Immunotherapy targeting neoantigen peptides derived from tumor transposon elements in glioblastoma

Abstract

The present disclosure provides shared neoantigen peptides derived from tumor-specific transposon element expression, as well as nucleic acids, vaccines, antibodies and immune cells useful in cancer treatment.

Description

Technical Field

[0001] The present disclosure provides shared neoantigenic peptides derived from expression of tumor-specific transposable elements, as well as nucleic acids, vaccines, antibodies and immune cells that can be used for cancer treatment. Background Art

[0002] Harnessing the immune system to generate an effective response against tumors is a central goal of cancer immunotherapy.

[0003] Part of an effective immune response involves T lymphocytes specific for tumor antigens. T cell activation requires interaction with antigen presenting cells (APCs), usually dendritic cells (DCs), which express TCR cognate peptides presented in the context of major histocompatibility molecules (MHC) and co-stimulatory signals. Tumors often contain infiltrating T lymphocytes that react with tumor cells. Subsequently, activated T cells can recognize peptide-MHC complexes presented by all cell types, even malignant cells.

[0004] It is widely accepted that T cells can control (and sometimes repel) solid tumors, especially after immune checkpoint blockade (ICB). In fact, the development of checkpoint blockade therapies has provided methods to bypass some of these mechanisms, thereby killing cancer cells more effectively. The promising results generated by this approach have opened new avenues for the development of T cell-based immunotherapies.

[0005] However, the nature of some of the tumor antigens targeted by these T cells remains unclear. Following the identification of differentiation antigens and tumor testis antigens decades ago (Boon et al., J Exp Med 1996, 183, 725-729, doi: 10.1084 / jem.183.3.725; Almeida et al., Nucleic Acids Res 2009, 37, D816-819, doi: 10.1093 / nar / gkn673; Simpson et al., Nat Rev Cancer 2005, 5, 615-625, doi: 10.1038 / nrc1669), new families of antigens have been discovered that originate from passenger tumor mutations. The set of mutations identified in single cells is amplified by clonal expansion of tumor cells before or after oncogenic transformation. This set of mutations, now expressed in multiple tumor cells, becomes "visible" to the immune system and triggers a T cell immune response. Unlike differentiation antigens and tumor-testis antigens, mutated neoantigens are by definition tumor-specific and are therefore recognized as "non-self" by the immune system. There is clear evidence, including high clinical response rates to ICB in patients with microsatellite instability (whose tumors carry a high number of point mutations) or a correlation between the median number of mutations in a cancer type and the response rate to ICB.

[0006] However, some evidence also suggests that point mutations are not the only antigens that T cells see on tumors. First, there are exceptions to the correlation between mutation frequency and response rate to ICB. For example, RCC has a mutation load of approximately 2 mutations per MB and a response rate to ICB of approximately 25%, while squamous non-small cell lung cancer (LUSC) has a mutation load of approximately 9 mutations per MB and a response rate to ICB of 17% (Yarchoan et al., N Engl J Med 2017, 377, 2500-2501, doi: 10.1056 / NEJMc1713444; Yarchoan et al., JCI Insight 2019, 4, doi: 10.1172 / jci.insight.126908). Second, at the individual patient level, the number of mutations does not predict clinical response to ICB. Third, tumor types with very low mutational burden (and limited genomic instability), such as rhabdomyosarcoma, have relatively high clinical response rates to ICB (McGrail et al., Ann Oncol 2021, 32, 661-672, doi:10.1016 / j.annonc.2021.02.006; Gromeier et al., Nat Commun 2021, 12, 352, doi:10.1038 / s41467-020-20469-6). Finally, there are multiple examples in the literature of patients with T cell responses to non-mutated antigens, including differentiation antigens and tumor testis antigens.

[0007] Noncoding genomic peptide antigens can also represent tumor-specific antigens. Recently, different groups have used proteomics (i.e., experimental approaches based on the combination of transcriptomic analysis and immunopeptidomic analysis) to randomly search for tumor-specific ORFs encoding peptides presented by MHC-I molecules on tumor cells (Laumont et al., Nature Communications (Nat Commun) 2016, 7, 10238, doi: 10.1038 / ncomms10238; Chong et al., Nature Communications (Nat Commun) 2020, 11, 1293, doi: 10.1038 / s41467-020-14968-9). Most of the identified peptides are from noncoding genomic regions. Some of these potential tumor antigens are present in several patients and can induce immune responses in vitro or in mouse models. However, there is no evidence to date that T cells are specific for shared tumor-specific neoantigens from the noncoding genome of cancer patients. Indeed, the identification of such tumor neoantigens will be of great interest and may improve the development of cancer therapies, especially in the context of vaccination and adoptive cell therapy.

[0008] A large part of the noncoding genome consists of transposable elements (TEs). TEs include three major classes of retrotransposons (short interspersed nuclear elements - SINEs, long interspersed nuclear elements - LINEs and long terminal repeats - LTRs) and DNA transposons (Grundy et al., FEBS J 2021, doi: 10.1111 / febs.15722; Burns, KH, Nat Rev Cancer 2017, 17, 415-424; Bourque et al., Genome Biol 2018, 19, 199, doi: 10.1186 / s13059-018-1577-z). Retrotransposition requires TEs to be transcribed, reverse transcribed into DNA and integrated into different genomic locations. Retrotransposition compromises genome stability, and mammalian cells protect themselves by epigenetically repressing TE transcription in adult tissues. Thus, TE transcription is relatively low (but detectable) in most adult cells, but more active in embryonic development, stem cells, and tumors. TE derepression in tumors occurs through multiple epigenetic changes at TE loci, including DNA and histone demethylation. Both of these epigenetic changes are associated with the carcinogenic process, which involves varying degrees of epigenetic dysregulation.

[0009] However, whether derepressed TEs in tumors can be a source of bona fide tumor-specific antigens has never been questioned.

[0010] Glioblastoma (GBM) remains one of the most challenging cases in clinical oncology. The gold standard treatment for GBM, tumor resection followed by radiation and chemotherapy (usually temozolomide), is limited in its effectiveness due to high recurrence rates, overall treatment tolerance, and devastating side effects.

[0011] Therefore, the identification of shared tumor-specific neoantigens will be of interest and may improve the development of cancer therapies, especially in the context of vaccination and adoptive cell therapy, thus holding great promise for treating glioblastoma in patients. Summary of the invention

[0012] The present disclosure relates to a method for identifying or screening tumor cell TE tags, the method comprising the following steps:

[0013] i. obtaining a single-cell transcriptome TE pattern of at least one tumor cell and a single-cell TE transcriptome pattern of at least one normal cell, and

[0014] ii. performing a differential expression analysis of the TE transcriptome pattern of the at least one tumor cell relative to the at least one normal cell, and

[0015] iii. selecting TE transcript sequences that are differentially expressed in the at least one tumor cell compared to the at least one normal cell, thereby obtaining a tumor cell TE signature.

[0016] Typically, in step i), single-cell transcriptome TE patterns are obtained by mapping single-cell transcriptomes to individual genomic TE occurrences.

[0017] The present disclosure also relates to a method for identifying TE-derived tumor neoantigenic peptides, the method comprising the following steps:

[0018] a) obtaining a tumor cell TE label according to the method for identifying a tumor cell TE label disclosed in the present invention;

[0019] b) Computer simulation translation of the TE transcript sequence of the tumor cell TE tag obtained in step a) to obtain TE-derived tumor peptides.

[0020] Typically, the method for identifying TE-derived tumor neoantigen peptides also includes step c) identifying TE-derived peptides that bind to at least one MHC molecule; in some embodiments, searching a library containing the TE-derived peptide sequence identified in step b) in the MHC ligand group from tumor cells, and wherein matching peptides are selected from the MHC ligand group, thereby identifying MHC-bound TE-derived peptides; in some embodiments, TE-derived MHC-binding peptides are further filtered against typical proteins.

[0021] Typically, the method for identifying TE-derived tumor neoantigenic peptides further comprises step d) selecting non-redundant TE-derived peptides; in some embodiments, this step is achieved by mapping the TE-derived peptides of step c) to individual TE genomic locations and selecting uniquely mapped TEs.

[0022] In some embodiments of the method for identifying TE-derived tumor neoantigenic peptides, a peptide having a concentration of less than 10 -5 M of K D The binding affinity binds to a TE-encoded peptide of at least one MHC class I or MHC class II molecule of the subject.

[0023] The present disclosure also includes an isolated tumor neoantigenic peptide sequence having at least 8 amino acids, wherein the neoantigenic peptide comprises a TE coding sequence and has a length of less than 10 -5 M of K D The binding affinity binds to at least one MHC class I or MHC class II molecule of the subject.

[0024] The neoantigenic peptides generally have one or more of the following characteristics:

[0025] -TE expression is suppressed in tumor cells compared with non-tumor cells;

[0026] - the peptide is encoded by a TE transcript sequence or a fragment thereof obtained according to the method for identifying a TETE tag of a tumor cell as defined above;

[0027] - the peptide is obtained according to the method for identifying TE-derived tumor neoantigenic peptides; and / or

[0028] - the peptide is encoded by a TE transcript of any one of SEQ ID NOs: 381 to 5020 or a fragment thereof; preferably, the peptide is encoded by a TE transcript of any one of SEQ ID NOs: 381 to 430 and 432 to 5020 or a fragment thereof; more preferably, the peptide is encoded by a TE transcript of any one of SEQ ID NOs: 381 to 393; 395 to 430 and 432 to 5020 or a fragment thereof;

[0029] Optionally, the peptide comprises at least 8 amino acids, particularly 8 or 9 to 15 amino acids, especially 12 to 15 amino acids and binds to at least one MHC class I molecule of the subject, or comprises 13 to 25 amino acids and binds to at least one MHC class II molecule of the subject.

[0030] In some embodiments, the neoantigenic peptide comprises or consists of any one of SEQ ID NO: 1 to 380 or a fragment thereof, and optionally, the peptide is encoded by a single genomic TE. In some preferred embodiments, the neoantigenic peptide comprises or consists of any one of SEQ ID NO: 1 to 26 and 28 to 380 or a fragment thereof; preferably, the neoantigenic peptide comprises or consists of any one of SEQ ID NO: 1 to 10; 12 to 26; 28 to 57; 59 to 242; 244 to 255; 257 to 319 and 321 to 380 or a fragment thereof; more preferably, the neoantigenic peptide is encoded by a single genomic TE.

[0031] In some embodiments of the present disclosure, the tumor is a glioblastoma.

[0032] Typically, TEs have one or more of the following characteristics:

[0033] -TE selected from more than 50×10 6 TE of the year; optionally, wherein the TE is selected from LINE-1, SVA and ERVK TE subfamilies; optionally, wherein the TE is selected from L1PA / B / x TE;

[0034] -TE selected from more than 50×10 6Year TE;

[0035] -TEs were selected from TEs carrying complete or nearly complete ORFs;

[0036] -TE is selected from intronic TE or intergenic TE;

[0037] -TE is encoded by chromosome 7.

[0038] The present disclosure also encompasses autologous dendritic cell populations or antigen presenting cell populations that have been pulsed with one or more of the above-defined TE-derived tumor neoantigenic peptides or transfected with polynucleotides encoding one or more of the peptides.

[0039] The present disclosure also includes a vaccine or immunogenic composition capable of eliciting a specific T cell response, comprising:

[0040] - one or more neoantigenic peptides as defined above;

[0041] - one or more polynucleotides encoding a neo-antigenic peptide as defined above, optionally encoding a neo-antigenic peptide linked to a heterologous regulatory nucleotide sequence; and / or

[0042] - A population of antigen presenting cells as defined above.

[0043] The present disclosure also includes an antibody or antigen-binding fragment thereof, a T cell receptor (TCR) or a chimeric antigen receptor (CAR), which specifically binds to the above-mentioned new antigen peptide, optionally associated with an MHC molecule, K d The affinity is about 10 -6 M or smaller;

[0044] Optionally, the antibody is a multispecific antibody further targeting at least one immune cell antigen,

[0045] Optionally, the immune cell is a T cell, a NK cell or a dendritic cell, and optionally, wherein the target antigen is CD3, CD16, CD30 or TCR; and / or

[0046] Optionally, the antibody is a multispecific antibody further targeting at least one immune cell antigen, optionally wherein the immune cell is a T cell, NK cell or dendritic cell, optionally wherein the target antigen is CD3, CD16, CD30 or TCR.

[0047] In some embodiments, the previously defined T cell receptor is made soluble and fused to an antibody fragment directed against a T cell antigen, optionally the target antigen is CD3 or CD16.

[0048] The present disclosure also includes a polynucleotide encoding a new antigenic peptide as defined herein or an antibody, CAR or TCR as defined herein. The present disclosure also includes a vector comprising the polynucleotide.

[0049] The present disclosure also includes an immune cell that specifically binds to one or more neoantigenic peptides defined herein; optionally, the immune cell is an allogeneic or autologous cell selected from T cells, NK cells, CD4+ / CD8+, TIL / tumor-derived CD8 T cells, central memory CD8+ T cells, Treg, MAIT and γδT cells.

[0050] The present disclosure also includes a T cell as defined above, comprising:

[0051] - a T cell receptor that specifically binds to one or more neoantigenic peptides as defined herein, and / or

[0052] - A TCR or CAR of the present disclosure.

[0053] The present disclosure also includes the neoantigenic peptides, dendritic cell populations, vaccines or immunogenic compositions, antibodies, antigen-binding fragments thereof, CARs, TCRs, polynucleotides, vectors or immune cells defined herein for use in treating cancer; optionally for inhibiting cancer cell proliferation, or for cancer vaccination therapy in a subject; optionally, the cancer is glioblastoma. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 Single-cell TE expression distinguishes all cell populations in a tumor

[0055] (A) Workflow showing a strategy for alignment and TE quantification using unique or multiple mapping reads. (B) Visualization of t-distributed random neighbor embedding (tSNE) of all single cells (n=3167) after filtering. Cell clusters are represented as different sorted cell populations and identified based on gene expression (left), TE subfamily expression (middle), and individual TE copy expression (right). (C) Violin plots showing TE-specific labels for tumor cells (top) and immune cells (bottom). (D) A graph showing TE subfamily enrichment analysis using all expressed TE (left), tumor (middle), and immune (right) labels. On the x-axis, the "adjusted p-value" is expressed as -log10 (adjusted p-value) using the Benjamini-Hochberg procedure. The ratio ratio (ratio in subset to genome ratio of RepeatMasker) is represented by a ratio circle. The dotted line indicates that the adjusted P value of the x-axis is <0.05. The length of the colored line indicates the adjusted p-value. Subfamilies are colored by category. The longer the colored line, the smaller the adjusted p-value. (E) Radar plots showing gene rates (top) and TE rates (bottom) on all chromosomes. The genome rates (black) and tumor signature rates (dark grey) based on RepeatMasker are shown. (F) Bar plots showing gene rates (first row) or TE rates (second row) on chromosome 10 (left) or chromosome 7 (right) on different feature subsets: all annotated features in the genome (genome), all expression features in the filtered dataset (expressed), and all differentially expressed features in tumor, immune, and OPC cell populations.

[0056] Figure 2 .TE expression in tumor cells is enriched in elements that are independent of their nearest gene. (A) Bar graph showing the distribution of different types of genomic regions for individual TE copies using RepeatMasker (4496056), all expressed TEs (130028), TEs from tumors (3428), and immune signatures (2920). (B) Bar graph showing the number of TEs in regions proximal or distal to the nearest protein-coding gene in tumor signatures and immune signatures. (C) Plot showing the distance of each type of TE to the nearest protein-coding gene comparing proximal (first row) and distal (second row) TEs in tumor signatures and immune signatures. (D) Summary of the distribution of TEs in tumor (top) signatures and immune signatures (bottom) Figure 1 Figure 2. The association between TEs and genes described in B. When TEs and genes are differentially expressed in the same cell population, TEs + Gene + The category indicates positive correlation. When TEs other than genes are differentially expressed, TE + Gene - The categories represent negative correlations. These categories are also divided based on proximal and distal status.

[0057] Figure 3 .Single-cell tumor TE signatures are highly enriched in the GBM cohort of TCGA compared to GTEx normal tissues. (AB) PCA and uniform manifold projection (UMAP) of the GBM TCGA cohort, GTEx normal brain, and other GTEx tissues based on single-cell tumor TE signatures, color-coded by dataset type. (C) Gene set enrichment analysis (GSEA) was performed to determine the specific enrichment of tumor signatures in GBM tumor samples and GTEx normal brain samples. Normalized enrichment scores (NES) and FDR are indicated in the figure. (D) Violin plots showing the average expression of single-cell tumor signatures in the GBM TCGA cohort, GTEx normal brain, and other GTEx datasets. (E) Violin plots showing the specific expression of individual TEs in tumor samples (bulk RNA-seq analysis, top) and tumor cells (single-cell analysis, bottom).

[0058] Figure 4 .Tumor-enriched TE-derived peptides are presented on HLA-I molecules and are immunogenic. (A) Workflow for identifying TE-derived peptides using mass spectrometry-based immunopeptidomics. (B) Boxplots showing peptide spectrum identification scores (SEQUEST scores) from annotated peptides and TE-derived peptides. (C) Binding to HLA-A02*01 and HLA-B*07:01 measured as the percentage of peptide-HLA-I complex formation compared to positive control. (D) Total frequency of multimer-positive populations of HLA-A*02:01 predicted peptides or MS-derived peptides and HLA-B*07:02MS-derived peptides in each evaluated donor. The total frequency was calculated based on the total number of multimer-positive cells in all replicates among all CD8+ T cells evaluated for each donor. The bottom line indicates the mix of peptides used for each donor. P#: predicted TE-derived peptides; pMS#: MHC-I peptidome-derived peptides; Melan-A mutation sequencing and N#: normal proteome-derived peptides.

[0059] Figure 5 TE-derived peptides in long ORFs starting with canonical and atypical start codons. The bar graph shows the quantification of LINE and LTR TEs with complete ORFs recorded in the gEVE database for different subsets.

[0060] Figure 6 .The redundancy of TE-derived peptides depends on the age of the TE. A graph showing TE family enrichment analysis using TEs encoding all assigned (left) or single assigned (right) peptides. The "log2 ratio" (ratio in subset to ratio in RepeatMasker) is represented on the x-axis. The significance of the hypergeometric test is represented by a proportional circle. The larger the circle, the smaller the adjusted p-value (-log10 adjusted p-value).

[0061] Figure 7 .TE-derived peptides are overexpressed in GBM tumor samples. The log2 ratio between the relative expression of total RNA of GBM and GTEx TE-derived peptides has been determined. Age information, redundancy, and TE class were taken into account. Tissues from GTEx were divided into the five normal tissue classes defined by Bradley et al. (Nat Commun 2020, 11, 5332). TEs were ranked using hierarchical clustering and two groups (Group 1 and Group 2). A graph showing the median age of TEs encoding peptides in each group. DETAILED DESCRIPTION

[0062] The inventors used single-cell transcriptomics (scRNAseq) of tumor samples to identify patterns of individual TEs that are selectively expressed in tumor cells, particularly in total glioblastoma (GBM) tumor cells. They further demonstrated that peptides encoded by these selectively expressed TEs are not only presented by HLA-I molecules in cancer cells and are immunogenic, but are also shared between patients. They also demonstrated that Peptides encoded by single TE (non-redundant TE) have stronger tumor specificity.

[0063] Their results also showed that differentially expressed TEs in GBM tumors were biased toward TEs encoded on chromosome 7, which is consistent with the known recurrent amplification of this chromosome in GBM cancers. TE-derived peptides presented by MHC-I were enriched for peptides derived from specific subfamilies, including the young LINE-1 and SVA elements.

[0064] Thus, the included results indicate that scRNAseq-guided, TE-centric proteomics is a powerful tool for identifying tumor-specific antigens and that TE-derived peptides recurrently presented on HLA-I molecules of GBM tumor cells are primarily encoded by young LINE-1 elements that are selectively derepressed in such GBM tumor cells.

[0065] Since the peptides identified according to the methods disclosed herein are immunogenic in healthy patients and presented to HLA-I, they are a source of shared tumor-specific neoantigens, which can be used to produce various cancer therapies, including antigen presenting cells and immunogenic compositions, especially for personalized vaccination strategies, and can also be used to construct CAR or TCR and produce modified immune cells containing them, or generate antibodies that can be used to treat cancer. Identifying truly specific epitopes expressed in many cancer patients will help to follow these treatments more effectively and greatly reduce costs. In the case of TCR adoptive therapy, identifying TCRs that are specific to shared neoepitopes will help develop better autologous or even allogeneic cell therapies. Antibodies that are specific to the presented shared HLA peptide complexes can also be developed for ADC or CAR-T cell approaches.

[0066] definition

[0067] According to the present disclosure, the term "normal" refers to a healthy state or condition of a healthy subject, tissue or cell, i.e., a non-pathological state, wherein "healthy" preferably refers to non-cancerous. Typically, in some embodiments, healthy cells refer to "non-tumor cells" or "non-malignant cells."

[0068] Cancer (medical term: malignancy) is a class of diseases in which a group of cells exhibit uncontrolled growth (dividing beyond normal limits), invasion (invading and destroying neighboring tissues), and sometimes metastasise (spreading to other parts of the body via the lymph or blood). These three malignant characteristics of cancer distinguish it from benign tumors, which are self-limited and do not invade or metastasize. Most cancers form tumors, but some, such as leukemia, do not.

[0069] Malignancy is essentially synonymous with cancer. Malignancy, malignant neoplasm, and malignant tumor are essentially synonymous with cancer.

[0070] As used herein, the term "tumor" or "neoplastic disease" refers to an abnormal growth of cells (referred to herein as neoplastic cells or tumor cells), preferably forming a swelling or lesion. "Tumor cell" refers to an abnormal cell that grows by rapid, uncontrolled cell proliferation and continues to grow after the stimulus that initiated the new growth has ceased. Tumors exhibit a partial or complete lack of structural organization and functional coordination with normal tissues, and usually form a distinct mass of tissue, which can be benign, precancerous or malignant.

[0071] Benign tumors are tumors that lack all three malignant properties of cancer. Therefore, by definition, benign tumors do not grow in an unrestricted, aggressive manner, do not invade surrounding tissues, and do not spread to non-adjacent tissues (metastasize).

[0072] A neoplasm is an abnormal mass of tissue that forms as a result of neoplasia. Neoplasia (Greek for new growth) is an abnormal proliferation of cells. The growth of cells exceeds and is out of sync with the growth of the normal tissue around them. Even after the stimulus is stopped, the growth continues in the same excessive manner. It usually results in a lump or tumor. Neoplasms can be benign, precancerous or malignant.

[0073] Cancer or tumors can affect any of the following tissues or organs: breast; liver; kidney; heart, mediastinum, pleura; floor of mouth; lips; salivary glands; tongue; gums; oral cavity; palate; tonsils; larynx; trachea; bronchi; lungs; pharynx, hypopharynx, oropharynx, nasopharynx; esophagus; digestive organs such as stomach, intrahepatic bile ducts, bile ducts, pancreas, small intestine, colon; rectum; urinary organs such as bladder, gallbladder, ureters; rectosigmoid junction; anus, anal canal; skin; bones; joints, limb joints cartilage; eye and appendages; brain; peripheral nerves, autonomic nervous system; spinal cord, cranial nerves, meninges; and parts of the central nervous system; connective tissue, subcutaneous tissue, and other soft tissues; retroperitoneum, peritoneum; adrenal glands; thyroid gland; endocrine glands and related structures; female reproductive organs, such as ovaries, uterus, cervix; uterine body, vagina, vulva; male reproductive organs, such as penis, testicles, and prostate; hematopoietic and reticuloendothelial systems; blood; lymph nodes; thymus. Tumors or cancer types according to the present disclosure also include leukemia, seminoma, melanoma, teratoma, lymphoma, neuroblastoma, glioma, rectal cancer, endometrial cancer, kidney cancer, adrenal cancer, thyroid cancer, blood cancer, skin cancer, brain cancer, cervical cancer, intestinal cancer, liver cancer, colon cancer, gastric cancer, intestinal cancer, head and neck cancer, gastrointestinal cancer, lymph node cancer, esophageal cancer, colorectal cancer, pancreatic cancer, ear, nose and throat (ENT) cancer, breast cancer, prostate cancer, uterine cancer, ovarian cancer and lung cancer and their metastases. In some embodiments, cancer or tumors are associated with de-repressed TE (see in particular references Kong, Y., Rose, CM, Cass, AA. et al., Transposable element expression in tumors is associated with immune infiltration and increased antigenicity, "Nat Commun" 10, 5228 (2019)). In some embodiments, the tumor or cancer is selected from a gastric tumor, a bladder tumor, a liver tumor, and a head and neck tumor. In a specific embodiment, the tumor is a glioblastoma.

[0074] "Growth of a tumor" or "tumor growth" according to the present disclosure relates to the tendency of a tumor to enlarge and / or the tendency of tumor cells to proliferate.

[0075] For the purposes of this disclosure, the terms "cancer" and "cancer disease" are used interchangeably with the terms "tumor" or "tumor disease."

[0076] Cancers are classified based on the type of cells that resemble the tumor and therefore based on the tissue from which the tumor is presumed to have originated. These are histology and location.

[0077] "Metastasis" refers to the spread of cancer cells from the primary site to another part of the body. The formation of metastasis is a very complex process that depends on the shedding of malignant cells from the primary tumor, invading the extracellular matrix, penetrating the endothelial basement membrane into the body cavity and blood vessels, and then infiltrating the target organ after being transported by the blood. Finally, the growth of new tumors (i.e., secondary tumors or metastatic tumors) at the target site depends on angiogenesis. Tumor metastasis often occurs even after the primary tumor is resected because tumor cells or components may remain and have metastatic potential. In one embodiment, the term "metastasis" according to the present disclosure relates to "distant metastasis", which involves metastasis away from the primary tumor and the regional lymph node system.

[0078] A recurrence or relapse occurs when an individual is once again affected by a disease that affected them in the past. For example, if a patient previously suffered from a neoplastic disease and then developed the disease again after being successfully treated, the new disease may be considered a recurrence or relapse. However, according to the present disclosure, a recurrence or relapse of a neoplastic disease may, but need not necessarily, occur at the site of the original neoplastic disease. Thus, for example, if a patient previously suffered from an ovarian tumor and was successfully treated, a recurrence or relapse may be the appearance of an ovarian tumor or the appearance of a tumor at a site other than the ovary. A recurrence or relapse of a tumor also includes situations where the tumor occurs at a site other than the original tumor site as well as at the original tumor site. Preferably, the original tumor for which the patient was treated is a primary tumor, and the tumor at a site other than the original tumor site is a secondary or metastatic tumor.

[0079] "Treatment" refers to administering a compound or composition described herein to a subject to prevent or eliminate a disease, including reducing the size or number of a tumor in a subject; stopping or slowing a disease in a subject; inhibiting or slowing the development of a new disease in a subject; reducing the frequency or severity of symptoms and / or recurrences in a subject currently suffering from or previously suffering from a disease; and / or prolonging, i.e., increasing, the life span of a subject. In particular, the term "disease treatment" includes curing, shortening the duration, ameliorating, preventing, slowing or inhibiting the progression or worsening of a disease or its symptoms, or preventing or delaying the onset.

[0080] "At risk" means that a subject, i.e., a patient, is identified as having a higher than normal chance of developing a disease, particularly cancer, compared to the general population. In addition, a subject who has had or currently has a disease, particularly cancer, is a subject at increased risk of developing the disease because such a subject may go on to develop the disease. A subject who currently has or has had cancer is also at increased risk of developing cancer metastasis.

[0081] The therapeutically active agents or products, vaccines and compositions described herein may be administered via any conventional route, including injection or infusion.

[0082] The agents described herein are administered in an effective amount. "Effective amount" refers to the amount that achieves the desired response or desired effect alone, together with other doses, or together with other therapeutic agents. In the case of treating a specific disease or a specific condition, the desired response preferably involves inhibiting the course of the disease. This includes slowing down the progression of the disease, particularly blocking or reversing the progression of the disease. When treating a disease or condition, the desired response can also be to delay the onset of the disease or condition or to prevent the onset of the disease or condition. The effective amount of the agents described herein will depend on the condition to be treated, the severity of the disease, the individual parameters of the patient (including age, physiological condition, body shape and body weight), the duration of treatment, the type of concomitant therapy (if present), a specific route of administration and similar factors. Therefore, the dosage of the agents described herein can depend on multiple of these parameters. If the patient's response to the initial dose is insufficient, a higher dose (or an effective higher dose achieved by a different, more localized route of administration) can be used.

[0083] The pharmaceutical compositions described herein are preferably sterile and contain an effective amount of the therapeutically active substance to produce the desired response or desired effect.

[0084] The pharmaceutical compositions described herein are generally administered in pharmaceutically compatible amounts and pharmaceutically compatible formulations. The term "pharmaceutically compatible" refers to non-toxic materials that do not interact with the action of the active ingredients in the pharmaceutical composition. Such formulations may generally contain salts, buffer substances, preservatives, carriers, supplementary immune enhancing substances such as adjuvants, such as CpG oligonucleotides, cytokines, chemokines, saponins, GM-CSF and / or RNA, and, where appropriate, other therapeutically active compounds. When used in a drug, the salt should be pharmaceutically compatible.

[0085] As used herein, a "transposable element (TE, transposon or jumping gene)" is a repetitive DNA sequence that is able to move from one location in the genome to another by RNA copies generated by reverse transcriptase (class I TE, retrotransposon) or by excision of itself from its original location (class II TE or DNA transposon).

[0086] There are many more retrotransposons, and their characteristics are similar to those of retroviruses (such as HIV). Retrotransposons work by reverse transcription through an RNA intermediate replication mechanism. They are generally divided into three major categories: retrotransposons with long terminal repeats (LTRs) on both sides of the main body of the retrotransposon, which encode reverse transcriptase, similar to retroviruses; retrotransposons with long interspersed nuclear elements (LINE, LINE-1 or L1), which encode reverse transcriptase but lack LTR and are transcribed by RNA polymerase II; and retrotransposons with short interspersed nuclear elements (SINEs), which do not encode reverse transcriptase and are transcribed by RNA polymerase III. DNA transposons have a transposition mechanism that does not involve RNA intermediates. Transposition is catalyzed by several transposases. LTRs include endogenous retroviruses (ERVs), while non-LTR TEs are subdivided into long interspersed nuclear elements (LINEs) and short interspersed nuclear elements (SINEs), non-autonomous transposons mobilized by the LINE integration mechanism. These lineages consist of phylogenetically related families that further branch into multiple subfamilies, each derived from a precursor copy. Over time, the accumulation of mutations leads to differences in the consensus sequence within each subfamily member. For a review of TE retrotransposons, see Richardson, Sandra R et al., "The Influence of LINE-1 and SINE Retrotransposons on Mammalian Genomes." Microbiology spectrum, Vol. 3, 2 (2015): MDNA3-0061-2014.

