A method for screening extracellular tissue peptides in mammalian tissue samples for health status assessment, disease diagnosis, and novel antigen discovery.

Abstract

This invention relates to a method for screening extracellular tissue peptides in mammalian tissue samples for health status assessment, disease diagnosis, and discovery of nascent antigens. The invention includes the preparation and analysis of tissue samples from solid tumors and focuses on extracellular peptidomics and major histocompatibility complex (MHC) class I immunopeptidomics. This method is antibody-free and utilizes amino acid sequencing and tandem mass spectrometry. It is a simple, inexpensive, and rapid method for comprehensively profiling solid tumor peptidomics (extracellular peptidomics and MHC class I immunopeptidomics). This method may be used for screening and as a biomarker for health status assessment, disease diagnosis, discovery of nascent antigens, and prognosis prediction in salivary gland tumors. Furthermore, tissue MHC class I immunopeptidomics can also be applied to the discovery of nascent antigens, vaccine development, and design of immunotherapies.

Description

Technical Field

[0001] The present invention relates to methods for preparing, qualitatively and quantitatively analyzing samples of extracellular peptide mixtures and major histocompatibility complex (MHC) class I immunopeptide mixtures from solid tumors without using antibodies. Extracellular peptide mixtures and immunopeptide mixtures from solid tumors based on amino acid sequencing using tandem mass spectrometry. The present invention is simple, cost-effective, rapid and comprehensive in solid tumor peptide mixture profiling. Furthermore, these methods enable the discovery of complex extracellular peptide mixtures that can be used for the evaluation of health status, disease diagnosis, and prognosis prediction. MHC class I immunopeptide mixtures for neoantigen discovery and vaccine development / immunotherapy design. Therefore, the present invention provides valuable insights for understanding basic and applied biological problems. A neoantigen is a peptide presented by cancer-specific MHC class I (or II) molecules. The present invention enables the effective screening of MHC class I immunopeptides / peptides from tissues. Therefore, any MHC class I peptide can be used for neoantigen discovery from MHC class I immunopeptides / peptides that are specific to cancer cells and can induce an immune response that kills cancer cells.

Background Art

[0002] Cell surface proteins, or surface proteins, are crucial mediators of cells and their extracellular environment. Surface protein analysis from clinical specimens such as tumor tissue is in high demand in the scientific community. Because surface proteins contain complex biological information and can be the origin of novel antigens, this information can be used for various applications in the medical field. Extracellular peptidomics and immunopeptidomics comprise the entire spectrum of surface proteins in patient-derived cells / tissues. Since normal cells and cancer cells have different surface proteins, surface protein analysis is extremely important in cancer biology and immunological research. Cancer cells are often highly dividing, vary greatly in size and shape, have a small cytoplasmic volume, and lack the regular characteristics of a cell surface. Therefore, surface analysis (extracellular peptidomics and immunopeptidomics) provides valuable insights into cell behavior, disease processes, biomarker discovery, and potential therapeutic targets. The concept of peptidomics has been known for at least 20-30 years, initially applied to specific bioactive peptides such as neuropeptides. Subsequently, with advancements in analytical techniques, peptides in bodily fluids and various secretions are also being analyzed. The peptidophyme [1] is thought to consist of low molecular weight peptides / proteins (less than 15 kilodaltons (kDa)). The extracellular peptidophyme of tissues contains a diverse range of peptides that play potential roles in various physiological and pathological processes (intercellular communication, signal transduction). Extracellular peptidophyme analysis of tissues focuses primarily on the analysis of peptides present in the extracellular space. Thus, these peptides indicate various biological processes such as cell communication and disease processes, and reflect human health or disease conditions. Furthermore, they enable the study of the disease microenvironment, providing insights into the discovery of biomarkers and therapeutic targets and drug development.

