Methods for optimizing the selection of effective vaccine formulations

Abstract

This invention relates to a method for ranking immunotherapies composed of or encoding multiple potential human leukocyte antigen (HLA) ligands to provide optimal immunological treatment for patients. The method includes: matching a patient's HLA profile with the potential HLA ligands; and, based thereon and optionally based on the expression patterns of the HLA ligands in the patient, ranking the immunotherapies according to the probability that they contain a preselected number of true HLA ligands.

Description

Technical Field

[0001] This invention relates to the field of cancer immunotherapy. Specifically, it relates to improved means and methods for optimizing precision anti-cancer vaccination, wherein the vaccination targets the expression product of a genomic sequence that is not expressed or is expressed only to a very limited extent in normal tissues but is present in the cancerous tissues of an individual patient. This invention also relates to a cancer treatment method and a computer system. Background Technology

[0002] Traditionally, the treatment of malignant tumors in patients has focused on eradicating / removing malignant tissue through surgery, radiotherapy, and / or chemotherapy using cytotoxic or cell-inhibiting drugs, the dosing regimens of which are designed to kill malignant cells rather than non-malignant cells.

[0003] In addition to using cytotoxic drugs, more recent approaches focus on targeting specific biological markers in cancer cells to reduce the systemic adverse effects of classic chemotherapy. Monoclonal antibody therapies targeting cancer-associated antigens have proven quite effective in prolonging life expectancy in a variety of malignancies. Although monoclonal antibodies are successful drugs, due to their nature, they can only be developed to target known expression products present in multiple patients. This means that the vast majority of cancer-specific antigens cannot be addressed by such therapies, as many cancer-specific antigens are only present intracellularly or in the tumors of only a single patient (see below).

[0004] The theory of immune surveillance, proposed as early as the late 1950s, posits that lymphocytes recognize and eliminate their own cells (such as cancer cells exhibiting altered antigenic determinants). It is now widely accepted that the immune system largely suppresses cancer development. However, immune surveillance is not 100% effective, and developing cancer therapies aimed at improving / stimulating the immune system's ability to eradicate cancer cells remains an ongoing task.

[0005] One approach is to induce immunity against cancer-associated antigens, but despite its potential, this approach shares the same limitation as antibody therapy: it can only target a limited number of antigens.

[0006] Many (if not all) tumors express mutations. These mutations may generate novel targetable antigens (neoantigens), which, if identified within a clinically relevant timeframe, could potentially be used for specific T-cell immunotherapy. Using existing technologies, it is possible to perform full sequencing of the cell genome within days and analyze for alterations or new expression products, thus enabling the design of personalized vaccines based on neoantigens and their novel epitopes within that timeframe.

[0007] Several bioinformatics workflows exist for predicting / identifying novel epitopes from patient-derived sequencing data (see Hundal, J. et al., 2016; Hundal, J. et al., Cancer Immunol. Res. 2020 8(3):409-420. DOI: 10.1158 / 2326-60662020; Bjerregaard, AM et al., 2017; Bais, P. et al., 2017; Rubinsteyn, A. et al., 2017; Schenck, RO et al., 2019; Diao, K. et al., Int. J.Mol. Sci. 2022 23(19):11624. DOI: 10.3390 / ijms2319116242022). Each workflow considers different sets of characteristics when selecting or ordinating novel epitopes, highlighting that the problem of novel epitope selection remains unresolved.

[0008] WO 2022 / 023521 discloses a method for selecting epitopes for inclusion in personalized cancer vaccines; its focus is on identifying and utilizing novel epitopes encoded by somatic variants of genes expressed in cancer cells. Therefore, the method disclosed in WO 2022 / 023521 relies on identifying short peptides present in the expression product that differ from the patient's normal expression product, thus requiring a personalized assessment of the potential usefulness of such short peptides.

[0009] WO 2023 / 111306 discloses a cancer treatment method that relies on immunization against expression products of genomic sequences (typically endogenous viral elements (“EVEs”), such as endogenous retroviral sequences (“ERVs”), wherein the expression products are present in a small or negligible proportion of normal tissue samples.

[0010] To date, it appears that no one has provided a cancer treatment approach that combines targeting multiple expression products derived from EVE, where the selection of the targeted antigens is optimized relative to the individual patient. In other words, the existing technology has not yet provided any such approach: gaining the advantage of specifically targeting multiple novel epitopes (e.g., based on precision) when targeting EVE expression products. Summary of the Invention

[0011] Purpose of the invention

[0012] The purpose of this invention is to provide an optimized method for treating cancer patients with immunotherapy, and in particular, to provide a method for evaluating the suitability of each of a series of immunotherapies for an individual patient, said immunotherapies being tailored to contain a variety of potential HLA ligands. Invention Overview

[0014] The inventors have discovered that by rationally selecting immunotherapies from those composed of immunotherapies that contain or express a variety of potential HLA ligands, it may be possible to optimize the treatment of cancer patients.

[0015] Therefore, in a first aspect, the present invention relates to a method for selecting at least one immunomodulatory agent suitable for active specific immunotherapy of a patient's malignant tumor.

[0016] The at least one of the immunomodulators is part of a group of distinct immunomodulators, each comprising: a) an amino acid sequence capable of being expressed from genomic DNA in malignant tumor cells but only at a predetermined low level in cells of normal patient tissue, or b) a nucleic acid encoding the amino acid sequence.

[0017] The method includes:

[0018] 1) For each immunoassay agent, identify potential human leukocyte antigen (HLA) ligands that match the HLA profile of the patient, wherein the potential HLA ligands consist of an amino acid sequence contained in the immunoassay agent or an amino acid sequence encoded by the nucleic acid of the immunoassay agent, wherein each such potential HLA ligand exhibits a probability P1 of constituting a true HLA ligand in the patient;

[0019] 2) For each immunizing agent, determine a quantitative index of the probability P2 that at least X potential HLA ligands contained in or encoded by it are true HLA ligands, where X is a predetermined integer ≥2 and ≤Y, where Y is the maximum number of potential HLA ligands contained in or encoded by the immunizing agent; and

[0020] 3) Select from the immunizing agents:

[0021] a) One or more immunomodulators that exhibit the highest probability of having at least X potential HLA ligands as true ligands in step 2, or

[0022] b) One or more immunomodulators that exhibit the lowest predetermined probability that at least X potential HLA ligands are true HLA ligands in step 2.