[0087] The typical L1 element is about 6000 base pairs (bp) long and consists of two non-overlapping open reading frames (ORFs) flanked by untranslated regions (UTRs) and target site repeats. LINE-1 retrotransposons have been amplified in mammalian genomes for more than 160 million years. In humans, the vast majority of LINE-1 sequences have been amplified since the divergence of the ancestral mouse and human lineages approximately 65 to 75 million years ago. Sequence comparisons between individual genomic LINE-1 sequences and consensus sequences derived from modern active LINE-1s can be used to estimate the age of genomic LINE-1s (Khan H, Smit A, Boissinot S; Genome Res. 2006 Jan;16(1):78-87). L1 subfamilies are generally classified into old subfamilies (L1M, AluJ), middle-aged subfamilies (L1P, L1PB, AluS), young subfamilies (L1HS, L1PA, AluY), and related subfamilies (HAL, FAM). In humans, the only family with autonomous activity is the long interspersed nuclear element-1 (LINE-1 or L1), although a few L1 copies still have retrotransposition ability and they all belong to the youngest human-specific L1HS subfamily.

[0088] SVA elements comprise an evolutionarily young family of non-autonomous retrotransposons that arose in the primate lineage approximately 25 million years ago (Hancks DC, Kazazian HH Jr, Semin Cancer Biol. 2010 Aug;20(4):234-45). The typical SVA element is approximately 2000 bp in length and its composite structure consists of: 1) hexameric CCCTCT repeats; 2) inverted Alu-like element repeats; 3) a set of GC-rich variable nucleotide tandem repeats (VNTRs); 4) a SINE-R sequence with homology to HERVK-10 (an inactive LTR retrotransposon); and 5) a typical cleavage and polyadenylation specificity factor (CPSF) binding site followed by a poly(A) channel. The youngest SVA subfamilies include the SVA-D, SVA-E, SVA-F, and SVA-F1 subfamilies.

[0089] In both class I and class II TEs, transposition can also be classified as “autonomous” or “non-autonomous.” Autonomous TEs can move on their own, whereas non-autonomous TEs require the presence of another TE to move.

[0090] The evolutionary age of a TE can be estimated based on the degeneration of its signature motifs, as described by Choudhary, Mayank Nk et al., Genome Biol, vol. 21, 116, 24 January 2020. More specifically, the evolutionary age of a TE can be estimated by dividing the percentage of divergence of extant copies from the consensus sequence by the neutral substitution rate of the species (i.e., in humans: 2.2 × 10 -9 ). Jukes-Cantor and Kimura distances can be calculated by aligning each TE with its consensus sequence and counting all possible mutations. Single nucleotide substitution counts are normalized by the length of the genomic TE minus the number of insertions (gaps in the consensus sequence). These mutation rates are then used to calculate the Jukes-Cantor and Kimura distances for each genomic TE. For most TE subfamilies, consensus sequences can be retrieved from the RepBase library. Full-length LINE consensus can be reconstructed as detailed in Choudhary et al., 2020.

[0091] The complete open reading frame (ORF) position can be retrieved from the gEVE database. The complete ORF and individual TE coordinates are usually matched to assign the complete ORF to the individual TE in case of overlapping coordinates. 30517 individual TEs overlapped with the complete ORF, most of which were L1 (mainly L1PA / B / x) and ERV (mainly ERV1, ERVK, ERVL) elements. In order to determine the amino acid sequence similarity between typical TE proteins in the gEVE database and peptides in the immunopeptidomics results, blastp can usually be performed between the gEVE protein sequence and the immunopeptidomics sequence. Usually, no threshold for Evalue is set, and the similarity is usually estimated and divided into 3 categories: (1) 100% match: no mismatch, no gap, and the query coverage of each HSP is 100%; (2) at most 1 mismatch: 1 mismatch, no gap, and the query coverage of each HSP is above 85%; (3) at most 2 mismatches: 2 mismatches, no gap, and the query coverage of each HSP is above 85%.

[0092] A "representative genome" (also called a reference genome or assembly) is a digital nucleic acid sequence database assembled by scientists as a representative example of a species' gene set. Because a reference genome is usually assembled by sequencing DNA from multiple donors, it cannot accurately represent the gene set of any single individual (animal or human). Instead, the reference genome provides a haploid mosaic of different DNA sequences from each donor.

[0093] An "exon" is any portion of a gene that will encode a portion of the final mature RNA produced by that gene after introns are removed by RNA splicing. The term "exon" refers both to DNA sequences within a gene and to corresponding sequences in RNA transcripts.

[0094] "Messenger RNA (mRNA)" is a single-stranded RNA molecule that corresponds to the genetic sequence of a gene and is read by the ribosome in the process of producing proteins. mRNA is produced during the process of transcription, in which the enzyme RNA polymerase converts the gene into a primary transcript mRNA (also called pre-mRNA). This pre-mRNA usually still contains introns, which are regions that will not go on to code for the final amino acid sequence. These are removed during the RNA splicing process, leaving only the exons, which are the regions that code for the protein. This exon sequence constitutes the mature mRNA. The mature mRNA is then read by the ribosome, and using the amino acids carried by the transfer RNA (tRNA), the ribosome produces the peptide sequence, a process called translation.

[0095] "Transcript" as referred to herein is a messenger RNA (or mRNA) or a portion of an mRNA expressed by an organism, particularly in a particular tissue, or even in a particular tissue. The expression of a transcript varies due to many factors. In particular, the expression of a transcript in a cancer cell may be modified compared to a normal healthy cell. In the present disclosure, a transcript may be provided in the form of its corresponding genomic sequence.

[0096] " Transcriptome " referred to herein is all messenger RNA or mRNA molecules expressed by an organism. In some embodiments, the term " transcriptome " or " transcriptome pattern " can also be used to describe the mRNA transcript array produced in a specific cell or tissue type. Contrary to the genome characterized by stability, the transcriptome actively changes. In fact, the transcriptome of an organism varies due to many factors, including developmental stage and environmental conditions. Usually, similarly, compared with corresponding (i.e., cells of the same type usually from the same species) normal healthy cells, the transcriptome in cancer cells is modified. Usually, the transcriptome referred to herein is the human transcriptome. When referring to individual cells, the terms " transcriptome pattern (transcriptomic pattern) " and " transcriptome (transcriptome) " are used as synonyms in this article.

[0097] Reading frame (RF) is a method of dividing the nucleotide sequence in a nucleic acid (DNA or RNA) molecule into a set of consecutive, non-overlapping triplets.

[0098] The open reading frame (ORF) is the part of the reading frame that can be translated into a peptide. The ORF is a stretch of consecutive codons containing a start codon (e.g., AUG) and a stop codon (e.g., UAA, UAG, or UGA) after the transcription start site (TSS). The ATG codon within the ORF (not necessarily the first) can indicate where translation begins. The transcription termination site is located after the ORF, after the translation stop codon. In eukaryotic genes with multiple exons, the ORF spans intron / exon regions, which can be spliced ​​together after the ORF is transcribed to produce the final mRNA for protein translation.

[0099] The "classical ORF" referred to herein is a protein coding sequence with a specific reading frame within an mRNA sequence described or annotated in a database, such as the Ensembl genome / transcriptome / proteome database collection (usually hg19). Typically, a classical ORF is an annotated (in a reference database) ORF for a given exon in a normal healthy cell.

[0100] "Non-annotated or non-classical transcripts or mRNAs" as referred to herein are protein coding sequences with specific reading frames within mRNA sequences that are not described (i.e., unannotated) in genome databases (e.g., Ensembl genome / transcriptome / proteome databases). The term "canonical protein" as referred to herein refers to a protein encoded by a classical or annotated reading frame. In some embodiments, some unannotated mRNA sequences may be minor mRNAs whose expression levels in normal healthy cells are less than 5%, particularly less than 2%, less than 1%, less than 0.5%, less than 0.2%, or less than 0.1% of total cell mRNA.

[0101] RNA-Seq (short for RNA sequencing) is a sequencing technology that uses next-generation sequencing (NGS) to reveal the presence and quantity of RNA (usually messenger RNA, mRNA) in a biological sample and generates a large number of raw sequencing reads (usually at least tens of millions). Single-cell RNA sequencing (scRNA-Seq) provides expression profiles of individual cells. A read refers to the RNA sequence of one RNA fragment from a biological sample or a single cell. The sequenced RNA sample is called an RNA library. Therefore, RNA sequencing data is often referred to as RNA reads. In RNA-seq data, there are two main methods for measuring the expression of transcripts (especially TE transcripts in this case):

[0102] The count is simply the number of reads that overlap with a given genomic position.

[0103] "TPM" ("transcripts per million") and FPKM (fragments per kilobase of exon model per million mapped reads) are also commonly used units reported to estimate gene expression based on RNA-seq data. Both units are calculated based on the number of reads mapped to each specific gene sequence, and the calculation of both units takes into account two important factors in RNA-seq:

[0104] -The number of reads from a gene depends on its length. One would expect more reads to be generated from longer genes.

[0105] -The number of reads for a gene depends on the sequencing depth, which is the total number of reads sequenced. One would expect more reads to be generated from samples with greater sequencing depth.

[0106] FPKM (proposed by Trapnell, C., Williams, B., Pertea, G., et al., Nat Biotechnol 28, 511-515 (2010)) is calculated by the following formula:

[0107]

[0108] where q i is the raw count (number of reads mapped to each gene), l i is the gene length, and the total number of mapped reads is the total number of mapped reads. The interpretation of FPKM is that if you sequence the RNA sample again, you would expect to see FPKMi reads for gene i divided by the length of gene i, which is over a thousand, divided by the total number of mapped reads, which is over a million.

[0109] The TPM unit was introduced by Li and Dewey in 2011 (Li, B., Dewey, CNRSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome, BMC Bioinformatics, 12, 323 (2011)), and Pachter, 2011 ( arXiv:1104.3889 [q-bio.GN] "Transcript quantification model based on RNA-Seq") establishes the relationship between these two units. TPM can be calculated from FPKM as follows:

[0110]

[0111] For the definition of TPM, please also refer to the following definition: Wagner et al., Theory Biosci. 2012 Dec; 131(4): 281-5.

[0112] For example, in the EMBL Expression Atlas database (which contains thousands of selected microarray and RNA sequencing data, which have been manually curated and annotated with the ontology term baseline expression results), the baseline expression level is set as follows and represented by different colors (see https: / / www.ebi.ac.uk / gxa / FAQ.html ):

[0113] -Gray box: expression level is below the critical value (0.5 FPKM or 0.5 TPM)

[0114] - Light blue box: low expression level (0.5 to 10 FPKM or 0.5 to 10 TPM)

[0115] -Medium blue box: medium expression level (11 to 1000 FPKM or 11 to 1000 TPM)

[0116] - Dark blue box: high expression level (more than 1000 FPKM or more than 1000 TPM)

[0117] If not otherwise specified, the above reference expression levels can be used as references, or thresholds, in the methods and definitions of the present disclosure. However, in some embodiments, other thresholds can be used. For example, based on the average expression of transcripts in samples or cells (usually cancer cells) from the disease of interest, the expression threshold or critical value can be set to 7.5 TPM or 10 TPM.

[0118] "Fold change" is a measure that describes how much a quantity has changed between an original measurement and a subsequent measurement. It is defined as the ratio between two quantities and is often used to measure changes in gene expression levels, or in this case, changes in TE expression levels in tumor cells compared to non-tumor cells. Logarithmic ratios are often used to analyze and visualize fold changes. The most commonly used is the logarithm with base 2.

[0119] In this specification, the term "peptide or polypeptide" is used interchangeably with "neoantigenic peptide or polypeptide" and refers to a series of residues (usually L-amino acids) connected to each other usually by peptide bonds between the α-amino and carboxyl groups of adjacent amino acids. Polypeptides or peptides can have various lengths, can be in neutral (uncharged) form or salt form, and can be free of modifications such as glycosylation, side chain oxidation or phosphorylation, or can contain these modifications, but the premise is that the modification does not destroy the biological activity of the polypeptide described herein.

[0120] According to the present application, "tumor neoantigen peptides" refer to peptides that can be recognized by T cells and can induce T cell reactivity once presented by specific MHC alleles. Generally, neoantigen peptide-specific T cells have functional affinity that can reach the affinity strength of antiviral T cells (see: Lennerz V et al., Mutation-specific CD4+ T cells in human melanoma). + Cancer immunotherapy based on mutation-specific CD4 + T cells in human melanoma, Nat Med 2015;21:81-5).

[0121] In some embodiments, the neoantigen peptide is completely absent (e.g., cannot be detected to be expressed) in the normal peptide group (particularly the human peptide group, such as the UNIPROT database and / or healthy cells). Generally, tumor-specific neoantigen peptides cannot be detected to be expressed in normal healthy cells or samples and are referred to herein as "tumor-specific".

[0122] According to the present disclosure, the statement that a neoantigenic peptide or TE transcript is "specifically expressed" in a tumor cell type means that the peptide or TE transcript is expressed, more particularly upregulated, in tumor cells with a statistical difference (p value adjusted by the Wilcoxon test equal to or lower than 0.05, particularly equal to or lower than 0.01) compared to non-tumor cells. In some embodiments, a log 2 fold change threshold of 0.25 for the expression of TE transcripts in tumor cells compared to non-tumor cells can also be used. Therefore, in some embodiments, the peptide is encoded by a TE transcript or a fragment thereof, and the TE transcript or a fragment thereof is expressed in tumor cells with a log 2 fold change of at least 0.25, particularly at least 0.5, at least 0.75, at least 1, at least 1.25, at least 1.5, at least 1.75 or at least 2 compared to non-tumor cells. In some embodiments, the TE transcript is expressed only in one or more tumor cells and is not significantly detected in normal non-tumor cells or samples (e.g., normal samples from the Genotype-Tissue Expression (GTEx) database).

[0123] Typically, the subject of the present application is a mammal, particularly a human. Therefore, typically, the representative or reference genome or transcriptome is a human genome or transcriptome.

[0124] In the present application, "MHC molecule" or "HLA molecule" refers to at least one MHC / HLA class I molecule or at least one MHC / HLA class II molecule. MHC class I proteins form functional receptors on most nucleated cells of the body. There are three major MHC class I genes in HLA: HLA-A, HLA-B, HLA-C and three minor genes HLA-E, HLA-F and HLA-G. 32-microglobulin binds to the major and minor gene subunits to produce heterodimers. Class I MHC molecules are composed of heavy chains and light chains, and can bind peptides of about 8 to 11 amino acids, but if the peptide has a suitable binding motif, it is usually a peptide of 8 or 9 amino acids, and present it to cytotoxic T lymphocytes. The binding of the peptide is stabilized at its two ends by contacts between atoms in the peptide backbone and invariant sites in the peptide binding groove of all MHC class I molecules. There are invariant sites at both ends of the groove, which bind the amino terminus and carboxyl terminus of the peptide. Variations in peptide length are accommodated by kinks in the peptide backbone, typically at proline or glycine residues that achieve the desired flexibility. Peptides bound by class I MHC molecules are typically derived from endogenous protein antigens. As an example, in humans, the heavy chains of class I MHC molecules are typically HLA-A, HLA-B or HLA-C monomers, while the light chains are beta-2-microglobulin. There are three major MHC class II proteins encoded by HLA and two minor MHC class II proteins. Class II genes bind to form heterodimeric (αβ) protein receptors, which are typically expressed on the surface of antigen presenting cells. Peptides bound by class II MHC molecules are typically derived from extracellular or exogenous protein antigens. As an example, in humans, α and β chains are particularly HLA-DR, HLA-DQ and HLA-DP monomers. MHC class II molecules are able to bind peptides of about 8 to 20 amino acids, particularly 10 to 25 amino acids or 13 to 25 amino acids, if the peptide has a suitable binding motif, and present them to helper T cells. The peptide is in an extended conformation along the MHC II peptide binding groove, which (unlike the MHC class I peptide binding groove) is open at both ends. It is held in place primarily by contacts between backbone atoms and conserved residues along the peptide binding groove.

[0125] The term "peptide group" refers to the entire set of peptides expressed by a particular genome or present in a particular organism or cell type (e.g., cancer cells). Thus, proteomic analysis (proteomics) refers to the separation, identification, and quantification of the entire set of peptides or proteins expressed by a genome, cell, or tissue at a specific time point.

[0126] Proteomic analysis is usually based on two main techniques, namely two-dimensional gel electrophoresis (2-DGE) (Harper S et al., Coligan JE, Dunn BM, Speicher DW, Wing-field PT, eds., Current Protocols in Protein Science John Wiley & Sons; Hoboken, NJ: 1998, pp. 10.4.1-10.4.36.) and mass spectrometry (MS) (Aebersold and Mann, 2003), both of which are effective methods for analyzing complex protein mixtures. HPLC is an alternative separation technique for proteomic studies, especially in the separation and identification of low molecular weight proteins and peptides (Garbis et al., 2005). MS can determine the molecular mass of a protein or peptide based on the mass-to-charge ratio (m / z) of ions in the gas phase. The terms "gel-based" or "gel-free" proteomics relate to the separation techniques applied, 2-DGE or HPLC; proteomics approaches can also be "bottom-up" or "top-down", which essentially identify proteins from protease (e.g. trypsin) digests or as a whole by mass spectrometry, respectively.

[0127] Bottom-up proteomics is a common approach used to identify proteins from biological samples (tissues or cells) and characterize their amino acid sequence and post-translational modifications by proteolytic digestion of proteins followed by analysis by mass spectrometry. Crude protein extracts are enzymatically digested and peptides are separated in one or more dimensions, usually by liquid chromatography coupled to mass spectrometry, a technique known as shotgun proteomics. Peptides are identified by comparing the masses of proteolytic peptides or their tandem mass spectra to the masses predicted by sequence databases or annotated peptide spectra in peptide spectral libraries, and multiple peptide identifications are assembled into protein identifications.

[0128] In top-down proteomics, intact proteins are purified prior to digestion and / or cleavage within a mass spectrometer or by two-dimensional electrophoresis. Top-down proteomics uses either an ion trap mass spectrometer to store isolated protein ions for mass measurement and tandem mass spectrometry (MS / MS) analysis or other protein purification methods such as two-dimensional gel electrophoresis coupled to MS / MS.

[0129] Based on the data generated by MS, proteins can be sequenced de novo by manual mass spectrometry analysis or automatically processed by sequence search engines such as SEQUEST, Mascot, Phenyx, X!Tandem, and OMSSA. These algorithms are developed based on the correlation between experimental and theoretical MS / MS data; the latter are generated by in silico digestion of protein databases such as UniProt / Swiss-Prot (Deutsch, Lam, & Aebersold, 2008).

[0130] The term "immunopeptidome", also often referred to as "immunopeptidomic pattern", "pMHC repertoire" or "MHC-ligand group" or "HLA ligand group" refers to the complete set of peptides that bind to at least one MHC / HLA molecule on the cell surface within a specific cell type. Accordingly, "immunopeptidomics" has become a term to describe the analysis of MHC / HLA ligand groups. The most common immunopeptidomic methods rely on mass spectrometry (MS). Immunopeptidomic samples are typically prepared by isolating MHC from lysed cells or tissues, for example by using allele-specific antibodies, pan-specific antibodies, or engineered affinity tag systems. The separated complexes are acid eluted, and the peptides are purified from the MHC molecules using molecular weight cutoff filtration (MWCO), solid phase extraction, or other techniques, and subsequently analyzed by MS (see, for example, LEStopfer et al., Immuno-Oncology and Technology, Vol. 11, 2021, 100042 for review).

[0131] Unless expressly stated or obvious from the context, as used herein, the term "about" should be understood as within the normal tolerance range in the art, such as within 2 standard deviations of the mean. "About" can be understood as within 20%, 15%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.05% or 0.01% of the value. Unless the context clearly indicates otherwise, all numerical values ​​provided herein are modified by the term "about".

[0132] Methods for identifying TE signatures of tumor cells

[0133] The method for identifying tumor cell TE tags disclosed in the present invention comprises the following steps:

[0134] i. obtaining a single-cell transcriptome TE pattern of at least one tumor cell and a single-cell TE transcriptome pattern of at least one non-tumor cell, and

[0135] ii. performing a differential expression analysis of the TE transcriptome pattern of the at least one tumor cell relative to the at least one normal cell, and

[0136] iii. selecting TE transcript sequences that are differentially expressed in the at least one tumor cell compared to the at least one normal cell, thereby obtaining a tumor cell TE signature.

[0137] Typically, one or more tumor cells (also referred to herein as tumor cells) are from the same patient, and / or obtained from the same tumor location and / or the same type of tumor, and / or obtained from the same tumor sample.

[0138] The same type of cancer or tumor as used herein refers to tumors or cancers affecting the same organ (skin, breast, lung, brain, bladder, kidney, stomach, intestine, spleen, pancreas, prostate, uterus, thyroid, ovaries, endocrine glands, uterus, testicles, tongue, esophagus, liver, gallbladder, rectum, skin, etc.) or the same tissue (e.g., carcinoma, sarcoma, myeloma, leukemia, lymphoma, etc.).

[0139] Similarly, one or more non-tumor (e.g., "normal" or "healthy") cells are typically obtained from the same patient, and / or from a sample near the tumor from the same or different patients. Non-tumor cells can typically be tumor-infiltrating cells or cells from the environment near the tumor. According to the present disclosure, non-tumor cells can be from one or more types, including tumor-infiltrating immune cells (e.g., macrophages) and non-immune cells from the environment near the tumor. For example, when the tumor is a glioblastoma, non-tumor cells from the tumor microenvironment (i.e., from a sample near the tumor) include immune cells (usually macrophages), oligodendrocytes and their precursors (OPC), neurons, astrocytes, and vascular cells.

[0140] The transcriptome profile of these cells can be obtained by performing high-depth single-cell RNA sequencing (RNA-seq) (for an example of a detailed procedure, see Darmanis, Spyros et al., Cell Reports, Vol. 21, 5 (2017): 1399-1410). In short, the Smart-seq2 protocol can be used (see also Picelli S, Faridani OR, AK, Winberg G, Sagasser S, Sandberg R., Nat Protoc. 2014 Jan;9(1):171-81) Analysis of single cell suspensions (reverse transcription followed by PCR amplification).

[0141] In some embodiments, short reads can be used, which are typically less than 400 base pairs (bp), particularly less than 200 bp or even less than 100 bp, while preferably at least 50 bp, particularly at least 75 bp, or at least 100 bp. In some other embodiments, long reads or reads exceeding 10 kbp-15 kbp can also be used. Typically, cells can be sequenced using, for example, 75 bp long double-end reads on a NextSeq instrument (Illumina) and High-Output v2 kit (Illumina).

[0142] Alternatively, public single cell (sc) RNAseq data (e.g., from the Sequence Read Archive (SRA) bioinformatics database, the largest publicly available repository of high-throughput sequencing data) may also be used, particularly when data are available for tumor cells and non-tumor cells from the tumor microenvironment, and / or cells infiltrating the tumor.

[0143] Step ii) typically involves aligning the reads to a reference genome and assembling the alignments into full-length transcripts, quantifying the expression level of each gene and transcript, normalizing the mapping data, and calculating the differential expression of all TEs in tumor cells versus non-tumor cells.

[0144] Raw RNA reads can be aligned (i.e., mapped) to a human genome (e.g., human genome assembly hg19 or hg38) as described in the accompanying Results (but see also Darmanis S et al., PNAS 2015 Jun 9;112(23):7285-90), typically using a software aligner such as Spliced ​​Transcripts Alignment to a Reference (STAR) software (Dobin, Alexander et al., Bioinformatics (Oxford, UK) 29:1 (2013):15-21).

[0145] scRNAseq reads can be mapped to transposable element (TE) subfamilies (e.g., as described in Kong et al., Nature-Communication (Nat Commun) 2019, 10, 5228) and / or individual genomic TE loci or occurrences. Typically, according to the present disclosure, scRNAseq reads are mapped to individual genomic TE occurrences. In addition, in order to accurately estimate the expression of older TE subfamilies and the youngest TE subfamilies (the locations mapped to individual genomes may be more particularly affected by the high conservation of their repeat motifs), multi-mapped TE reads (i.e., TE sequencing reads mapped to more than one position in the genome) and uniquely mapped TE reads (mapped / aligned to only one position in the genome) are generally considered (see in particular Lanciano, S. and Cristofari, G. (2020), Measuring and interpreting transposable element expression, Nature Review: Genetics (Nat Rev Genet) 21, 721-736). 。 ).

[0146] For TE mapping, files annotating TE positions can be added. Thus, transposable element annotations can generally be retrieved from various databases and merged when necessary (e.g., as shown in the attached examples) to generally obtain information about TEs, such as classes, families, subfamilies, differences, and / or coordinates. The embodiments section of this disclosure also describes detailed example processes. Typically, raw RNA reads (75 bp paired-end unstranded reads) can be mapped to the human genome sequence (hg19) using the 2-pass mode of STAR (e.g. version 2.7.1.a) with the following parameters: --quantMode GeneCounts, --twopassModeBasic, --alignSJDBoverhangMin 1, --bamRemoveDuplicatesType UniqueIdentical, --winAnchorMultimapNmax 1000, --outFilterMultimapNmax 1000, --outFilterScoreMinOverLread 0.33, --outFilterMatchNminOverLread 0.33, --outFilterMismatchNoverLmax 0.04, --outMultimapperOrder Random, --sjdbOverhang76.

[0147] In a specific embodiment, TEs that are completely contained in exons are deleted from the single-cell transcriptome TE pattern. This means that the single-cell transcriptome TE pattern obtained in (step (i)) does not contain TEs that are completely contained in exons.

[0148] Gene and TE expression can be quantified according to classical methods in the field, as exemplified in the "Materials and Methods" of this article. For example, to quantify TE and gene expression, featureCounts from Subread (v1.6.4) can be calculated based on each genome mapping read file (a very suitable method is described in particular in Teissandier, A., Servant, N., Barillot, E. and Bourc'his, D. (2019) Tools and best practices for retrotransposon analysis using high-throughput sequencing data, Mobile DNA (Mob DNA) 10, 52). For example, depending on the analysis, the following parameters can be used in featureCounts: (1) for gene expression: -p-ignoreDup-g gene_id using gencode gtf annotation file; (2) for TE expression on a single copy (a) with only uniquely mapped reads: -p-ignoreDup-g transcript_id using TEtranscript hg19 gtf annotation file; (b) with both unique and multiply mapped reads: -p-ignoreDup-g transcript_id-M--primary; (3) for TE expression in a subfamily with both unique and multiply mapped reads: -p-ignoreDup-g gene_id-M--primary. The cell count files can then be merged into a matrix using a regular Python script (Python 3.6).

[0149] Exploratory analysis, visualization, and statistical modeling are also typical steps after assembling and quantifying transcripts. The R programming language and the Bioconductor software suite can generally be used in accordance with the present disclosure and provide a set of tools from plotting raw data to normalization to downstream statistical modeling. In fact, the scater package is an open source R / Bioconductor package that implements a convenient data structure for representing scRNA-seq data and contains preprocessing, quality control, normalization, and visualization functions. It provides a workflow that converts raw read sequences into data sets ready for higher-level analysis in the R programming environment. RNA-seq data analysis often requires scaling normalization to eliminate bias caused by differences in sequencing depth, capture efficiency, or compositional effects between samples. Commonly used scaling normalization methods include the truncated mean of M values ​​(Robinson MD et al., Genome Biol. 2010, 11, R25), relative logarithmic expression (Anders S et al., Genome Biol. 2010, 11, R106) and the upper quartile method (Bullard JH et al., BMC Bioinformatics 2010, 247, 1-62). The scran package of seater implements a method for calculating size factors using cell pooling and deconvolution, which is also very suitable for seRNA-seq data according to the present disclosure (Lun ATL et al., Genome Biol. 2016b 17, 75). For more details, see Davis J McCarthy, Kieran R Campbell, Aaron TL Lun and Quin F Wills, Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R, Bioinformatics, vol. 33, issue 8, 15 April 2017, pp. 1179–1186.

[0150] Optionally, considering the uniquely mapped reads TE matrix, individual TEs with an average count of less than 1 per cell can be removed, and for multiply mapped reads, individual TEs with counts of less than 5 in at least 20 cells can be removed to account for expression in small populations. Low-quality cells are also typically removed from the analysis. For example, a cell can be considered low-quality if its library size is below 100,000 reads; and / or it expresses less than 5,000 genes; and / or its spike ratio is above 10%; and / or its mitochondrial ratio is above 10%.

[0151] Typically, to visualize the transcriptome landscape of various sequenced single cells, dimensionality reduction can be used to generate a two-dimensional (2D) plot. In short, the genes with the highest overdispersion are selected and used to construct a matrix of cell-to-cell dissimilarity. Then, t-distributed stochastic neighbor embedding (tSNE) can be performed on the resulting distance matrix to create a 2D plot of the cells. K-means clustering can also be used on the 2D tSNE plot.

[0152] Figure 1 An example workflow is specifically shown in A, which illustrates an alignment strategy based on individual TE occurrences (rather than family level), and TE quantification using unique or multiple mapped reads.

[0153] In step iii), a TE transcript sequence that is differentially expressed (i.e., upregulated) in at least one tumor cell compared to the at least one non-tumor cell is selected, and a tumor cell TE tag is obtained. According to the present disclosure, TE transcripts that are statistically differentially expressed in tumor cells compared to non-tumor cells are selected, and the adjusted p-value is usually equal to or lower than 0.05, particularly equal to or lower than 0.01. Alternatively, or additionally, in some embodiments, TE transcripts expressed in tumor cells with an average log 2 fold change of 0.25 compared to non-tumor cells can be selected. In some embodiments, the peptide is encoded by a TE transcript or a fragment thereof, which is expressed in tumor cells with a log 2 fold change of at least 0.25, particularly at least 0.5, at least 0.75, at least 1, at least 1.25, at least 1.5, at least 1.75 or at least 2 compared to non-tumor cells. In some embodiments, the TE signature according to the present disclosure comprises at least 30, particularly at least 25, at least 20, at least 15, at least 10, at least 5 most differentially expressed TEs of tumor cells compared to at least one other non-tumor cell.

