presented by patient HLA alleles in matched samples of AML blasts obtained at diagnosis and at relapse after allogeneic hematopoietic cell transplantation was significantly higher than that in matched samples of AML blasts obtained at diagnosis and at relapse after allogeneic hematopoietic cell transplantation. 高 No differences were observed when comparing the RNA expression of TSA (Figure 4H). Because leukemia stem cells (LSCs) are the primary mediator of relapse (Shlush et al., 2017), TSA 高 Expression of HLA and HLA RNA was also assessed in RNA-seq data from LSCs and sorted blast cells from another study (Corces et al., 2016), and no differences were found between the two cell populations (Figures 4I-J). Nevertheless, by using gene set enrichment analysis (GSEA), we were able to identify a large number of TSAs. 高 Patients expressing TSA were also found to express higher levels of the well-established LSC gene signature (Eppert et al., 2011) (Figure 4K). Collectively, these results support the notion that TSA 高 These results further support the highly immunogenic nature of TSAs and demonstrate that they can be targeted in nearly all AML patients, either at diagnosis or at relapse. 高 It can be concluded that immune targeting of LSCs could be envisaged at any stage of the disease and would have the potential to eliminate LSCs.

[0217] Example 6: Multiple TSAs 高 Presentation of IL-1 correlates with better survival. Second, TSA at diagnosis relative to patient survival 高 Surprisingly, the highest number (top quartile) of TSAs 高 Patients expressing multiple TSAs had significantly better survival than the rest of the cohort (Figure 5A). 高 The survival benefit associated with presenting with HLA-positive tumors remained significant in multivariate analysis, along with other known prognostic factors such as age, cytogenetic risk, and NPM1 and FLT3-ITD mutations (Figure 5B). Importantly, HLA-positive tumors were associated with a survival benefit of 100%.pred The same comparison, performed independently of presentation, showed no difference between high and low expressers (Fig. 11A, B). 高 This means that the protective effect of TAA, HSA, or TSA is HLA-restricted. 低 The same analysis performed on TSA showed no significant effect on survival (Figure 11C-H). 高 These results suggest that IgG4 is sufficiently immunogenic to induce spontaneous anti-AML immune responses.

[0218] TSA 高 The survival advantage provided by their cumulative HLA pred High TSA to demonstrate that the presentation is due 高 Patients and low TSA 高 The log-rank p-values ​​of patients were calculated from the analysis to determine whether increasing TSA 高 The number of TSAs (1–29 out of 58) was calculated after random removal (1000 random permutations / number). 高 The probabilistic removal of HLA-TSA was significantly higher in high expressers (HLA-TSA) compared with low expressers (all other patients). 高 The significant survival advantage of the TSA (top quartile of TSA counts) was rapidly lost (Figure 5C-D). 高 This incremental decrease in survival advantage with the subtraction of TSA 高 This suggests that a large proportion of TSA presented by a larger proportion of patients contributes to this survival advantage. 高 had the greatest effect on the p-value (Figure 5E). Similarly, removing common HLA alleles (shared by >5% of patients) from the log-rank analysis had a greater effect on the p-value than removing low-frequency alleles (Figure 5F). Collectively, these data suggest a significant effect of TSA on patient survival. 高 In the next experiment, we demonstrate that the benefits of TSA presentation are HLA-restricted. 高 We considered the simplest explanation for the survival advantage associated with TSA. 高 teeth 、 Induce a spontaneous anti-AML protective immune response.

[0219] Example 7: TSA高 Presentation elicits a cytotoxic T cell response TSA 高 As a prerequisite for evaluating the immunogenicity (i.e., their ability to induce an immune response) of TAA and other MOIs, we used Repitope, a machine learning algorithm that relies on public TCR databases to predict the probability of a T cell response (Ogishi and Yotsuyanagi, 2019). Using MAPs presented by thymic epithelial cells (Adamopoulou et al., 2013) or MAPs derived from HIV as negative and positive controls, respectively, Repitope's predictions indicated that TAA were largely non-immunogenic, while the other three MOIs were immunogenic, similar to HIV peptides (Figure 6A). Accordingly, TAA showed high expression (approximately 12.1 rphm) in mTECs compared with the other three MOIs and with the set of 1,411 MAPs reported as immunogenic in the IEDB (Figure 6B). All non-TAA MOIs showed very low RNA expression in mTECs compared with other immunogenic peptides, supporting their immunogenicity. To validate the Repitope prediction, an in vitro T cell assay was performed and HLA-A*02:01-presenting TSAs were predicted to be most immunogenic. 高 :Started with ALPVALPSL. Epitope [Sequence Table 1] 945 was used as a positive control in IFN-γ ELISpot because it is one of the most immunogenic human MAPs (Dutoit et al., 2002; Hesnard et al., 2016). [Sequence Table 2] The results were similar to those of JPEG0007822953000028.jpg944 (Figure 6C). 高 ELISpot of these TSAs also supported their immunogenicity (Fig. 6D). 高Cytokine secretion assays and dextramer staining were also performed for TSA, which confirmed the ELISpot results and supported the specificity of the immune response (Figure 6E-F). 高 To further demonstrate that TSA can induce spontaneous and specific T cell clonotype proliferation, 高 A functional proliferation of specific T cells (FEST) assay was performed (Danilova et al., 2018), in which short-term cultures of peripheral blood T cells stimulated with different pools of TSAs were analyzed by TCR sequencing. 高 Each pool induced the specific proliferation of 9–10 different clonotypes, supporting their spontaneous immunogenicity (Figure 6G and Table 5). [Table 7] All TSAs that can be presented by the HLA alleles of healthy donors: HLA-A*02:01, HLA-A*29:02, HLA-B*15:01, HLA-B27:05, HLA-C*01:02, and HLA-C*03:04 高 TCR-seq was generated by Adaptive Biotechnologies, and raw data were processed with the FEST analysis tool (http: / / www.stat-apps.onc.jhmi.edu / FEST). Here, the number of pools, TSAs present in each pool, and 高 , the sequences of each significantly expanded clonotype in each pool are reported, as well as the FDR and odds ratios provided by the FEST analysis tool.