[0023] In a second aspect, the present invention relates to a method for treating a human patient suffering from a malignant tumor, the method comprising: providing the patient with a human leukocyte antigen (HLA) profile, selecting at least one immunomodulator from a group of dissimilar immunomodulators according to any embodiment of the method according to the first aspect of the invention, and subsequently immunizing the patient once or multiple times with the immunomodulator so selected. Attached Figure Description

[0024] Figure 1: A schematic diagram of the expression plasmid selection method according to the present invention.

[0025] Figure 2 IFNγ ELISPOT in spleen cells.

[0026] For the A20 tumor model (A) and the CT26 tumor model (B), IFNγ ELISpot assays were performed on spleen cells from immunized BALB / c mice after restimulation with immune-related peptide pools, and IFNγ ELISpot assays were performed using plasmids designed for other tumors in another mouse strain (C57BL / 6) (C). Mice were immunized with plasmids designed to contain 20–24 mERV hotspots, and the mERV hotspots were divided into two pools, each containing 10–12 corresponding mERV hotspot peptides, for restimulation assays (i.e., peptide pool 1 and peptide pool 2). Data not shown: Mean SFU value of unstimulated cells (DMSO): (A) per 10 6 3.3 SFU per cell, (B) per 10 6 0 SFU per cell, and (C) per 10 6 Two SFUs per cell. This assay was performed using biological replicates (n=4 mice per group in Part A, n=6 mice per group in Part B, and n=5-6 mice per group in Part C). The bars represent the group mean. SFU: Spot-forming units.

[0027] Figure 3 : A graph showing the results of the A20 tumor study.

[0028] Each group of n=13 BALB / c mice underwent prophylactic immunization with 25 μg of personalized (A20) DNA plasmid before subcutaneous inoculation with A20 tumor cells on day 0, followed by electroporation. Mice were housed together with untreated tumor-bearing control mice, and tumor growth was compared with that of the untreated tumor-bearing control mice. The mean tumor growth curve for each group (in mm) is shown. 3 The standard error (SEM) of the mean (calculated) ± the mean. Detailed Implementation

[0029] definition

[0030] Endogenous retrotransposon elements ("EREs") are genetic elements. EREs comprise nearly 50% of the human genome. These elements are present in almost all organisms and are believed to be remnants of transposon elements that integrated into germline cells millions of years ago. Most ERE sequences contain mutated or truncated open reading frames and have lost their ability to transpose within the genome. EREs include short-scattered and long-scattered retrotransposon elements (SINEs and LINEs), collectively referred to as non-LTR elements. The remaining endogenous retrotransposon elements include LTR-binding elements, which comprise two main groups occupying comparable portions of the genome: endogenous retroviruses (ERVs) and mammalian dominant LTR retrotransposons (MaLRs) (Kassiotis & Stoye, 2016).

[0031] "Endogenous viral elements" ("EVEs") are a type of ERE that consists of members of a subset of genes obtained through a computer-simulated filtering process based on the presence of viral motifs in the genes. Therefore, this group primarily consists of EREs, but may also include members from other different subclasses.

[0032] A “novel or unannotated open reading frame” (nuORF) is a genomic sequence that is not typically the source of translation products, but immunopeptidomics analysis has revealed the presence of MHC-binding peptides derived from malignant tissues (Ouspenskaia T et al. 2021, Nature Biotechnology, doi.org / 10.1038 / s41587-021-01021-3).

[0033] "Malignant neoplasms" (also known as cancer or malignant tumors) refer to a group of cells in a multicellular organism that exhibit uncontrolled, invasive growth and often have the ability to metastasize.

[0034] "Cancer-specific" antigens are antigens that do not appear as expression products in an individual's non-malignant somatic cells, but do appear as expression products in that individual's cancer cells. This contrasts with "cancer-associated" antigens, which also appear in normal somatic cells (albeit at lower abundance), but are present at higher levels in at least some malignant tumor cells. Generally, peptides identified according to this invention are considered cancer-specific.

[0035] The term "adjuvant" has its common meaning in the field of vaccine technology, namely, a substance or combination of substances that 1) cannot elicit a specific immune response against the vaccine immunogen on its own, but 2) can still enhance the immune response against the immunogen. In other words, vaccination with an adjuvant alone does not provide an immune response against the immunogen, vaccination with the immunogen may or may not produce an immune response against the immunogen, but the immune response against the immunogen induced by combined vaccination with an immunogen and an adjuvant is stronger than the immune response induced by the immunogen alone.

[0036] MHC molecules (major histocompatibility molecules) are tissue antigens expressed by nucleated vertebrate cells that bind to peptide antigens and present (“present”) the antigen to T cells carrying T cell receptors. MHC class I molecules are expressed by all nucleated cells and primarily present proteolytically degraded protein fragments derived from intracellular proteins. MHC class II molecules are expressed by professional antigen-presenting cells, which typically take up extracellular proteins, degrade them with lysosomal proteases, and present the protein fragments to the cell surface. In humans, MHC molecules are encoded by human leukocyte antigen (HLA) sites; in this invention, HLA molecules are preferred for evaluating MHC binding.

[0037] A "T-cell epitope" is an MHC-binding peptide that is recognized as a foreign substance (not an autosome) by T cells in vertebrates due to the specific binding between T-cell receptors and cells carrying MHC-peptide complexes on their surfaces. Therefore, a peptide that constitutes a T-cell epitope in one individual may not be a T-cell epitope in different individuals of the same species. First, two individuals with different MHC molecules that bind different groups of peptides may not present the same MHC-complexed peptide. Furthermore, if a peptide is autosome-like in one individual, it may not bind to any T-cell receptors.