[0154] According to the present invention, a tumor cell transposable element (TE) signature corresponds to a TE transcript specifically expressed by tumor cells, particularly in some embodiments, the TE transcript selected in the tumor cell signature is not found in a single cell transcriptome pattern of TE transcripts obtained from at least one non-tumor cell.

[0155] Typically, one or more tumor cells from one or more tumor samples of one or more patients are analyzed differentially with one or more non-tumor cells from one or more samples of one or more subjects. As previously mentioned, tumor cells can be from the same or different tumors or tumor types. Non-tumor cells may or may not be from the same tissue sample, including immune cells, such as tumor-infiltrating immune cells (e.g., macrophages) and non-tumor cells from tumor microenvironment (e.g., samples near tumors). In some embodiments, differential analysis is performed between cells from the same patient, particularly from the same sample or from samples collected from the same type of tumor (from one or more patients) and samples from the vicinity of the tumor of one or more subjects (i.e., tumor adjacent samples). In some embodiments, one or more tumor cells can be obtained from tumor samples of the same patient at different times T.

[0156] The method disclosed herein can obtain a set of TE transcripts that are differentially expressed in tumor cells compared to non-tumor cells, also referred to herein as a tumor cell "TE signature" or a tumor cell "transcriptome TE pattern".

[0157] For example, differential analysis can be performed as described in detail in the Materials and Methods section of the Examples section.

[0158] Method for identifying or screening TE-derived tumor neoantigenic peptides

[0159] The method includes obtaining a single tumor cell TE signature (or transcriptome TE pattern), and then translating the TE transcript sequence from the tumor cell TE signature by computer simulation to obtain TE-derived tumor peptides.

[0160] The method includes the step of identifying an open reading frame (ORF) sequence from a TE-tagged transcript. In some embodiments, the transcript is then translated by computer simulation using a six-frame translation (forward and reverse), and the resulting amino acid sequence is then fragmented at all stop codons to obtain TE-encoded tumor peptide sequences, which can then be grouped to form a TE-derived tumor peptide library.

[0161] In some embodiments, the method further comprises a step of identifying TE-derived peptides that bind to at least one MHC molecule. According to such embodiments, a library (tumor TE library) containing TE-encoded tumor peptide sequences obtained from TE tumor tags as described above is usually compared with an MHC / HLA-ligand group obtained from more tumor cells from one or more samples (including tumor cells from the same and / or different tumor types, such as glioblastoma cells). Peptides that match the tumor TE library are usually selected from the MHC / HLA-ligand group. This step can clearly identify tumor neoantigen peptides encoded by TE presented by HLA / MHC molecules.

[0162] Typically, a tumor TE library as described above is combined with human protein sequences (ie: a human annotated proteome - e.g., Uniprot / SwissProt).

[0163] According to the present disclosure, identification of TE-derived peptides that bind to at least one MHC / HLA molecule is typically achieved by a proteomic approach, wherein mass spectrometry (MS)-based proteomic (particularly immunoproteomic) data is matched with a peptide library obtained from a tumor TE library as described above, more particularly, open reading frames derived from de novo assembled transcripts (e.g., a previously defined tumor TE library) are searched based on immunopeptide-based MS / MS spectra (obtained from tissue samples or cells, including cell lines, such as tumor samples and tumor cells, particularly tumor samples or cell lines).

[0164] Therefore, MHC-ligandomes are usually in the form of raw mass spectrometry (MS) data (i.e., spectra) obtained by MS-based proteomics (especially immunoproteomics) techniques, such as bottom-up proteomics (shot-gun proteomics) and top-down proteomics from one or more tissue samples or cells (e.g., tumor samples and tumor cells).

[0165] Immunopeptidomics approaches are typically based on immunoaffinity purification (IP) of HLA / MHC complexes, typically from mild detergent-solubilized lysates, followed by extraction of HLA / MHC peptides (HLA / MHCp). The extracted peptides are then separated by chromatography and directly injected into a mass spectrometer. Tumor MHC / HLA-ligandomes are typically obtained by first purifying surface MHC-binding (i.e., HLA-I or HLA-2 molecules) peptides and then characterizing their amino acid sequences. Typically, MHC / HLA ligandomes are obtained from one or more tumor samples (e.g., biopsies or tissues) or tumor cells (e.g., glioblastoma cells) from a tumor cell line. For example, MHC / HLA-binding molecules can be purified from cell lysates by immunoprecipitation using antibodies specific for the desired MHC / HLA species (e.g., using MHC / HLA-IP). MHC / HLA-associated peptides can be separated from larger MHC / HLA components, and the peptide fraction can be further analyzed by liquid chromatography tandem mass spectrometry (LC-MS / MS). Peptide sequences can be identified by spectral interpretation. Large-scale data acquired from high-resolution mass spectrometers are typically interpreted using algorithms that can assign mass spectra to amino acid sequences.

[0166] A variety of software can be used by those skilled in the art to interpret MS fragmentation spectra (see, for example, Purcell, AW, Ramarathinam, SH, and Ternette, N. Mass spectrometry-based identification of MHC-binding peptides for immunopeptidomics). - Based identification of MHC-bound peptides for immunopeptidomics), Nat Protoc 14, 1687-1707 (2019) or Prianichnikov, Nikita et al., "MaxQuant Software for Ion Mobility Enhanced Shotgun Proteomics." Molecular & Cellular Proteomics: MCP 19, 6 (2020): 1058-1069). MS-based immunopeptidomic analysis is also described in detail in Forlani, Greta et al., Molecular & cellular proteomics (MCP), vol. 20, 100032, January 6, 2021; and Chong, Chloe et al., "High-throughput and Sensitive Immunopeptidomics Platform Reveals Profound Interferonγ-Mediated Remodeling of the Human Leukocyte Antigen (HLA) Ligandome." Molecular & cellular proteomics (MCP), vol. 17, 3 (2018): 533-548, referring to the use of the MaxQuant computational proteomics platform to search the peak list against the UniProt database - see Cox, Jürgen and Matthias Mann, Nature Biotechnology (Nat Biotechnol. Biotechnol. Vol. 26, 12 (2008): 1367-72.). Additional references are also provided in the results of this application, which describe very suitable protocols for obtaining MS raw data that can be used according to the present disclosure. According to the methods of the present disclosure, public MS data can be used, as shown in the results included in this article, for example.

[0167] In some embodiments, the selected TE-encoded tumor neoantigen MHC binding peptides are further filtered according to classical proteins, typically classical proteins from the human proteome (e.g., typically obtained from Swiss-Prot and TrEMBL databases). UniProtKB / TrEMBL is a computer-annotated protein sequence database that supplements the UniProtKB / Swiss-Prot protein knowledge base. UniProtKB / TrEMBL contains translations of all coding sequences (CDS) present in the EMBL / GenBank / DDBJ nucleotide sequence databases and protein sequences extracted from the literature or submitted to UniProtKB / Swiss-Prot. The database is enriched by automatic classification and annotation.

[0168] In some embodiments, non-redundant peptides (i.e., peptides encoded by a single genomic TE) are further selected. This selection can be achieved by mapping the tumor neoantigen MHC binding peptides encoded by the identified TE to the corresponding TE in the TE tag, as shown in the results included herein.

[0169] In other embodiments, redundant peptides are further selected. Redundant peptides encoded by TEs with low genomic TE incidence (e.g., encoded by less than 100, particularly less than 50, particularly less than 10 genomic TE occurrences), whose expression is highly upregulated in tumor cells (log2 fold change of at least 0.25, particularly at least 0.5, at least 1, at least 1.5) and / or not expressed in normal cells or samples (e.g., using the GTEx database), are of particular relevance.

[0170] Binding of putative neoantigenic peptides obtained from tumor cell TE signatures (and in particular MHC / HLA binding peptides identified in the above method) to at least one MHC molecule can also be determined by computer simulation.

[0171] When performed on a human sample, the method may include the step of determining the patient's class I or class I major histocompatibility complex (MHC, also known as human leukocyte antigen (HLA) alleles).

[0172] An MHC allele database is established by analyzing known MHC I and MHC II sequences and determining the allelic variability of each domain. This can usually be determined by computer simulation using appropriate software algorithms well known in the field. Several tools have been developed to obtain HLA allele information from whole genome sequencing data (whole exome, whole genome and RNA sequencing data), including OptiType, Polysolver, PHLAT, HLAreporter, HLAforest, HLAminer and seq2HLA (see Kiyotani K et al., Immunopharmacogenomics towards personalized cancer immunotherapy targeting neoantigens; Cancer Science 2018; 109: 542-549). For example, the seq2hla tool (see Boegel S, Lower M, Schafer M, et al., HLA typing from RNA-Seq sequence reads, Genome Med. 2012; 4: 102), which is well designed to perform the methods disclosed herein, is a computer simulation method written in python and R that takes standard RNA-Seq sequence reads in fastq format as input, uses a bowtie index containing all HLA alleles (Langmead B, et al., Ultrafast and memory-efficient alignment of short DNA sequences to the human genome., Genome Biol. 2009, 10: R25-10.1186 / gb-2009-10-3-r25), and outputs the most likely HLA Class I and II genotypes (4-digit resolution), p-values ​​for each call, and expression for each category.

[0173] Computational methods can be used to determine the affinity of all possible peptides encoded by each transcript sequence of each MHC allele of a subject by computer simulation to predict the binding affinity of the peptide to the HLA molecule. In fact, accurate prediction methods are based on having predicted IC 50For example, NetMHCpan software, which is a modification of NetMHC to predict peptide binding to alleles for which ligands have not yet been reported, is well suited for implementing the methods disclosed herein (Lundegaard C et al., NetMHC-3.0: accurate web accessible predictions of human, mouse and monkey MHC class I affinities for peptides of length 8-11; Nucleic Acids Res 2008;36:W509-W512; Nielsen M et al., NetMHCpan, a method for quantitative predictions of peptide binding to any HLA-A and -B locus protein of known sequence, PLoS One. One. 2007;2:e796, but see also Kiyotani K et al., Immunopharmacogenomics towards personalized cancer immunotherapy targeting neoantigens. Cancer Science. 2018;109:542-549 and Yarchoan M et al., Nat Rev Cancer. 2017;17(4):209-222, and see also Reynisson B., Barra C., Kaabinejadian S., Hildebrand WH, Peters B., Nielsen M., J. Proteome Res. 2020;19:2304-2315 and Jurtz V, Paul S, Andreatta M, Marcatili P, Peters B, Nielsen M, who published NetMHCpan-4.0. M. J Immunol. 2017 Nov 1;199(9):3360-3368).NetMHCpan software uses an artificial neural network (ANN) to predict peptide binding to any MHC molecule of known sequence. The method is trained on a combination of over 180,000 quantitative binding data and MS-derived MHC eluted ligands. Binding affinity data cover 172 MHC molecules from human (HLA-A, B, C, E), mouse (H-2), bovine (BoLA), primate (Patr, Mamu, Gogo) and porcine (SLA). MS eluted ligand data cover 55 HLA and mouse alleles. NetMHCpan-4.0 version also provides improved peptide-MHC class I interaction predictions integrating eluted ligand and peptide binding affinity data.

[0174] In an exemplary embodiment, the Kd affinity score of the predicted peptides from the tumor cell TE signature disclosed herein and for the MHC alleles is less than 10 -4 , 10 -5 , 10 -6 , 10 -7 TE-encoded peptides with a concentration of less than 500 nM or a level less than 2% (usually depending on the netMHCpan version) were selected as tumor neoantigen peptides.

[0175] MixMHCpred (v.2.0.2) (see Bassani-Sternberg M. et al., PLoS Comput. Biol. 2017; 13) and MixMHC2pred (v.1.0) (see Racle J. et al., Nat Biotechnol 2019; 37: 1283-1286) can also be used to predict peptide binding to patient HLA / MHC class I alleles and patient HLA / MHC class II alleles, respectively, as described by Forlani, Greta et al. (Molecular & cellular proteomics: MCP, Vol. 20, 2021: 100032).

[0176] In some embodiments, peptides binding to MHC class I molecules can therefore be predicted using, for example, the NetMHCpan 4.1 suite (http: / / www.cbs.dtu.dk / services / NetMHCpan / ) using "HLA allele" = A2, "peptide length" = 8-11, and "rank threshold for strong binding" = 0.5%, an example protocol of which is detailed in the results included in the "Examples" section.

[0177] Typically, the predicted K values ​​for MHC alleles are derived from tumor cell TE signatures. dTE-encoded peptides with an affinity score of less than 50 nM or a rank of less than 0.5% (usually depending on the netMHCpan version) are selected as tumor neoantigenic peptides. Thus, in some embodiments, the neoantigenic peptides encoded by TEs according to the present disclosure, generally identified according to the method, bind to at least one HLA / MHC molecule with an affinity sufficient to present the peptide as an antigen on the cell surface. Generally, the IC of the neoantigenic peptide to at least one HLA / MHC molecule (usually a molecule of the subject with cancer or tumor) is 50 Affinity less than 10 -4 , or 10 -5 , or 10 -6 , or 10 -7 Or less than 500 nM, at least less than 250 nM, at least less than 200 nM, at least less than 150 nM, at least less than 100 nM, at least less than 50 nM or less (the smaller the number, the higher the binding affinity).

[0178] Neoantigenic peptides, polynucleotides and vectors

[0179] The present disclosure also includes isolated tumor neoantigenic peptides having at least the following characteristics:

[0180] i. It has at least 8 amino acids and comprises a TE coding sequence.

[0181] ii. Its value is less than 10 -5 M of K D The binding affinity binds to at least one MHC class I or class II molecule of the subject and / or it is presented by an MHC molecule of the subject.

[0182] As previously described, MHC binding of the peptides disclosed herein can be assessed by computer simulation. As shown in the examples, Kd affinity for at least one MHC / HLA molecule can also be determined or predicted in vitro by using tetramer preparations. In short, suitable commercial kits (e.g., of Kit, which can be used according to its training guide) to prepare HLA 02:01 / peptide multimers and to bind to human CD8 prepared from healthy donors + Tetramer-CD8 + Cell binding. Typically, the binding affinity can be determined based on the percentage of binding to a positive control. Typically, peptides are selected that exhibit a percentage of binding to a positive control of at least 30%, particularly at least 40% or even at least 50%. Typically, the neoantigenic peptides disclosed herein and typically obtainable according to the present methods bind to at least one HLA / MHC molecule with an affinity sufficient to present the peptide as an antigen on the cell surface. In some embodiments, the IC of the neoantigenic peptide is 50Less than 10 -4 , or 10 -5 or 10 -6 or 10 -7 Or less than 500 nM, at least less than 250 nM, at least less than 200 nM, at least less than 150 nM, at least less than 100 nM, at least less than 50 nM or less (the smaller the number, the higher the binding affinity). For example, the IC 50 In some embodiments, the novel antigenic peptides disclosed herein are present in an amount of less than about 10 -4 , 10 -5 , 10 -6 , 10 -7 , 10 -8 or 10 -9 M (smaller numbers indicate higher binding affinity), especially 10 -4 M to 10 -9 M, especially 10 -4 M to 10 -8 M, especially 10 -4 M to 10 -7 The binding affinity, Kd, ​​of M binds to both MHC class I and class II molecules.

[0183] In some embodiments, the neoantigenic peptides of the present disclosure bind to MHC class I molecules with a binding affinity of less than 2% percentile rank score, for example, as predicted by NetMHCpan 4.0. In some embodiments, the neoantigenic peptides of the present disclosure bind to MHC class II molecules with a binding affinity of less than 10% percentile rank score, for example, as predicted by NetMHCpanII 3.2.

[0184] The presentation of the neoantigenic peptides according to the present disclosure by MHC / HLA molecules can also be assessed by interrogating the tumor immune peptidome using the neoantigenic peptide sequences previously detailed.

[0185] The tumor neoantigen peptides disclosed herein also exhibit one or more of the following characteristics:

[0186] iii. The expression of TE is derepressed in tumor cells compared with non-tumor cells.

[0187] De-repression in tumor cells refers herein to the expression of TE transcript sequences in tumor cells that are statistically significantly upregulated (as defined above) compared to normal healthy cells. In some embodiments, TE expression is de-repressed in tumor cells from a given type of cancer. In some embodiments, compared to one or more non-tumor cells, TE transcripts are expressed in tumor cells with an average of at least 0.25, particularly at least 0.5, at least 0.75, at least 1, at least 1.25, at least 1.5, at least 1.75 or at least 2 log 2 fold changes. For example, in some embodiments, TE is de-repressed in glioblastoma. The validation of TE-specific tumor cell expression can be evaluated by performing RNAseq analysis, as shown in the Examples section. For example, compared to scRNAseq from non-tumor cells (e.g., including tumor infiltrating cells, particularly immune cells, such as macrophages) and / or TCGA tumor vicinity batch RNAseq samples (usually from the same tumor as the tumor cells used for tumor single cell analysis), TE transcript sequences according to the present disclosure are overexpressed in scRNAseq from one or more tumor cells. In some embodiments of the present disclosure, the TE transcript sequence is not expressed in non-tumor cells (including tumor-infiltrating cells), samples from normal tissues, and / or tumor-adjacent samples (eg, obtained from the TCGA database).

[0188] iv.TE is selected from more than 50×10 6 (50.10 6 ) years of TE;

[0189] v. TE is selected from LINE-1, SVA and ERVK TE subfamilies; more particularly, TE is selected from L1PA / B / x TE;

[0190] vi. TEs are selected from TEs carrying complete or almost complete ORFs (usually the mismatches between the classic TE proteins from the gEVE database and the peptide sequences retrieved from the immunopeptidomics map are no more than 2, especially no more than 1);

[0191] vii. TEs are selected from TEs encoding unique peptides;

[0192] viii. TEs were selected from introns or intergenic TEs (usually distal TEs located more than 2 kb away from the nearest gene).

[0193] ix.TE is encoded by chromosome 7.

[0194] Generally, the neoantigenic peptides of the present disclosure are obtained according to the methods previously described in detail.

[0195] In some embodiments, the tumor cell TE tag is a glioblastoma cell TE tag, and the peptide sequence is obtained from a glioblastoma cell TE tag comprising a transcript sequence of SEQ ID NOs: 381-5020.

[0196] In some embodiments, the tumor cell TE signature, in particular the glioblastoma cell TE signature, excludes TEs that are completely contained in exons. In some more specific embodiments, the new antigenic peptide sequence is obtained from a glioblastoma cell TE signature comprising transcript sequences of SEQ ID NOs: 381 to 430 and 432 to 5020; preferably, the new antigenic peptide sequence is obtained from a glioblastoma cell TE signature comprising transcript sequences of SEQ ID NOs: 381 to 393; 395 to 430 and 432 to 5020.

[0197] In some embodiments, the neoantigenic peptide is encoded by an ORF sequence or a fragment thereof from a transcript of any one of SEQ ID NOs: 381 to 5020. In some specific embodiments, the neoantigenic peptide is encoded by an ORF sequence or a fragment thereof from a transcript of any one of SEQ ID NOs: 381 to 430 and 432 to 5020; more particularly, the neoantigenic peptide is encoded by an ORF sequence or a fragment thereof from a transcript of any one of SEQ ID NOs: 381 to 393; 395 to 430 and 432 to 5020. Typically, the transcripts are translated in six-frame translation (forward and reverse), and the resulting amino acid sequence is then fragmented at all stop codons to obtain the TE-encoded (tumor-specific neoantigen) peptide sequence.

[0198] In some embodiments, the neoantigenic peptide comprises a sequence of any one of SEQ ID NOs: 1 to 380 or a fragment thereof, in particular SEQ ID NOs: 1, 2, 9, 11, 13, 18, 22, 23, 27, 30 to 32, 35, 36, 38 to 40, 42, 45, 48 to 50, 54, 57, 58, 60, 61, 63-66, 68, 70 to 73, 76, 78, 79, 82, 83, 88, 89, 91, 93 to 95, 98, 104 to 107, 110, 111, 114, 115, 117 to 124, 126, 127, 131, 133, 138, 139, 141, 143, 144, 150 to 153, 157, 159, 161, 162, 164, 165, 167, 172, 173, 177, 179 to 182, 188, 190, 193, 198, 199, 206, 208, 212, 214, 215, 217, 218, 222, 223, 228, 238, 239, 243 to 246, 248, 251, 253, 256, 257, 259, 260, 262, 265, 267, 275, 277, 279, 281 to 283, 285 to 288, 290 to 292, 294 to 302, 304, 305, 307, 311 to 315, 317, 318, 320, 323, 325, 326, 328, 329, 331, 333 to 335, 337, 343, 344, 346, 350, 352 to 356, 359 to 362, 365, 367, 369 or 370 (non-redundant; see Table 3).

[0199] In some specific embodiments, the new antigenic peptide comprises a sequence or a fragment thereof of any one of SEQ ID NOs: 1 to 26 and 28 to 380; preferably, the new antigenic peptide comprises a sequence or a fragment thereof of any one of SEQ ID NOs: 1 to 10; 12 to 26; 28 to 57; 59 to 242; 244 to 255; 257 to 319 and 321 to 380, in particular SEQ ID NOs: 1, 2, 9, 11, 13, 18, 22, 23, 30 to 32, 35, 36, 38 to 40, 42, 45, 48 to 50, 54, 57, 60, 61, 63-66, 68, 70 to 73, 76, 78, 79, 82, 83, 88, 89, 91, 93 to 95, 98, 104 to 107, 110, 111 1, 114, 115, 117 to 124, 126, 127, 131, 133, 138, 139, 141, 143, 144, 150 to 153, 157, 159, 161, 162, 164, 165, 167, 172, 173, 177, 179 to 182, 188, 190, 193, 198, 199 , 206, 208, 212, 214, 215, 217, 218, 222, 223, 228, 238, 239, 244 to 246, 248, 251, 253, 257, 259, 260, 262, 265, 267, 275, 277, 279, 281 to 283, 285 to 288, 290 to 292, 294 to 302, 304, 305, 307, 311 to 315, 317, 318, 323, 325, 326, 328, 329, 331, 333 to 335, 337, 343, 344, 346, 350, 352 to 356, 359 to 362, 365, 367, 369 or 370 (non-redundant; see Table 3).

[0200] In some embodiments, the isolated tumor-specific neoantigenic peptide comprises at least 8 amino acids, particularly 8 or 9 amino acids, and binds to at least one MHC class I molecule of the subject previously defined, or comprises 13 to 25 amino acids and binds to at least one MHC class II molecule of the subject previously defined.

[0201] In some embodiments, the tumor neoantigen peptides according to the present disclosure bind to MHC molecules present in at least 1%, 5%, 10%, 15%, 20%, 25% or more of the subjects. Notably, the tumor neoantigen peptides disclosed herein are expressed in at least 1%, 5%, 10%, 15%, 20%, 25% of the subjects in a population of subjects with a given type of cancer or tumor (e.g., in a population of subjects with glioblastoma).

[0202] More particularly, the tumor neoantigen peptides disclosed herein can elicit an immune response against tumors present in at least 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95% or even 99% of a population of subjects suffering from cancer or tumors, and more particularly a population of subjects suffering from a given type of tumor (e.g., glioblastoma).

[0203] In some embodiments, the isolated tumor neoantigen peptide comprises at least 8, 9, 10, 11 or 12 amino acids, which is encoded by a portion of an open reading frame (ORF) of a TE transcript from SEQ ID NOs: 381 to 5020, or comprises a sequence of any one of SEQ ID NOs: 1 to 380 or a fragment thereof. In some specific embodiments, the isolated tumor neoantigen peptide comprises at least 8, 9, 10, 11 or 12 amino acids, which are encoded by a portion of an open reading frame (ORF) of a TE transcript of SEQ ID NO: 381 to 430 and 432 to 5020, preferably SEQ ID NO: 381 to 393; 395 to 430 and 432 to 5020; or a sequence or a fragment thereof comprising any one of SEQ ID NO: 1 to 26 and 28 to 380, preferably SEQ ID NO: 1 to 10; 12 to 26; 28 to 57; 59 to 242; 244 to 255; 257 to 319 and 321 to 380. The length of the peptide can be 8-9, 8-10, 8-11, 12-25, 13-25, 12-20 or 13-20 amino acids. Thus, the N-terminus of a peptide of at least 8 amino acids can generally be encoded by a triplet codon starting at any one of nucleotide positions 1, 4, 7, 10, 13, 16, 19 (forward and reverse).

[0204] Generally, the tumor-specific neoantigenic peptides according to the present disclosure may exhibit one or more of the following properties:

[0205] - When administered to a subject, it does not induce an autoimmune response and / or induces immune tolerance. The tolerance mechanism involves clonal deletion, ignorance, anergy or suppression of the reduction in the number of high-affinity autoreactive T cells in the host.

[0206] -It is specifically expressed in tumor cells, in some embodiments, it is only expressed in one or more tumor cells, but not expressed in healthy cells (e.g., expression cannot be detected). For example, the lack of expression of the new antigenic peptide in healthy cells can be tested using, in particular, the Basic Local Alignment Search Tool (BLAST) and aligning the sequence of the new antigenic peptide with the transcriptome of healthy cells.

[0207] In some embodiments, the peptide is encoded by a single genomic TE (ie, the peptide is non-redundant).

[0208] In other embodiments, the peptide is encoded by more than one TE (i.e., the peptide is redundant). In more specific embodiments, the peptide is highly repetitive (typically encoded by more than 200 genomic TEs) and is non-tumor specific, while in other specific embodiments, the peptide has low redundancy (typically encoded by less than 100, particularly less than 50 or less than 10 genomic TEs) and is encoded by a TE that is highly upregulated in tumor cells and / or is not expressed in normal cells or samples (e.g., only expressed in at least one tumor cell, particularly glioblastoma cells).

[0209] Typically, immunization with a tumor neoantigen peptide according to the present disclosure elicits a T cell response (i.e., is immunogenic). The immunogenicity of the neoantigen peptide can be assessed using an in vitro vaccination assay, such as described in the "Examples" section. + Assessment of T cells can be achieved by flow cytometry (Flow Cytometry and Fluorescence-Activated Cell Sorting, FACS) using multimer staining.

[0210] The new antigenic peptide can also be modified by extending or reducing the amino acid sequence of the compound, for example, by adding or deleting amino acids. The peptide can also be modified by changing the order or composition of certain residues. It is easy to understand that certain amino acid residues that are essential for biological activity (such as amino acid residues or conserved residues at key contact sites) cannot usually be changed, otherwise the biological activity will be adversely affected. Non-critical amino acids are not necessarily limited to amino acids naturally occurring in proteins, such as L-α-amino acids or their D-isomers, but can also include non-natural amino acids, such as β-γ-δ-amino acids, and many derivatives of L-α-amino acids.

[0211] Typically, a series of peptides with single amino acid substitutions are used to determine the effects of electrostatic charge, hydrophobicity, etc. on binding. For example, a series of positively charged (e.g., Lys or Arg) or negatively charged (e.g., Glu) amino acid substitutions are performed along the length of the peptide, which show different sensitivity patterns to various MHC molecules and T cell receptors. In addition, multiple substitutions using small and relatively neutral parts such as Ala, Gly, Pro or similar residues can be used. The substitution can be a homo-oligomer or a hetero-oligomer. The number and type of residues substituted or added depends on the necessary spacing between the basic contact points and certain functional attributes sought (e.g., hydrophobicity and hydrophilicity). Compared with the affinity of the parent peptide, such substitutions can also achieve an increase in binding affinity to MHC molecules or T cell receptors. In any case, such substitutions should use amino acid residues or other molecular fragments to avoid, for example, spatial and charge interferences that may destroy binding.

[0212] Amino acid substitutions are usually single residue amino acid substitutions. Substitution, deletion, insertion or any combination thereof can be combined to obtain the final peptide. The substituted variant is a variant in which at least one residue of the peptide is removed and a different residue is inserted in its place. When it is necessary to finely adjust the characteristics of the peptide, such substitutions are usually carried out according to Table 1 below.

[0213]

[0214]

[0215] Table 1

[0216] Significant changes in function (e.g., affinity for MHC molecules or T cell receptors) are achieved by selecting substitutions that are less conservative than those in Table 1 above, i.e., selecting residues that have a more pronounced difference in effect in maintaining (a) the structure of the peptide backbone in the region of the substitution (e.g., sheet or helical conformation), (b) the charge or hydrophobicity of the molecule at the target site, or (c) the bulk of the side chain. Substitutions that are generally expected to produce the greatest changes in the properties of the peptide will be those in which (a) a hydrophilic residue (e.g., seryl) replaces a hydrophobic residue (e.g., leucyl, isoleucyl, phenylalanyl, valyl, or alanyl) (or is replaced by a hydrophobic residue); (b) a residue with a positively charged side chain (e.g., lysyl, arginyl, or histidyl) replaces a negatively charged residue (e.g., glutamyl or aspartyl) (or is replaced by a negatively charged residue); or (c) a residue with a bulky side chain (e.g., phenylalanine) replaces a residue without a side chain (e.g., glycine) (or is replaced by a residue without a side chain).

[0217] Peptides and polypeptides may also comprise isosteres of two or more residues in a neoantigenic peptide or polypeptide. An isostere as defined herein is a sequence of two or more residues that can replace a second sequence because the stereo conformation of the first sequence is suitable for having a specific binding site for the second sequence. The term specifically includes peptide backbone modifications well known to those skilled in the art. Such modifications include modifications of amide nitrogen, α-carbon, amide carbonyl, complete replacement of amide bonds, extension, deletion or backbone cross-linking. See generally Spatola, Chemistry and Biochemistry of Amino Acids, Peptides and Proteins, Volume VII (Edited by Weinstein, 1983).

[0218] In addition, the neoantigenic peptides can be conjugated to carrier proteins, ligands or antibodies. The half-life of the peptides can be improved by PEGylation, glycosylation, polysialylation, HESylation, recombinant PEG mimics, Fc fusion, albumin fusion, nanoparticle attachment, nanoparticle encapsulation, cholesterol fusion, iron fusion or acylation.