[0220] Next, for "in vivo veritas," we performed a detailed analysis of transcriptome data from 437 Leucegene patients to examine TSA by T cells. 高 We evaluated the in vivo recognition potential of TSAs. First, we evaluated the diversity of TCR repertoires of T cells using the TRUST4 algorithm (Zhang et al., 2019). In contrast to TAAs (used herein as non-immunogenic controls), elevated TSAs were detected. 高 number of pred Presented by anti-TSA高 This was associated with a decline in TCR repertoire diversity, suggesting clonotype expansion (Figure 6H). To demonstrate the specificity of this expansion, we used the ERGO algorithm to predict MOI-TCR interactions (Springer et al., 2020). To identify anti-MOI clonotypes, we used a large number of TSAs with ERGO probability >80%. 高 Patients with anti-TSA also had the highest CDR3 count among all detected CDR3s. 高 The frequency of clonotypes was higher (Figure 6I). A similar correlation was not observed for TAAs (Figure 6J). Next, the MOI presented by each AML sample was compared. pred The proportion of anti-MOI clonotypes capable of recognizing the TCR (i.e., frequency of cognate TCR-MOI interactions) was calculated. pred Normalized according to the number of MOIs (otherwise, a higher percentage of antibodies would naturally be obtained with more MOIs presented). pred This is because the MOI clonotypes are detected. 高 pred We showed that presentation was associated with a dramatically higher frequency of specific T cell recognition than TAAs ( Fig. 6K ).

[0221] Anti-TSA 高 In light of T cell recognition, TSA 高 pred It was reasoned that presentation must be associated with the infiltration of activated CD8 T cells. 高 Transcript diversity was inversely correlated with CD8A and CD8B expression in AML samples; pred This was not correlated with the diversity of presentation (Fig. 6L-M). 高 We suggested that the high transcript diversity reflects a slightly higher blast purity in AML samples (as expected for TSA). To avoid this possible bias, we also compared HLA-TSA 高 The number of expressed TSA 高 Because it is mathematically related to the number of pred Presented by TSA 高The number of expressed TSAs 高 Normalized to the number of transcripts and normalized pred Differential gene expression was analyzed in patients with presentations above or below the median. 高 pred Some of the 123 genes positively associated with presentation, including CD8A, CD8B, GZMA, GZMB, IL2RB, PRF1, and ZAP70, were associated with T cell activation and cytolysis (Figure 6N). Notably, the GO terms associated with these 123 genes were exclusively related to T cell activation and differentiation (Figure 6O). The CD4 gene was not differentially expressed, and none of the GO terms could be significantly associated with the downregulated genes. Therefore, TSA 高 pred It was concluded that presentation was associated with a higher abundance of activated CD8 T cells.

[0222] Example 8: TSA 高 RNA expression is associated with signatures of immunoediting, AML driver mutations, and epigenetic abnormalities TSA 高 Given the potential therapeutic value of TSAs, it is desirable to gain insight into their biogenesis. 高 (HE-TSA 高 ) counts for each Leucegene patient (i.e., a given TSA 高 Across all patients with non-null expression of TSA 高 The counts of genes expressed at levels higher than their median expression were then compared to the expression of each protein-coding gene and HE-TSA. 高 Calculate the expression of specific genes by performing pairwise Pearson correlations between counts and TSA 高 This was done to assess whether the expression of genes involved in MAP presentation (HLA-A, HLA-B, HLA-C, B2M, and NLRC5) could be correlated with the expression of HE-TSA. 高 showed a consistent inverse correlation with the number of TSA 高This suggests the emergence of immunoediting in response to increased expression (Figures 7A and 12A-B). Immunoediting was also supported by a positive correlation with CD47 (an immune checkpoint molecule involved in inhibiting dendritic cell phagocytosis) (Majeti et al., 2009) and CD84 (promoting PD-L1 expression by leukemia cells) (Lewinsky et al., 2018). NPM1 mutations can regulate the expression of PD-L1 (CD274) (Greiner et al., 2017), so NPM1 変異型 and NPM1 野生型 AML patients were analyzed separately. This analysis showed that HE-TSA above the median 高 NPM1 with count 野生型 Patients with HE-TSA below the median 高 We found that patients with PD-L1 expressed significantly higher levels of PD-L1 than patients with PD-L1 (Figure 7B).

[0223] Next, we analyzed gene pathways correlated with HE-TSA counts (Figure 7C and Table 6). Negatively correlated pathways included biological processes involved in cell proliferation (including transport and cellular organization), mitochondrial OXPHOS, and proteasome-mediated protein catabolism. Interestingly, inhibition of mitochondrial activity has been shown to reduce MHC-I expression and can be used as an immune evasion mechanism by cancer cells (Charni et al., 2010). Similarly, inhibition of protein degradation may result in fewer peptides available for presentation by MHC-I molecules, thus reducing TSA counts. 高 (Tripathi et al., 2016). Finally, a reduction in mitosis-related processes could be a side effect of MHC-I downregulation, as both processes are regulated by NLRC5 (Wang et al., 2019). Collectively, these data support the notion that TSA 高 We show that expression of IL-1 is associated with various responses that may function as an immunoediting mechanism for AML cells.

[0224] In contrast to the negatively correlated pathways, the positively correlated pathways were restricted to regulatory processes (Figure 7D). Thus, 16.1% of the positively correlated genes (vs. 2.5% of the negatively correlated genes) were associated with TSA. 高 These were transcription factors that can directly mediate the transcription of TSA. Among them, the most correlated gene was ZNF445 (Figure 7A), a regulator of genomic imprinting (i.e., an epigenetic process associated with DNA methylation) (Takahashi et al., 2019). Because ZNF445 function depends on DNA methylation, it is typically associated with TSA. 高 We investigated possible associations between expression and AML mutations associated with aberrant DNA methylation. The three most frequent AML driver mutations (NPM1 変異型 , FLT3-ITD, and DNMT3A 変異型 ) were tested first, and all three showed high (above the median) HE-TSA 高 We found that the HE-TSA gene was significantly enriched in patients expressing the CR1 gene (Figure 7E). Furthermore, patients with two or three coexisting mutations were significantly more likely to have HE-TSA than patients with one or none. 高 Twelve of the 19 AML specimens used for MS analysis presented either FLT3-ITD or NPM1 mutations. Regarding other frequent AML mutations, IDH2 and biallelic CEBPA mutations were also frequently observed in HE-TSA. 高 We found that mutations in NPM1, DNMT3A, IDH2, and CEBPA were positively associated with elevated counts, whereas mutations in ASXL1, SRSF2, and U2AF1 were negatively associated, and mutations in FLT3-TKD, IDH1, RUNX1, TET2, TP53, and WT1 were not (Figure 12D). 両変異型 Because mutations are associated with abnormal methylation profiles ( Figueroa et al., 2010a , Figueroa et al., 2010b , Ley et al., 2013 ), HE-TSA 高 The correlation with the rise in counts is that TSA 高 Supporting the significance of epigenetic dysregulation in expression. [Table 8-1] [Table 8-2] [Table 8-3] [Table 8-4] [Table 8-5] [Table 8-6]