[0038] A "novel epitope" is an antigenic determinant (usually an MHC class I or II restriction epitope) that, due to the lack of a gene encoding it, is not present as an expression product in normal somatic cells of an individual, but is present as an expression product in mutant cells (such as cancer cells) of the same individual. Therefore, from an immunological perspective, although a novel epitope originates from the individual, it is truly non-self and can thus be characterized as a tumor-specific antigen in that individual, where it constitutes an expression product. Because it is non-self, a novel epitope has the potential to elicit a specific adaptive immune response in an individual, where the elicited immune response is specific to the antigen and cells containing that novel epitope. On the other hand, novel epitopes are individual-specific because the chance of the same novel epitope becoming an expression product in other individuals is extremely small. Therefore, several characteristics contrast novel epitopes with epitopes such as tumor-specific antigens: the latter are often present in multiple cancers of the same type (because they can be the expression product of activated oncogenes) and / or they will be present (albeit in trace amounts) in non-malignant cells because cancer cells overexpress the relevant genes.

[0039] A “novel peptide” is a peptide (i.e., a polyamino acid compound of up to about 50 amino acid residues) whose sequence contains a novel epitope as defined herein. Noval peptides are typically “natural,” meaning that the entire amino acid sequence of the peptide constitutes a fragment of the expression product that can be isolated from an individual. However, novel peptides can also be “artificial,” meaning that they consist of a novel epitope sequence and one or two additional amino acid sequences, at least one of which is not naturally associated with the novel epitope. In the latter case, the additional amino acid sequence can simply serve as a carrier of the novel epitope, or it can even enhance the immunogenicity of the novel epitope (e.g., by promoting the processing of the novel peptide by antigen-presenting cells, improving the biological half-life of the novel peptide, or altering its solubility).

[0040] The term "amino acid sequence" refers to the order in which amino acid residues linked by peptide bonds in the chain of peptides and proteins. Sequences are conventionally listed from the N-terminus to the C-terminus.

[0041] An "immunogenic carrier" is a molecule or part to which an immunogen or hapten can be conjugated to enhance or enable it to elicit an immune response against the immunogen / hapten. In the classic case, immunogenic carriers are relatively large molecules (such as tetanus toxoid, KLH, diphtheria toxoid, etc.) that can fuse or conjugate with immunogens / haptens that are insufficiently immunogenic to the body. Typically, immunogenic carriers elicit a strong T helper lymphocyte response against the complex of the immunogen and the carrier, which in turn provides an enhanced response from B lymphocytes and cytotoxic lymphocytes against the immunogen. In recent years, large molecular carriers have been partially replaced by so-called broad-spectrum T helper epitopes, which are shorter peptides recognized by most HLA haplotypes in the population that elicit a T helper lymphocyte response.

[0042] The “T helper lymphocyte response” is an immune response based on peptides that can bind to MHC class II molecules (such as HLA class II molecules) in antigen-presenting cells. This immune response stimulates T helper lymphocytes in animal species due to the complex between the T cell receptor recognition peptide and the MHC class II molecule that presents the peptide.

[0043] An "immunogen" is a substance that can induce an adaptive immune response in a host whose immune system encounters it. Therefore, an immunogen is a subset of a larger class of "antigens," which are substances that can be specifically recognized by the immune system (e.g., when bound by antibodies, or alternatively, when an antigen fragment bound to an MHC molecule is recognized by a T-cell receptor), but not necessarily induce immunity—however, an immunogen can always trigger immunity, meaning that a host that has developed memory immunity against an immunogen will elicit a specific immune response against that immunogen.

[0044] "Adaptive immune response" is an immune response that occurs in response to an encounter with an antigen or immunogen, wherein the immune response is specific to the antigenic determinants of the antigen / immunogen. Examples of adaptive immune responses include the induction of antigen-specific antibodies or the antigen-specific induction / activation of T helper lymphocytes or cytotoxic lymphocytes.

[0045] A “protective adaptive immune response” is an antigen-specific immune response induced in a subject as a reaction to immunization (artificial or natural) with an antigen, wherein the immune response protects the subject from subsequent attack by the antigen or a pathologically relevant factor containing the antigen. Typically, prophylactic vaccination aims to establish a protective adaptive immune response against one or more pathogens. In the context of this invention, the immune response induced by the identified peptide is typically a therapeutic immune response against a patient’s cancer.

[0046] "Immune system stimulation" refers to the general, non-specific immunostimulatory effect of a substance or combination of substances. Many adjuvants and putative adjuvants (such as certain cytokines) share the ability to stimulate the immune system. The result of using immunostimulants is increased immune system "alertness," meaning that the immune response induced by simultaneous or subsequent immunization with an immunogen is significantly more effective than that induced by the immunogen alone.

[0047] In the context of this invention, the term "peptide" is intended to refer to short peptides of 2 to 50 amino acid residues, oligopeptides of 50 to 100 amino acid residues, and polypeptides of more than 100 amino acid residues. Furthermore, the term is also intended to include proteins, i.e., functional biomolecules containing at least one polypeptide; when containing at least two polypeptides, they may form a complex, be covalently linked, or may be non-covalently linked. Polypeptides in proteins may be glycosylated and / or lipidized and / or contain prosthetic groups. An "HLA ligand" is a peptide defined by an amino acid sequence whose length and amino acid distribution allow it to bind to at least one HLA molecule.

[0048] "Hotspot sequences" refer to amino acid sequences that contain a high density of HLA ligand amino acid sequences, that is, a higher number of HLA ligand amino acid sequences per base pair than normal.

[0049] A “potential HLA ligand” (also known as a “predicted” HLA ligand) is a peptide defined by an amino acid sequence whose length and amino acid distribution allow it to bind to at least one HLA molecule in a patient.

[0050] "True HLA ligands" are peptides that bind to HLA molecules in a given patient and are presented by antigen-presenting cells.