[0219] Modification of peptides and polypeptides with various amino acid mimetics or unnatural amino acids is particularly useful for improving the stability of peptides and polypeptides in vivo. Stability can be determined in a variety of ways. For example, peptidases and various biological media (e.g., human plasma and serum) have been used to test stability. See, for example, Verhoef et al., Eur. J. Drug Metab Pharmacokin. 11: 291-302 (1986). The half-life of the peptides disclosed herein is conveniently determined using a 25% human serum (v / v) assay. The protocol is generally as follows. Pooled human serum (AB type, non-heat inactivated) is defatted by centrifugation prior to use. The serum is then diluted to 25% with RPMI tissue culture medium and used to test peptide stability. A small amount of the reaction solution is removed at predetermined time intervals and added to a 6% trichloroacetic acid aqueous solution or ethanol. The turbid reaction sample is cooled (4° C.) for 15 minutes and then spun to pellet the precipitated serum proteins. The presence of the peptide is then determined by reverse phase high performance liquid chromatography (HPLC) using stability-specific chromatographic conditions.

[0220] Peptides and polypeptides can be modified to provide the desired attributes except improving serum half-life. For example, the ability of peptide inducing CTL activity can be enhanced by being connected to a sequence containing at least one epitope that can induce helper T cell response. Particularly preferred immunogenic peptide / helper T cell conjugates are connected by spacer molecules. Spacers are usually composed of relatively small neutral molecules, such as amino acids or amino acid mimetics, which are basically uncharged under physiological conditions. Spacers are usually selected from neutral spacers such as Ala, Gly or other non-polar amino acids or neutral polar amino acids. It should be understood that the spacers that are optionally present do not have to be composed of the same residues, so they can be heterologous or homologous oligomers. When present, spacers usually contain at least one or two residues, more commonly containing three to six residues. Alternatively, peptides can be connected to T helper peptides without spacers.

[0221] The new antigenic peptide can be linked to the T helper peptide directly or through a spacer at the amino or carboxyl terminus of the peptide. The amino terminus of the new antigenic peptide or the T helper peptide can be acylated. Exemplary T helper peptides include tetanus toxoid 830-843, influenza 307-319, malaria circumsporozoite protein 382-398 and 378-389.

[0222] Protein or peptide can be prepared by any technology known to those skilled in the art, including expressing protein, polypeptide or peptide by standard molecular biology techniques, isolating protein or peptide from natural sources or chemically synthesizing protein or peptide. Nucleotide and protein, polypeptide and peptide sequences corresponding to various genes have been previously disclosed, and can be found in computerized databases known to those of ordinary skill in the art. One of the databases is Genbank and GenPept databases of the National Center for Biotechnology Information on the website of the National Institutes of Health. The coding region of known genes can be amplified and / or expressed using technology disclosed herein or technology known to those of ordinary skill in the art. Alternatively, various commercial preparations of proteins, polypeptides and peptides known to those skilled in the art.

[0223] In another aspect, the present disclosure provides a nucleic acid (e.g., polynucleotide) encoding a new antigenic peptide disclosed herein. The polynucleotide can be selected from DNA, cDNA, PNA, CNA, RNA, and can be single-stranded and / or double-stranded, or a natural or stable form of polynucleotide, such as a polynucleotide with a thiophosphate backbone, or a combination thereof, and as long as the encoded peptide may contain or not contain introns. Only peptides containing naturally occurring amino acid residues connected by naturally occurring peptide bonds can be encoded by polynucleotides. In some embodiments, the polynucleotide can be connected to a heterologous regulatory sequence (e.g., a heterologous transcription and / or translation regulatory nucleotide sequence well known in the art).

[0224] Another aspect of the present disclosure provides expression vectors capable of expressing the new antigenic peptides disclosed herein. Expression vectors for different cell types are well known in the art and can be selected without undue experimentation. Typically, DNA is inserted into an expression vector, such as a plasmid, in the correct orientation and the correct reading frame for expression. The expression vector will contain appropriate heterologous transcription and / or translation regulatory nucleotide sequences recognized by the desired host. The polynucleotide encoding the tumor new antigen peptide can be linked to such heterologous regulatory nucleotide sequences, or can be non-adjacent but operably linked to such heterologous regulatory nucleotide sequences. The vector is then introduced into the host by standard techniques. Guidance can be found, for example, in Sambrook et al. (1989) Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.

[0225] Antigen presenting cells (APC)

[0226] The present disclosure also includes a population of antigen presenting cells that have been pulsed with one or more peptides defined previously and / or that can be obtained by the methods described previously. Preferably, the antigen presenting cell is a dendritic cell (DC) or an artificial antigen presenting cell (aAPC) (see Neal, Lillian R et al., "The Basics of Artificial Antigen Presenting Cells in T Cell-Based Cancer Immunotherapies." Journal of immunology research and therapy Vol. 2, 1 (2017): 68-79). Dendritic cells (DC) are professional antigen presenting cells (APC) with the extraordinary ability to stimulate naive T cells and initiate primary immune responses to pathogens. In fact, the main role of mature DC is to sense antigens and produce mediators that activate other immune cells (particularly T cells). DC is a potent stimulator of lymphocyte activation because they express MHC molecules that trigger TCR (signal 1) and co-stimulatory molecules (signal 2) on T cells. In addition, DCs secrete cytokines that support T cell expansion. T cells need antigens presented in the form of processed peptides to recognize foreign pathogens or tumors. The presentation of peptide epitopes derived from pathogen / tumor proteins is achieved through MHC molecules. MHC class I (MHC-I) and MHC class II (MHC-II) molecules present processed peptides to CD8+T cells and CD4+T cells, respectively. Importantly, DCs go to inflammatory sites containing large populations of T cells to promote immune responses. Therefore, DCs can be a key component of any immunotherapy approach because they are closely related to the activation of adaptive immune responses. In the context of vaccines, DC therapy can enhance the T cell immune response of healthy volunteers or patients with infectious diseases or cancer to the desired target. In one embodiment, the APC is an artificial APC that is genetically modified to express the desired T cell co-stimulatory molecules, human HLA alleles and / or cytokines. Such artificial antigen presenting cells (aAPCs) can provide sufficient T cell engagement, co-stimulation, and sustained release of cytokines, thereby achieving controlled T cell expansion. These cells are not constrained by time and limited availability and can be stored in small aliquots for subsequent use in generating T cell lines from different donors, thus being ready-made reagents for immunotherapy applications. The expression of potent co-stimulatory signals on these aAPCs confers a higher efficiency to this system, thereby increasing the potency of adoptive immunotherapy.In addition, aAPCs can be engineered to express genes that direct the release of specific cytokines to promote preferential expansion of desired T cell subsets for adoptive transfer; for example, long-lived memory T cells (see Hasan AH et al., Artificial Antigen Presenting Cells: An Off the Shelf Approach for Generation of Desirable T-Cell Populations for Broad Application of Adoptive Immunotherapy, Adv Genet Eng. 2015;4(3):130, Kim JV, Latouche JB, Rivière I, Sadelain M., The ABCs of artificial antigen presentation, Nat Biotechnol 2004;22:403-410 or Wang C, Sun W, Ye Y, Bomba HN, Gu ... Z., Bioengineering of Artificial Antigen Presenting Cells and Lymphoid Organs, Theranostics, 2017;7(14):3504-3516.

[0227] Typically, dendritic cells are autologous dendritic cells pulsed with the new antigenic peptides disclosed herein. The peptide can be any suitable peptide that causes an appropriate T cell response. Antigen presenting cells (or stimulator cells) typically have MHC class I or MHC class II molecules on their surface, and in one embodiment, they themselves are substantially unable to load MHC class I and MHC class II molecules with selected antigens. MHC class I or class II molecules can be easily loaded with selected antigens in vitro.

[0228] As an alternative, the antigen presenting cell may contain an expression construct encoding the tumor neoantigen peptide disclosed herein. The polynucleotide may be any suitable polynucleotide as previously defined, and preferably, it is capable of transducing dendritic cells, thereby achieving peptide presentation and immunity induction.

[0229] Thus, the present disclosure includes APC populations that can be pulsed or loaded with the neoantigenic peptides disclosed herein, APC populations that are genetically modified to express at least one neoantigenic peptide disclosed herein, or APC populations that contain expression constructs encoding tumor neoantigenic peptides disclosed herein, and methods of producing the same. Typically, the APC population is pulsed or loaded, modified to express or contain at least one, at least 5, at least 10, at least 15, or at least 20 different neoantigenic peptides or expression constructs encoding the same.

[0230] The present disclosure also includes compositions comprising the APCs disclosed herein. The APCs can be suspended in any known physiologically compatible pharmaceutical carrier, such as cell culture medium, physiological saline, phosphate buffered saline, cell culture medium, etc., to form a physiologically acceptable aqueous pharmaceutical composition. Parenteral vehicles include sodium chloride solution, Ringer's dextrose, dextrose and sodium chloride, lactated Ringer's solution. Other substances, such as antibacterial agents, may be added as needed. As used herein, "carrier" refers to any substance suitable as a vehicle for delivering APCs to a suitable in vitro or in vivo site of action. Therefore, the carrier can be used as an excipient for the preparation of therapeutic or experimental reagents containing APCs. Preferred carriers are capable of maintaining the APCs in a form capable of interacting with T cells. Examples of such carriers include, but are not limited to, water, phosphate buffered solutions, saline, Ringer's solutions, glucose solutions, serum-containing solutions, Hank's solutions, and other physiologically balanced aqueous solutions or cell culture media. The aqueous carrier may also contain suitable auxiliary substances required to approach the physiological conditions of the receptor, such as enhancing chemical stability and isotonicity. Suitable auxiliary substances include, for example, sodium acetate, sodium chloride, sodium lactate, potassium chloride, calcium chloride, sorbitan monolaurate, triethanolamine oleate and other substances which are used for the production of phosphate buffers, Tris buffers and bicarbonate buffers.

[0231] Vaccine composition

[0232] The present disclosure also includes a vaccine or immunogenic composition capable of eliciting a specific T cell response, comprising:

[0233] - one or more neoantigenic peptides as defined herein,

[0234] - one or more polynucleotides encoding a neoantigenic peptide as defined herein; and / or

[0235] - A population of antigen presenting cells as described above (eg autologous dendritic cells or artificial APCs).

[0236] A suitable vaccine or immunogenic composition will preferably contain 1 to 20 neoantigenic peptides, more preferably 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25 different neoantigenic peptides, further preferably 6, 7, 8, 9, 10, 11, 12, 13 or 14 different neoantigenic peptides, and most preferably 12, 13 or 14 different neoantigenic peptides.

[0237] The new antigenic peptide can be connected to a carrier protein. When the composition contains two or more new antigenic peptides, the two or more (e.g., 2 to 25) peptides can be linearly connected by a spacer molecule as described above, for example, a spacer comprising 2 to 6 non-polar or neutral amino acids.

[0238] In one embodiment of the present disclosure, different neoantigenic peptides, encoding polynucleotides, vectors or APCs are selected so that a vaccine or immunogenic composition comprises neoantigenic peptides that can associate with different MHC molecules (e.g., different MHC class I molecules). Preferably, such neoantigenic peptides are capable of associating with the most common MHC class I molecules, for example, different fragments that can associate with at least 2 preferred, more preferably at least 3 preferred, even more preferably at least 4 preferred MHC class I molecules. In some embodiments, the composition comprises peptides, encoding polynucleotides, vectors or APCs that can associate with one or more MHC class II molecules. MHC is optionally HLA-A, -B, -C, -DP, -DQ or -DR.

[0239] The vaccine or immunogenic composition is capable of eliciting a specific cytotoxic T cell response and / or a specific helper T cell response.

[0240] Therefore, in a specific embodiment, the present disclosure also relates to a new antigenic peptide as described above, wherein the new antigenic peptide has a tumor-specific new epitope and is contained in a vaccine or immunogenic composition. Vaccine compositions should be understood to refer to compositions used to generate immunity to prevent and / or treat diseases. Therefore, vaccines are drugs that contain or generate antigens and are intended to be used in humans or animals through vaccination to generate specific defense and protection substances. "Immunogenic composition" should be understood to refer to compositions that contain or generate antigens and are capable of eliciting antigen-specific humoral or cellular immune responses (e.g., T cell responses).

[0241] In a preferred embodiment, the length of the new antigenic peptide according to the present disclosure is 8 or 9 residues, or 13 to 25 residues. When the peptide is less than 20 residues, in order to obtain a peptide more suitable for in vivo immunization, the new antigenic peptide can be optionally flanked by additional amino acids to obtain an immune peptide with more amino acids, usually more than 20.

[0242] Pharmaceutical compositions (i.e., vaccines or immunogenic compositions) comprising the peptides described herein can be administered to individuals who already have cancer or tumors. In therapeutic applications, the composition is administered to the patient in an amount sufficient to elicit an effective CTL response to the tumor antigen and to cure or at least partially inhibit symptoms and / or complications. An amount sufficient to achieve this purpose is defined as a "therapeutically effective dose". The effective amount for this purpose will depend, for example, on the peptide composition, the mode of administration, the stage and severity of the disease being treated, the patient's weight and overall health, and the judgment of the prescribing physician, but for a 70 kg patient, the initial immunization (i.e., for therapeutic or prophylactic administration) typically ranges from about 1.0 μg to about 50,000 μg of peptide, and then, depending on the patient's response and condition, by measuring the specific CTL activity in the patient's blood, the dose is boosted or about 1.0 μg to about 10,000 μg of peptide according to a boosting regimen within a few weeks to months. It must be remembered that the peptides and compositions of the present invention can generally be used in serious disease states, i.e., life-threatening or potentially life-threatening situations, especially when the cancer has metastasized. In such cases, the treating physician may, and may deem it desirable to, administer a large excess of these peptide compositions in view of the minimization of foreign matter and the relatively nontoxic nature of the peptides.

[0243] For therapeutic use, administration should begin upon detection of a tumor or surgical removal of the tumor, with subsequent booster doses until symptoms are at least substantially alleviated and for a period thereafter.

[0244] The vaccine or immunogenic composition for treatment is intended for parenteral, topical, nasal, oral or topical administration. Preferably, the pharmaceutical composition is administered parenterally, for example, intravenously, subcutaneously, intradermally or intramuscularly. The composition can be administered at the site of surgical resection to induce a local immune response to the tumor.

[0245] The vaccine or immunogenic composition can be a pharmaceutical composition, which additionally comprises a pharmaceutically acceptable adjuvant, an immunostimulant, a stabilizer, a carrier, a diluent, an excipient and / or any other material well known to those skilled in the art. Such materials should be non-toxic and should not interfere with the effectiveness of the active ingredient. The carrier is preferably an aqueous carrier, but the exact nature of the carrier or other material will depend on the route of administration. Various aqueous carriers can be used, for example, water, buffered water, 0.9% saline, 0.3% glycine, hyaluronic acid, etc. These compositions can be sterilized by conventional, well-known sterilization techniques, or can be aseptically filtered. The resulting aqueous solution can be packaged and used as is, or it can be lyophilized, and the lyophilized preparation is mixed with a sterile solution before administration. The composition can also contain pharmaceutically acceptable auxiliary substances to approach physiological conditions, such as pH regulators and buffers, tension regulators, wetting agents, etc., such as sodium acetate, sodium lactate, sodium chloride, potassium chloride, calcium chloride, sorbitol monolaurate, triethanolamine oleate, etc., see, for example, Butterfield, BMJ. 2015;22;350 Discussion on cancer vaccines.

[0246] Examples of adjuvants that increase or expand the host's immune response to antigenic compounds include emulsifiers, muramyl dipeptides, avridine, aqueous adjuvants such as aluminum hydroxide, chitosan-based adjuvants, saponins, oils, Amphigen, LPS, bacterial cell wall extracts, bacterial DNA, CpG sequences, synthetic oligonucleotides, cytokines, and combinations thereof. Emulsifiers include potassium, sodium, and ammonium salts of lauric and oleic acid, calcium, magnesium, and aluminum salts of fatty acids, organic sulfonates such as sodium dodecyl sulfate, hexadecyl triethylammonium bromide, glycerides, polyoxyethylene glycol esters and ethers, sorbitan fatty acid esters and polyoxyethylene thereof, gum arabic, gelatin, lecithin, and / or cholesterol. Adjuvants comprising oil components include mineral oils, vegetable oils, or animal oils. Other adjuvants include Freund's complete adjuvant (FCA) or Freund's incomplete adjuvant (FIA). Cytokines that may be used as additional immunostimulants include interferon alpha, interleukin-2 (IL-2), and granulocyte-macrophage colony stimulating factor (GM-CSF), or a combination thereof.

[0247] The concentration of the peptides described herein in the vaccine or immunogenic formulation can vary widely, i.e., from less than about 0.1 wt %, typically about 2 wt % or at least about 2 wt % to as much as 20 wt % to 50 wt % or more, and will be selected primarily by fluid volume, viscosity, etc., depending on the selected mode of administration.

[0248] The peptides described herein can also be administered via liposomes, which target the peptides to specific cell tissues, such as lymphoid tissue. Liposomes can also be used to extend the half-life of peptides. Liposomes include emulsions, foams, micelles, insoluble monolayers, liquid crystals, phospholipid dispersions, sheets, etc. In these preparations, the peptide to be delivered is combined with molecules that bind to receptors ubiquitous in lymphocytes (e.g., monoclonal antibodies that bind CD45 antigens) as part of the liposome or is combined with other therapeutic or immunogenic compositions. Therefore, liposomes filled with the desired peptides of the present invention can be guided to lymphocyte sites, and then the liposomes deliver the selected therapeutic / immunogenic peptide composition there. The liposomes used in the present invention are formed by standard vesicle-forming lipids, which typically include neutral and negatively charged phospholipids and sterols, such as cholesterol. The selection of lipids typically takes into account, for example, liposome size, acid instability, and the stability of liposomes in the bloodstream. There are many methods for preparing liposomes, such as described in, for example, Szoka et al., Ann. Rev. Biophys. Bioeng. 9; 467 (1980), U.S. Pat. Nos. 4,235,871; 4,501,728; 4,837,028; and 5,019,369.

[0249] For targeting immune cells, the ligand to be incorporated into the liposomes may include, for example, antibodies or fragments thereof specific for cell surface determinants of the desired immune system cells. The liposomal suspension containing the peptide may be administered intravenously, topically, externally, etc., with dosages varying, inter alia, depending on the mode of administration, the peptide being delivered, and the stage of the disease being treated.

[0250] For solid compositions, conventional or nanoparticulate non-toxic solid carriers can be used, including, for example, pharmaceutical grade mannitol, lactose, starch, magnesium stearate, sodium saccharin, talc, cellulose, glucose, sucrose, magnesium carbonate, etc. For oral administration, a pharmaceutically acceptable non-toxic composition is formed by incorporating any commonly used excipients (such as those listed above) and generally 10%-95% of the active ingredient (i.e., one or more peptides of the present invention, more preferably at a concentration of 25%-75%).

[0251] For aerosol administration, the immunogenic peptide is preferably provided in finely divided form with a surfactant and a propellant. The typical percentage of the peptide is 0.01% by weight-20% by weight, preferably 1% by weight-10% by weight. Of course, the surfactant must be nontoxic and preferably soluble in the propellant. Representatives of such agents are esters or partial esters of fatty acids (e.g., caproic acid, caprylic acid, lauric acid, palmitic acid, stearic acid, linoleic acid, linolenic acid, cholesterol acid and oleic acid) containing 6 to 22 carbon atoms and aliphatic polyols or their cyclic anhydrides. Mixed esters can be used, such as mixed or natural glycerides. The surfactant can account for 0.1%-20% of the composition weight, preferably 0.25-5%. The remainder of the composition is usually a propellant. A carrier can also be included as required, such as the lecithin for intranasal delivery.

[0252] Cytotoxic T cells (CTLs) recognize antigens in the form of peptides bound to MHC molecules, rather than the complete foreign antigens themselves. The MHC molecules themselves are located on the cell surface of antigen presenting cells. Therefore, CTLs can only be activated when there is a trimeric complex of peptide antigens, MHC molecules and antigen presenting cells (APCs). Accordingly, if not only peptides are used to activate CTLs, but also APCs with corresponding MHC molecules are added, the immune response can be enhanced. Therefore, in some embodiments, the vaccine or immunogenic composition according to the present disclosure may alternatively or additionally contain at least one antigen presenting cell, preferably an APC population.

[0253] Thus, the vaccine or immunogenic composition can be delivered in the form of cells (e.g., antigen presenting cells), for example as dendritic cell vaccines. Antigen presenting cells (e.g., dendritic cells) can be pulsed or loaded with the neoantigenic peptides disclosed herein, can contain expression constructs encoding the neoantigenic peptides disclosed herein, or can be genetically modified (via DNA or RNA transfer) to express one, two or more neoantigenic peptides disclosed herein, for example at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 neoantigenic peptides.

[0254] Suitable vaccines or immunogenic compositions can also be in the form of DNA or RNA associated with the neoantigenic peptides described herein. For example, DNA or RNA encoding one or more neoantigenic peptides or proteins derived therefrom can be used as a vaccine, for example, by direct injection to a subject. For example, DNA or RNA encoding at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24 or 25 neoantigenic peptides or proteins derived therefrom.

[0255] Several methods are conveniently used to deliver nucleic acids to patients. For example, nucleic acids can be delivered directly as "naked DNA". This method is described, for example, in Wolff et al., Science 247: 1465-1468 (1990) and U.S. Patents 5,580,859 and 5,589,466. Nucleic acids can also be administered using ballistic delivery, such as described in U.S. Patent 5,204,253. Particles consisting only of DNA can be administered. Alternatively, the DNA can be attached to particles, such as gold particles.

[0256] Nucleic acids can also be delivered in combination with cationic compounds (e.g., cationic lipids). Lipid-mediated gene delivery methods are described in, for example, WO 96 / 18372; WO 93 / 24640; Mannino & Gould-Fogerite, BioTechniques 6 (7): 682-691 (1988); U.S. Pat. No. 5,279,833; WO 91 / 06309; and Felgner et al., Proceedings of the National Academy of Sciences (PNAS) 84: 7413-7414 (1987).

[0257] Delivery system can optionally include cell penetrating peptide, nanoparticle encapsulation, virus-like particle, liposome or its any combination.Cell penetrating peptide includes TAT ​​peptide, herpes simplex virus VP22, cell membrane-penetrating peptide (transportan), Antp.Liposome can be used as delivery system.Listeria vaccine or electroporation can also be used.

[0258] One or more new antigenic peptides can also be delivered via a bacterial or viral vector containing a DNA or RNA sequence encoding one or more new antigenic peptides. DNA or RNA can be delivered as a vector itself, or it can be delivered in an attenuated bacterial virus or a live attenuated virus (e.g., vaccinia or fowlpox). This method involves using vaccinia virus as a vector to express a nucleotide sequence encoding the peptide of the present invention. After introduction into an acutely or chronically infected host or an uninfected host, the recombinant vaccinia virus expresses an immunogenic peptide, thereby causing a host CTL response. For example, U.S. Pat. No. 4,722,84 describes vaccinia vectors and methods that can be used for immunization regimens. Another vector is BCG (bacillus Calmette-Guérin). BCG vectors are described in Stover et al. (Nature 351: 456-460 (1991)). Based on the description herein, various other vectors (e.g., Salmonella typhi vectors, etc.) that can be used for therapeutic administration or immunization of the peptides of the present invention will be apparent to those skilled in the art.

[0259] A suitable method for administering nucleic acids encoding peptides described herein involves the use of minigene constructs encoding multiple epitopes. In order to create a DNA sequence encoding a selected CTL epitope (minigene) for expression in human cells, the amino acid sequence of the epitope is reverse translated. The human codon usage table is used to guide the codon selection for each amino acid. These DNA sequences encoding epitopes are directly adjacent to form a continuous polypeptide sequence. In order to optimize expression and / or immunogenicity, additional elements can be incorporated into the minigene design. Examples of amino acid sequences that can be reverse translated and included in the minigene sequence include helper T lymphocytes, epitopes, leader (signal) sequences, and endoplasmic reticulum retention signals. In addition, the MHC presentation of CTL epitopes can be improved by adding synthetic (e.g., polyalanine) or naturally occurring flanking sequences adjacent to the CTL epitopes.

[0260] The minigene sequence is converted to DNA by assembling oligonucleotides encoding the plus and minus strands of the minigene. Overlapping oligonucleotides (30-100 bases long) are synthesized, phosphorylated, purified, and annealed under appropriate conditions using well-known techniques. The ends of the oligonucleotides are ligated using T4 DNA ligase. The synthetic minigene encoding the CTL epitope polypeptide can then be cloned into the desired expression vector.

[0261] The vector contains standard regulatory sequences well known to those skilled in the art to ensure expression in target cells. Therefore, the DNA or RNA encoding the new antigenic peptide can generally be operably linked to one or more of the following:

[0262] -Promoters that can be used to drive the expression of nucleic acid molecules. AAV ITR can be used as a promoter and is conducive to eliminating the need for additional promoter elements. For pan-expression, the following promoters can be used: CMV (particularly human cytomegalovirus immediate early promoter (hCMV-IE)), CAG, CBh, PGK, SV40, RSV, ferritin heavy chain or light chain, etc. For brain expression, the following promoters can be used: Synapsinl is used for all neurons, CaMKIIalpha is used for excitatory neurons, GAD67 or GAD65 or VGAT is used for GABA neurons, etc. Promoters for driving RNA synthesis can include: Pol III promoters, such as U6 or HI. The use of Pol II promoters and intron boxes can be used to express guide RNA (gRNA). Typically, the promoter includes a downstream cloning site for small gene insertion. For examples of suitable promoter sequences, please refer in particular to U.S. Patents 5,580,859 and 5,589,466.

[0263] - Transcriptional transactivators or other enhancer elements that can also increase transcriptional activity, for example, the regulatory R region from the 5' long terminal repeat (LTR) of human T-cell leukemia virus type 1 (HTLV-1) (which has been shown to induce higher cellular immune responses when combined with the CMV promoter).

[0264] - Translation optimization sequences, e.g. Kozak sequences flanking the AUG start codon (ACCAUGG) in mRNA, and codon optimization.

[0265] Additional vector modifications may be required to optimize minigene expression and immunogenicity. In some cases, introns are required for efficient gene expression, and one or more synthetic or naturally occurring introns can be incorporated into the transcribed region of the minigene. The addition of mRNA stabilizing sequences to increase minigene expression can also be considered. Recently, it has been proposed that immunostimulatory sequences (ISS or CpG) play a role in the immunogenicity of DNA vaccines. If these sequences are found to enhance immunogenicity, they can be added to the vector, outside the minigene coding sequence.

[0266] In some embodiments, bicistronic expression vectors can be used to produce a minigene-encoded epitope and add a second protein to enhance or reduce immunogenicity.

[0267] The DNA vaccines or immunogenic compositions described herein can be enhanced by co-delivering cytokines (e.g., IL-2, IL-12, IL-18, GM-CSF, and IFNγ) that promote cell-mediated immune responses. CXC chemokines (e.g., IL-8) and CC chemokines (e.g., macrophage inflammatory protein (MIP)-1α, MIP-3α, MIP-3β, and RANTES) can enhance the effectiveness of immune responses. The immunogenicity of DNA vaccines can also be enhanced by co-delivering plasmid-encoded cytokine inducing molecules (e.g., LeIF), costimulatory and adhesion molecules (e.g., B7-1 (CD80) and / or B7-2 (CD86)). Auxiliary (HTL) epitopes can be combined with intracellular targeting signals and expressed separately from CTL epitopes. In this way, HTL epitopes can be directed to cell compartments different from CTL epitopes. If necessary, this can help HTL epitopes to enter the MHC II class pathway more effectively, thereby improving CTL induction. In contrast to CTL induction, specific reduction of immune responses by co-expression of immunosuppressive molecules such as TGF-β may be beneficial in certain diseases.

[0268] Once the expression vector is selected, the minigene is cloned into the polylinker region downstream of the promoter. This plasmid is transformed into an appropriate strain of E. coli and DNA is prepared using standard techniques. Restriction mapping and DNA sequence analysis are used to confirm the orientation and DNA sequence of the minigene, as well as all other elements included in the vector. Bacterial cells carrying the correct plasmid can be stored as master and working cell banks.

[0269] Purified plasmid DNA can be prepared for injection using various formulations. The simplest of these is to reconstruct lyophilized DNA in sterile phosphate buffered saline (PBS). Various methods have been described, and new technologies may be available. As mentioned above, nucleic acids are conveniently formulated with cationic lipids. In addition, glycolipids, fusogenic liposomes, peptides and compounds (collectively referred to as protection, interaction, non-condensation (PINC)) can also be compounded with purified plasmid DNA to affect variables such as stability, intramuscular dispersibility or transportation to specific organs or cell types.

[0270] Vaccines or immunogenic compositions comprising peptides can be co-administered with vaccines or immunogenic compositions comprising polynucleotides encoding the peptides. For example, peptide vaccines and DNA vaccines can be administered alternately in a prime-boost regimen. For example, it is contemplated to prime with a peptide immunogenic composition and boost with a DNA immunogenic composition, just as it is contemplated to prime with a DNA immunogenic composition and boost with a peptide immunogenic composition.

[0271] The present disclosure also includes a method for preparing a vaccine composition, comprising the following steps:

[0272] a) optionally, identifying at least one neoantigenic peptide according to the methods described previously;

[0273] b) preparing the at least one neoantigenic peptide, at least one polypeptide encoding the neoantigenic peptide, or at least one vector comprising the polypeptide described herein; and

[0274] c) optionally adding a physiologically acceptable buffer, excipient and / or adjuvant, and preparing a vaccine with the at least one neoantigenic peptide, polypeptide or carrier.

[0275] Another aspect of the present disclosure is a method for preparing a DC vaccine, wherein the DC presents at least one neoantigenic peptide disclosed herein or expresses at least one expression construct encoding a tumor neoantigenic peptide disclosed herein.

[0276] Antibody TCR, CAR and their derivatives

[0277] The present disclosure also relates to an antibody or antigen-binding fragment thereof that specifically binds to the neoantigenic peptide defined herein.

[0278] In some embodiments, the neoantigenic peptide is associated with an MHC or HLA molecule.