[0225] Finally, TSA 高 We investigated whether expression could be associated with other clinical features that allow us to predict their presence in AML patients, such as the French-American-British (FAB) subtype (Figure 12E-H). Surprisingly, high counts of HE-TSA 高 Patients expressing TSA were over- and under-represented in M1 and M5 AML, respectively. Thus, patients with differentiated AML without a normal karyotype had the highest levels of TSA. 高 Therefore, this is due to an overestimation of FAB M1 AML in the samples used to detect TSA (9 out of 19), and the 高 Most of the TSAs were located in intron regions (37 / 58, ERE-derived TSAs located in introns). 高 Since introns are located in the AML subtype (including the AML subtype), it was hypothesized that this could be explained by the existence of different intron retention patterns among FAB types. Accordingly, unsupervised consensus clustering performed on introns specifically retained in AML showed clear clustering according to the patients' FAB type (Figure 7G). Collectively, these data support the conclusion that TSA 高 Expression is shown to be associated with AML subtype-specific intron retention patterns.

[0226] While the present invention has been described above by specific embodiments thereof, modifications can be made without departing from the spirit and nature of the invention, as defined in the appended claims. In the claims, the word "comprising" is used as an open-ended term that is substantially equivalent to the phrase "including but not limited to." The singular forms "a," "an," and "the" include the corresponding plural referents unless the context clearly dictates otherwise. References Adamopoulou, E., Tenzer, S., Hillen, N., Klug, P., Rota, IA, Tietz, S., Gebhardt, M., Stevanovic, S., Schild, H., Tolosa, E., et al. (2013). Exploring the MHC-peptide matrix of central tolerance in the human thymus. Nat Commun4, 2039. Andreatta, M., and Nielsen, M. (2016). Gapped sequence alignment using artificial neural networks: application to the MHC class I system. Bioinformatics32, 511-517. Audemard, EO, Gendron, P., Feghaly, A., Lavallee, VP, Hebert, J., Sauvageau, G., and Lemieux, S. (2019).Targeted variant detection using unaligned RNA-Seq reads.Life Sci Alliance2. Audoux,J.,Philippe,N.,Chikhi,R.,Salson,M.,Gallopin,M.,Gabriel,M.,Le Coz,J.,Drouineau,E.,Commes,T.,and Gautheret,D.(2017).DE-kupl:exhaustive capture of biological variation in RNA-seq data through k-mer decomposition.Genome Biology18,243. Avigan,D.,and Rosenblatt,J.(2018).Vaccine therapy in hematologic malignancies.Blood131,2640-2650. Bamezai,S.,Rawat,V.P.,and Buske,C.(2012).Concise review:The Piwi-piRNA axis:pivotal beyond transposon silencing.Stem Cells30,2603-2611. Berlin,C.,Kowalewski,D.J.,Schuster,H.,Mirza,N.,Walz,S.,Handel,M.,Schmid-Horch,B.,Salih,H.R.,Kanz,L.,Rammensee,H.G.,et al.(2015).Mapping the HLA ligandome landscape of acute myeloid leukemia:a targeted approach toward peptide-based immunotherapy.Leukemia29,647-659. Boegel,S.,Lower,M.,Bukur,T.,Sorn,P.,Castle,J.C.,and Sahin,U.(2018).HLA and proteasome expression body map.BMC Med Genomics11,36. Bouwmeester,R.,Gabriels,R.,Hulstaert,N.,Martens,L.,and Degroeve,S.(2020).DeepLC can predict retention times for peptides that carry as-yet unseen modifications.bioRxiv,2020.2003.2028.013003. Bray,N.L.,Pimentel,H.,Melsted,P.,and Pachter,L.(2016).Near-optimal probabilistic RNA-seq quantification.Nature Biotechnology34,525-527. Chapuis,A.G.,Egan,D.N.,Bar,M.,Schmitt,T.M.,McAfee,M.S.,Paulson,K.G.,Voillet,V.,Gottardo,R.,Ragnarsson,G.B.,Bleakley,M.,et al.(2019).T cell receptor gene therapy targeting WT1 prevents acute myeloid leukemia relapse post-transplant.Nat Med25,1064-1072. Charni,S.,de Bettignies,G.,Rathore,M.G.,Aguilo,J.I.,van den Elsen,P.J.,Haouzi,D.,Hipskind,R.A.,Enriquez,J.A.,Sanchez-Beato,M.,Pardo,J.,et al.(2010).Oxidative Phosphorylation Induces De Novo Expression of the MHC Class I in Tumor Cells through the ERK5 Pathway.The Journal of Immunology185,3498-3503. Chen,J.,Brunner,A.D.,Cogan,J.Z.,Nunez,J.K.,Fields,A.P.,Adamson,B.,Itzhak,D.N.,Li,J.Y.,Mann,M.,Leonetti,M.D.,and Weissman,J.S.(2020).Pervasive functional translation of noncanonical human open reading frames.Science367,1140-1146. Chong,C.,Muller,M.,Pak,H.,Harnett,D.,Huber,F.,Grun,D.,Leleu,M.,Auger,A.,Arnaud,M.,Stevenson,B.J.,et al.(2020).Integrated proteogenomic deep sequencing and analytics accurately identify non-canonical peptides in tumor immunopeptidomes.Nat Commun11,1293. Corces,M.R.,Buenrostro,J.D.,Wu,B.,Greenside,P.G.,Chan,S.M.,Koenig,J.L.,Snyder,M.P.,Pritchard,J.K.,Kundaje,A.,Greenleaf,W.J.,et al.(2016).Lineage-specific and single-cell chromatin accessibility charts human hematopoiesis and leukemia evolution.Nat Genet48,1193-1203. Cosma,G.L.,and Eisenlohr,L.C.(2019).Impact of epitope density on CD8(+)T cell development and function.Mol Immunol113,120-125. Coulie,P.G.,Van den Eynde,B.J.,van der Bruggen,P.,and Boon,T.(2014).Tumour antigens recognized by T lymphocytes:at the core of cancer immunotherapy.Nat Rev Cancer14,135-146. Courcelles,M.,Durette,C.,Daouda,T.,Laverdure,J.P.,Vincent,K.,Lemieux,S.,Perreault,C.,and Thibault,P.