[0051] Specific embodiments of the present invention

[0052] The first aspect of the present invention and its embodiments

[0053] A first aspect of the invention relates to a method for selecting at least one immunotherapy agent suitable for active, specific immunotherapy against a patient's malignant tumor, wherein the at least one immunotherapy agent is part of a group of distinct immunotherapy agents, each comprising: a) an amino acid sequence expressible from genomic DNA in malignant tumor cells but expressed only at predetermined low levels in cells of the patient's normal tissue, or b) a nucleic acid encoding the amino acid sequence.

[0054] The method includes: 1) for each immunoassay agent, identifying potential HLA ligands that match the HLA profile of the patient, wherein the potential HLA ligands are composed of an amino acid sequence contained in the immunoassay agent or an amino acid sequence encoded by a nucleic acid of the immunoassay agent, wherein each such potential HLA ligand exhibits a probability P1 of constituting a true HLA ligand in the patient; 2) for each immunoassay agent, determining a quantitative index of the probability P2 of containing or encoding at least X potential HLA ligands as true HLA ligands, wherein X is a predetermined integer ≥2 and ≤Y, where Y is the maximum number of potential HLA ligands contained or encoded by the immunoassay agent; and 3) selecting from the immunoassay agents: a) one or more immunoassay agents that exhibit the highest probability of at least X potential HLA ligands as true ligands in step 2, or b) one or more immunoassay agents that exhibit the lowest predetermined probability of at least X potential HLA ligands as true HLA ligands in step 2.

[0055] Therefore, this invention relates to a situation where a patient with a malignant tumor can be treated with a vaccine (alone or as part of a combination therapy) capable of actively inducing specific immunity against a malignant tumor antigen. The group of immunotherapies is pre-prepared, each immunotherapy in the group being designed to incorporate a series of potential HLA ligands by including such potential HLA ligands that bind to multiple HLA molecules in distant subgroups (such as ethnic groups or disease subgroups). By performing the method of the first aspect, the immunotherapies in the group are sorted to identify those most likely to induce a beneficial immune response in the patient by matching the patient's HLA profile with the ability of the immunotherapy's potential HLA ligands (or, if the immunotherapy is a nucleic acid-based vaccine, the potential HLA ligands encoded by the immunotherapy) to be presented by the patient's HLA molecules. In addition to simply matching HLA ligands to the HLA profile, it is also valuable to further consider the expression levels of the genetic material encoding the potential HLA ligands in the patient.

[0056] As previously stated, the at least one immunomodulator may be at least one expression vector, such as at least one plasmid vector or viral vector, or a composition comprising the at least one expression vector. Expression vectors are preferred because they allow for the design of very compact coding sequences containing high-density coding regions for potential HLA ligands. The composition of the potential HLA ligands encoded by the expression vector is preferably such that the at least one expression vector: 1) encodes a polypeptide containing multiple potential HLA ligands capable of being expressed from DNA in malignant tumor cells but only at predetermined low levels in cells of normal patient tissue, or 2) encodes multiple peptides collectively containing the multiple potential HLA ligands. This selection of potential HLA ligands ensures that immunization with the vector will exhibit a low probability of inducing undesirable adverse events by inducing an immune response against normal patient cells.

[0057] The at least one immunoassay agent may also be a polypeptide or a group of peptides, or a composition comprising a polypeptide or a group of peptides. The considerations for selecting a potential HLA ligand contained in such immunoassay agents are the same as those for selecting a potential HLA ligand encoded by the expression vector described above.

[0058] As described above, when evaluating this group of immunostimulants, each potential HLA ligand is preferably an empirically proven expression product derived from malignant tumor cells. Therefore, in this embodiment of the first aspect of the invention, even a potential HLA ligand exhibiting a high probability of P1 and thus contributing to a high probability of P2 will be ignored if the source expression product of that HLA ligand is not expressed in the patient's malignant cells. Similarly, if an immunostimulant is ultimately found to not contain or express an HLA ligand found as an expression product in the patient, that immunostimulant will be ranked lower among the immunostimulants.

[0059] To assign a probability P2, it must be greater than 0. This means that the selection of the number X among the at least X potential HLA ligands used for evaluation in steps 2 and 3 will allow for a meaningful ranking of the individual immunostimulants. For example, if the value of X is set too low, the probability that all immunostimulants contain X true HLA ligands will appear so high that they cannot be ranked. Similarly, if the value of X is chosen too high, the probability that all immunostimulants contain X true HLA ligands will appear to be zero. Therefore, the number X is determined empirically based on the ranking of immunostimulants (which in turn depends on the number and composition of potential HLA ligands in or encoded by the immunostimulant). However, X is typically set to a value such that the probability P2 determined in step 2 is at least 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 potential HLA ligands are true ligands.

[0060] As shown in Example 1, the probability of the presence of each of at least 1 to Y true HLA ligands contained in or encoded by an immunizing agent can be conveniently determined for each immunizing agent by simulation or precise calculation, thus allowing for the determination of the quantitative index in step 2. Also as described in the examples, the quantitative index for each immunizing agent can conveniently be expressed as the area under the curve (AUC) of the determined probability, but it is also possible to express the quantitative index for each immunizing agent as the probability that at least Z potential HLA ligands are true HLA ligands, where Z is a predetermined integer >1, preferably predetermined to be 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 or even higher. Similarly, as described above with respect to X, this Z value must be selected to provide a meaningful evaluation of the immunizing agent.

[0061] Preferably, the amino acid sequence that can be expressed from genomic DNA in malignant tumor cells but only at a predetermined low level in cells of normal tissue in patients is an amino acid sequence encoded by normally non-coding DNA, such as an endogenous viral element (EVE; particularly human endogenous retrovirus (hERV)), but the amino acid sequence may also be an amino acid sequence of a tumor-associated antigen (TAA), a tumor-specific antigen (TSA), or a neoantigen.

[0062] However, preferably, an EVE is an amino acid sequence that can be expressed from genomic DNA in malignant tumor cells but is expressed only at a predetermined low level in the cells of normal tissue in a patient, especially those EVEs that contain or consist of DNA sequences of human endogenous retroviruses (hERV).