[0279] Typically, the antibody or antigen-binding fragment thereof binds to a neoantigenic peptide as defined herein, alone or optionally in association with an MHC or HLA molecule. d The binding affinity is 10 -7 M or lower, 10 -8 M or lower, 10 -9 M or lower, 10 -10 M or lower or 10 -11 M or lower.

[0280] To promote the infiltration and recognition of tumor cells by lymphocytes T (LT), another strategy is to use antibodies that can recognize more than one antigen target and more specifically two antigen targets at the same time. Bispecific antibodies come in many forms. BiTE (bispecific T cell engager) was the first to be developed. These are fusion proteins composed of two scFvs (variable domain heavy chain VH and light chain VL) from two antibodies connected by a binding peptide: one recognizes the LT marker (CD3 + ), and the other recognizes a tumor antigen. The goal is to promote the recruitment and activation of LTs in contact with the tumor, leading to tumor cell lysis (reviewed by Patrick A. Baeuerle and Carsten Reinhardt; Bispecific T-Cell Engaging Antibodies for Cancer Therapy; Cancer Res 2009;69:(12), June 15, 2009; and Galaine et al., Innovations & Therapeutiques en Oncologie, Vol. 3-No. 3-7, May-August 2017).

[0281] In a specific embodiment, the antibody is a bispecific T cell engager that targets a tumor neoantigenic peptide as defined herein (optionally associated with an MHC or HLA molecule) and also targets at least one immune cell antigen. Typically, the immune cell is a T cell, NK cell, or dendritic cell. In this case, the targeted immune cell antigen may be, for example, CD3, CD16, CD30, or TCR.

[0282] The term "antibody" herein is used in the broadest sense, including polyclonal antibodies and monoclonal antibodies, including complete antibodies and functional (antigen-binding) antibody fragments, including fragment antigen-binding (Fab) fragments, F(ab')2 fragments, Fab' fragments, Fv fragments, recombinant IgG (rlgG) fragments, variable heavy chain (VH) regions capable of specific binding to antigens, single-chain antibody fragments, including single-chain variable fragments (scFv) and single-domain antibodies (e.g., VHH antibodies, sdAbs, sdFvs, nanobodies) fragments. The term includes genetically engineered and / or other variant modified forms of immunoglobulins, such as intracellular antibodies, peptide antibodies, chimeric antibodies, fully human antibodies, humanized antibodies and heterologous conjugate antibodies, multispecific antibodies, for example, bispecific antibodies, double antibodies, three antibodies and four antibodies, tandem two-scFv, tandem three-scFv. Unless otherwise indicated, the term "antibody" should be understood to include functional antibodies and fragments thereof. The term also includes intact antibodies or full-length antibodies, including antibodies of any class or subclass, including IgG and its subclasses, IgG1, IgG2, IgG3, IgG4, IgM, IgE, IgA and IgD. In some embodiments, the antibody comprises a light chain variable domain and a heavy chain variable domain, for example, in the form of scFv (single chain antibody fragment).

[0283] Antibodies include variant polypeptide species having one or more amino acid substitutions, insertions or deletions in the native amino acid sequence, provided that the antibody retains or substantially retains its specific binding function. Conservative substitutions of amino acids are well known and described above.

[0284] The present disclosure also includes a method for preparing an antibody or an antigen-binding fragment thereof, comprising the steps of selecting an antibody that binds to a tumor neoantigen peptide as defined herein, optionally associated with an MHC or HLA molecule, d The binding affinity is about 10 -6 M or lower, 10 -7 M or lower, 10 -8 M or lower, 10 -9 M or lower, 10 -10 M or lower or 10 -11 M or lower.

[0285] In some embodiments, the antibody is selected from a library of human antibody sequences. In some embodiments, the antibody is generated by immunizing an animal with a polypeptide comprising a neoantigenic peptide, optionally associated with an MHC or HLA molecule, followed by a selection step.

[0286] Antibodies (including chimeric antibodies, humanized antibodies or human antibodies) can be further affinity matured and selected as described above. Humanized antibodies contain CDR regions derived from rodent sequences; generally, rodent CDRs are implanted into human frameworks, and some of the human framework residues can be backmutated to the original rodent framework residues to retain affinity, and / or one or more CDR residues can be mutated to increase affinity. Fully human antibodies have no murine sequences and are usually produced via phage display technology of human antibody libraries or immunization of transgenic mice (whose natural immunoglobulin loci have been replaced by fragments of human immunoglobulin loci).

[0287] The present disclosure also includes antibodies produced by the methods and immune cells expressing such antibodies or fragments thereof.

[0288] The present disclosure also includes pharmaceutical compositions, which contain one or more antibodies disclosed herein alone, or a combination of one or more antibodies disclosed herein and at least one other agent (e.g., a stabilizing compound), which can be administered by any sterile, biocompatible pharmaceutical carrier, and optionally formulated with sterile pharmaceutically acceptable buffers, diluents, and / or excipients. Pharmaceutically acceptable carriers typically enhance or stabilize the composition, and / or can be used to promote the preparation of the composition. Pharmaceutically acceptable carriers include solvents, dispersion media, coatings, antibacterial and antifungal agents, isotonic agents, and absorption delay agents, etc., which are physiologically compatible and, in some embodiments, pharmaceutically inert.

[0289] Administration of pharmaceutical compositions comprising antibodies disclosed herein can be accomplished orally or parenterally. Methods of parenteral delivery include topical, intra-arterial (directly to the tumor), intramuscular, spinal, subcutaneous, intramedullary, intrathecal, intraventricular, intravenous, intraperitoneal or intranasal administration.

[0290] Therefore, in addition to the active ingredients, these pharmaceutical compositions may contain suitable pharmaceutically acceptable carriers, including excipients and adjuvants, which facilitate processing of the active compounds into pharmaceutically usable preparations. For more details on formulation and administration techniques, see the latest edition of Remington's Pharmaceutical Sciences (Editor, Mark Publishing Co, Easton, Pa.).

[0291] Depending on the route of administration, the active compound (ie, antibody, bispecific molecule and multispecific molecule) may be coated in a material to protect the compound from acids and other natural conditions that may inactivate the compound.

[0292] Compositions are normally sterile and preferably fluid. For example, by using a coating (e.g., lecithin), maintaining the required particle size in the case of dispersions and using a surfactant, suitable fluidity can be maintained. In many cases, it is preferred to add isotonic agents, such as sugar, polyols (e.g., mannitol or sorbitol) and sodium chloride in the composition. The long-term absorption of injectable compositions can be achieved by adding a delayed absorption agent (e.g., aluminum monostearate or gelatin) in the composition.

[0293] Pharmaceutical compositions for oral administration can be formulated with a dosage suitable for oral administration using pharmaceutically acceptable carriers well known in the art. Such carriers enable the pharmaceutical composition to be formulated as tablets, pills, dragees, capsules, liquids, gels, syrups, slurries, suspensions, etc. for patient intake.

[0294] The pharmaceutical compositions of the present disclosure can be prepared according to methods well known and routinely practiced in the art. See, for example, Remington: The Science and Practice of Pharmacy, Mark Publishing Company, 20th edition, 2000; and Sustained and Controlled Release Drug Delivery Systems, J R. Robinson, ed., Marcel Dekker, Inc., New York, 1978. The pharmaceutical compositions are preferably manufactured under GMP conditions.

[0295] The present disclosure also includes a T cell receptor (TCR) targeting a neoantigenic peptide as defined herein associated with an MHC or HLA molecule.

[0296] The present disclosure also includes a method for preparing a TCR or an antigen-binding fragment thereof, comprising the step of selecting a TCR that binds to a tumor neoantigen peptide as defined herein, optionally associated with an MHC or HLA molecule, optionally K d The binding affinity is about 10 -6 M or lower, 10 -7 M or lower, 10 -8 M or lower, 10 -9 M or lower, 10 -10 M or lower or 10 -11 M or lower.

[0297] Nucleic acids encoding TCRs can be obtained from a variety of sources, such as by polymerase chain reaction (PCR) amplification of naturally occurring TCR DNA sequences, then expressing antibody variable regions, and then performing the above-mentioned selection steps. In some embodiments, TCRs are obtained from T cells isolated from patients or cultured T cell hybridomas. In some embodiments, TCR clones for target antigens have been generated in transgenic mice engineered with human immune system genes (e.g., human leukocyte antigen system or HLA). See, e.g., tumor antigens (see, e.g., Parkhurst et al. (2009) Clin Cancer Res. 15:169-180 and Cohen et al. (2005) J Immunol. 175:5799-5808. In some embodiments, phage display is used to isolate TCRs against target antigens (see, e.g., Varela-Rohena et al. (2008) Nat Med. 14:1390-1395 and Li (2005) Nat Biotechnol. 23:349-354).

[0298] "T cell receptor" or "TCR" refers to a molecule containing variable α and β chains (also referred to as TCRRa and TCRp, respectively) or variable γ and δ chains (also referred to as TCRγ and TCRδ, respectively) and capable of specifically binding to an antigenic peptide bound to an MHC receptor. In some embodiments, the TCR is in the αβ form. Typically, TCRs in the αβ and γδ forms are generally similar in structure, but the T cells expressing them may have different anatomical locations or functions. The TCR may be present on the cell surface or in a soluble form. Typically, the TCR is located on the surface of a T cell (or T lymphocyte) and is generally responsible for recognizing antigens bound to a major histocompatibility complex (MHC) molecule. In some embodiments, the TCR may also include a constant domain, a transmembrane domain, and / or a short cytoplasmic tail region (see, e.g., Janeway et al., Immunobiology: The Immune System in Health and Disease, 3rd edition, Current Biology Publications, 4:33, 1997). For example, in some aspects, each chain of TCR can have an N-terminal immunoglobulin variable domain, an immunoglobulin constant domain, a transmembrane region, and a short cytoplasmic tail region at the C-terminus. In some embodiments, TCR associates with the invariant protein of the CD3 complex involved in mediating signal transduction. Unless otherwise indicated, the term "TCR" should be understood to include its functional TCR fragment. The term also includes complete or full-length TCR, including TCRs in αβ form or γδ form.

[0299] Therefore, for the purposes of this article, reference to a TCR includes any TCR or functional fragment, such as an antigen binding portion of a TCR, which binds to a specific antigen peptide bound in an MHC molecule, i.e., an MHC-peptide complex. "Antigen binding portion" or "antigen binding fragment" of a TCR is used interchangeably and refers to a molecule that contains a portion of a TCR domain but binds to an antigen bound to a complete TCR (e.g., an MHC-peptide complex). In some cases, the antigen binding portion contains a variable domain of a TCR, such as a variable α chain and a variable β chain of a TCR, sufficient to form a binding site that binds to a specific MHC-peptide complex, such as a position where each chain typically contains three complementary determining regions.

[0300] In some embodiments, the variable domains of the TCR chains associate to form loops, or complementarity determining regions (CDRs) similar to immunoglobulins, which confer antigen recognition and determine peptide specificity by forming the binding site of the TCR molecule. Typically, like immunoglobulins, CDRs are separated by framework regions (FRs) (see, e.g., Jores et al., Proceedings of the National Academy of Sciences of the United States of America (Pwc. Nat'l Acad. Sci. USA) 87:9138, 1990; Chothia et al., Journal of the European Molecular Biology Association (EMBO J.) 7:3745, 1988; see also Lefranc et al., Development and Comparative Immunology (Dev. Comp. Immunol.) 27:55, 2003). In some embodiments, CDR3 is the primary CDR responsible for recognizing processed antigens, although CDR1 of the α chain has also been shown to interact with the N-terminal portion of the antigenic peptide, while CDR1 of the β chain interacts with the C-terminal portion of the peptide. CDR2 is thought to recognize MHC molecules. In some embodiments, the variable region of the beta chain may comprise an additional hypervariable (HV4) region.

[0301] In some embodiments, the TCR chain comprises a constant domain. For example, as with immunoglobulins, the extracellular portion of the TCR chain (e.g., α chain, β chain) can contain two immunoglobulin domains, a variable domain at the N-terminus (e.g., Va or Vp; typically amino acids 1 to 116 based on Kabat numbering, Kabat et al., "Sequences of Proteins of Immunological Interest, US Dept. Health and Human Services, Public Health Service National Institutes of Health, 1991, 5th edition). ed.)), and a constant domain adjacent to the cell membrane (e.g., an α-chain constant region or Ca, typically amino acids 117 to 259 based on Kabat, and a β-chain constant region or Cp, typically amino acids 117 to 295 based on Kabat). For example, in some cases, the extracellular portion of the TCR formed by the two chains comprises two membrane-proximal constant domains, and two membrane-distal variable domains containing CDRs. The constant domain of the TCR domain contains a short linker sequence in which cysteine ​​residues form a disulfide bond, thereby forming a connection between the two chains. In some embodiments, the TCR can have additional cysteine ​​residues in each of the α and β chains so that the TCR contains two disulfide bonds in the constant domain.

[0302] In some embodiments, the TCR chain may contain a transmembrane domain. In some embodiments, the transmembrane domain is positively charged. In some cases, the TCR chain contains a cytoplasmic tail region. In some cases, this structure allows the TCR to associate with other molecules (e.g., CD3). For example, a TCR containing a constant domain with a transmembrane region can anchor the protein to the cell membrane and associate with the invariant subunit of the CD3 signaling device or complex.

[0303] Typically, CD3 is a multiprotein complex that can have three different chains (γ, δ, and ε) and a ζ chain in mammals. For example, in mammals, the complex can contain a homodimer of a CD3y chain, a CD35 chain, two CD3s chains, and a CD3ζ chain. CD3y, CD35, and CD3s chains are highly related cell surface proteins in the immunoglobulin superfamily that contain a single immunoglobulin domain. The transmembrane regions of CD3y, CD35, and CD3s chains are negatively charged, a property that allows these chains to associate with positively charged T cell receptor chains. The intracellular tails of CD3y, CD35, and CD3s chains all contain a single conserved motif, called an immunoreceptor tyrosine-based activation motif or ITAM, while each CD3ζ chain has three conserved motifs. Typically, ITAMs are involved in the signal transduction ability of the TCR complex. These auxiliary molecules have negatively charged transmembrane regions and play a role in propagating signals from the TCR into cells. The CD3 chain and the ζ chain together with the TCR form the so-called T cell receptor complex.

[0304] In some embodiments, the TCR can be a heterodimer of two chains α and β (or alternatively γ and δ), or can be a single-chain TCR construct. In some embodiments, the TCR is a heterodimer containing two separate chains (α and β chains or γ and δ chains) connected by, for example, one or more disulfide bonds.

[0305] While T cell receptors (TCRs) are transmembrane proteins and do not naturally exist in a soluble form, antibodies can be both secreted and membrane-bound. Importantly, the advantage of TCRs over antibodies is that, when presented in the context of MHC molecules, they can in principle recognize peptides generated from all degraded cellular proteins (intracellular and extracellular). Therefore, TCRs have significant therapeutic potential.

[0306] The present disclosure also relates to soluble T cell receptors (sTCRs) containing antigen recognition portions for the tumor neoantigenic peptides disclosed herein (see Walseng E, S, Fallang LE, Yang W, Vefferstad A, Areffard A, et al. (2015). Soluble T-Cell Receptors Produced in Human Cells for Targeted Delivery. PLoS One 10(4): e0119559. In a particular embodiment, the soluble TCR can be fused to an antibody fragment against a T cell antigen, optionally wherein the targeted antigen is CD3 or CD16 (see, e.g., Boudousquie, Caroline et al., "Polyfunctional response by ImmTAC (IMCgp100) redirected CD8+ and CD4+ T cells." Immunology, Vol. 152, 3 (2017): 425-438, doi: 10.1111 / imm.12779).

[0307] The present disclosure also includes a chimeric antigen receptor (CAR) for a tumor neoantigen peptide disclosed herein. CAR is a fusion protein comprising an antigen binding domain, usually derived from an antibody, connected to a signaling domain of a TCR complex. CAR can be used to guide immune cells (eg, T cells or NK cells) against previously defined tumor neoantigen peptides and select a suitable antigen binding domain.

[0308] The antigen binding domain of CAR is usually based on scFv (single-chain variable fragment) derived from antibodies. In addition to the extracellular antibody binding domain at the N-terminus, CAR can also generally include a hinge domain (which plays the role of a spacer, extending the antigen binding domain away from the plasma membrane of the immune effector cell on which it is expressed), a transmembrane (TM) domain, an intracellular signaling domain (e.g., a signaling domain of the ζ chain (CD3ζ) of the CD3 molecule from the TCR complex, or an equivalent) and optionally one or more costimulatory domains, which can assist in the signaling or functionality of cells expressing CAR. Signaling domains from costimulatory molecules, including CD28, OX-40 (CD134), ICOS-1, CD27, GITR, CD28, DAP10, and 4-1BB (CD137), can be added alone (second generation) or in combination (third generation), to improve the survival rate of CAR-modified T cells and increase proliferation.

[0309] Therefore, CAR may include:

[0310] (1) In its extracellular portion, one or more antigen-binding molecules, such as one or more antigen-binding fragments, domains or portions of an antibody, or one or more antibody variable domains (heavy and / or light chains), and / or antibody molecules.

[0311] (2) In its transmembrane portion, a transmembrane domain derived from human T cell receptor-α or -β chain, CD3ζ chain, CD28, CD3-ε, CD45, CD4, CD5, CD8, CD9, CD16, CD22, CD33, CD37, CD64, CD80, CD86, CD134, CD137, ICOS, CD154 or GITR. In some embodiments, the transmembrane domain is derived from CD28, CD8 or CD3-ζ.

[0312] (3) one or more costimulatory domains, such as costimulatory domains derived from human CD28, 4-1BB (CD137), ICOS-1, CD27, OX40 (CD137), DAP10 and GITR (AITR). In some embodiments, CAR comprises costimulatory domains of CD28 and 4-1BB.

[0313] (4) in its intracellular signaling domain, one or more intracellular signaling domains comprising one or more ITAMs, for example: an intracellular signaling domain or a portion thereof from CD3-ζ, or a variant thereof lacking one or two ITAMs (e.g., ITAM3 and / or ITAM2), FcRγ, FcRβ, CD3γ, CD3δ, CD3ε, CDS, CD22, CD79a, CD79b and / or CD66d, particularly selected from the intracellular domain of CD3-ζ, or a variant thereof lacking one or two ITAMs (e.g., ITAM3 and / or ITAM2), or an intracellular signaling domain of FcεRIγ or a variant thereof.

[0314] CARs can be designed to recognize tumor neoantigen peptides alone or in association with HLA or MHC molecules.

[0315] Exemplary antigen receptors, including CARs and recombinant TCRs, and methods for engineering receptors and introducing them into cells include, for example, International Patent Application Publication Nos. WO2000 / 14257, WO2013 / 126726, WO2012 / 129514, WO2014 / 031687, WO2013 / 166321, WO2013 / 071154, WO2013 / 123061, U.S. Patent Application Publication Nos. US2002131960, US2013287748, US201301 49337, U.S. Pat. Nos. 6,451,995, 7,446,190, 8,252,592, 8,339,645, 8,398,282, 7,446,179, 6,410,319, 7,070,995, 7,265,209, 7,354,762, 7,446,191, 8,324,353, and 8,479,118, and European Patent Application No. EP2537416, and / or those described by Sadelain et al., Cancer Discovery Discov. April 2013; 3(4): 388-398; Davila et al. (2013) PLoS One 8(4): e61338; Turtle et al. Curr. Opin. Immunol. October 2012; 24(5): 633-39; Wu et al. Cancer March 2012, 18(2): 160-75. In some aspects, the genetically engineered antigen receptors include CARs as described in U.S. Pat. No. 7,446,190, and those described in International Patent Application Publication No. WO2014 / 055668.

[0316] The present disclosure also includes polynucleotides encoding the previously described antibodies, antigen-binding fragments or derivatives thereof, TCRs and CARs, and vectors comprising the polynucleotides.

[0317] Immune cells

[0318] The present disclosure also includes immune cells that target one or more of the previously described tumor neoantigenic peptides.

[0319] As used herein, the term "immune cell" includes cells of hematopoietic origin and that play a role in immune response. Immune cells include lymphocytes, such as B cells and T cells, natural killer cells, myeloid cells, such as monocytes, macrophages, eosinophils, mast cells, basophils and granulocytes.

[0320] As used herein, the term "T cell" includes cells carrying a T cell receptor (TCR), in particular a TCR for a tumor neoantigen peptide disclosed herein. T cells according to the present disclosure may be selected from the group consisting of: inflammatory T lymphocytes, cytotoxic T lymphocytes, regulatory T lymphocytes, mucosal-associated constant T cells (MAIT), Yδ T cells, tumor infiltrating lymphocytes (TIL) or helper T lymphocytes including type 1 and type 2 helper T cells and Th17 helper cells. In another embodiment, the cell may be derived from CD4 + T lymphocytes and CD8 + The immune cells may be derived from a healthy donor or from a subject suffering from cancer or a tumor.

[0321] Immune cells can be extracted from blood or derived from stem cells. Stem cells can be adult stem cells, embryonic stem cells, more particularly non-human stem cells, umbilical cord blood stem cells, progenitor cells, bone marrow stem cells, induced pluripotent stem cells, totipotent stem cells or hematopoietic stem cells. Representative human cells are CD34 + cell.

[0322] T cells can be obtained from a number of non-limiting sources, including peripheral blood mononuclear cells, 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, any number of techniques known to those skilled in the art (e.g., FICOLL TM T cells are obtained from a unit of blood collected from a subject by apheresis (separation). In one embodiment, cells from the circulating blood of a subject are obtained by apheresis. In certain embodiments, T cells are isolated from PBMCs. PBMCs can be isolated from whole blood by density gradient centrifugation (e.g., by LYMPHOPREP TM Gradient, PERCOLL TM GRADIENT OR FICOLL TM T cells can be isolated from the buffy coat obtained by centrifugation of a gradient. T cells can be isolated by depletion of monocytes, for example by using CD14 Isolation from PBMCs. In some embodiments, red blood cells may be lysed prior to density gradient centrifugation.

[0323] In another embodiment, the cells may be derived from a healthy donor, from a subject diagnosed with cancer or a tumor, in particular with a glioblastoma. The cells may be autologous or allogeneic.

[0324] In allogeneic immune cell therapy, immune cells are collected from healthy donors rather than from the patient. Typically these are HLA-matched to reduce the likelihood of graft-versus-host disease. Alternatively, generic “off-the-shelf” products that may not require HLA matching include modifications designed to reduce graft-versus-host disease, such as disruption or removal of the TCR α β receptor. See review by Graham et al., Cell 2018 Oct;7(10):155. Since a single gene encodes the α chain (TRAC), rather than two genes encoding the β chains, the TRAC locus is a typical target for removal or disruption of TCR α β receptor expression. Alternatively, inhibitors of TCR α β signaling can be expressed, such as a truncated form of CD3ζ that can be used as a TCR inhibitory molecule. Disruption or removal of HLA class I molecules has also been employed. For example, Torikai et al., Blood 2013; 122: 1341-1349 used ZFN to knock out the HLA-A locus, while Ren et al., Clin Cancer Res. 2017; 23: 2255-2266 knocked out beta-2 microglobulin (B2M) required for HLA class I expression. Ren et al. simultaneously knocked out TCRαβ, B2M, and the immune checkpoint PD1. Typically, immune cells are activated and expanded for adoptive cell therapy. The immune cells disclosed herein can be expanded in vivo or ex vivo. Immune cells, particularly T cells, can generally be activated and expanded using methods known in the art. Typically, T cells are expanded by contact with a surface to which agents that stimulate signals associated with the CD3 / TCR complex and ligands that stimulate co-stimulatory molecules on the surface of T cells are attached.

[0325] In one embodiment of the present disclosure, immune cells can be modified to target the tumor neoantigen peptides previously described. In a specific embodiment, the immune cells can express recombinant antigen receptors for the neoantigen peptides on their cell surface. "Recombinant" refers to an antigen receptor that is not encoded by the cell in its native state, that is, it is heterologous and non-endogenous. Therefore, it can be seen that the expression of recombinant antigen receptors introduces new antigen specificity to immune cells, allowing the cells to recognize and bind to the peptides previously described. Antigen receptors can be isolated from any useful source. In some embodiments, the cell contains one or more nucleic acids introduced via genetic engineering, which encode one or more antigen receptors, wherein the antigen includes at least one tumor neoantigen peptide according to the present disclosure.

[0326] Antigen receptors according to the present disclosure include genetically engineered T cell receptors (TCRs) and components thereof, as well as functional non-TCR antigen receptors, such as the chimeric antigen receptors (CARs) described previously.

[0327] Methods by which immune cells can be genetically modified to express recombinant antigen receptors are well known in the art. Nucleic acid molecules encoding antigen receptors can be introduced into cells in the form of, for example, vectors or any other suitable nucleic acid constructs. Vectors and components required therefor are well known in the art. Nucleic acid molecules encoding antigen receptors can be generated using any method known in the art, for example, molecular cloning using PCR. Antigen receptor sequences can be modified using conventional methods (e.g., site-directed mutagenesis).

[0328] The present disclosure also relates to a method for providing a T cell population targeting the tumor neoantigen peptides disclosed herein.

[0329] T cell populations can include CD8 + T cells, CD4 + T cells or CD8 + and CD4 + T cells.

[0330] The T cell population prepared according to the present disclosure can be enriched for T cells that are specific (i.e., targeted) to the tumor neoantigen peptides disclosed herein. That is, the T cell population prepared according to the present disclosure will have an increased number of T cells that target one or more tumor neoantigen peptides. For example, compared to the T cells in the sample isolated from the subject, the T cell population of the present disclosure will have an increased number of T cells that target tumor neoantigen peptides. That is, the composition of the T cell population will be different from the "natural" T cell population (i.e., a population that has not undergone the recognition and amplification steps discussed herein) because the percentage or proportion of T cells targeting tumor neoantigen peptides will increase.

[0331] The T cell population prepared according to the present disclosure can be enriched for T cells that are specific (i.e., targeted) to tumor neoantigen peptides. That is, the T cell population prepared according to the present disclosure will have an increased number of T cells that target one or more tumor neoantigen peptides of the present disclosure. For example, compared to the T cells in the sample isolated from the subject, the T cell population of the present disclosure will have an increased number of T cells that target tumor neoantigen peptides. That is, the composition of the T cell population will be different from the "natural" T cell population (i.e., a population that has not undergone the recognition and amplification steps discussed herein) because the percentage or proportion of T cells targeting tumor neoantigen peptides will increase.

[0332] According to the present disclosure, the T cell population can have at least about 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95 or 100% of T cells targeting the tumor neoantigen peptides disclosed herein. For example, the T cell population can have about 0.2%-5%, 5%-10%, 10-20%, 20-30%, 30-40%, 40-50%, 50-70% or 70-100% of T cells targeting the tumor neoantigen peptides disclosed herein.

[0333] The expanded tumor neoantigen peptide reactive T cell population can have higher activity than the unamplified T cell population, for example, using tumor neoantigen peptides. Reference to "activity" can refer to the response of a T cell population to restimulation of a tumor neoantigen peptide (e.g., a peptide corresponding to a peptide used for amplification, or a mixture of tumor neoantigen polypeptides). Suitable methods for determining the response are known in the art. For example, the production of cytokines (e.g., the production of IL2 or IFNγ can be measured). Reference to "higher activity" includes, for example, an increase in activity of 1-5 times, 5-10 times, 10-20 times, 20-50 times, 50-100 times, 100-500 times, 500-1000 times. In one aspect, the activity can be more than 1000 times higher.

[0334] In some embodiments, the present disclosure provides a plurality of T cells or T cell populations, wherein the plurality of T cells or T cell populations include at least a T cell that recognizes a clonal tumor neoantigen peptide and at least another T cell that recognizes a different clonal tumor neoantigen peptide. Therefore, the present disclosure provides a plurality of T cells that recognize different clonal tumor neoantigen peptides. A plurality of different T cells or different T cell populations may alternatively have different TCRs that recognize the same tumor neoantigen peptide.

[0335] In some embodiments, the number of clonal tumor neoantigen peptides recognized by the plurality of T cells is from 2 to 1000. For example, the number of recognized clonal neoantigens can be 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 50, 100, 150, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 950 or 1000, preferably 2 to 100. There can be a plurality of T cells with different TCRs but recognizing the same clonal neoantigen.

[0336] The T cell population can be composed entirely or predominantly of CD8 + T cells, either entirely or predominantly CD8+ T cells and CD4 + A mixture of T cells, either entirely or predominantly CD4 + T cell composition.

[0337] In a specific embodiment, the T cell population is generated from T cells isolated from a subject with a tumor. For example, the T cell population can be generated from T cells in a sample isolated from a subject with a tumor. The sample can be a tumor sample, a peripheral blood sample, or a sample from other tissues of the subject.

[0338] In a specific embodiment, a T cell population is generated from a tumor sample in which a tumor neoantigen peptide is identified. In other words, a T cell population is isolated from a sample derived from a tumor of a patient to be treated. Such T cells are referred to herein as "tumour infiltrating lymphocytes" (TIL).

[0339] T cells can be isolated using methods well known in the art. For example, based on CD3 + 、CD4 + or CD8 + Expression of T cells. T cells can be purified from a single cell suspension generated from a sample and can be enriched from the sample by Ficoll plaque gradient.

[0340] Cancer treatment and diagnosis methods

[0341] In any embodiment, the cancer therapeutic products described herein can be used in methods for inhibiting cancer cell proliferation. The cancer therapeutic products described herein can also be used to treat the previously listed cancers or tumors, or for the prophylactic treatment of such cancers in patients at risk of such cancers or tumors.

[0342] Cancers that can be treated using the therapies described herein include any solid or non-solid tumor. In a specific embodiment of the present disclosure, the tumor is a glioblastoma.

[0343] Cancer also includes cancers that are refractory to treatment with other chemotherapeutic drugs. The term "refractory" as used herein refers to cancer (and / or its metastasis) that does not show an antiproliferative response or only shows a weaker antiproliferative response (e.g., no or only weaker tumor growth inhibition) after treatment with another chemotherapeutic agent. These are cancers that cannot be satisfactorily treated with other chemotherapy. Refractory cancers include not only (i) cancers in which one or more chemotherapeutic drugs have failed during the patient's treatment, but also (ii) cancers that can be proven to be refractory by other means (e.g., biopsy and culture in the presence of chemotherapeutic drugs).