(2020).MAPDP:A Cloud-Based Computational Platform for Immunopeptidomics Analyses.J Proteome Res. Danilova,L.,Anagnostou,V.,Caushi,J.X.,Sidhom,J.W.,Guo,H.,Chan,H.Y.,Suri,P.,Tam,A.,Zhang,J.,Asmar,M.E.,et al.(2018).The Mutation-Associated Neoantigen Functional Expansion of Specific T Cells(MANAFEST)Assay:A Sensitive Platform for Monitoring Antitumor Immunity.Cancer Immunol Res6,888-899. Daouda,T.,Perreault,C.,and Lemieux,S.(2016).pyGeno:A Python package for precision medicine and proteogenomics.F1000Res 5,381. Di Stasi,A.,Jimenez,A.M.,Minagawa,K.,Al-Obaidi,M.,and Rezvani,K.(2015).Review of the Results of WT1 Peptide Vaccination Strategies for Myelodysplastic Syndromes and Acute Myeloid Leukemia from Nine Different Studies.Frontiers in Immunology 6. Dobin,A.,Davis,C.A.,Schlesinger,F.,Drenkow,J.,Zaleski,C.,Jha,S.,Batut,P.,Chaisson,M.,and Gingeras,T.R.(2013).STAR:ultrafast universal RNA-seq aligner.Bioinformatics29,15-21. Dutoit,V.,Rubio-Godoy,V.,Pittet,M.J.,Zippelius,A.,Dietrich,P.Y.,Legal,F.A.,Guillaume,P.,Romero,P.,Cerottini,J.C.,Houghten,R.A.,et al.(2002).Degeneracy of antigen recognition as the molecular basis for the high frequency of naive A2 / Melan-a peptide multimer(+)CD8(+)T cells in humans.J Exp Med196,207-216. Dvinge,H.,and Bradley,R.K.(2015).Widespread intron retention diversifies most cancer transcriptomes.Genome medicine7,45-45. Efremova,M.,Finotello,F.,Rieder,D.,and Trajanoski,Z.(2017).Neoantigens Generated by Individual Mutations and Their Role in Cancer Immunity and Immunotherapy.Front Immunol8,1679. Egen,J.G.,Ouyang,W.,and Wu,L.C.(2020).Human Anti-tumor Immunity:Insights from Immunotherapy Clinical Trials.Immunity52,36-54. Ehx,G.,and Perreault,C.(2019).Discovery and characterization of actionable tumor antigens.Genome Medicine11,29. Elias,J.E.,and Gygi,S.P.(2010).Target-decoy search strategy for mass spectrometry-based proteomics.Methods Mol Biol604,55-71. Eng,J.K.,Hoopmann,M.R.,Jahan,T.A.,Egertson,J.D.,Noble,W.S.,and MacCoss,M.J.(2015).A deeper look into Comet--implementation and features.J Am Soc Mass Spectrom26,1865-1874. Eppert,K.,Takenaka,K.,Lechman,E.R.,Waldron,L.,Nilsson,B.,van Galen,P.,Metzeler,K.H.,Poeppl,A.,Ling,V.,Beyene,J.,et al.(2011).Stem cell gene expression programs influence clinical outcome in human leukemia.Nat Med17,1086-1093. Fennell,K.A.,Bell,C.C.,and Dawson,M.A.(2019).Epigenetic therapies in acute myeloid leukemia:where to from here? Blood134,1891-1901. Fergusson,J.R.,Morgan,M.D.,Bruchard,M.,Huitema,L.,Heesters,B.A.,van Unen,V.,van Hamburg,J.P.,van der Wel,N.N.,Picavet,D.,Koning,F.,et al.(2018).Maturing Human CD127+CCR7+PDL1+Dendritic Cells Express AIRE in the Absence of Tissue Restricted Antigens.Front Immunol9,2902. Figueroa,M.E.,Abdel-Wahab,O.,Lu,C.,Ward,P.S.,Patel,J.,Shih,A.,Li,Y.,Bhagwat,N.,Vasanthakumar,A.,Fernandez,H.F.,et al.(2010a).Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype,disrupt TET2 function,and impair hematopoietic differentiation.Cancer Cell18,553-567. Figueroa,M.E.,Lugthart,S.,Li,Y.,Erpelinck-Verschueren,C.,Deng,X.,Christos,P.J.,Schifano,E.,Booth,J.,van Putten,W.,Skrabanek,L.,et al.(2010b).DNA methylation signatures identify biologically distinct subtypes in acute myeloid leukemia.Cancer Cell17,13-27. Garrison,E.,and Marth,G.(2012).Haplotype-based variant detection from short-read sequencing.arXiv:Genomics. Gaujoux,R.,and Seoighe,C.(2010).A flexible R package for nonnegative matrix factorization.BMC Bioinformatics11,367. Greiner,J.,Hofmann,S.,Schmitt,M.,Gotz,M.,Wiesneth,M.,Schrezenmeier,H.,Bunjes,D.,Dohner,H.,and Bullinger,L.(2017).Acute myeloid leukemia with mutated nucleophosmin1:an immunogenic acute myeloid leukemia subtype and potential candidate for immune checkpoint inhibition.Haematologica 102,e499-e501. Gu,M.,Zwiebel,M.,Ong,S.H.,Boughton,N.,Nomdedeu,J.,Basheer,F.,Nannya,Y.,Quiros,P.M.,Ogawa,S.,Cazzola,M.,et al.(2020).RNAmut:robust identification of somatic mutations in acute myeloid leukemia using RNA-sequencing.Haematologica105,e290-e293. Gutierrez,S.E.,and Romero-Oliva,F.A.(2013).Epigenetic changes:a common theme in acute myelogenous leukemogenesis.J Hematol Oncol6,57. Hardy,M.-P.,Vincent,K.,and Perreault,C.(2019).The Genomic Landscape of Antigenic Targets for T Cell-Based Leukemia Immunotherapy.Frontiers in Immunology10,2934. He,H.,Kondo,Y.,Ishiyama,K.,Alatrash,G.,Lu,S.,Cox,K.,Qiao,N.,Clise-Dwyer,K.,St John,L.,Sukhumalchandra,P.,et al.(2020).Two unique HLA-A*0201 restricted peptides derived from cyclin E as immunotherapeutic targets in leukemia.Leukemia. Hesnard,L.,Legoux,F.,Gautreau,L.,Moyon,M.,Baron,O.,Devilder,M.C.,Bonneville,M.,and Saulquin,X.(2016).Role of the MHC restriction during maturation of antigen-specific human T cells in the thymus.Eur J Immunol46,560-569. Janelle,V.,Carli,C.,Taillefer,J.,Orio,J.,and Delisle,J.S.(2015).Defining novel parameters for the optimal priming and expansion of minor histocompatibility antigen-specific T cells in culture.J Transl Med13,123. Jung,N.,Dai,B.,Gentles,A.J.,Majeti,R.,and Feinberg,A.P.