[0063] The second aspect of the present invention and its embodiments

[0064] This aspect relates to a method for treating a human patient with a malignant tumor, the method comprising: providing the patient with a human leukocyte antigen (HLA) profile; selecting at least one immunotherapy agent from a group of dissimilar immunotherapy agents according to any embodiment of the method according to the first aspect of the invention; and subsequently immunizing the patient once or multiple times with the so-called selected immunotherapy agent. As mentioned above, this treatment may be the patient's sole treatment, but it is typically part of a combination therapy, wherein the patient is further treated with at least one additional cancer therapy, such as radiation therapy, cell-inhibiting drug therapy, cytotoxic drug therapy, immune checkpoint inhibitor therapy, immunomodulatory effector therapy, chimeric antigen receptor (CAR)-T therapy, adoptive T-cell therapy, ablation therapy, targeted therapy, tyrosine kinase inhibitor therapy, and one or more of cancer immunotherapies targeting surface expression determinants.

[0065] Example 1

[0066] ERV-based plasmid constructs were selected for vaccination of model patients.

[0067] Overall, the schematic diagram of this method is as follows: Figure 1 As shown.

[0068] Step 1

[0069] HLA typing of patients according to existing technical procedures: traditional serological methods can be used, but methods based on PCR or next-generation sequencing (NGS) are preferred.

[0070] Useful PCR methods include those using any of the following: 1) Sequence-specific oligonucleotides (SSO), which use fluorescently labeled oligonucleotide probes that hybridize to a specific HLA sequence to allow detection and analysis of the resulting fluorescent signal to determine the HLA type; 2) Sequence-specific primers (SSP), which use sequence-specific primers to amplify a target HLA gene by PCR. The presence or absence of the amplification product indicates the HLA type; and 3) Sequence-based genotyping (SBT), which involves PCR amplification of the HLA gene followed by direct sequencing of the amplicon.

[0071] Next-generation sequencing (NGS) can sequence millions of DNA fragments simultaneously, allowing for comprehensive and efficient HLA genotyping. NGS-based HLA genotyping can be performed using targeted gene sequencing or whole-genome sequencing.

[0072] In this embodiment, the patient was found to exhibit the following HLA class I type: HLA-A 03:01, HLA-A 11:01, HLA-B 35:37, HLA-B 52:01, HLA-C 12:02 and HLA-C 04:01, and the following HLA class II types: HLA-DRB1 01:01 and HLA-DRB1 04:04.

[0073] In addition to HLA typing of the patient, the following additional steps can be taken:

[0074] Tumor biopsy samples can be obtained from patients, and the mRNA within them can be sequenced. The mRNA reads are then mapped to a human reference genome containing ERV sequences, for example using the STAR program (see Alexander Dobin et al. (2013), Bioinformatics 29(1): 15-21, doi: 10.1093 / bioinformatics / bts635). The mRNA expression value (transcripts per million, TPM) is then determined using the RSEM program (see Bo Li and Colin N Dewey (2011), BMC Bioinformatics 12, article number: 323; Brian J Haas et al. (2013), Nature Protocols 8, 1494-1512), and ERVs with TPM values ​​above a selected threshold (e.g., TPM>1) are counted as expressed. Furthermore, germline and somatic variants can be identified from the RNA data (or optionally, DNA data). ERV oligomers that match the construct sequence 100% and are present in expressed ERVs are counted as expressed oligomers.

[0075] Step 2

[0076] Each plasmid construct in the selection set contains a certain number (e.g., >1) of "hotspot sequences" of variable length to obtain a certain (relatively high) number of predicted HLA ligands per base pair. These hotspot sequences are extracted from one or more ERV DNA sequences. In this example, 15 plasmids have been designed, each containing 5-25 hotspot sequences. The following are some hotspot sequences encoded on an example plasmid containing 20 hotspot sequences:

[0077] Hotspot sequence 1 (containing ligands derived from ERV transcript ID Hsap38.chr12.79864930.79865697.+): KLSLFTDDKI VYLQNPIVSA PNLLKLISNF SKFSGYKINV QKSQASLYTK (SEQ ID NO: 1), wherein the predicted ligands are present throughout the sequence: FTDDKIVYL (SEQ ID NO: 2), IVSAPNLLK (SEQ ID NO: 3), up to and including KSQASLYTK (SEQ ID NO: 4).

[0078] Hotspot sequence 2 (containing a ligand derived from ERV transcript ID Hsap38.chr12.79864930.79865697.+): LEAFHLKTAT RQGSPLSSLL FNIVLEVLAR AIRQEKAIKR IQIGREEVKL SLFTDDKIVY L (SEQ ID NO: 5), wherein the predicted ligand is present throughout the sequence: FTDDKIVYL (SEQ ID NO: 2).

[0079] Hotspot sequence 3 (containing ligands derived from ERV transcript ID Hsap38.chr12.43978860.43979192.-): HTLNDYQKLL GNINWLRPSL NITTDKLQNL FSIPKGNTTL DSL (SEQ ID NO: 6), wherein the predicted ligands are present throughout the sequence: YQKLLGNI (SEQ ID NO: 7), KLQNLFSIPK (SEQ ID NO: 8), up to and including SIPKGNTTL (SEQ ID NO: 9).

[0080] Hotspot sequence 20 (containing ligands derived from ERV transcript IDs Hsap38.chr1.167319501.167319884.-, Hsap38.chr3.185014621.185015133.-, Hsap38.chr3.185108708.185109298.+ and Hsap38.chrY.13257234.13257596.+): NAPCYTSQAL (SEQ ID NO: 10), wherein the predicted ligands are present throughout the sequence: NAPCYTSQAL (SEQ ID NO: 10).

[0081] In this embodiment, it was found that ERV transcripts corresponding only to the aforementioned hotspot sequences 1, 2, and 20 were not expressed in the model patients.