[0003] Over the past few decades, advanced technologies have enabled comprehensive peptidemix analysis. In particular, label-free liquid chromatography / tandem mass spectrometry (LC-MS / MS) routinely identifies and quantifies thousands of peptides in a single run, providing an unprecedented opportunity to investigate changes in the extracellular peptidophysical profile of solid tissues or organisms. Furthermore, downstream multi-bioinformatics reveals a functional understanding. Two main reasons have driven this direction: firstly, the expectation for biomarkers and critical peptides at appropriate concentrations; and secondly, the limitations of the technologies available at the time. Unlike proteomics, extracellular tissue peptidemix presents a much higher requirement for accurate peptide identification in the peptidemix approach. Future applications are primarily expected to involve the discovery of novel biomarker peptides that may arise from unexpected sources, such as proteolysis, secretion, or translation of non-standard open reading frames (ORFs).

[0004] Furthermore, achieving this requires quantitative characterization of the entire tissue extracellular peptidophysics. The complexity of the tissue peptide mixture and the abundance of specific types of tissue peptides must be addressed by the analytical method being used. Firstly, a high dynamic range must be achieved to detect and quantify peptides in mixtures containing large amounts of peptides. Secondly, the LC / MS (liquid chromatography and tandem mass spectrometry) method applied must handle peptides with different masses and hydrophobicities, enabling good chromatographic separation and, as much as possible, acquisition of further MS data. Currently, the methods applied mainly rely on differences in solubility in organic solvents or partial denaturation of proteins by heat treatment [2], [3]. These methods suffer from the problem of low recovery rates of extracellular peptides. This study presents a method to significantly overcome this problem and provides guidelines for sample preparation methods that do not involve the lysis or denaturation of tissue cells, i.e., direct peptidophysics analysis using readily available amounts of tissue material. Furthermore, extracellular peptidophysics analysis of tissues requires further development, mainly due to challenges in extracting low molecular weight peptides from the cell surface. The potential applications of extracellular peptidomics stem directly from the novelty of this field. Most importantly, it enables the discovery of novel tissue surface biomarkers of molecules that are similar to proteins but potentially stimulating and important. This will allow for the detection of new drug targets, including those involved in antibody-dependent cell-mediated cytotoxicity (ADCC), complement-dependent cell-mediated cytotoxicity (CDC), immune complexes, and multi-step targeting, which are easier and more specific targets for drug discovery. Furthermore, extracellular peptides outside of tissues may also serve as biomarkers in assessing health status, diagnosing diseases, predicting prognosis, or evaluating treatment efficacy.

[0005] The presentation of major histocompatibility complex (MHC) ligands on the cell surface is a crucial method for communicating the cell's state to the immune system and promoting immune recognition and activation. Any disruption in this process can be a potential cause of disease. For example, autoimmune diseases can develop when the immune system mistakenly recognizes cells as non-self and attacks them. Conversely, cancer cells can remain hidden from the immune system if the cell surface signals do not trigger the immune system to kill the altered cells. The identification and characterization of peptides presented on the cell surface in complex with major histocompatibility complex (MHC) molecules is also known as immunopeptidemic analysis. Furthermore, there is a strong need to develop methods for detecting mutant or non-standard peptide precursors presented by histocompatibility antigens (MHC) from patient samples such as tumor tissue. Moreover, elucidating the peptide presentation pathway from translation products to MHC ligands in various disease scenarios is essential for the development of vaccines, autoimmune disease treatments, cell therapies, and anticancer and antiviral drugs. Therefore, immunopeptidemic analysis provides valuable information about the peptide repertoire. Furthermore, in the context of neoantigen discovery and immunotherapy, it is necessary to discover specific antigens (also called neoantigens) present on the surface of cancer cells. Because neoantigens originate from tumor-specific mutations or abnormally expressed proteins, they are attractive targets for cancer immunotherapy. Therefore, neoantigen discovery is becoming an active research area, with the potential to revolutionize cancer treatment by leveraging the power of the immune system to target and eliminate tumor cells with high specificity. Immunotherapy is a rapidly evolving medical field that has revolutionized the treatment of various diseases. It utilizes the power of the immune system to recognize, attack, and destroy abnormal cells such as cancer cells and pathogen-infected cells. Immunotherapy includes immune checkpoint inhibitors, chimeric antigen receptor T-cell (CAR-T) therapy, cancer vaccines, and adoptive cell transplantation. In recent years, significant progress has been made in the treatment of cancer, infectious diseases, and autoimmune diseases. Continued research and development in immunotherapy focuses on optimizing treatment strategies, expanding the range of treatable diseases, and improving patient outcomes.