(2015).An LSC epigenetic signature is largely mutation independent and implicates the HOXA cluster in AML pathogenesis.Nature Communications6,8489. Knaus,H.A.,Berglund,S.,Hackl,H.,Blackford,A.L.,Zeidner,J.F.,Montiel-Esparza,R.,Mukhopadhyay,R.,Vanura,K.,Blazar,B.R.,Karp,J.E.,et al.(2018).Signatures of CD8+T cell dysfunction in AML patients and their reversibility with response to chemotherapy.JCI Insight3. Krokhin,O.V.(2006).Sequence-specific retention calculator.Algorithm for peptide retention prediction in ion-pair RP-HPLC:application to300-and100-A pore size C18 sorbents.Anal Chem78,7785-7795. Lamoliatte,F.,McManus,F.P.,Maarifi,G.,Chelbi-Alix,M.K.,and Thibault,P.(2017).Uncovering the SUMOylation and ubiquitylation crosstalk in human cells using sequential peptide immunopurification.Nat Commun8,14109. Larouche,J.D.,Trofimov,A.,Hesnard,L.,Ehx,G.,Zhao,Q.,Vincent,K.,Durette,C.,Gendron,P.,Laverdure,J.P.,Bonneil,E.,et al.(2020).Widespread and tissue-specific expression of endogenous retroelements in human somatic tissues.Genome Med12,40. Laumont,C.M.,Daouda,T.,Laverdure,J.P.,Bonneil,E.,Caron-Lizotte,O.,Hardy,M.P.,Granados,D.P.,Durette,C.,Lemieux,S.,Thibault,P.,and Perreault,C.(2016).Global proteogenomic analysis of human MHC class I-associated peptides derived from non-canonical reading frames.Nat Commun7,10238. Laumont,C.M.,Vincent,K.,Hesnard,L.,Audemard,E.,Bonneil,E.,Laverdure,J.P.,Gendron,P.,Courcelles,M.,Hardy,M.P.,Cote,C.,et al.(2018).Noncoding regions are the main source of targetable tumor-specific antigens.Sci Transl Med10. Lavallee,V.P.,Baccelli,I.,Krosl,J.,Wilhelm,B.,Barabe,F.,Gendron,P.,Boucher,G.,Lemieux,S.,Marinier,A.,Meloche,S.,et al.(2015).The transcriptomic landscape and directed chemical interrogation of MLL-rearranged acute myeloid leukemias.Nat Genet47,1030-1037. Lavallee,V.P.,Krosl,J.,Lemieux,S.,Boucher,G.,Gendron,P.,Pabst,C.,Boivin,I.,Marinier,A.,Guidos,C.J.,Meloche,S.,et al.(2016).Chemo-genomic interrogation of CEBPA mutated AML reveals recurrent CSF3R mutations and subgroup sensitivity to JAK inhibitors.Blood127,3054-3061. Lawrence,M.S.,Stojanov,P.,Polak,P.,Kryukov,G.V.,Cibulskis,K.,Sivachenko,A.,Carter,S.L.,Stewart,C.,Mermel,C.H.,Roberts,S.A.,et al.(2013).Mutational heterogeneity in cancer and the search for new cancer-associated genes.Nature499,214-218. Legat,A.,Maby-El Hajjami,H.,Baumgaertner,P.,Cagnon,L.,Abed Maillard,S.,Geldhof,C.,Iancu,E.M.,Lebon,L.,Guillaume,P.,Dojcinovic,D.,et al.(2016).Vaccination with LAG-3Ig(IMP321)and Peptides Induces Specific CD4 and CD8 T-Cell Responses in Metastatic Melanoma Patients--Report of a Phase I / IIa Clinical Trial.Clin Cancer Res22,1330-1340. Lewinsky,H.,Barak,A.F.,Huber,V.,Kramer,M.P.,Radomir,L.,Sever,L.,Orr,I.,Mirkin,V.,Dezorella,N.,Shapiro,M.,et al.(2018).CD84 regulates PD-1 / PD-L1 expression and function in chronic lymphocytic leukemia.J Clin Invest128,5465-5478. Ley,T.J.,Miller,C.,Ding,L.,Raphael,B.J.,Mungall,A.J.,Robertson,A.,Hoadley,K.,Triche,T.J.,Jr.,Laird,P.W.,Baty,J.D.,et al.(2013).Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia.N Engl J Med368,2059-2074. Li,B.,Li,T.,Wang,B.,Dou,R.,Zhang,J.,Liu,J.S.,and Liu,X.S.(2017).Ultrasensitive detection of TCR hypervariable-region sequences in solid-tissue RNA-seq data.Nat Genet49,482-483. Li,H.(2011).A statistical framework for SNP calling,mutation discovery,association mapping and population genetical parameter estimation from sequencing data.Bioinformatics27,2987-2993. Li,H.,Handsaker,B.,Wysoker,A.,Fennell,T.,Ruan,J.,Homer,N.,Marth,G.,Abecasis,G.,and Durbin,R.(2009).The Sequence Alignment / Map format and SAMtools.Bioinformatics25,2078-2079. Li,S.,Garrett-Bakelman,F.E.,Chung,S.S.,Sanders,M.A.,Hricik,T.,Rapaport,F.,Patel,J.,Dillon,R.,Vijay,P.,Brown,A.L.,et al.(2016).Distinct evolution and dynamics of epigenetic and genetic heterogeneity in acute myeloid leukemia.Nat Med22,792-799. Loffler,M.W.,Mohr,C.,Bichmann,L.,Freudenmann,L.K.,Walzer,M.,Schroeder,C.M.,Trautwein,N.,Hilke,F.J.,Zinser,R.S.,Muhlenbruch,L.,et al.(2019).Multi-omics discovery of exome-derived neoantigens in hepatocellular carcinoma.Genome Med 11,28. Logtenberg,M.E.W.,Scheeren,F.A.,and Schumacher,T.N.(2020).The CD47-SIRPα Immune Checkpoint.Immunity52,742-752. Luo,K.,Yuan,J.,Shan,Y.,Li,J.,Xu,M.,Cui,Y.,Tang,W.,Wan,B.,Zhang,N.,Wu,Y.,and Yu,L.(2006).Activation of transcriptional activities of AP1 and SRE by a novel zinc finger protein ZNF445.Gene367,89-100. Macrae,T.,Sargeant,T.,Lemieux,S.,Hebert,J.,Deneault,E.,and Sauvageau,G.(2013).RNA-Seq reveals spliceosome and proteasome genes as most consistent transcripts in human cancer cells.PLoS One8,e72884. Maere,S.,Heymans,K.,and Kuiper,M.(2005).BiNGO:a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks.Bioinformatics21,3448-3449. Maiga,A.,Lemieux,S.,Pabst,C.,Lavallee,V.P.,Bouvier,M.,Sauvageau,G.,and Hebert,J.(2016).Transcriptome analysis of G protein-coupled receptors in distinct genetic subgroups of acute myeloid leukemia:identification of potential disease-specific targets.Blood Cancer J6,e431. Marcais,G.,and Kingsford,C.(2011).A fast,lock-free approach for efficient parallel counting of occurrences of k-mers.Bioinformatics27,764-770. Maslak,P.G.,Dao,T.,Bernal,Y.,Chanel,S.M.,Zhang,R.,Frattini,M.,Rosenblat,T.,Jurcic,J.G.,Brentjens,R.J.,Arcila,M.E.,et al.