[0082] Step 3

[0083] Each ligand contained in the construct can bind to multiple HLA alleles. Furthermore, due to the high redundancy in the ERV, each ligand may exist in more than one ERV hotspot contained in the plasmid. Therefore, a list of potential ligand-HLA pairs is first obtained based on the hotspot sequences in the construct and the patient's HLA profile.

[0084] Next, we obtain the probability that a ligand is a true ligand for the entire list of potential ligands, as shown in the table below for some potential ligands:

[0085]

[0086] In the table above, the ligands in the last two rows are not expressed in the patients' RNA data.

[0087] Optionally, the ligand list can be limited to ERVs expressed only in patients, i.e., only ligands #1-5 in this embodiment. It is also possible to combine ERV expression values ​​with probability scores to provide combined HLA ligands and expression probabilities, as described in WO 2022 / 023521 A2. Here, the ligand probability p(L) is defined as the probability that a peptide is presented on the cell surface (i.e., that the peptide is a true ligand) by taking into account the probability that the peptide is a true MHC ligand and the expression of the source protein containing that peptide (e.g., by measuring the mRNA of said source protein). If expression is included in the ligand probability, the “Expression” column in the table above is irrelevant.

[0088] The construct in this embodiment contains 7 unique peptides: HLA pairs, 5 of which are expressed in tumor biopsy samples.

[0089] Step 4

[0090] Selecting one or more of the best plasmid constructs from the set depends on sorting the plasmid constructs based on simulation or exact calculation:

[0091] Sorting is performed through simulation.

[0092] For each ligand encoded by the plasmid construct, a random number between 0 and 1 is sampled. If this number is less than the predicted probability that the potential ligand encoded by the plasmid is the true ligand in the patient, the potential ligand is marked as a "hit" and the hit count is counted.

[0093]

[0094] In the table above, the number of "hit" numbers is 3 for the 5 assigned random numbers.

[0095] The process of sampling random numbers and labeling peptides as hits is repeated a large number of times (e.g., 10,000 times) to collect statistics on the predicted ligand pool for the patient; the statistics are then calculated as the proportion of rounds that produce n or more hits, where n is an integer ≥0 and ≤ the number of potential ligands evaluated. For example (in a simplified version), performing 10 rounds of experiments on the above 5 ligands might result in a hit count of [0,1,1,1,2,2,2,2,3,3], producing the following hit distribution:

[0096]

[0097] The values ​​in the “Proportion Column” can be interpreted as the respective probabilities that the plasmid construct contains at least n true HLA ligands.

[0098] Such tables can be summarized using area under the curve (AUC) calculations (AUC=1.7 in this example), or by indicating an n value that exceeds a certain threshold (n value would be 1.5 in the simplified case above with only 5 potential ligands if the threshold is 0.75), or by indicating a proportion at a given n value (the proportion would be 0.6 in the simplified embodiment above if n value = "≥2").

[0099] Finally, the plasmid constructs are then sorted according to the same principles (AUC, threshold-related n value, or proportion at a specific n), and one or more plasmid constructs exhibiting the highest AUC value, n value, or proportion are selected.

[0100] Sorting is performed through precise calculations.

[0101] This sorting can be performed as follows, where a table summarizes the exact probabilities of n or more ligands:

[0102] First, populate an n×n probability table, where n is the number of potential ligands + 1. Each cell represents the probability of observing exactly j "true ligands" when testing 0 to i ligands. The table is initially filled with 1.0 in [i=0, j=0] (i.e., it is 100% certain that there are 0 true ligands in 0 trials), and with 0.0 for i>j (it is impossible to find i true ligands in j trials when i>j):

[0103]

[0104] Then, column j=0 (no ligand is a true ligand) can be filled by multiplying the probability that the ligand is not a true ligand by the probability that other ligands are also not true ligands, for example:

[0105] D[0, 1] = 1.0 × (1 – 0.394) = 0.606,

[0106] D[0, 2] = 0.606 × (1 – 0.539) = 0.279,

[0107] D[0, 3] = 0.279 × (1 – 0.448) = 0.154

[0108] D[0, 4] = 0.154 × (1 – 0.469) = 0.082

[0109] D[0, 5] = 0.082 × (1 – 0.468) = 0.044

[0110]

[0111] The remaining cells can be filled with the following sum:

[0112] 1) The probability that j true ligands exist in i-1 trials, therefore the i-th trial must fail; and 2) The probability that j-1 true ligands exist in i-1 trials, therefore the i-th trial must succeed.

[0113]

[0114]

[0115] Finally, using the last row of the table, we calculate the probability that n or more potential ligands are true ligands by summing from k to n, in this example:

[0116] The probability that ≥1 potential ligand is a true ligand:

[0117] 0.192 + 0.333 + 0.288 + 0.123 + 0.021 = 0.956

[0118] The probability that ≥2 potential ligands are true ligands:

[0119] 0.333 + 0.288 + 0.123 + 0.021 = 0.765,

[0120] The probability that ≥3 potential ligands are true ligands:

[0121] 0.288 + 0.123 + 0.021 = 0.431

[0122] The probability that ≥4 potential ligands are true ligands:

[0123] 0.123 + 0.021 = 0.144, and

[0124] The probability that 5 potential ligands are true ligands:

[0125] 0.021.

[0126] Similarly, a set of calculations is performed for each plasmid construct to select one or more plasmid constructs that exhibit the highest computational probability of being the true ligand in the patient, with at least n (pre-selected threshold) encoded potential ligands.

[0127] Example 2

[0128] Immunization / Attack Experiment

[0129] plasmid DNA vaccine design

[0130] Three different design schemes have been used to develop mouse endogenous retrovirus (mERV)-based vaccine designs:

[0131] - "Personalized solutions," where vaccines are customized to target specific tumors.

[0132] - "Precision therapy," in which vaccines target tumor subgroups, and finally,

[0133] - The "shared approach" targets the entire tumor population. In this test case, four mouse tumor cell lines, A20, CT26, C1498, and B16, were selected to represent the entire tumor population.