[0344] The therapies described herein are also suitable for the treatment of patients in need thereof who have not previously received treatment.

[0345] The subject according to the present disclosure is typically a patient in need who has been diagnosed with a tumor. The subject is typically a mammal, particularly a human, dog, cat, horse or any animal in need of a tumor-specific immune response.

[0346] The present disclosure also relates to a neoantigenic peptide, an APC population, a vaccine or immunogenic composition, a polynucleotide encoding a neoantigenic peptide or a vector as previously defined for use in cancer vaccination therapy in a subject or for treating cancer in a subject, wherein the peptide binds to at least one MHC molecule of the subject.

[0347] The present disclosure also provides a method for treating cancer in a subject, comprising administering a vaccine or immunogenic composition as described herein to the subject in a therapeutically effective amount to treat the subject. The method may also include the step of identifying a subject with cancer or a tumor, particularly a glioblastoma.

[0348] The present disclosure also relates to a method of treating cancer, typically glioblastoma, comprising preparing an antibody or an antigen-binding fragment thereof by the method described herein, and administering the antibody or antigen-binding fragment thereof in a therapeutically effective amount to a subject having cancer or a tumor, or administering an immune cell expressing the antibody or antigen-binding fragment thereof to treat the subject.

[0349] The present disclosure also relates to antibodies (including variants and derivatives thereof), T cell receptors (TCRs) (including variants and derivatives thereof) or CARs (including variants and derivatives thereof) against the tumor neoantigen peptides described herein, optionally associated with MHC or HLA molecules, for use in cancer therapy, particularly glioblastoma therapy, in a subject, wherein the tumor neoantigen peptide binds to at least one MHC molecule of the subject.

[0350] The present disclosure also relates to antibodies (including variants and derivatives thereof), T cell receptors (TCRs) (including variants and derivatives thereof) or CARs (including variants and derivatives thereof) directed against the tumor neoantigen peptides described herein, optionally associated with MHC or HLA molecules, or immune cells targeting neoantigen peptides as previously defined, for adoptive cell therapy or CAR-T cell therapy of a subject, wherein the tumor neoantigen peptides bind to at least one MHC molecule of the subject.

[0351] Generally, one skilled in the art is able to select an appropriate antigen receptor that binds and recognizes a previously defined tumor neoantigen peptide, thereby redirecting immune cells for cancer cell therapy, in particular glioblastoma cell therapy. In a specific embodiment, the immune cells used in the disclosed method are redirected T cells, for example, redirected CD8+ and / or CD4 + T cells.

[0352] The inventors here provide a method for identifying or screening population-specific TE tags, particularly tumor cell-specific TE tags. This discovery has great potential in diagnosis. In fact, it provides a tumor-specific biomarker shared by patients, which can distinguish tumor cells from other cell populations from core tumors and / or tumor microenvironments and tumor cells from different types of tumors.

[0353] Therefore, the present disclosure also includes a method for diagnosing a tumor (eg, glioblastoma). The method comprises identifying the tumor cell-specific TE signature defined herein in a tumor sample obtained from a patient according to the method disclosed herein.

[0354] The present application also comprises a method for treating a patient suffering from a tumor, in particular a tumor associated with derepressed TE, in particular a glioblastoma, comprising the steps of diagnosing the tumor according to the method defined above and administering a treatment specifically directed against the identified tumor.

[0355] In some embodiments, the present application relates to a method for treating a patient suffering from a tumor, in particular a tumor associated with derepressed TE, in particular a glioblastoma, comprising (i) a step of diagnosing the tumor according to the method defined above and (ii) a step of administering any one or a combination of the cancer treatment products described herein.

[0356] In some embodiments, the cancer treatment, vaccination therapy and / or adoptive cell cancer therapy as described above are co-administered with additional cancer therapies. In some embodiments, the cancer treatment, vaccination therapy and / or adoptive cell cancer therapy as described above are co-administered with targeted therapy, immunotherapy (e.g., immune checkpoint therapy and immune checkpoint inhibitors), co-stimulatory antibodies, chemotherapy and / or radiotherapy.

[0357] Immune checkpoint therapies, such as checkpoint inhibitors, include but are not limited to programmed death receptor-1 (PD-1) inhibitors, programmed death ligand-1 (PD-L1) inhibitors, programmed death ligand-2 (PD-L2) inhibitors, lymphocyte activation gene 3 (LAG3) inhibitors, T cell immunoglobulin and mucin domain-containing protein 3 (TIM-3) inhibitors, T cell immunoreceptor with Ig and ITIM domains (TIGIT) inhibitors, B and T lymphocyte attenuator (BTLA) inhibitors, V-domain Ig inhibitory factor of T cell activation (VISTA) inhibitors, cytotoxic T lymphocyte-associated protein 4 (CTLA4) inhibitors, indoleamine 2,3-dioxygenase (IDO) inhibitors, killer cell immunoglobulin-like receptor (KIR) inhibitors, KIR2L3 inhibitors, KIR3DL2 inhibitors and carcinoembryonic antigen-related cell adhesion molecule 1 (CEACAM-1) inhibitors. In particular, checkpoint inhibitors include anti-PD1, anti-PD-L1, anti-CTLA-4, anti-TIM-3, anti-LAG3 antibodies. Co-stimulatory antibodies deliver positive signals through immunomodulatory receptors, including but not limited to ICOS, CD137, CD27, OX-40 and GITR.

[0358] Examples of anti-PD1 antibodies include, but are not limited to, nivolumab, cemiplimab (REGN2810 or REGN-2810), tislelizumab (BGB-A317), tislelizumab, spartalizumab (PDR001 or PDR-001), ABBV-181, JNJ-63723283, BI 754091, MAG012, TSR-042, AGEN2034, pidilizumab, nivolumab (ONO-4538, BMS-936558, MDX1106, GTPL7335 or Opdivo), pembrolizumab (MK-3475, MK03475, lambrolizumab, SCH-900475 or Keytruda) and antibodies described in international patent applications WO2004004771, WO2004056875, WO2006121168, WO2008156712, WO2009014708, WO2009114335, WO2013043569 and WO2014047350.

[0359] Examples of anti-PD-L1 antibodies include, but are not limited to, LY3300054, atezolizumab, durvalumab, and avelumab.

[0360] Examples of anti-CTLA-4 antibodies include, but are not limited to, ipilimumab (see, e.g., U.S. Pat. Nos. 6,984,720 and 8,017,114), tremelimumab (see, e.g., U.S. Pat. Nos. 7,109,003 and 8,143,379), single-chain anti-CTLA4 antibodies (see, e.g., International Patent Applications WO1997020574 and WO2007123737), and antibodies described in U.S. Pat. No. 8,491,895.

[0361] Examples of anti-VISTA antibodies are described in U.S. patent application US20130177557.

[0362] Examples of LAG3 receptor inhibitors are described in U.S. Pat. No. 5,773,578.

[0363] An example of a KIR inhibitor is IPH4102, which targets KIR3DL2.

[0364] As used herein, the term "chemotherapy" has its general meaning in the art and refers to the treatment of administering a chemotherapeutic agent to a patient. The chemotherapeutic entity used herein refers to an entity that is destructive to cells, i.e., the entity reduces the viability of cells. The chemotherapeutic entity can be a cytotoxic drug. Chemotherapeutic agents include, but are not limited to, alkylating agents such as thiotepa and cyclophosphamide; alkyl sulfonates such as busulfan, improsulfan, and piposulfan; aziridines such as benzodopa, carboquinone, meturedopa, and uredopa; ethyleneimine and methylmelamine, including hexamethylmelamine, triethylenemelamine, triethylenephosphoramide, triethylenethiophosphoramide, and trihydroxymethylmelamine; acetylspermine (acetogenins) (particularly bullatacin and bullatacinone); camptothecins (including the synthetic analog topotecan); bryostatin; callystatin; CC-1065 (including its synthetic analogs adozelesin, carzelesin, and bizelesin); cryptophycins (particularly cryptophycin 1 and cryptophycin 8); dolastatin; duocarmycin (including synthetic analogs, KW-2189 and CB1-TM1); eleutherobin; pancratistatin; sarcodictyin; spongistatin; nitrogen mustards, such as chlorambucil, naphthyl mustard, clofosamide, estramustine, ifosfamide, bischloroethyl, nitrogen oxide mustard hydrochloride, melphalan (m elphalan), novembichin, phenylephrine, prednimustine, trofosfamide, uracil mustard; nitrosoureas, such as carmustine, chlorozotocin, fotemustine, lomustine, nimustine and ranimnustine; antibiotics, such as enediyne antibiotics (e.g. alicheamicin, in particular alicheamicin gamma ll and alicheamicin omega ll; dynemicins, including dynemicin A; bisphosphonates, such as clodronate; esperamicin;and the neocarcinogenins and related chromoprotein enediyne antibiotic chromophores, aclacinomycins, actinomycins, authrarnycins, azaserines, bleomycins, actinomycin C, carabicin, caminomycin, carzinophilin, chromomycins, actinomycin D, daunorubicin, detorubicin, 6-diazo-5-oxo-L-norleucine, doxorubicin (including morpholino doxorubicin, cyanomorpholino doxorubicin, 2-pyrro-doxorubicin and deoxydoxorubicin), epirubicin, ethoxybacillus, chloramphenicol ... esorubicin, idarubicin, marcellomycin, mitomycins such as mitomycin C, mycophenolic acid, nogalamycin, olive mycin, peplomycin, potfiromycin, puromycin, quelamycin, rodorubicin, streptomycin, streptozotocin, tubercidin, ubenimex, zinostatin, zorubicin; Metabolites such as methotrexate and 5-fluorouracil (5-FU); folic acid analogs such as dimethoate, methotrexate, pteropterin, trimetrexate; purine analogs such as fludarabine, 6-mercaptopurine, thiopurine, thioguanine; pyrimidine analogs such as ancitabine, azacitidine, 6-azauridine, carmofur, cytarabine, dideoxyuridine, doxifluridine, enocitabine, floxuridine; androgens such as calustratone, drostanolone propionate, cyclothiocarb, melastane, testolactone; antiadrenal drugs such as aminoglutethimide, mitotane, trilostane, lostane; folic acid supplements such as folinic acid; aceglucuronolide; aldophosphamide glycoside; aminolevulinic acid; eniluracil; amsacrine; bestrabucil; bisantrene; edatraxate; fosfoamide; colchicine; diazocone; elformithine; elliptonium acetate; epothilone; etoglucid; gallium nitrate; hydroxyurea; lentinan; lonidainine; maytansine alkaloids such as maytansine and ansamitocin; mitoguanidine; mitoxantrone; mopidanmol;diamine nitraerine; pentostatin; methamine mustard; pirarubicin; losoxantrone; podophyllic acid; 2-ethylhydrazine; methylhydrazine derivatives including N-methylhydrazine (MIH) and procarbazine; PSK polysaccharide complex); razoxane; rhizoxin; sizofuran; spirogermanium; tenuisporic acid; triazoquinone; 2,2′,2″-trichlorotriethylamine; trichothecenes (particularly T-2 toxin, verracurin A, roridin A, and anguidine); urethane ( urethan; vindesine; dacarbazine; mannitol mustard; dibromomannitol; dibromodulanol; pipobroman; gacytosine; cytarabine ("Ara-C"); cyclophosphamide; thiotepa; taxanes, such as paclitaxel and doxetaxel; chlorambucil; gemcitabine; 6-thioguanine; mercaptopurine; methotrexate; platinum coordination complexes, such as cisplatin, oxaliplatin, and carboplatin; vinblastine; platinum; etoposide (VP-16); ifosfamide; mitoxantrone; vincristine; vinorelbine; novantron e; teniposide; edatrexate; daunorubicin; aminopterin; xeloda; ibandronate; irinotecan (e.g., CPT-11); topoisomerase inhibitor RFS2000; difluoromethylornithine (DMFO); retinoids, such as retinoic acid; capecitabine; anthracyclines, nitrosoureas, antimetabolites, epipodophyllotoxins, enzymes, such as L-asparaginase; anthracenediones; hormones and antagonists, including adrenocortical steroid antagonists, such as prednisone (prednisone); isone) and equivalents, dexamethasone and aminoglutethimide; progestogens such as hydroxyprogesterone caproate, medroxyprogesterone acetate and megestrol acetate; estrogens such as diethylstilbestrol and ethinyl estradiol equivalents; antiestrogens such as tamoxifen; androgens including testosterone propionate and fluoxymesterone / equivalents; antiandrogens such as flutamide, gonadotropin-releasing hormone analogs and leuprolide; and nonsteroidal antiandrogens such as flutamide; biological response modifiers such as IFNa, IL-2, G-CSF and GM-CSF; and pharmaceutically acceptable salts, acids or derivatives of any of the above. ;

[0365] Suitable examples of radiotherapy include, but are not limited to, external beam radiotherapy (e.g., superficial X-ray therapy, orthovoltage X-ray therapy, megavoltage X-ray therapy, radiosurgery, stereotactic radiotherapy, fractionated stereotactic radiotherapy, cobalt therapy, electron therapy, fast neutron therapy, neutron capture therapy, proton therapy, intensity modulated radiation therapy (IMRT), three-dimensional conformal radiation therapy (3D-CRT), etc.), brachytherapy, non-sealed source radiation therapy, helical tomotherapy, etc. Gamma rays are another form of photons used in radiotherapy. Gamma rays are spontaneously generated due to the release of radiation from certain elements (e.g., radium, uranium, and cobalt 60) when they decompose or decay. In some embodiments, the radiotherapy can be proton radiotherapy or proton microbeam radiotherapy. Proton radiotherapy is an ultra-precise form of radiation therapy that uses proton beams (Prezado Y, Jouvion G, Guardiola C, Gonzalez W, Juchaux M, Bergs J, Nauraye C, Labiod D, DeMarzi L, Pouzoulet F, Patriarca A, Dendale R, Tumor control in rats with RG2 glioma: a comparison of proton microbeam therapy with standard proton therapy, Int J Radiat Oncol Biol Phys. 2019 Jun 1;104(2):266-271. doi:10.1016 / j.ijrobp.2019.01.080; Prezado Y, Jouvion G, Patriarca A, Nauraye C, Guardiola C, Juchaux M, Lamirault C, Labiod D, Jourdain L, Sebrie C, Dendale R, Gonzalez W, Pouzoulet F., Proton microbeam radiotherapy broadens the therapeutic index of high-grade gliomas, Scientific Reports 2018 Nov 7;8(1):16479. doi:10.1038 / S41598-018-34796-8). Radiotherapy can also be FLASH radiotherapy (FLASH-RT) or FLASH proton irradiation. FLASH radiotherapy involves ultra-fast radiation therapy at dose rates several orders of magnitude higher than current routine clinical practice (ultra-high dose rate) (Favaudon V, Fouillade C, Vozenin MC, FLASH radiotherapy saves healthy tissue, Medical Sciences (Paris) 2015;31:121-123. DOI:10.1051 / medsci / 20153102002); Patriarca A, Fouillade CM, Martin F, Pouzoulet F, Nauraye C, et al., Experimental setup for FLASH proton irradiation of small animals using a clinical system, International Journal of Radiation Oncology Biology Physics 102 (2018), pp.619-626. doi:10.1016 / j.ijrobp.2018.06.403. Epub Date: July 11, 2018 (Prezado Y, Jouvion G, Guardiola C, Gonzalez W, Juchaux M, Bergs J, Nauraye C, Labiod D, De Marzi L, Pouzoulet F, Patriarca A, Dendale R. Tumor Control in RG2 Glioma-Bearing Rats: A ComparisonBetween Proton Minibeam Therapy and Standard Proton Therapy.Int J RadiatOncol Biol Phys.2019 Jun 1;104(2):266-271.doi:10.1016 / j.ijrobp.2019.01.080; Prezado Y, Jouvion G, Patriarca A, Nauraye C, Guardiola C, Juchaux M, Lamirault C, Labiod D, Jourdain L, Sebrie C, Dendale R, Gonzalez W, Pouzoulet F. Proton minibeamradiation therapy widens the therapeutic index for high-grade gliomas. SciRep. 2018 Nov 7;8(1):16479. doi: 10.1038 / s41598-018-34796-8). Radiotherapy may also be FLASH radiotherapy (FLASH-RT) or FLASH proton irradiation.FLASHradiotherapy involves the ultra-fast delivery of radiation treatment at doserates several orders of magnitude greater than those currently in routine clinical practice (ultra-high dose rate) (Favaudon V, Fouillade C, Vozenin MC. The radiotherapy FLASH to save healthy tissues. Med Sci (Paris) 2015; 31: 121-123. DOI: 10.1051 / medsci / 20153102002); Patriarca A., Fouillade CM, Martin F., Pouzoulet F., Nauraye C., et al. Experimental set-up for FLASH protonirradiation of small animals using a clinical system. Int J Radiat Oncol BiolPhys, 102(2018), pp.619-626.doi:10.1016 / j.ijrobp.2018.06.403.Epub 2018 Jul11)). .

[0366] "In combination" may refer to administration of an additional therapy before, simultaneously with, or after administration of a T cell composition according to the present disclosure.

[0367] In addition, or as an alternative to checkpoint blockade, gene editing techniques (including but not limited to TALEN and Crispr / Cas) can also be used to genetically modify the T cell composition of the present disclosure to make it resistant to immune checkpoints. Such methods are known in the art, see, for example, US20140120622. Gene editing techniques can be used to prevent the expression of immune checkpoints expressed by T cells (see checkpoint inhibitors listed above), and more particularly but not limited to PD-1, Lag-3, Tim-3, TIGIT, BTLA CTLA-4 and combinations thereof. The T cells discussed herein can be modified by any of these methods.

[0368] T cells according to the present disclosure may also be genetically modified to express molecules that increase tumor homing and / or deliver inflammatory mediators to the tumor microenvironment, including but not limited to cytokines, soluble immunomodulatory receptors and / or ligands.

[0369] In some embodiments, the tumor neoantigen peptides disclosed herein are used in combination with another immunotherapy (e.g., immune checkpoint therapy), more particularly with an anti-checkpoint antibody (e.g., the antibodies exemplified above, particularly but not limited to anti-PD1, anti-PDL1, anti-CTLA-4, anti-TIM-3, anti-LAG3, anti-GITR antibodies) for cancer vaccination therapy.

[0370] The present disclosure also includes the use of tumor cell TE tags defined herein as cancer cell biomarkers and / or as biomarkers for the effectiveness of immune checkpoint therapy. In some embodiments, the cancer is glioblastoma, and the tumor cell TE tag comprises SEQ ID NO: 1 to 5020, and is therefore a glioblastoma biomarker. In some specific embodiments, the cancer is glioblastoma, and the tumor cell TE tag comprises SEQ ID NO: 1 to 26, 28 to 5020; preferably SEQ ID NO: 1 to 10; 12 to 26, 28 to 430 and 432 to 5020; more preferably SEQ ID NO: 1 to 10, 12 to 26, 28 to 57, 59 to 242, 244 to 255, 257 to 319, 321 to 393, 395 to 430 and 432 to 5020, and is therefore a glioblastoma biomarker.

[0371] The ability to distinguish between self and non-self is a core principle of immunity. Invading pathogens must be recognized as non-self to trigger an adequate response, while self-antigens must be tolerated to avoid autoimmunity. Innate detection of pathogens depends on the recognition of pathogen-associated molecular patterns (PAMPs) by pattern recognition receptors (PRRs). Recognition of foreign nucleic acids is a key step in sensing pathogens, however, host nucleic acid sensors recognize nucleic acids in a non-sequence-specific manner. The fact that nucleic acid sensing is not sequence-specific blurs the fundamental distinction between self and non-self. In fact, the expression of TEs can generate nucleic acids that serve as endogenous PAMPs and may drive harmful immune responses. Infectious viruses that are antigenically similar to endogenous retroviral proteins may trigger this autoimmune response. The loss of tolerance to endogenous viral proteins in transgenic mice after viral infection also illustrates this possibility (see Benihoud K et al., Oncogene (2002) 21, 5593-5600).

[0372] Thus, there is increasing evidence for a relationship between autoimmune and / or inflammatory manifestations and the presence of endogenous or exogenous retroviral sequences. Systemic autoimmune diseases are characterized by a defect in immune tolerance to self-antigens, which may include endogenous retroviral sequences and products of retrotransposons. (Herrmann M et al., (1998) Curr. Opin. Rheumatol. 10, 347-354.; Nakagawa K and Harrison LC., (1996) Immunol. Rev. 152, 193-236; CA Thomas et al., Cell Stem Cell 21, 319-331. e31 (2017), Tokuyama, Maria et al., Proceedings of the National Academy of Sciences of the United States of America, Vol. 115, 50 (2018): 12565-12572; Zhang X, Zhang R, Yu J., Frontiers in Cell and Developmental Biology 2020; 8: 657 Published on August 7, 2020 (Herrmann M et al., (1998). Curr. Opin. Rheumatol., 10, 347-354.; Nakagawa K and Harrison LC.(1996).Immunol.Rev., 152, 193-236; CAThomas et al., Cell stem cell 21, 319-331.e318(2017), Tokuyama, Maria et al. "PNAS vol.115,50(2018): 12565-12572; Zhang X, ZhangR, Yu J., Front Cell Dev Biol. 2020;8:657. Published 2020 Aug 7)).

[0373] In this context, the TE-derived peptides of the present disclosure can be used for tolerance-inducing cell therapy, including vaccination or induction of tolerogenic dendritic cells (tolDCs) or regulatory T cells (Tregs). Such cell therapies have indeed generated a great deal of interest in the treatment or prevention of autoimmune diseases (see Flórez-Grau, Georgina et al. “Tolerogenic Dendritic Cells as a Promising Antigen-Specific Therapy in the Treatment of Multiple Sclerosis and Neuromyelitis Optica From Preclinical to Clinical Trials.” Frontiers in Immunology vol. 9 1169.31 May.2018; and Cauwels, Anje and Jan Tavernier, “Tolerogenic Strategies in the Treatment of Autoimmune Diseases: From Ex Vivo to In Vivo Strategies.” Frontiers in Immunology vol. 11 674.31 May.2018;and Cauwels, Anje, and Jan Tavernier. "Tolerizing Strategies for the Treatment of Autoimmune Diseases: From ex vivo to in vivo Strategies." Frontiers in immunology vol. 11674. 14 May. 2020)). Highly suitable TE-derived peptides according to the present disclosure include peptides of any of the following: SEQ ID NO: 3 to 8, 10, 12, 14 to 17, 19 to 21, 24 to 26, 28, 29, 33, 34, 37, 41, 43, 44, 46, 47, 51 to 53, 55, 56, 59, 62, 67, 69, 74, 75, 77, 80, 81, 84 to 87, 90, 92, 96, 97, 99 to 103 , 108, 109, 112, 113, 116, 125, 128 to 130, 132, 134 to 137, 140, 142, 145 to 149, 154 to 156, 158, 160, 163, 166, 168 to 171, 174 to 176, 178, 183 to 187, 189, 191, 192 , 194 to 197, 200 to 205, 207, 209 to 211, 213, 216, 219 to 221, 224 to 227, 229 to 237, 240 to 242, 247, 249, 250, 252, 254, 255, 258, 261, 263, 264, 266, 268 to 274, 276 , 278, 280, 284, 289, 293, 303, 306, 308 to 310, 316, 319, 321, 322, 324, 327, 330, 332, 336, 338 to 342, 345, 347 to 349, 351, 357, 358, 363, 364, 366, and 368 (redundant;See Table 3), in particular SEQ ID NO: 3 to 7, 10, 12, 14 to 17, 19 to 21, 24 to 26, 28, 29, 33, 34, 37, 41, 43, 46, 52 to 53, 55, 56, 59, 62, 69, 74, 75, 77, 80, 92, 97, 99 to 102, 108, 109, 112, 113, 116, 128 to 130, 132, 134 to 137, 142, 145, 146, 148, 149, 154 to 156, 160, 163, 166, 168 to 171, 174 to 176, 178, 183 to 187, 189, 191, 194 to 197, 200 to 205, 207, 208 09 to 211, 213, 216, 219, 221, 224 to 227, 229 to 237, 240 to 242, 247, 249, 250, 252, 255, 261, 263, 266, 268, 271, 273, 274, 276, 278, 280, 284, 293, 303, 306, 308 to 310, 316, 319, 324, 327, 332, 336, 338 to 342, 345, 348, 349, 357, 358, 363 and 368 (redundancy group 1; see Table 3), which are redundantly expressed by a large number of TEs in the genome (usually encoded by more than 200 genomic TEs). Generally, such TE-encoded peptides are not tumor-specific. In some embodiments, the TR-derived peptide is a LINE-1 peptide, particularly a young L1HS, L1PAx-derived peptide, and L1PBx-derived peptide. ;

[0374] In some embodiments, the expression of one or more TEs (particularly encoding peptides as described above) or preferably a combination thereof can be used as a biomarker for the diagnosis of immune diseases.

[0375] Example

[0376] Materials and methods

[0377] Transposable element annotation

[0378] Classification and TE metadata

[0379] Transposable element annotations have been retrieved from two different databases: the Homer repeats gtf annotation file (v4.11.1) based on the hg19 (v6.4) UCSC annotation; and the hg19 gtf annotation file from TEtranscript (Jin et al., 2015, doi:10.1093 / bioinformatics / btv422, e-published July 23, 2015). Both annotations were based on the RepeatMasker database and have been merged based on the same coordinates to obtain the following information for each repeat: class, family, subfamily, divergence, coordinates). The L1 family was subdivided into 2 families: (1) L1PA / B / x, which includes TEs from the closely related L1HS, L1PA(x), L1PB(x), L1P(x) subfamilies; and (2) Other L1, which recombines all other L1 TEs not present in L1PA / B / x. All DNA transposon TEs were classified as DNA. Homer's annotatePeaks.pl was run to obtain the genomic position (intron, exon, 3'UTR, 5'UTR, intergenic, other) of each individual TE. The distance of each TE to the nearest protein-coding gene in the gencode gtf annotation file (Release 19GRCh37.p13) was retrieved using the closest and intersect tools in bedtools (v2.29.2).

[0380] Age of TE

[0381] The repeat age of human repeats was calculated using the divergence percentage using the following formula: divergence / (2.2*10 -9 ), according to the formula of the article (Choudhary et al., Genome Biol 2020, 21, 16).

[0382] Complete ORF

[0383] Complete open reading frame (ORF) positions were retrieved from the gEVE database (Nakagawa, S. and Takahashi, MU Database (Oxford) 2016). For analysis of the human genome version hg19, the hg38 gEVE annotations were formatted and adapted for hg19 using the UCSC “Lift Genome annotations” tool, available here: https: / / genome.ucsc.edu / cgi-bin / hgLiftOver. The coordinates of the complete ORFs from the gEVE annotation and all individual TEs from the genome were matched to assign the complete ORFs to individual TEs in case of overlapping coordinates. 30,517 individual TEs overlapped with the complete ORFs, most of which were L1 (mainly L1PA / B / x) and ERV (mainly ERV1, ERVK, ERVL) elements. To determine the amino acid sequence similarity between typical TE proteins in the gEVE database and peptides in the immunopeptidomics results, blastp was performed between the gEVE protein sequences and the immunopeptidomics sequences. No threshold for Evalue was set, and the similarity was estimated and divided into 3 categories: (1) 100% match: no mismatch, no gap, and the query coverage of each HSP was 100%; (2) at most 1 mismatch: 1 mismatch, no gap, and the query coverage of each HSP was above 85%; (3) at most 2 mismatches: 2 mismatches, no gap, and the query coverage of each HSP was above 85%.

[0384] Retrieve TE nucleotide sequences

[0385] The fasta sequences were obtained from each TE using getfasta (bedtools version 2.30.0). Due to the getfasta processing step, the first nucleotide was not considered, so the length of the sequence was minus 1 nucleotide.

[0386] Analysis of known TE proteins

[0387] LTR and LINE proteins

[0388] LTR TEs encoding peptides that overlapped with the complete ORF were classified as Env, Gag, Pol, or Pro using gEVE's RetroTector annotation. For LINE elements, blastp was performed between LINE-derived peptides and the ORF1p and ORF2p protein sequences found in Uniprot (accession numbers Q9UN81 and O00370). Allowing up to 1 mismatch, 28 matches from ORF1p and ORF2p were identified in our LINE-derived peptides. LINE and LTR TEs encoding peptides were also compared to the gEVE HMM profile annotation to classify TE protein motifs found in these TEs.

[0389] TE ORF annotation

[0390] ORFs were identified and annotated from TE sequences using a homemade R script. Details: (1) TE nucleotide sequences were formatted to obtain six frames using the R package Biostrings (v2.58.0) and its functions DNAStringSet and reverseComplement; (2) The sequences of the six frames were translated using the translate function in Biostrings; (3) Stop codons and methionine were detected using the matchPDict function in Biostrings; (4) Peptides from immunopeptidomics results were also found using the matchPDict function; (5) ORFs of at least 30 bp (start codon of 3, sequence of 8AA*3, stop codon of 3) were detected using the ORFik R package (v1.10.13) and only the longest ORFs were retained. Two different start codon patterns were submitted to detect ORFs: “ATG” for classical start codons and “ATG|CTG|GTG|TTG” for classical and non-classical start codons. ORFs found using only the second pattern were classified as “CTG|GTG|TTG”; (6) The length of the ORF was calculated using the start and end positions; (7) The positions of all identified ORFs, stop codons, methionine, and peptides in all five frames of TE were represented using the R package ggplot2.

[0391] Single-cell data analysis

[0392] Download data and align reads to genome

[0393] Smart-seq2 data (GEO accession number: GSE84465) were downloaded from the Sequence Read Archive (SRA) database using pre-reads from the SRA Toolkit (v2.10.0). SRA files were converted to fastq files using fastq-dump. Fastq files are 75 bp paired-end unstranded reads. Raw RNA reads were mapped to the human genome sequence (hg19) using the 2-pass mode of STAR (version 2.7.1.a) (parameters: --quantMode GeneCounts, --twopassMode Basic, --alignSJDBoverhangMin 1, --bamRemoveDuplicatesType UniqueIdentical, --winAnchorMultimapNmax 1000, --outFilterMultimapNmax 1000, --outFilterScoreMinOverLread 0.33, --outFilterMatchNminOverLread 0.33, --outFilterMismatchNoverLmax 0.04, --outMultimapperOrder Random, --sjdbOverhang76).