(2018).Phase2 trial of a multivalent WT1 peptide vaccine(galinpepimut-S)in acute myeloid leukemia.Blood Adv2,224-234. Merico,D.,Isserlin,R.,Stueker,O.,Emili,A.,and Bader,G.D.(2010).Enrichment map:a network-based method for gene-set enrichment visualization and interpretation.PLoS One5,e13984. Middleton,R.,Gao,D.,Thomas,A.,Singh,B.,Au,A.,Wong,J.J.,Bomane,A.,Cosson,B.,Eyras,E.,Rasko,J.E.,and Ritchie,W.(2017).IRFinder:assessing the impact of intron retention on mammalian gene expression.Genome Biol18,51. Ntziachristos,P.,Abdel-Wahab,O.,and Aifantis,I.(2016).Emerging concepts of epigenetic dysregulation in hematological malignancies.Nat Immunol17,1016-1024. Ogishi,M.,and Yotsuyanagi,H.(2019).Quantitative Prediction of the Landscape of T Cell Epitope Immunogenicity in Sequence Space.Frontiers in Immunology10. Pabst,C.,Bergeron,A.,Lavallee,V.P.,Yeh,J.,Gendron,P.,Norddahl,G.L.,Krosl,J.,Boivin,I.,Deneault,E.,Simard,J.,et al.(2016).GPR56 identifies primary human acute myeloid leukemia cells with high repopulating potential in vivo.Blood127,2018-2027. Papaemmanuil,E.,Gerstung,M.,Bullinger,L.,Gaidzik,V.I.,Paschka,P.,Roberts,N.D.,Potter,N.E.,Heuser,M.,Thol,F.,Bolli,N.,et al.(2016).Genomic Classification and Prognosis in Acute Myeloid Leukemia.N Engl J Med374,2209-2221. Pearson,H.,Daouda,T.,Granados,D.P.,Durette,C.,Bonneil,E.,Courcelles,M.,Rodenbrock,A.,Laverdure,J.P.,Cote,C.,Mader,S.,et al.(2016).MHC class I-associated peptides derive from selective regions of the human genome.J Clin Invest126,4690-4701. Qazilbash,M.H.,Wieder,E.,Thall,P.F.,Wang,X.,Rios,R.,Lu,S.,Kanodia,S.,Ruisaard,K.E.,Giralt,S.A.,Estey,E.H.,et al.(2017).PR1 peptide vaccine induces specific immunity with clinical responses in myeloid malignancies.Leukemia31,697-704. Quinlan,A.R.,and Hall,I.M.(2010).BEDTools:a flexible suite of utilities for comparing genomic features.Bioinformatics26,841-842. Rashidi,A.,and Walter,R.B.(2016).Antigen-specific immunotherapy for acute myeloid leukemia:where are we now,and where do we go from here? Expert Rev Hematol9,335-350. Ritchie,M.E.,Phipson,B.,Wu,D.,Hu,Y.,Law,C.W.,Shi,W.,and Smyth,G.K.(2015).limma powers differential expression analyses for RNA-sequencing and microarray studies.Nucleic Acids Res43,e47. Robinson,J.T.,Thorvaldsdottir,H.,Winckler,W.,Guttman,M.,Lander,E.S.,Getz,G.,and Mesirov,J.P.(2011).Integrative genomics viewer.Nat Biotechnol29,24-26. Robinson,M.D.,McCarthy,D.J.,and Smyth,G.K.(2010).edgeR:a Bioconductor package for differential expression analysis of digital gene expression data.Bioinformatics26,139-140. Rogers,M.F.,Shihab,H.A.,Mort,M.,Cooper,D.N.,Gaunt,T.R.,and Campbell,C.(2018).FATHMM-XF:accurate prediction of pathogenic point mutations via extended features.Bioinformatics34,511-513. Sarkizova,S.,Klaeger,S.,Le,P.M.,Li,L.W.,Oliveira,G.,Keshishian,H.,Hartigan,C.R.,Zhang,W.,Braun,D.A.,Ligon,K.L.,et al.(2020).A large peptidome dataset improves HLA class I epitope prediction across most of the human population.Nat Biotechnol38,199-209. Schulz,W.A.,Steinhoff,C.,and Florl,A.R.(2006).Methylation of endogenous human retroelements in health and disease.Curr Top Microbiol Immunol310,211-250. Shannon,P.,Markiel,A.,Ozier,O.,Baliga,N.S.,Wang,J.T.,Ramage,D.,Amin,N.,Schwikowski,B.,and Ideker,T.(2003).Cytoscape:a software environment for integrated models of biomolecular interaction networks.Genome Res13,2498-2504. Shao,W.,Pedrioli,P.G.A.,Wolski,W.,Scurtescu,C.,Schmid,E.,Vizcaino,J.A.,Courcelles,M.,Schuster,H.,Kowalewski,D.,Marino,F.,et al.(2018).The SysteMHC Atlas project.Nucleic Acids Res46,D1237-d1247. Shlush,L.I.,Mitchell,A.,Heisler,L.,Abelson,S.,Ng,S.W.K.,Trotman-Grant,A.,Medeiros,J.J.F.,Rao-Bhatia,A.,Jaciw-Zurakowsky,I.,Marke,R.,et al.(2017).Tracing the origins of relapse in acute myeloid leukaemia to stem cells.Nature547,104-108. Smart,A.C.,Margolis,C.A.,Pimentel,H.,He,M.X.,Miao,D.,Adeegbe,D.,Fugmann,T.,Wong,K.-K.,and Van Allen,E.M.(2018).Intron retention is a source of neoepitopes in cancer.Nature Biotechnology36,1056-1058. Smith,C.C.,Selitsky,S.R.,Chai,S.,Armistead,P.M.,Vincent,B.G.,and Serody,J.S.(2019).Alternative tumour-specific antigens.Nat Rev Cancer19,465-478. Springer,I.,Besser,H.,Tickotsky-Moskovitz,N.,Dvorkin,S.,and Louzoun,Y.(2020).Prediction of Specific TCR-Peptide Binding From Large Dictionaries of TCR-Peptide Pairs.Front Immunol11,1803. Szolek,A.,Schubert,B.,Mohr,C.,Sturm,M.,Feldhahn,M.,and Kohlbacher,O.(2014).OptiType:precision HLA typing from next-generation sequencing data.Bioinformatics30,3310-3316. Takahashi,N.,Coluccio,A.,Thorball,C.W.,Planet,E.,Shi,H.,Offner,S.,Turelli,P.,Imbeault,M.,Ferguson-Smith,A.C.,and Trono,D.(2019).ZNF445 is a primary regulator of genomic imprinting.Genes Dev33,49-54. Tamura,H.,Dan,K.,Tamada,K.,Nakamura,K.,Shioi,Y.,Hyodo,H.,Wang,S.D.,Dong,H.,Chen,L.,and Ogata,K.(2005).Expression of functional B7-H2 and B7.2 costimulatory molecules and their prognostic implications in de novo acute myeloid leukemia.Clin Cancer Res11,5708-5717. Tate,J.G.,Bamford,S.,Jubb,H.C.,Sondka,Z.,Beare,D.M.,Bindal,N.,Boutselakis,H.,Cole,C.G.,Creatore,C.,Dawson,E.,et al.