[0134] Two tumor subpopulations were selected for the mouse strains based on tumor origin: A20 and CT26 from BALB / c, and C1498 and B16 from C57BL / 6. A total of seven vaccines were designed: four personalized vaccines (one for each tumor cell line), two precision vaccines (one for each mouse strain), and one shared vaccine covering all four mouse tumor cell lines.

[0135] The vaccine was designed based on ERV (mERV) expression in tumor-bearing mice and the MHC type of the selected mouse strain. For the BALB / c (H2-K) mouse strain... d,H2-D d, H2-L d and H2-IA d) and C57BL / 6 (H2-K b, H2-D b and H2-IA (b) MHC morphology was readily available online, while tumor mERV expression was estimated based on internal RNA sequencing data. Using STAR version 2.7.10b, tumor RNA sequencing data from four mouse tumor cell lines—A20, CT26, C1498, and B16—were mapped to the reference genome GRCm38, using gene annotations from Ensembl release 102, including mouse endogenous retrovirus (mERV) annotations from gEVE database version 1.1. For each of the mouse tumors A20, CT26, C1498, and B16, mERV expression was quantified in terms of transcripts per million (TPM) using the RNA sequencing mapping results and RSEM version 1.3.1.

[0136] Personalized vaccine design for each mouse tumor was developed by extracting all mERV sequences expressed by that tumor. Ligands for the relevant mouse strain MHC type (BALB / c for A20 and CT26, and C57BL / 6 for C1498 and B16) in the mERV sequences were predicted using EvaxMHC4 (an in-house developed MHC ligand prediction tool). MHC ligand-rich mERV subsequences, hereinafter referred to as “hotspots,” were extracted and sorted by their ranking score (in the form of area under the curve (AUC), see the description of ranking by simulation in Example 1). An iterative optimization process was run to select the set of hotspots that shared the highest probable score. The vaccine plasmid product was created by constructing a DNA insert containing the set of hotspots identified during the optimization process and inserting it into the DNA plasmid backbone.

[0137] In the development of precision vaccine design and shared vaccine design, the same procedure was followed with minor modifications. For precision vaccine design, mERV sequences expressed in tumor subpopulations were extracted (from BALB / c tumor populations A20 and CT26, and from C57BL / 6 tumor populations C1498 and B16), and MHC ligand prediction was performed for each tumor subpopulation using its corresponding mouse strain's MHC type. For shared vaccine design, mERV sequences expressed in any of the tumors A20, CT26, C1498, and B16 were used, and MHC ligand prediction was performed using both BALB / c and C57BL / 6 MHC types.

[0138] The scores for each developed vaccine design in each mouse tumor have been calculated and are shown in the table below:

[0139]

[0140] The ranking score for each vaccine design against each mouse tumor. The "-" marker indicates that the vaccine design does not include the presence of... Cases where scores cannot be calculated due to the presence of MHC ligands in mouse tumors.

[0141] This score is the AUC determined based on the predicted probability of each MHC ligand in the hotspot set, see Example 1.

[0142] As expected, the highest score for each mouse tumor was for its personalized design, and the second highest score was for the precise design containing a given mouse tumor. Interestingly, the personalized designs also showed reasonable scores for mouse tumors derived from the same mouse strain. This can be explained by the fact that both designs were optimized for the same MHC type, coupled with slight overlap in mERV expression. Finally, the shared designs showed reasonable scores for all mouse tumors targeted in their development.

[0143] In vivo study design

[0144] BALB / c mice (8-12 weeks old) were prophylactically immunized weekly in the left and right tibialis anterior muscles (intramuscular) with 25 μg of plasmid DNA vaccine (50 μl per leg), followed by electroporation (EP). Five immunizations were administered, starting two weeks prior to tumor cell inoculation (i.e., on days -14, -7, 1, 7, and 14). On the day of tumor cell inoculation (defined as study day 0), in vitro expanded CT26 or A20 cells were harvested from the culture flasks by trypsin digestion (CT26) or pipette collection (A20) and washed in serum-free medium. The dosage of CT26 cells per mouse was 2 × 10-10. 5 100 μl of culture medium per cell; 2 × 10⁶ cells per mouse inoculated with A20 cells. 6 100 cells / 100 μl culture medium. Tumor cells were subcutaneously (sc) injected into the right ventral region of mice. After tumor establishment, the tumor diameter was measured three times a week using a digital caliper. Tumor volume was calculated using the following formula:

[0145]

[0146] Where d1 and d2 are the orthogonal diameters of the tumor. Mice were euthanized by cervical dislocation when most tumors in the control group reached the maximum permissible size of 15 mm in diameter in any direction, or when the humane endpoint was reached.

[0147] After euthanasia, spleens were isolated from 7 mice in each group. Spleens were collected in cold RPMI supplemented with 10% FCS, followed by treatment with GentleMACS (Miltenyi Biotec, tube C #130-096-334 and separator #130-093-235) and a 70 mm filter (Corning, CLS431751) to obtain a single-cell suspension. Spleen cells were cryopreserved in FCS containing 10% DMSO (Merck, #D8418).

[0148] Peptide restimulation and IFNγ enzyme-linked immunospot assay (ELISpot)

[0149] The PVDF membrane plate (Merck Millipore, #MAIP4510), which was first activated with 35% v / v ethanol, was coated overnight with 5 μg / ml anti-IFNγ capture antibody (BD, #51-2525KZ, 1:200). 5 × 10⁶ membranes were deposited per well. 5Spleen cells were collected and stimulated with or without stimulation (DMSO) at a total volume of 200 μl in R10 medium with 5 mg / ml synthetic peptide (purchased from Pepscan, Lelystad, Netherlands). Cells were incubated overnight at 37°C and 5% CO2. To detect IFNγ-secreting cell spots, anti-IFNγ detection antibody (BD, #51-1818KA, 1:250), streptavidin-HRP enzyme (BD, #557630), and AEC chromogenic substrate (BD, #551951) were applied sequentially according to the manufacturer's instructions. ELISpot plates were imaged, and IFNγ spots were counted using an ELISpot reader (Cellular Technology, Ltd).