[0394] Quantification of gene and TE expression

[0395] To calculate TE and gene expression quantification, featureCounts of Subread (v1.6.4) were calculated on each genome mapped read file. Different parameters were used depending on the analysis: (1) for gene expression: -p-ignoreDup-g gene_id using gencode gtf annotation file; (2) for TE expression on single copy (a) only with uniquely mapped reads: -p-ignoreDup-g transcript_id using TEtranscript hg19 gtf annotation file; (b) with unique and multiple mapped reads: -p-ignoreDup-g transcript_id-M--primary (3) for TE expression in subfamilies with unique and multiple mapped reads: -p-ignoreDup-g gene_id-M--primary. Cell count files were merged into matrices using a homemade python script (Python 3.6).

[0396] Filtering features and cells, normalization, batch correction

[0397] Cell metadata and feature raw count matrices were imported into R (version 4.0.3) to create SingleCellExperiment R objects. CPM, FPKM, and TPM values ​​for gene and TE expression were calculated from raw counts using the cuttle R package (v1.0.4) and its functions (calculateCPM, calculateFPKM, calculateTPM) before any filtering. Cells with low count numbers and low feature numbers (1 / 3 of MAD) were removed using the Scater and Scran packages. Considering the uniquely mapped reads TE matrix (1): individual TEs with an average count of less than 1 per cell were removed [22,000 individual TEs remaining]; for multiply mapped reads (2): individual TEs with less than 5 counts in at least 20 cells were removed to account for expression in small populations [130,028 individual TEs]; for gene expression (3): genes with less than 5 counts in at least 20 cells were removed [19,867 genes remaining]; for subfamily expression: no filtering was performed [992 subfamilies]. The raw count matrix was then normalized using the logNormCounts function in the scater R package. After multiple validations, a batch effect associated with the cell plate ID was identified. To correct for this, the removeBatchEffect function in the limma R package was used with the plate ID as the batch and the cell type as the design.

[0398] Dimensionality reduction

[0399] A single Seurat object was created and the raw, normalized, and normalized+corrected feature matrices were imported into the different assays. CPM, FPKM, and TPM matrices were also imported. Seurat v3 was used for uniquely mapped reads analysis; Seurat v4 was used for multiply mapped reads analysis, subfamily analysis, and gene analysis. In Seurat, FindVariableFeatures was performed to distinguish the 5000 most variable genes or individual TEs; ScaleData was used to scale feature expression, RunPCA was used to compute 75 principal components, and RunTSNE was used to perform t-SNE dimensionality reduction on 50 principal components. Dimensionality reduction steps were performed on the normalized+corrected assays.

[0400] Differential expression analysis and enrichment testing

[0401] In Seurat, FindAllMarkers was performed on annotated cell types with a threshold of 0.25 fold change (natural log for Seurat v3 or log2 for v4) for features expressed by at least 10% of all cells in 1 cell type. Based on the results of FindAllMarkers, differentially expressed signatures were designed for genes, subfamilies, and individual TE signatures using adjusted p-values ​​less than or equal to 0.05. Using the signatures of interest, signature scores were calculated by the Seurat function AddModuleScore. This function calculated the mean expression of each feature in the signature of each individual cell and subtracted the aggregate expression of the control signature set. TE subfamily enrichment was performed using all annotated individual TEs in the genome (4.6 million TEs) as a reference and all expressed TEs or individual TE signatures in each population as queries. Hypergeometric tests were calculated using phyper in the stats R package (v4.0.3). False discovery rate correction was then applied using p.adjust in the stats R package.

[0402] chart

[0403] Most of the graphs were made using R (v4.0.3). Pie charts, lollipop plots, bar charts, violin plots, boxplots, jitter plots, volcano plots, density plots, scatter plots, and dimensionality reduction plots were made using the ggplot2 R package (v3.3.3) or functions from the Seurat package. Pie-donut plots were made using the PieDonut function from the webR package (v0.1.6). Heatmaps were created using the Pheatmap R package (v1.0.12) and ComplexHeatmap (v2.6.2). The clustering method used was ward.D2. IGV (v2.8.10) was used to visualize the read coverage of bulk RNA samples.

[0404] Radar plot and chromosome distribution

[0405] Radar plots representing the distribution of features on chromosomes were made using the radarchart function in the fmsb R package (v0.7.1). Genomic proportions were calculated using all annotated genes and individual TEs from the gencode and TEtranscript annotations, respectively.

[0406] Bulk RNA sequencing data analysis

[0407] Download, alignment to genome, and quantification

[0408] Approximately 50 samples were randomly selected from each GTEx tissue, and their fastq read files were downloaded using preread and fasterq-dump of sratoolkit (v2.10.0). Fastq reads of the TCGA-GBM project were downloaded using gdc-client (v1.6.1). Alignment and feature quantification (genes, individual TEs, subfamilies) were performed following the same protocol described for Smart-seq2 analysis. Expression was normalized using estimateSizeFactors in the DESeq2 R package (v1.30.1) to obtain normalized counts. TPM values ​​were also calculated using the calculateTPM function of scuttle. Two subsets of TE expression matrices were obtained for each database: (1) an expression matrix with only TEs from tumor single-cell TE signatures; (2) an expression matrix with only TEs considered to be expressed. A TE was considered to be expressed if at least 5 counts could be observed in 20% of the samples (considering all samples from the TCGA or GTEx database, respectively). TCGA samples retained 130,640 TEs, while GTEx samples retained 192,243 TEs, of which 103,585 TEs were shared by both databases.

[0409] Downstream analysis of bulk RNA sequencing samples

[0410] The merged tumor signature-specific matrix was imported into a Seurat object along with all samples from TCGA and GTEx. Both DESeq2 normalized counts and TPM values ​​were imported. Using normalized counts, ScaleData, RunPCA, and RunUMAP were applied to obtain UMAP representations. To assess signature expression in samples, the average expression of all TEs in the tumor signature was calculated using TPM values.

[0411] Gene set enrichment analysis

[0412] Gene set enrichment analysis (GSEA) was performed using the DESeq2 normalized count matrix of co-expressed TEs (103,585 TEs) between the TCGA and GTEx databases to detect the enrichment of single-cell tumor signatures in normal or tumor samples. GSEA (v4.1.0) was run with default parameters. GSEA results were imported into R and presented using ggplot2.

[0413] Read coverage

[0414] Mapped read bam files from tumor single cells, immune single cells, TCGA tumor samples, and TCGA normal samples were merged using samtools (v1.9) and its merge function. The merged bam files were indexed using samtools' index. Read coverage for each merged bam file was calculated using deepTools (v3.3.1)'s bamCoverage and the following parameters: --outFileFormat bigwig --normalizeUsing CPM. The results were visualized using IGV (v2.8.10).

[0415] Binding of peptides to HLA-4*02:01 and HLA-B*07:02 and formation of multimers

[0416] The predicted peptides were synthesized by GeneCust with a purity of >98%. HLA-A*0201 monomers were purchased from Immunaware (Copenhagen, Denmark) in the form of easYmers. According to the manufacturer's instructions, the predicted and mass spectrometry (MS) TE-derived peptides bound to HLA-A*0201 were measured as HLA-I complex formation by FACS. In brief, biotinylated monomers were incubated with synthetic peptides (100 mM) at 18°C ​​for 48 hours, then bound to streptavidin-coated microbeads and stained with PE-conjugated anti-β2 microglobulin. As positive controls for HLA-I complex formation, CMV peptides pp65 495-503 (NLVPMVATV:: SEQ ID NO: 5021), CMV pp65 417-426 (TPRVTGGGAM:: SEQ ID NO: 5022) and CMV IE1 99-107 (RIKEHMLKK:: SEQ ID NO: 5023) were used for HLA-A*02:01, HLA-B*07:01. Melan-A mutant sequence (ELAGIGILTV:: SEQ ID NO: 502), a peptide known to bind well to HLA-A*0201, was also included as a second positive control for HLA-I complex formation of this monomer. Binding rates are expressed as a percentage of HLA-I complex formation relative to the CMV positive control. Peptides with at least 50% HLA-I complex formation relative to the positive control were used in the in vitro vaccination experiments.

[0417] To form multimers, peptide-HLA-I complexes were tetramerized using different combinations of streptavidin conjugated to fluorochromes (PE, APC BV421, BV711, PE-CF549, and PECy5) at a final concentration of 8 mg / ml. All tetramers were stored at 4°C and used within 2 months.

[0418] Polysome staining and analysis

[0419] After the in vitro vaccination experiment, total cells were stained for multimers by combining 1 μl of each tetramer specificity and two different SA-labeled tetramers of each specificity. Staining was performed at room temperature for 20 minutes in a final volume of 100 μl PBS 1% BSA / 1M cells. Then, 100 μl of a surface antibody mixture containing anti-CD3 BV650 and anti-CD8 PECy7 (BD Biosciences) was added at a final dilution of 1 / 200 and incubated for another 20 minutes at 4°C. Finally, the cells were washed twice with PBS-1% BSA and analyzed by flow cytometry. Dead cells were excluded using Live / Dead Aqua-405nm (ThermoFisher). Data were collected using a ZE5 cell analyzer (Bio-Rad) and analyzed using FlowJo v10.3.

[0420] According to the strategy described by Andersen et al. (Andersen et al., "Nature-Laboratory Guide (Nat Protoc.)" 2012, 7891-902), live single cells, CD3+CD8+ cells were subjected to multimer analysis. According to the double multimer staining criteria, the expansion was considered positive. The amplified population of each peptide is expressed as the frequency of total CD8+ cells in each replication, or as the total multimer frequency in total CD8+T cells evaluated in all replications of a donor.

[0421] In vitro vaccination trials

[0422] According to INSERM's Code of Ethics, Buffy coats from healthy donors were obtained from École des Sciences et Médecins du Sang (Paris, France). According to the French Public Health Code (Articles L 1121-1-1, L 1121-1-2), written consent and IRB approval were not required for non-interventional studies in human subjects.

[0423] PBMCs were obtained by density gradient separation using Lymphprep (StemCell Technologies) and phenotyped by FACS using anti-HLA-A2 antibody (clone BB7.2, BD Biosciences) and anti-HLA-B7 antibody (clone BB7.1, Biolegend). Only HLA-A2+ and HLA-B7+ donors were used. Monocytes and lymphocytes from the same donor were purified by positive selection for CD14+ using magnetic beads (Miltenyi Biotec). + , CD4+ and CD8 +Cells were cultured at 10 °C in RPMI-1650 / lutamax (Gibco) supplemented with recombinant human IL-4 (50 ng / ml) and GM-CSF (10 ng / ml), 10% FBS, penicillin (100 U / ml) / streptomycin (100 μg / ml). 6 The isolated CD4+ fraction was differentiated into CD14+ fraction within 5 days to obtain monocyte-derived dendritic cells (mo-DCs). + and CD8 + T cells were cryopreserved during mo-DC differentiation.

[0424] After differentiation, mo-DCs were cultured at 1x10 6 Cells / ml were seeded in culture medium in 24-well plates and matured OVNs with LPS (100 ng / ml). Afterwards, the medium was removed and LPS-treated mo-DCs were pulsed with a mixture of selected good binders TE-derived peptides (predicted peptides or MS-derived peptides from HLA-I peptidomics data) for 3 h at 37°C. The final concentration of each peptide was 1 μg / mL. Finally, peptide-loaded mo-DCs were harvested, pelleted, and counted. Cryopreserved lymphocyte fractions were thawed and cultured by mixing 1x10 6 CD8+T cells and 0.1x10 6 CD4+T cells and 0.1x10 6 Peptide-loaded mo-DC (CD8-CD4-mo-DC ratio: 10:1:1) were co-cultured. Each well was considered an independent replicate. The total number of replicates was determined by the total number of CD8+T cells. Without interfering with the cells, half of the culture medium was replaced after 5 days and then monitored every 3 days until day 15-20. The expansion of specific CD8+T cell populations was assessed by FACS using multimer staining. X-vivo 15 medium (Lonza) supplemented with penicillin (100U / ml) / streptomycin (100μg / ml) (Gibco), 10% FBS, 10U / ml IL-2 (Novartis) and 10ng / ml IL-7 (PeproTech) was used as a culture medium. As a negative control, in the case of MS-derived peptides, mo-DC non-peptide pulses were used for replication. For HLA-A2+ donors, mo-DC pulsed only with Melan-A peptide (ELAGIGILTV) was used as a positive control for T cell expansion (1 or 2 replicates). In the mixture of MS-derived HLA-A2+ peptides, three HLA-A*02:01 binding peptides derived from the classical sequence of normal proteins (present in the Uniprot normal proteome) were included.

[0425] Mass spectrometry-based immunopeptidomics

[0426] Mass spectrometry data analysis

[0427] Mass spectrometry-based immunopeptidomics files were obtained from PXD020079, PXD008127, PXD003790, and MSV000084442 and analyzed with ProteomeDiscoverer 2.5 (ThermoFisher) using the following parameters: no enzyme, precursor mass tolerance 20 ppm, and fragment mass tolerance 0.02 Da. Methionine and N-acetylation were enabled as variable modifications. Using Percolator, a false discovery rate (FDR) of 1% was applied at the peptide level, while no FDR was used at the protein level. Spectra were searched against human Uniprot / SwissProt in the in silico translated tumor-enriched TE database with isoforms linked to 6 reading frames (updated on June 3, 2020). The identified potential TE-derived peptides were subsequently filtered using the UniProt / TrEMBL database, considering leucine-isoleucine and lysine-glutamine as equivalents, respectively. Finally, the spectra of the identified TE-derived peptides were manually verified.

[0428] Calculation of peptide hydrophobicity index (HI)

[0429] For comparison of retention time and hydrophobicity, HI was predicted using the SSRCalc (Krokhin et al., 2004) web server. http: / / hs2.proteome.ca / SSRCalc / SSRCalcX.html ).

[0430] Single assignment and all assignment definitions

[0431] Since multiple TEs can encode the same peptide, they were divided into two different categories in order to observe the characteristics of TE-encoded peptides. All assignments correspond to all TEs that encode a certain peptide (all 568 TEs for 370 peptides). Single assignments correspond to a random selection of each peptide for the individual TEs that can encode the corresponding peptide (370 TEs for 370 peptides).

[0432] Identification of potential peptide-encoding TEs

[0433] To identify or filter all TEs containing peptide sequences, peptide sequences were aligned to all annotated individual TEs in the genome in all six frames using tblastn (v2.11.0). Using TETranscript gtf processed into BED format, getfasta from bedtools (v2.30.0) was used to retrieve the sequences of all TEs in the genome. No restrictions on Evalue were required. No restrictions on Evalue were required. All matches with 0 number of mismatches, 0 number of gap openings, and 100 query coverage per HSP were retained and considered as peptide-encoding TEs, except for those in the tumor signature identified with ProteomeDiscoverer.

[0434] Spectroscopic validation with synthetic peptides

[0435] To validate the spectra, 24 identified peptides were synthesized (GeneCust) with 95% HPLC purity and injected into the Velos Orbitrap (CID). The raw files were analyzed with ProteomeDiscoverer 2.5 (ThermoFisher). Spectra were exported and compared with those obtained from the immunopeptidomics analysis. Only synthetic and endogenous PSMs with identical charge and without modification were analyzed. Where possible, the same fragmentation type (CID or HCD) between the two spectra was given priority.

[0436] Identification of tumor-enriched TE-derived peptides

[0437] The TPM expression of all possible TEs in the genome that could encode the identified peptides was retrieved and the 90th percentile value was calculated for each tissue. The TEs encoding each specific peptide were selected and their 90th percentile values ​​were summed to obtain the total transcript expression associated with these peptides. For non-redundant peptides, the associated transcript expression was directly the 90th percentile value of the TE encoding the peptide. The log2 ratio of the peptide-related expression in GBM samples and each GTEx tissue was then performed to assess whether the relative expression of these peptides in GBM samples was higher than that in normal tissues. Using the median value of TPM expression in GBM samples as a threshold, the percentage of expression with the same or higher expression in normal samples was also calculated for each tissue. The Pheatmap function of the ComplexHeatmap R package (v2.6.2) was then used to represent the log2 ratio, 90th percentile value, and the percentage of expression in normal samples. The clustering method with log2 ratios used in the heatmap was ward.D2.

[0438] Statistical analysis

[0439] The R package ggpubr (version 0.4.0) and its function stat_compare_means were used to perform Wilcoxon tests (1) to compare the distances to the nearest gene between immune and tumor signatures; (2) to compare the average expression of tumor signatures in bulk RNA-seq samples; and (3) to compare the lengths of canonical and non-canonical TE-derived peptide ORFs. Pearson correlation scores were calculated using ggpubr's stat_cor to (1) assess the correlation between TEs and their nearest protein-coding genes; and (2) to assess the correlation between the median age of peptide-encoding TEs and the number of TEs that can encode that peptide. Two proportion z tests were calculated to compare the proportions of LINEs in different subsets of individual TEs. The p-values ​​corresponding to the symbols are as follows: ns: p>0.05; *: p<=0.05; **: p<=0.01; ***: p<=0.001; ****: p<=0.0001.

[0440] result

[0441] Single-cell TE expression profiles all cell populations in a tumor

[0442] It is inferred that an effective method to identify TEs specifically expressed in tumor cells would be to compare TE expression in tumors and tumor-infiltrating cells of the same patient. To this end, single-cell transcriptomics (scRNAseq) of all cells present in the tumor microenvironment was used. The study is based on a public dataset that includes tumor and tumor-adjacent samples from four GBM patients analyzed by SMARTseq2 (Darmanis et al., Cell Rep 2017, 21, 1399-1410). Consistent with the analysis performed in the original article, dimensionality reduction and t-SNE visualization based on gene expression resolved seven sorted cell populations from the tumor core and surrounding tissues: immune cells (mainly macrophages), tumor cells, and oligodendrocyte precursor cells (OPCs) were the largest in number ( Figure 1 B).

[0443] To investigate TE expression in single cells, scRNAseq reads were mapped to TE subfamilies (as previously described by Kong et al., Nat Commun, 2019, 10, 5228) or to individual genomic TEs ( Figure 1A). Because mapping of TEs to single genomic locations can be affected by the high conservation of their repetitive motifs, particularly in young TE subfamilies, the use of uniquely and multiply mapped RNAseq reads was compared. Uniquely mapped reads accurately estimate expression of older TE subfamilies but underestimate expression of the youngest TE subfamilies compared to multiply mapped reads, which more accurately reflect expression of young TE subfamilies (Lanciano and Cristofari, Nat Rev Genet, 2020, 21721-736). To quantify TE expression, FeatureCounts with --primary and random reporter positions (-M, for multiple alignments) was used as recommended by Teissandier et al. (Teissandier et al., MobDNA, 2019, 10, 52). As with gene expression, tSNE based on expression of 992 TE subfamilies or the 5000 most variable individual TEs in single cells was used to resolve all cell populations in the tumor microenvironment ( Figure 1 B middle panel). Tumor cells and OPCs were predominantly present in tumor and peri-tumor samples, respectively, whereas immune cells were present in both tumor and peri-tumor samples as expected (Darmanis et al., 2017). Individually mapped TEs better resolved distinct cell populations compared to TE subfamilies ( Figure 1 B right panel). These results demonstrate that the expression of individual TEs can be resolved at the single-cell level and is sufficient to distinguish different cell populations within the tumor microenvironment.

[0444] TE subfamilies are differentially expressed in tumor cells and immune cells

[0445] To better understand the nature of these TEs, differential expression (DE) analysis of TEs in each cell population relative to all other cell populations was performed to define population-specific TE signatures. These signatures are specific for tumor cells (Table 2), immune cells ( Figure 1C) and each of the other cell populations present in the tumor microenvironment. Unsupervised clustering heatmaps of the 20 most differentially expressed TEs in each cell type based on average log2 fold change schematically show selective expression in each cell population, including tumor cells (not shown). To further investigate the nature of the differentially expressed TEs in each cell population, each tag was compared with all TEs expressed in the dataset (130028). TEs differentially expressed in tumor cells were depleted in SINEs (51.68% vs. 44.52%) and enriched in LTRs (8.33% vs. 12.11%), while TEs in immune cells were depleted in LINEs (30.29% vs. 26.47%) and LTRs (8.33% vs. 5.62%) and enriched in SINEs (51.68% vs. 59.18%), confirming the results of direct mapping of TE subfamilies. Statistical analysis by subfamily revealed that several LTR subfamilies were strongly enriched in tumor cells (mostly HERV), while immune cells differentially expressed several SINE subfamilies (mostly Alu) (Figure ID). Thus, different cell types present in the tumor environment express distinct patterns of TE subfamilies, which can be analyzed from individually mapped TEs by single-cell transcriptomics.

[0446] We next investigated the relationship between TE expression and genomic copy number changes. Gain of chromosome 7 and loss of chromosome 10 are recurrent events in GBM (Kurscheid et al., Genome Biol 2015, 16, 16). Genes and TEs were mapped to their respective chromosomes in each cell type-specific signature. Figure 1 As shown in E, TEs that are differentially expressed in tumor cells but not in other cell populations show a clear preference for chromosome 7 ( Figure 1 E and Figure 1 F). In tumor cells, TEs have an even stronger preference for chromosome 7 than genes (17.91% of expressed TEs are encoded in chromosome 7, compared to only 9.14% of genes) ( Figure 1 F). In contrast, the loss of chromosome 10 was similar in TE (0.93% vs. 4.55% of the genome) and gene signature (1.43 vs. 3.88% of the genome) ( Figure 1 F) Thus, individual TEs can be accurately mapped by scRNAseq and, as expected, show a chromosome 7 bias selectively in neoplastic GBM cells.

[0447] TE expression in tumor cells is enriched at elements independent of their nearest genes

[0448] To better understand the control of TE expression in different cell populations, we first analyzed TE genomic locations. TEs differentially expressed in tumor cells showed a decrease in intronic locations compared to all expressed TEs in the dataset (77% vs. 38.74%), including a decrease in intronic locations when compared to the proportion of differentially expressed TEs in immune cells (68.77%). Figure 2 A). TEs encoded by 3′UTRs were also significantly increased in tumor TEs (25.29%) compared to all expressed TEs (5.02%) or immune cell TEs (11.27%) ( Figure 2 A) These results indicate that, while differentially expressed TEs in immune cells are primarily intronic TEs, differential expression of intergenic and 3′UTR TEs is more frequent in tumor cells.

[0449] Consistent with these results, the proportion of TEs located more than 2 Kb (distal) from the nearest protein-coding gene was higher in the tumor cell signature (22.32%) than in the immune cell signature (12.98%, Figure 2 B). t-SNE analysis based on distal TEs resolved all cell populations, indicating that cell type-specific TE expression may not be entirely attributable to gene-driven transcription. Consistently, TEs differentially expressed in tumor cells have increased TE-gene distances compared to those differentially expressed in immune cells, especially LINE and LTR ( Figure 2 C). The larger distances to the nearest genes of TEs selectively expressed in tumor cells may reflect gene-independent TE expression, including enhancer-dependent or long noncoding RNA (lnc) RNA-dependent readthrough transcription. Therefore, we next analyzed the correlation between the expression of TEs and their nearest genes in tumor and immune single cells. Quantification of the proportion of proximal and distal TEs (expressed together or independently of their nearest genes) showed that the proportion of proximal and distal TEs expressed when their nearest genes were silenced (TE+ genes) was higher in tumor cells (39%) than in immune cell TE signatures (24%) ( Figure 2 D) These results indicate that a higher proportion of TEs differentially expressed in tumor cells are distant from their nearest gene neighbors and are transcribed independently of them, suggesting a higher degree of autonomy of TE transcription in GBM cells.

[0450] Validation of single-cell tumor TE signatures in an independent GBM patient cohort

[0451] To validate the single-cell-based TE signature, we next analyzed bulk RNAseq from TCGA (155 GBM patients and 5 tumor-adjacent samples) and GTEx (1080 healthy samples from 25 tissues). The muscle GTEx cohort was excluded due to its small library size compared to other cohorts. RNAseq reads were mapped to the human genome, and TE expression was quantified using RepeatMasker annotations. Principal component analysis (PCA) and uniform manifold approximation and projection (UMAP) based on the GBM TE signature indicated that the GBM sample cluster was far away from the normal tissue GTEx samples ( Figure 3 A and 3B). Heatmap Z scores in TCGA and GTEx samples show that the top 2000 TEs of the single-cell GBM signature are highly expressed in TCGA GBM samples, while their expression is reduced in healthy tissues (not shown). Gene set enrichment analysis (GSEA) analysis showed that compared with normal brain samples (NES = 1.67 and FDR < 0.05, Figure 3 C) and the expression of scRNAseq GBM TE signatures is highly enriched in GBM compared to other normal tissue samples in GTEx. The average scRNAseq GBM TE signature expression level is also higher in GBM samples compared to normal tissue GTEx samples ( Figure 3 D). Notably, a portion of healthy brain tissue samples expressed high levels of GBM TE signatures. Figure 3 E, Examples of individual TEs overexpressed in scRNAseq (top panel) and scRNAseq (bottom panel) illustrate the specific expression of certain TEs in GBM cells. Thus, analysis of individual TEs in scRNAseq is accurate and can identify recurrent tumor-specific TEs.

[0452] TE-derived peptides are presented on HLA-I and are immunogenic in vitro

[0453] To investigate whether TE-derived peptides are presented by HLA-I molecules in GBM cells, 30 mass spectrometry-based immunopeptidomics samples from GBM primary tumors and cell lines were used (Forlani et al., Molecular & cellular proteomics: MCP, 2021, 20, 100032; Sarkizova et al., Nature: Biotechnology, 2020, 38, 199-209; Shraibman et al., Molecular & cellular proteomics: MCP, 2018, 17, 2132-2145; Shraibman et al., Molecular & cellular proteomics: MCP, 2016, 15, 3058-3070) Figure 4 A). Two different in silico translated TE databases were generated from multiply mapped (3428) or uniquely mapped (1945) differentially expressed TE coding reads. Sequences of all TEs were in silico translated in all 6 reading frames (forward and reverse). The resulting translated TE sequences were combined with the human annotated proteome and tested in an HLA-I peptidomics sample using Proteome Discoverer. The identified TE-derived peptides were then filtered based on canonical proteins (Swissprot+TrEMBL) and the spectra were manually inspected ( Figure 4 A). A total of 178 to 13,720 peptides were identified per sample, of which 370 were TE-derived peptides (Table 3), including 63 peptides predicted from both signatures, 147 peptides predicted from multiple mapped read signatures alone, and 160 peptides predicted from unique mapped read signatures alone. Heatmap of all identified TE-derived peptides shows that the number of peptides varied between samples and that the same peptides recurred in several different patients and cell lines (not shown).

[0454] TE-derived peptides exhibited similar SEQUEST quality scores and peptide length distributions to the Uniprot-annotated peptide set, indicating that they are reliable identifications ( Figure 4B). HLA-A3 binding TE-derived peptides (n=96) contained expected binding motifs obtained from the Immune Epitope Database (IEDB) (not shown). In addition, TE-derived peptides maintained the correlation between hydrophobicity and retention time (not shown). These results indicate that the TE-derived peptide set is reliable and contains features similar to the typical peptide set. 23 TE-derived peptides were synthesized and validated by comparison with endogenous sequences (a total of 24 were tested). The robustness of the process was confirmed, and the identified peptides (using unique and multi-mapped tags) were similar to the TE signature and were preferentially encoded by TEs from chromosome 7. Therefore, differentially expressed TEs in GBM tumor cells are the source of peptides presented on HLA-I molecules.

[0455] To investigate the possibility that TE-encoded peptides could encode potential tumor antigens, T cell precursors were searched in healthy donors. TEs differentially expressed in tumor cells were translated in silico and HLA-A2 binding peptides (strong and weak binders) were predicted using NetMHC. -50 ) and average log fold change (higher than 2.5) were selected. Using a tetramer formation assay, the binding of 7 peptides from immunopeptidomics and 17 peptides predicted by NetMHC to the in silico translated GBM T tag of HLA-A*02:01 (and 2 peptides from immunopeptidomics) was first tested experimentally ( Figure 4 C) 19 peptides were confirmed as HLA-I binders and used to test immunogenicity in vitro. Peptide-loaded monocyte-derived dendritic cells were inoculated with autologous CD4 + and CD8 + T cells were co-cultured to test immunogenicity, and tetramer staining was used as a readout. A mutated melanocyte antigen (Melan-A) peptide (a strong binder of HLA-A*02:01 with high T cell precursor frequency in most healthy donors) (Pitt et al., J Exp Med 190, 1999, 705-715) was used as a positive control for cell expansion. Three HLA-A*02:01 binding peptides from proteins not specifically expressed in GBM tumors and derived from the classical proteome were also used as negative controls. Amplification of tetramer-positive CD8+ cells was observed in at least one donor for 14 TE-derived peptides (including 5 from immunopeptidomics identification, Table 4). + Although the melanocyte antigen (Melan-A)-derived peptide, which is also a non-TE-derived, non-GBM-specific protein, induced a high T cell response, the three peptides derived from typical proteins induced very weak or no responses ( Figure 4D) In ​​conclusion, a subset of TEs differentially expressed in GBM may encode HLA-I binding peptides that are immunogenic in vitro in healthy donors and may be a source of tumor antigens.

[0456] L1, LTR, and SVA subfamilies are the main sources of TE-derived HLA-I peptides

[0457] To investigate the nature of tumor-enriched TEs encoding HLA-I-presented peptides in GBM, we next mapped peptide sequences from all differentially expressed TEs from the single-cell GBMTE signature. In the process, we realized that although 85.41% of the 347 peptides were encoded by a single TE, the remaining 15% of the peptides could be encoded by 2 to hundreds of TEs differentially expressed in this GBM-TE signature ( Figure 4 A). These peptides will be referred to as "single TE encoded peptides" or "multiple TE encoded peptides". For further analysis, when the same peptide can be redundantly encoded by multiple TEs (because it is impossible to determine which TE encodes the peptide), all TEs carrying the peptide encoding nucleotide sequence ("all assignment") or only one of the TEs (arbitrarily selected) per peptide ("single assignment") were considered. The genomic position of the TE encoding the peptide relative to the nearest gene was first analyzed.