(2018).COSMIC:the Catalogue Of Somatic Mutations In Cancer.Nucleic Acids Research47,D941-D947. Thomsen,M.,Lundegaard,C.,Buus,S.,Lund,O.,and Nielsen,M.(2013).MHCcluster,a method for functional clustering of MHC molecules.Immunogenetics65,655-665. Toffalori,C.,Zito,L.,Gambacorta,V.,Riba,M.,Oliveira,G.,Bucci,G.,Barcella,M.,Spinelli,O.,Greco,R.,Crucitti,L.,et al.(2019).Immune signature drives leukemia escape and relapse after hematopoietic cell transplantation.Nat Med25,603-611. Tripathi,S.C.,Peters,H.L.,Taguchi,A.,Katayama,H.,Wang,H.,Momin,A.,Jolly,M.K.,Celiktas,M.,Rodriguez-Canales,J.,Liu,H.,et al.(2016).Immunoproteasome deficiency is a feature of non-small cell lung cancer with a mesenchymal phenotype and is associated with a poor outcome.Proc Natl Acad Sci U S A113,E1555-1564. van der Lee,D.I.,Reijmers,R.M.,Honders,M.W.,Hagedoorn,R.S.,de Jong,R.C.,Kester,M.G.,van der Steen,D.M.,de Ru,A.H.,Kweekel,C.,Bijen,H.M.,et al.(2019).Mutated nucleophosmin 1 as immunotherapy target in acute myeloid leukemia.J Clin Invest129,774-785. Vasu,S.,Kohlschmidt,J.,Mrozek,K.,Eisfeld,A.K.,Nicolet,D.,Sterling,L.J.,Becker,H.,Metzeler,K.H.,Papaioannou,D.,Powell,B.L.,et al.(2018).Ten-year outcome of patients with acute myeloid leukemia not treated with allogeneic transplantation in first complete remission.Blood Adv2,1645-1650. Vizcaino,J.A.,Csordas,A.,del-Toro,N.,Dianes,J.A.,Griss,J.,Lavidas,I.,Mayer,G.,Perez-Riverol,Y.,Reisinger,F.,Ternent,T.,et al.(2016).2016 update of the PRIDE database and its related tools.Nucleic Acids Res44,D447-456. Wang,E.,Lu,S.X.,Pastore,A.,Chen,X.,Imig,J.,Chun-Wei Lee,S.,Hockemeyer,K.,Ghebrechristos,Y.E.,Yoshimi,A.,Inoue,D.,et al.(2019a).Targeting an RNA-Binding Protein Network in Acute Myeloid Leukemia.Cancer Cell35,369-384.e367. Wang,Q.,Ding,H.,He,Y.,Li,X.,Cheng,Y.,Xu,Q.,Yang,Y.,Liao,G.,Meng,X.,Huang,C.,and Li,J.(2019b).NLRC5 mediates cell proliferation,migration,and invasion by regulating the Wnt / beta-catenin signalling pathway in clear cell renal cell carcinoma.Cancer Lett444,9-19. Whiteway,A.,Corbett,T.,Anderson,R.,Macdonald,I.,and Prentice,H.G.(2003).Expression of co-stimulatory molecules on acute myeloid leukaemia blasts may effect duration of first remission.Br J Haematol120,442-451. Wong,J.J.L.,Gao,D.,Nguyen,T.V.,Kwok,C.-T.,van Geldermalsen,M.,Middleton,R.,Pinello,N.,Thoeng,A.,Nagarajah,R.,Holst,J.,et al.(2017).Intron retention is regulated by altered MeCP2-mediated splicing factor recruitment.Nature Communications8,15134. Wu,T.D.,and Nacu,S.(2010).Fast and SNP-tolerant detection of complex variants and splicing in short reads.Bioinformatics26,873-881. Yang,L.,Rau,R.,and Goodell,M.A.(2015).DNMT3A in haematological malignancies.Nat Rev Cancer15,152-165. Zhang,J.,Hu,X.,Wang,J.,Sahu,A.D.,Cohen,D.,Song,L.,Ouyang,Z.,Fan,J.,Wang,B.,Fu,J.,et al.(2019).Immune receptor repertoires in pediatric and adult acute myeloid leukemia.Genome Med11,73. Zhao,Q.,Laverdure,J.P.,Lanoix,J.,Durette,C.,Cote,C.,Bonneil,E.,Laumont,C.M.,Gendron,P.,Vincent,K.,Courcelles,M.,et al.(2020).Proteogenomics Uncovers a Vast Repertoire of Shared Tumor-Specific Antigens in Ovarian Cancer.Cancer Immunol Res. Zhi,H.,Ning,S.,Li,X.,Li,Y.,Wu,W.,and Li,X.(2014).A novel reannotation strategy for dissecting DNA methylation patterns of human long intergenic non-coding RNAs in cancers.Nucleic Acids Res42,8258-8270. Zhou,J.,and Chng,W.J.(2017).Aberrant RNA splicing and mutations in spliceosome complex in acute myeloid leukemia.Stem Cell Investig4,6. Zhou,Y.,Lu,Y.,and Tian,W.(2012).Epigenetic features are significantly associated with alternative splicing.BMC Genomics13,123. Wilson CS,Davidson GS,Martin SB,Andries E,Potter J,Harvey R,et al.Gene expression profiling of adult acute myeloid leukemia identifies novel biologic clusters for risk classification and outcome prediction.Blood2006;108(2):685-696. Liu L,Xu F,Chang C-K,He Q,Wu L-Y,Zhang Z,et al.MYCN contributes to the malignant characteristics of erythroleukemia through EZH2-mediated epigenetic repression of p21.Cell Death&Disease2017 2017 / 10 / 01;8(10):e3126-e3126. Yang X,Lu B,Sun X,Han C,Fu C,Xu K,et al.ANP32A regulates histone H3 acetylation and promotes leukemogenesis.Leukemia2018 2018 / 07 / 01;32(7):1587-1597. de Sa Machado Araujo G,da Silva Francisco Junior R,dos Santos Ferreira C,Mozer Rodrigues PT,Terra Machado D,Louvain de Souza T,et al.Maternal 5mCpG Imprints at the PARD6G-AS1 and GCSAML Differentially Methylated Regions Are Decoupled From Parent-of-Origin Expression Effects in Multiple Human Tissues.Frontiers in Genetics2018 2018-March-01;9(36). Lagus H,Klaas M,Juteau S,Elomaa O,Kere J,Vuola J,et al.Discovery of increased epidermal DNAH10 expression after regeneration of dermis in a randomized with-in person trial-reflections on psoriatic inflammation.Scientific Reports2019 2019 / 12 / 13;9(1):19136. Zhang X,Dong W,Zhou H,Li H,Wang N,Miao X,et al.alpha-2,8-Sialyltransferase Is Involved in the Development of Multidrug Resistance via PI3K / Akt Pathway in Human Chronic Myeloid Leukemia.IUBMB Life2015 Feb;67(2):77-87. Johnson KD,Kong G,Gao X,Chang Y-I,Hewitt KJ,Sanalkumar R,et al.Cis-regulatory mechanisms governing stem and progenitor cell transitions.Science advances2015;1(8):e1500503-e1500503.