[0150] result

[0151] Vaccination with DNA plasmids encoding the most specific mERV epitope hotspots for a given tumor cell line elicits the strongest immune response.

[0152] Spinal cells from BALB / c mice immunized with personalized, precise, and shared mERV hotspot DNA plasmids showed varying degrees of response to immune-related peptides, as indicated by IFNγ ELISpot. Figure 2 In both A20 and CT26 tumor environments (tumors derived from BALB / c mice, thus haplotype-matched with BALB / c), when restimulated with their homologous peptides, the highest-scoring personalized mERV hotspot-designed vaccine correspondingly generated the highest level of immune response compared to precisely designed vaccines and shared-design vaccines. Figure 2 (A and 2B).

[0153] When BALB / c mice were immunized with a DNA plasmid designed to contain mERV hotspots associated with a tumor cell line derived from a different haplotype mouse strain (C57BL / 6 mice), subsequent peptide restimulation and IFNγ ELISpot assays showed no immune response against the peptide. Figure 2 C). This observation corresponds well with the low scores of these vaccines.

[0154] Vaccination with DNA plasmids encoding mERV epitope hotspots can inhibit tumor growth.

[0155] Compared with untreated tumor-bearing control mice, prophylactic immunization of BALB / c mice with a personalized (A20) DNA plasmid resulted in a reduction in mean tumor volume. Figure 3 This effect on tumor growth and the lower terminal tumor volume confirm the concept of mERV hotspot selection and formulation as an anticancer vaccine.

Claims

1. A method for selecting at least one immunomodulator, said immunomodulator being suitable for active specific immunotherapy against a patient's malignant tumor. The at least one of the immunomodulators is part of a group of distinct immunomodulators, each comprising: a) an amino acid sequence capable of being expressed from genomic DNA in malignant tumor cells but only at a predetermined low level in cells of normal patient tissue, or b) a nucleic acid encoding the amino acid sequence. The method includes: 1) For each immunoassay agent, identify potential human leukocyte antigen (HLA) ligands that match the HLA profile of the patient, wherein the potential HLA ligands consist of an amino acid sequence contained in the immunoassay agent or an amino acid sequence encoded by the nucleic acid of the immunoassay agent, wherein each such potential HLA ligand exhibits a probability P1 of constituting a true HLA ligand in the patient; 2) For each immunizing agent, determine a quantitative index of the probability P2 that at least X potential HLA ligands contained in or encoded by it are true HLA ligands, where X is a predetermined integer ≥2 and ≤Y, where Y is the maximum number of potential HLA ligands contained in or encoded by the immunizing agent; and 3) Select from the immunizing agents: a) One or more immunomodulators that exhibit the highest probability of having at least X potential HLA ligands as true ligands in step 2, or b) One or more immunomodulators that exhibit the lowest predetermined probability that at least X potential HLA ligands are true HLA ligands in step 2.

2. The method of claim 1, wherein the at least one immunomodulator is at least one expression vector, such as at least one plasmid vector or viral vector, or a composition comprising the at least one expression vector.

3. The method according to claim 2, wherein the at least one expression vector: - Encoding a polypeptide containing multiple potential HLA ligands, which can be expressed from DNA in malignant tumor cells but only at predetermined low levels in cells of normal patient tissues, or - Encodes multiple peptides that collectively contain the aforementioned potential HLA ligands.

4. The method according to claim 1, wherein the at least one immunomodulator is a polypeptide or a group of peptides, or a composition comprising the polypeptide or a group of peptides.

5. The method according to any one of the preceding claims, wherein each potential HLA ligand is further an empirically proven expression product derived from malignant tumor cells.

6. The method according to any one of the preceding claims, wherein the probability P2 must be > 0, and the at least X potential HLA ligands are at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least 21, at least 22, at least 23, at least 24, at least 25, at least 26, at least 27, at least 28, at least 29, or at least 30 potential HLA ligands.

7. The method according to any one of the preceding claims, wherein the probability of the presence of each of at least 1 to Y true HLA ligands contained in or encoded by the immunizing agent is determined by simulation or precise calculation, and a quantitative indicator in step 2 is determined for each immunizing agent.

8. The method of claim 7, wherein the quantitative index of each immunizing agent is expressed as the area under the curve (AUC) of the determined probability.

9. The method of claim 7, wherein the quantitative index of each immunoassay agent is expressed as the probability that at least Z potential HLA ligands are true HLA ligands, wherein Z is a predetermined integer >1, and wherein Z is preferably predetermined to be 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30.

10. The method according to any one of the preceding claims, wherein the amino acid sequence expressed from genomic DNA in malignant tumor cells but only at a predetermined low level in cells of normal patient tissue is an amino acid sequence encoded by normally non-coding DNA, preferably an endogenous viral element (EVE) such as an endogenous retrovirus (ERV), or an amino acid sequence of a tumor-associated antigen (TAA), tumor-specific antigen (TSA), or neoantigen.

11. The method according to any one of the preceding claims, wherein the amino acid sequence expressed from genomic DNA in malignant tumor cells but only at a predetermined low level in cells of normal patient tissue is a DNA sequence comprising human endogenous retrovirus (hERV) or an EVE consisting of a human endogenous retrovirus (hERV) DNA sequence.

12. A method for treating a human patient suffering from a malignant tumor, the method comprising: Provide the patient with a human leukocyte antigen (HLA) profile, select at least one immunizing agent from a group of different immunizing agents according to the method of any one of the preceding claims, and subsequently immunize the patient once or multiple times with the so-called selected immunizing agent.

13. The method of claim 12, wherein the patient is further treated with at least one additional cancer therapy, such as radiation therapy, cell inhibitory therapy, cytotoxic therapy, immune checkpoint inhibitor therapy, immunomodulatory effector therapy, chimeric antigen receptor (CAR)-T therapy, adoptive T-cell therapy, ablation therapy, targeted therapy, tyrosine kinase inhibitor therapy, and cancer immunotherapy targeting surface expression determinants.

Images