[0458] Among TEs encoding HLA-I presented peptides, 37.85% and 31.89% (for all assignments and single assignments, respectively) were located distally (more than 2 Kb from their nearest gene) compared to all expressed TEs (12.11%) or TEs differentially expressed by tumors (22.32%). Analysis of the genomic location of peptide-encoding TEs showed that the majority were intergenic (35.04% and 28.92% for all assignments and single assignments, respectively, compared to 15.17% in the GBM-TE signature). The proportion of intronic TEs also increased, but not much (50% and 50.7% for all assignments and single assignments, respectively, compared to 38.74% in TEs expressed in tumor cells). Among peptide-encoding TEs, 3'UTR TEs were less frequent, accounting for 25.29% of the tumor differentially expressed TEs, while among peptide-encoding TEs, all assigned and single assigned 3'UTR TEs were only 5.81% and 7.03%, respectively. These results confirm the selectivity of the genomic location of peptide-encoding TEs, which are preferentially intergenic TEs or intronic TEs and are not found in 3'UTRs.

[0459] Then the identified peptides are studied whether they are preferentially derived from certain TE categories. According to all assignments and single assignments, peptide-encoding TEs are significantly enriched with LINE elements (accounting for about 30% of all expressed or tumor differentially expressed TEs, respectively, and 52% to 64% for all assignments and single assignments of peptide-encoding TEs, respectively). These TE category analyses also show that TEs classified as "others" are also enriched (see below). These TE category analyses also show that TEs classified as "others" are also enriched (see below). In the "other" category, SVA elements and other types of repetitive sequences encoded as RC, RNA, satellites (Satellite) and unknown (Unknown) are included. Of all TE-derived peptides in this category, about half are from SVA elements (23 of 51). Regarding SINE elements, it has been observed that they are depleted in peptide-generating TEs (from 51.68% and 44.52% in all expressed and differentially expressed TEs, to about 11% in TE-encoded peptides). Thus, GBM differentially expressed LINE elements are the major source of TE-derived peptides presented on HLA-I in GBM.

[0460] TEs in each category are divided into families and subfamilies. The evolutionary "age" of these subfamilies can be estimated by the degradation of their characteristic repeat motifs (Choudhary et al., Genome Biol 2020, 21, 16). Several of the newest subfamilies include TEs that encode complete viral protein ORFs, some of which can still be "active" in terms of retrotransposition (Burns, Science 2017, 348, 803-808; Rodic et al., Nat Med. 2015, 21, 1060-1064; Scott et al., Genome Res. 2012, 26, 745-755). Some peptides can be redundantly encoded by multiple TEs, a finding that may be due to the presence of conserved sequences in young TEs from the same TE subfamily. Therefore, the age of the TE subfamily of each peptide-encoding TE was analyzed. The median age of peptide-encoding SINE and DNA TEs is similar to all genomic TEs annotated in RepeatMasker and all expressed and differentially expressed TEs. For LTR, the proportion of younger TEs in peptide-encoding TEs increases (compared with other categories, the median age of peptide-encoding TEs decreases), but older TEs are also presented on HLA-I. For LINE and "other" (see below for a more detailed analysis of this category) categories, a bimodal distribution was observed, in which peptides encoded by TEs from young subfamilies (less than 50 million years old) were significantly enriched, which was rare in RepeatMasker, all expressed and tumor differentially expressed TEs. Therefore, in the LINE and LTRTE categories, the most recent TEs are more likely to provide peptides for HLA-I presentation.

[0461] Ancient viral proteins are the source of HLA-presenting peptides

[0462] We next investigated whether peptides from TEs were derived from annotated endogenous viral elements (EVEs) documented and validated in the gEVE database (Nakagawa and Takahashi, Database (Oxford) 2016). These EVEs consisting of at least 80 amino acids were identified processing RepeatMasker annotations and conserved known motifs from viral proteins such as Gag and Pol. Mapping peptide-encoding TEs to gEVEs showed that for both LINEs and LTRs, TEs mapping annotated EVEs were significantly enriched in peptide-encoding TEs (based on all assignments and single assignments) compared to RepeatMasker, all expressed, and differentially expressed TEs ( Figure 5). Consistent with these results, mapping of peptide-encoding TEs to their corresponding subfamilies showed selectivity for Alu in SINEs, L1PA / B / x and L2 in LINEs, ERV1, ERVK, ERVL and ERV-MaLR in LTRs, and SVA among others. Allowing one or two nucleotide mismatches (to account for possible mutations or polymorphisms) significantly increased the proportion of peptide-encoding TEs that mapped to annotated ORFs in gEVE, both classes and subfamilies, indicating that recently mutated TEs are also a major source of peptides for HLA-I presentation. Most peptides originated from ORFs carrying a start codon that was either ATG (classical) or CTG / GTG / TTG (non-classical).

[0463] Examples of peptides are the 3 peptides encoded in the SVA family member SVA_B_dup189. These 3 peptides are encoded on the positive strand in 2 different reading frames (RFs). The 2 peptides encoded in RF1 are present in ORFs longer than 30 amino acids, while the third peptide (encoded in RF3) was not found in any of the detected ORFs. This could be because the ORF is shorter than 30 amino acids, or the start codon of the ORF is not among the 4 ORFs used in the pipeline, or the start codon is outside of the TE. Analysis of the length of ORFs encoding HLA-presented peptides showed that in L1PA|B|x, the ORFs that generate peptides and contain the canonical ATG start codon are longer than those starting with the non-canonical ATG start codon, but this is not the case in other TE subfamilies. Among peptide-encoding TEs mapping to gEVE-annotated LTR ORFs, actual peptide-coding sequences can be found in all retroviral proteins and are enriched for Gag (10.6% of ORFs in gEVE and 28% in peptide-encoding TE ORFs). In the case of LINEs, Pol is the only gEVE-annotated protein. Blast of peptide-coding sequences showed that most LINE-encoded peptides do not originate from the two major LINE ORFs, ORF1p (3.1%) and ORF2p (10.8%). In summary, TEs from young subfamilies preferentially carry retroviral protein motifs and are more likely to provide peptides for presentation by HLA-I molecules in GBM cells. These peptides are encoded by ORFs carrying classical or alternative start codons and can be 10 to 1000 amino acids in length.

[0464] TE subfamilies share HLA-presented peptide coding sequences

[0465] To investigate whether certain TEs are more likely to provide HLA-binding peptides than others, the proportion of TE families differentially expressed in GBM (and used for peptide MS / MS searches) and the proportion of peptide-encoding TEs were compared. For LTRs, SINEs, and Other, the proportions of different families were similar in the GBM TE-signature and peptide-encoding TEs (both with all or single assignments). In contrast, for LINEs, peptides were preferentially derived from L1PA / B / x: 25.3% in the GBM TE-signature, and 76.6% or 49.7% for all and single assignments, respectively. Other LINE families were depleted in peptide-encoding TEs (particularly L2, which accounted for 25.1% of the GBM TE-signature but provided only 7.4% or 15.4% of peptide-encoding TEs, with all and single assignments, respectively). Statistical analysis showed that peptide-encoding TEs of L1PA / B / x and SVA, as well as ERVK, were significantly enriched compared to the GBM TE signature. In the RM category “Other”, XX was also enriched. Compared with the GBM TE-signature, L2, SINEs (including Alu and MIR), and ERVs (including ERVL and ERVL-MalR) were significantly depleted in peptide-encoding TEs ( Figure 6 ). L1PA / B / x includes L1HS (or L1PA1, one of the few TEs still active in humans) and its closely related subfamilies L1PA(x) and L1PB(x), which are relatively young subfamilies compared with other LINE-1 subfamilies. In summary, some recent TE families (mainly LINE-1) preferentially generate HLA-I presented peptides in GBM.

[0466] Since the most recent TEs have more conserved repeat motifs, the next attempt was to investigate whether the HLA-presented peptides encoded by multiple TEs correspond to shared subfamily motifs. The 152 TE subfamilies encoding 347 identified HLA peptides are represented in a two-dimensional diagram, and the intersections between the two subfamilies are colored according to the number of shared peptides (not shown). The green diagonal in the figure indicates that most subfamilies encode only one peptide. The red squares on the diagonal indicate that a TE subfamily can encode more than one peptide. The green squares outside the diagonal indicate that the peptides can be encoded by TEs from different subfamilies, while the red squares outside the diagonal indicate that two different subfamilies encode several shared peptides (up to 25). The category and age of the subfamily are indicated by a color scale on the side of the chart. The three main groups or redundant clusters of TE subfamilies encoding one or more peptides are shown as large squares and are enlarged. The first redundant cluster (upper left corner) corresponds to the L1HS and L1PA (x) groups, which are two young subfamilies of LINE-1 elements that share up to 25 peptides in pairs. The second cluster identifies relatively young SINE elements (mostly Alu) that share a single peptide (bottom right of the first cluster). The third cluster (bottom right of the zoomed-in panel) corresponds to a group of young subfamilies of SVA elements that share a variable number of peptides. Thus, redundancy occurs in multiple TEs from the same recently related subfamilies, each of which may encode multiple peptides presented on HLA-I molecules. Therefore, redundancy in peptide-encoding TEs is limited to a few recent TE subfamilies.

[0467] To further investigate the association between redundancy and age of TEs, the analysis was extended to all TEs in the genome (so far, redundancy was analyzed in the GBM TE-signature). Analysis of genomic TE redundancy showed that 49.46% of the 370 peptides identified by immunopeptidomics were encoded by only one TE in the genome (compared to 85.49% in the scRNAseqGBM TE-signature). Conversely, 15.95% of these peptides could be encoded by 201-13,500 TE occurrences in the genome. Each peptide was plotted according to the number of TEs that could potentially encode it and the age of the corresponding subfamily. In SINEs, Alu-derived peptides were highly redundant and derived from the most recent subfamily, while MIR-derived peptides were encoded by a single TE from an older subfamily. The same correlation was observed in LINE-1 peptides, where peptides derived from the young L1HS, L1PA(x), and L1PB(x) were encoded by multiple elements, while peptides from the older L2 and other L1 subfamilies were encoded by unique elements. A negative correlation (r = -0.61) was confirmed in all TE families between the number of TEs that could potentially encode a single peptide and the age of the corresponding TE subfamily. In summary, regardless of the TE class (LINE, SINE, LTR or DNA), subfamilies of young TEs carry shared (redundant) sequences that could encode the same HLA-I peptides, while peptides encoded by TEs from older, more degenerate subfamilies are largely derived from unique genomic sequences.

[0468] Single TE-encoded peptides have stronger tumor specificity

[0469] To investigate how the redundancy of TE-derived peptides affects tumor specificity, for each differentially expressed GBM TE from the scRNAseq dataset (not shown), its expression ratio in TCGA GBM samples versus all healthy tissues from GTEx is represented, with brown indicating higher expression in GBM and blue indicating the opposite. Unsupervised clustering of TEs identified two major groups of peptide-encoding TEs, Group 1 and Group 2, dominated by TEs overexpressed in GBM and GTEx, respectively. Group 1 (TEs overexpressed in GTEx) contained a higher proportion of LINEs and others (including all 23 peptide-encoding SVA elements), whereas Group 2 contained more LTRs and DNA transposons ( Figure 6 , right panel). In addition, group 1 contained the majority of redundant TEs (63.5%), while group 2 contained only 26.6% ( Figure 6 ). Consistently, the median age of TE in group 1 was much lower than that in group 2 ( Figure 7 These results suggest that non-redundant peptides from older TE subfamilies are more likely to be overexpressed in GBM compared with healthy tissues than TEs from younger subfamilies encoding redundant peptides.

[0470] We then asked whether tumor-specific TEs could be identified. Expression of the top 50 tumor-enriched peptide-encoding TEs in GBM and all GTEx healthy tissues was determined (as 90th percentile expression, left panel, and percentage of samples expressing above GBM median expression, right panel) (not shown). The most tumor-specific TEs came from different classes, but preferentially originated from ORFs containing classical start codons. Some of these TEs were expressed at varying levels in most GBM tumors and were undetectable in all or most GTEx healthy tissues, including the brain. For some of these TEs, in four patients in the scRNAseq dataset, more than 90% of cells expressing the TE were GBM tumor cells. In summary, a unique, non-redundant, peptide-encoding subset of TEs was highly tumor-specific and recurrent in cancer patients. These peptide-encoding non-redundant TEs represent interesting potential targets for immunotherapy.

[0471] discuss

[0472] Here, the inventors used a TE-centric proteomic approach to study HLA-I presentation of TE-derived peptides, searching for tumor-specific recurrent antigens. Two major innovative approaches were combined: i) the pipeline started with TE analysis of scRNAseq from total primary tumors, which can assign TE reads to GBM cells rather than hematopoietic or stromal cells; ii) the alignment of TE mapping was performed for individual TE occurrences, rather than for TE subfamilies as previously done (Kong et al., 2019). Differentially expressed TEs were also overexpressed in a cohort of 155 bulk RNAseq samples from GBM patients (TCGA) compared to all tissues (including brain tissue) from healthy donors (GTEx), validating this sc / individual TE transcriptome analysis. The tags showed a bias towards TEs encoded on chromosome 7, which are frequently amplified in GBM tumor cells, further validating this sc / individual TE strategy. TE tags were used to query an immunopeptidomics mass spectrometry database from 30 GBM primary tumors and cell lines. We identified a collection of 347 TE-derived peptides that were reliably characterized and had motif compliance to the HLA alleles of the corresponding samples. Analysis of these peptides, encoded by 568 TEs, revealed some new aspects of the biology of TE peptide presentation in GBM cells. However, not all of the identified peptides were derived from tumor-specific TEs. Further analysis of the peptide-encoding TEs could identify true tumor-specific individual TEs that actually provide HLA-presented peptides, thus providing a potential source of targets for immunotherapy.

[0473] This study mainly relied on scRNAseq mapping of TEs. Several recent papers have analyzed TEs on scRNAseq datasets, and even though some earlier studies (He et al., Nat Commun., 2021, 10, 5228. Shao and Wang, Genome Res., 2021, 3188-100) pointed out possible biases and limitations, there are now reliable processes and guidelines, which were followed in this study. These results also rely on different internal controls that support the robustness of these TE scRNAseq analyses. First, the results show that TEs expressed in tumor GBM cells are biased towards TEs encoded on chromosome 7, but not in other cell populations ( Figure 1 This corresponds to the known amplification of chromosome 7 in GBM and was also detected in coding genes. Second, the scRNAseq-based GBMTE-signature was overexpressed in the GBM bulk RNAseq patient cohort compared to healthy tissue ( Figure 3 C and Figure 3D). Similarly, HLA-I peptidomics performed with an in silico translated RNAseq database is delicate, with the potential for a large number of false positive identifications. Particular care was taken in validating these TE-derived peptides based on peptide length, recognition score, hydrophobicity / RT correlation analysis, and binding motif compliance. In addition, 23 of the peptides were validated using synthetic peptide comparisons. Importantly, the identified peptides also showed the same chromosome 7 bias, which further and independently validated the identification results. An original finding was that the proportion of intronic and intergenic TEs occurring in peptide-encoding TEs increased, while 3′UTR TEs decreased, compared to the corresponding proportions in the GBM TE-signature (a database used to identify peptides). Therefore, HLA-I presented peptides can originate from both gene-dependent and gene-independent transcription and translation, but the reason why intronic TEs provide a higher proportion of peptides than 3′UTR TEs deserves further analysis. Previous studies have found that 3′UTR can encode HLA-presenting peptides (Laumont et al., Nature Communications, 2016, 7, 10238; Ruiz Cuevas et al., Cell Reports, 2021, 34, 108815; Zhao et al., Cancer Immunol Res, 2020, 8, 544-555), but these studies did not consider TEs from other genomic locations as done in this paper. The study also found that LINE-1 elements are the main source of HLA-I-presenting peptides in GBM. LINE-1 accounts for approximately 30% of TEs in the human genome, all TEs expressed in GBM, and GBM TE-signatures, but accounts for more than 50% of TE-encoded peptides presented on HLA-I. SVA-derived peptides were also strongly enriched, while the proportion of SINE-derived peptides was reduced (compared to the genome, expressed, and differentially expressed SINEs in GBM). LINE-1 elements with and without complete ORFs were preferentially represented in TEs that generated peptides, and this bias was observed regardless of whether the TE was assigned to multiple or a single position, indicating that the bias is not due to TE mapping issues.

[0474] Another conclusion of this study is that peptides presented by HLA-I molecules can be encoded by one or more redundant TEs (carrying exactly the same nucleotide sequence encoding the peptide). Other peptides are encoded by TE sequences that occur only once in the genome. In most cases, redundancy occurs within TE subfamilies and, in some cases, in different subfamilies that are always from the same TE class. The most redundant TEs (occurring hundreds to thousands of times) are from L1PA / B / x and often carry complete annotated ORFs. Peptides from Alu (a member of the SINE family), ERV1 (a member of the LTR family), and SVA (a medium-length independent family), which are all the youngest TE families in humans, also have high representation and redundancy. Redundancy is negatively correlated with the age of the TE subfamily, suggesting that the recurrent sequences encoding HLA-I binding peptides were part of an ancestral TE insertion event that was subsequently degraded by mutations and lost over time as subfamily members diverged. This scenario is supported by the observation that the number of redundant TEs is greater if 1 or 2 nucleotide mismatches are allowed. This is an interesting observation, as it is not yet clear whether the peptides identified by mass spectrometry originate from multiple or unique TE loci. The observation that many of these redundant TEs are actively transcribed and differentially expressed in GBM suggests that redundant peptides are indeed encoded by multiple loci.

[0475] Analysis of TE ORFs encoding peptides showed that peptides are generally encoded in ORFs between 10 and 100 amino acids long (with the exception of approximately half of LINE-encoded peptides, which are derived from longer ORFs). In LTRs, peptides were derived from all viral ORFs, with a positive bias towards env-derived peptides compared to the proportion of env genes annotated in the database. Among LINE-derived peptides, only a small fraction (approximately 10%) was derived from the known ORF1p and ORF2p loci. TE-encoding ORFs carry either classical or alternative start codons, with the exception of longer LINE1 ORFs (over 100 amino acids), which are all driven by the classical ATG start codon.

[0476] So, can this knowledge be used to identify tumor-specific TE-derived antigens? Analysis of the relative expression of individual peptide-encoding TEs in GBM tumors and a range of healthy tissues showed that redundant TEs from younger subfamilies are generally less tumor-specific than unique TEs from older subfamilies. Because their expression is more promiscuous, the immune system is likely to be more tolerant to antigens from these TEs (although this will require specific solutions). Therefore, although vaccination with LINE-1 complete ORF has been shown to be immunogenic and safe for mice and monkeys, redundant TEs may not be the best candidates for tumor-specific targets for immunotherapy (Sacha et al., Immunology, 2012, 189, 1467-1479). However, these results also identified unique peptide-encoding TEs, which preferentially come from the oldest subfamilies of MIR, LINE-1 and -2, and some of the ERVs. Most of these non-redundant peptide-encoding TEs come from relatively ancient TE subfamilies (more than 50 million years old), and tBLASTn analysis shows that some of these sequences have only appeared once in the genome. Some of these TEs are from subfamilies that are recurrently and selectively derepressed in tumors, primarily through local DNA demethylation (Brocks et al., Nature Genet, 2017, 49, 10512-1060; Chiappinelli et al., Cell, 2017, 169, 361; Lavie et al., Journal of Virology, 2005, 79, 876-883; Ohtani et al., Cancer Research, 2020, 80, 2441-2450; Roulois et al., Cell, 2015, 1162, 961-973; Sacha et al., Immunology, 2012, 189, 1467-1479). Virol, 2005, 79, 876-883; Ohtani et al., Cancer Res, 2020, 80, 2441-2450; Roulois et al., Cell, 2015, 1162, 961-973; Sacha et al., Immunol, 2012, 189, 1467-1479)). It has been shown that some of these peptide-encoding TEs expressed in most GBM tumors are not detected or are detected at low frequency and / or low levels in healthy tissues.

[0477] The results of in vitro stimulation with some of these TE-derived peptides indicate that there is a TCR repertoire of TEs in healthy individuals, providing the possibility that these TEs are immunogenic in patients. However, previous studies have shown that T cells are reactive to TEs expressed by tumors, thus proving the view that TEs (including ERVs) can be immunogenic in cancer patients (Saini et al., Nature Communications (Nat Commun) 2020, 11, 5660; Smith et al.; Wang-Johaning et al., Cancer Research (Cancer Res) 2008, 68, 5869-5877). In this case, mapping the expression of individual TEs in single cells and bulk RNAseq in cancer patients proved that individual TEs that produce HLA-I presented peptides can be effectively defined. The tumor specificity and high recurrence of these peptide-generating TEs open up new prospects for immunotherapy in many cancer types with de-repressed TEs and other immunopathologies of TE dysregulation.

[0478] Tables 2 to 4

[0479] Table 2 relates to the detailed identification of TEs in the tumor signature in this study, corresponding to transcripts of SEQ ID NOs: 381 to 5020. The column numbers refer to the following:

[0480] -Column 1: TE ID

[0481] - Column 2: Analysis

[0482] - Column 3: Peptides discovered by immunopeptidomics

[0483] - Column 4: Peptide ID

[0484] - Column 5: Specificity

[0485] - Column 6: TE category

[0486] - Column 7: TE type

[0487] - Column 8: Genomic coordinates

[0488] - Column 9: Chain

[0489] - Column 10: Age (millions)

[0490] - Column 11: Length

[0491] - Column 12: nearest gene

[0492] - Column 13: Gene proximity

[0493] - Column 14: distance to the nearest gene

[0494] - Column 15: TE with gEVE

[0495] - Column 16: TE transcript sequence SEQ ID NO:

[0496] According to conventional nomenclature, TE transcript sequences are disclosed herein as the DNA sequence corresponding to the coding DNA.

[0497]

[0498]

[0499]

[0500]

[0501]

[0502]

[0503]

[0504]

[0505]

[0506]

[0507]

[0508]

[0509]

[0510]

[0511]

[0512]

[0513]

[0514]

[0515]

[0516]

[0517]

[0518]

[0519]

[0520]

[0521]

[0522]

[0523]

[0524]

[0525]

[0526]

[0527]

[0528]

[0529]

[0530]

[0531]

[0532]

[0533]

[0534]

[0535]

[0536]

[0537]

[0538]

[0539]

[0540]

[0541]

[0542]

[0543]

[0544]

[0545]

[0546]

[0547]

[0548]

[0549]

[0550]

[0551]

[0552]

[0553]

[0554]

[0555]

[0556]

[0557]

[0558]

[0559]

[0560]

[0561]

[0562]

[0563]

[0564]

[0565]

[0566]

[0567]

[0568]

[0569]

[0570]

[0571]

[0572]

[0573]

[0574]

[0575]

[0576]

[0577]

[0578]

[0579]

[0580]

[0581]

[0582] Table 3 relates to the detailed identification of peptides derived from the tumor-TE signature by immunopeptidomics, corresponding to the neoantigenic peptides of SEQ ID NO: 1 to 370. These peptides are identified by their SEQ ID NO:; for example, PEP:0001 corresponds to the peptide of SEQ ID NO: 1 in the attached sequence listing.

[0583] Column numbers refer to the following:

[0584] - Column 1: Peptide ID

[0585] - Column 2: Analysis

[0586] - Column 3: TE category

[0587] - Column 4: TE type

[0588] - Column 5: Median age (millions)

[0589] - Column 6: Specificity

[0590] - Column 7: Number of genomic TEs encoding peptides

[0591] - Column 8: Peptide sequence

[0592] - Column 9: Group

[0593] - Column 10: Immunogenicity ID

[0594]

[0595]

[0596]

[0597]

[0598]

[0599]

[0600] Table 4 relates to the detailed recognition of immunogenic peptides derived from the tumor-TE signature by HLA-I binding prediction, corresponding to the neoantigenic peptides of SEQ ID NOs: 371 to 380. These peptides are identified by their SEQ ID NOs: ; for example, PEP: 0371 corresponds to the peptide of SEQ ID NO: 371 in the attached sequence listing.

[0601]

[0602] Table 4

[0603] Sequence Description:

[0604]

Claims

1. A method for identifying TE tags of tumor cells, comprising the following steps: i. obtaining a single-cell transcriptome TE pattern of at least one tumor cell and a single-cell TE transcriptome pattern of at least one normal cell, and ii. performing a differential expression analysis of the TE transcriptome profile obtained from the at least one tumor cell relative to the at least one normal cell, and iii. Selecting TE transcript sequences that are differentially expressed in the at least one tumor cell compared to the at least one normal cell, thereby obtaining a tumor cell TE signature.

2. The method according to claim 1, wherein In step i), the single-cell transcriptome pattern is obtained by mapping the single-cell transcriptome to a single genomic TE occurrence.

3. A method for identifying TE-derived tumor neoantigen peptides, the method comprising the following steps: a) obtaining a tumor cell TE tag according to the method of any one of claims 1 or 2, and b) computer simulation translation of the TE transcript sequence of the tumor cell TE tag obtained in step a) to obtain TE-derived tumor peptides.

4. The method according to claim 3, further comprising the step c) identifying TE-derived peptides bound to at least one MHC molecule; Optionally, wherein searching a library comprising the TE-derived peptide sequence identified in step b) against a set of MHC ligands from tumor cells, and wherein matching peptides are selected from said set of MHC ligands, thereby identifying MHC-bound TE-derived peptides; Optionally, wherein the TE-derived MHC binding peptides are further filtered against canonical proteins; and / or Optionally, wherein the selection is less than 10 -5 M's K D The binding affinity binds to a TE-encoded peptide that binds to at least one MHC class I molecule or MHC class II molecule of a subject.

5. The method according to claim 4, further comprising step d) selecting non-redundant TE-derived peptides; optionally, wherein this step is achieved by mapping the TE-derived peptides of step c) to individual TE genomic positions and selecting uniquely mapped TEs.

6. An isolated tumor neoantigen peptide sequence having at least 8 amino acids, wherein: The new antigen peptide comprises a TE coding sequence and has a length of less than 10 -5 M's K D binds with affinity to at least one MHC class I molecule or MHC class II molecule of a subject, The new antigenic peptide also has one or both of the following characteristics: - the expression of said TE is suppressed in tumor cells compared to non-tumor cells; - the peptide is encoded by a TE transcript sequence obtained according to any one of claims 1 or 2 or a fragment thereof; - the peptide is obtained by the method according to any one of claims 3 to 5; and / or - the peptide is encoded by a TE transcript of any one of SEQ ID NOs: 381 to 5020 or a fragment thereof; Optionally, the peptide comprises at least 8 amino acids, particularly 8 to 15, especially 8 to 12 amino acids and binds to at least one MHC class I molecule of the subject, or comprises 13 to 25 amino acids and binds to at least one MHC class II molecule of the subject.

7. The neoantigenic peptide according to claim 6, comprising or consisting of any one of SEQ ID NOs: 1 to 26 and 28 to 380 or a fragment thereof, optionally wherein the peptide is encoded by a single genomic TE.

8. The method according to any one of claims 1 to 5 or the neoantigenic peptide according to claim 6 or 7, wherein The tumor is a glioblastoma tumor.

9. The neoantigenic peptide according to any one of claims 6 to 8, wherein The TE has one or more of the following characteristics: - the TE is selected from more than 50×10 6 TE of the year; optionally, wherein the TE is selected from the LINE-1, SVA and ERVK TE subfamilies; optionally, wherein the TE is selected from the L1PA / B / x TE; - the TE is selected from more than 50×10 6 Year TE; - the TE is selected from TEs carrying a complete or nearly complete ORF; - the TE is selected from an intronic TE or an intergenic TE; -The TE is encoded by chromosome 7.

10. An autologous dendritic cell population or antigen presenting cell population which has been pulsed with one or more peptides as defined in any one of claims 6 to 9 or transfected with a polynucleotide encoding one or more peptides as defined in any one of claims 6 to 9.

11. A vaccine or immunogenic composition capable of eliciting a specific T cell response, comprising: - one or more neoantigenic peptides as defined in any one of claims 6 to 9; - one or more polynucleotides encoding a neoantigenic peptide as defined in any one of claims 6 to 9, optionally linked to a heterologous regulatory nucleotide sequence; and / or -Antigen presenting cell population as defined in claim 10.

12. An antibody or antigen-binding fragment thereof, T cell receptor (TCR) or chimeric antigen receptor (CAR) that specifically binds to a neoantigenic peptide as defined in any one of claims 6 to 9, optionally associated with an MHC molecule, K d The affinity is about 10 -6 M or smaller; Optionally, wherein the antibody is a multispecific antibody further targeting at least one immune cell antigen; Optionally, the immune cells are T cells, NK cells or dendritic cells; Optionally, wherein the target antigen is CD3, CD16, CD30 or TCR; Optionally, wherein the antibody is a multispecific antibody further targeting at least one immune cell antigen; Optionally, wherein the immune cell is a T cell, a NK cell or a dendritic cell, optionally, wherein the target antigen is CD3, CD16, CD30 or TCR; and / or Optionally, wherein the T cell receptor is made soluble and fused to an antibody fragment directed against a T cell antigen, optionally wherein the target antigen is CD3 or CD16.

13. A polynucleotide encoding the neoantigenic peptide defined in any one of claims 6 to 9 or the antibody, CAR or TCR defined in claim 12, or a vector comprising the polynucleotide.

14. An immune cell that specifically binds to one or more neoantigen peptides as defined in any one of claims 6 to 9; optionally, wherein the immune cell is an allogeneic or autologous cell selected from T cells, NK cells, CD4+ / CD8+, TIL / tumor-derived CD8 T cells, central memory CD8+ T cells, Treg, MAIT and YδT cells; and / or optionally, wherein the T cell comprises a T cell receptor that specifically binds to one or more neoantigen peptides as defined in any one of claims 6 to 9 or the TCR or CAR of claim 12.

15. A neoantigenic peptide as defined in any one of claims 6 to 9, The dendritic cell population according to claim 10, the vaccine or immunogenic composition according to claim 11, the antibody, antigen-binding fragment thereof, CAR or TCR as defined in claim 12, the polynucleotide or vector as defined in claim 13, or the immune cell according to claim 14, for use in treating cancer; optionally for inhibiting cancer cell proliferation, or for cancer vaccination therapy in a subject; optionally, wherein the cancer is glioblastoma.

Images