Claims

1. An acute myeloid leukemia (AML) tumor-specific antigen (TSA) comprising the amino acid sequence set forth in SEQ ID NO:

123.

2. The AML TSA described in claim 1, consisting of the amino acid sequence set forth in SEQ ID NO:

123.

3. A composition comprising an AML TSA as defined in claim 1 or 2 and at least one additional TSA comprising one of the amino acid sequences set forth in SEQ ID NOs: 1-122 and 124-190.

4. A nucleic acid encoding the AML TSA of claim 1 or 2.

5. The nucleic acid of claim 4 which is mRNA or is contained in a viral vector.

6. 6. A vehicle or carrier comprising the AML TSA of claim 1 or 2, the composition of claim 3, or the nucleic acid of claim 4 or 5.

7. The vehicle of claim 6 which is a liposome.

8. A pharmaceutical composition comprising the AML TSA of claim 1 or 2, the composition of claim 3, the nucleic acid of claim 4 or 5, or the vehicle or carrier of claim 6 or 7, and a pharmaceutically acceptable carrier.

9. 10. A vaccine comprising the AML TSA of claim 1 or 2, the composition of claim 3, the nucleic acid of claim 4 or 5, the vehicle or carrier of claim 6 or 7, or the pharmaceutical composition of claim 8, and an adjuvant.

10. 1. An isolated cell expressing on its surface a major histocompatibility complex (MHC) class I molecule, the MHC class I molecule comprising the AML TSA of claim 1 or 2 within its peptide-binding groove.

11. The cell of claim 10, which is an antigen-presenting cell (APC).

12. The cell of claim 11 , wherein the APC is a dendritic cell.

13. A T cell receptor (TCR) that specifically recognizes an MHC class I molecule expressed on the surface of the cell according to any one of claims 10 to 12.

14. An isolated cell, which expresses the TCR of claim 13 on its cell surface.

15. CD8 ＋ 15. The isolated cell of claim 14, which is a T lymphocyte.

16. A cell population comprising at least 0.5% isolated cells as defined in claim 14 or 15.

17. 1. A pharmaceutical composition for use in treating acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS) in a subject, comprising: (i) an AML TSA according to claim 1 or 2; (ii) the composition of claim 3; (iii) a nucleic acid according to claim 4 or 5; (iv) a vehicle or carrier according to claim 6 or 7; (v) the vaccine of claim 9 ; (vi) a cell according to any one of claims 10 to 12, 14 and 15; (vii) a TCR according to claim 13, or (viii) The cell population according to claim 16 A pharmaceutical composition comprising:

18. 18. The pharmaceutical composition for use according to claim 17, which is a pharmaceutical composition for use in the treatment of AML.

19. 19. A pharmaceutical composition for use according to claim 17 or 18, for use in combination with at least one additional anti-tumour agent or therapy.

20. 20. The pharmaceutical composition for use according to claim 19, wherein the at least one additional anti-tumor agent or therapy is a chemotherapeutic agent, immunotherapy, immune checkpoint inhibitor, radiation therapy, or surgery.

Images