[0006] In immunopeptidomics analysis, immunoprecipitation (IP) [4] is widely used for screening novel antigens / immunopeptides. However, this method has several technical limitations: a) it is an antibody-dependent / biased method due to the specificity of antibody binding, and IP-based methods using pan-MHC antibodies also lose MHC subtype information [5]; b) IP requires long washing times to remove contaminants / nonspecific peptides and surfactants, which can lead to the loss of low-affinity MHC I-related peptides [6][7]; c) it is very costly when preparing a large number of samples due to its antibody dependence; d) it is time-consuming and requires highly skilled personnel; and e) it cannot be used for non-human models because there are no effective antibodies against other species. The present invention involves directly isolating the immunopeptideome on major histocompatibility (MHC) class I molecules from solid tumors using an antibody-free method and screening it qualitatively and quantitatively. Patient-derived tissue immunopeptides enable the use of screened neoantigens as vaccine candidates, allowing for the development of T-cell therapies for cancer. The approach of the present invention is a relatively rapid, high-throughput, and cost-effective pipeline that is globally applicable for the direct purification and quantification of neoantigens from tumor patient samples using unbiased mass spectrometry tools.

[0007] From the perspective of tissue extracellular peptidochemistry, the inventors compared commonly used methods and scientific literature, including existing tissue peptidochemistry and proteomics techniques. Only a few known methods have been reported to perform qualitative or quantitative extracellular tissue peptidochemistry. Therefore, the inventors examined tissue peptidochemistry extracted from human and non-human tissues. They observed various approaches, including tissue lysis, polypeptide size fractionation, removal of most proteins typically by organic solvents or heat precipitation, followed by ultracentrifugation. Interestingly, approaches involving precipitation only or filtering only have also been reported, suggesting that none of these steps are essential. Considering the amount of material used, sample preparation procedures, analytical type, and analytical depth, the inventors observed differences in the number of extracellular tissue peptides identified and quantified between published methods and the current inventive method. This indicates a lack of consensus or accumulated knowledge regarding efficient and reproducible performance for separating low molecular weight patient-derived extracellular tissue peptidochemistry. Methods for extracellular peptidochemistry analysis of tissue cells described to date are characterized by a small number of identified peptides, and quantitative methods for peptidochemistry have rarely been introduced. Several reasons limit this approach to comprehensive analysis: a) Patient-derived tissue surfaces are highly complex, containing a variety of elements such as proteins, lipids, metabolites, and signaling molecules. b) The presence of abundant proteins or lipids masks the detection of other proteins, especially small amounts of proteins and peptides. c) Small amounts of peptides have diagnostic potential, but also have a limit of detection (LOD). d) Sample preparation is complex and multi-step, and small amounts of proteins or peptides (hormones or small secretory peptides) are unstable. f) Cell lysis, denaturation, precipitation, and heat treatment methods to remove impurities and large amounts of protein are overwhelming, which can impair the number and yield of diagnostic cell peptides identified, and important biological information may be buried in background noise.

[0008] The invention of extratissue cell peptidomics is based on sequencing the amino acid sequences of natural (unmodified during sample preparation) extratissue peptides without modifying the proteins during sample preparation, such as through denaturation, lysis, reduction, alkylation, or enzymatic digestion. However, it differs significantly from proteomics / peptidomics approaches to peptide biomarker findings from clinical samples [8][9]

[10] in terms of sample preparation, time, and cost. Table 1 compares some of the limitations of existing methods as presented by the inventors. These limitations include complex or multi-step sample preparation and the limited number of extratissue peptides that can be identified / quantified. These are major limitations of existing methods. Furthermore, the method of the present invention is faster (requiring several hours for two-step sample preparation) and less expensive (eliminating the need for additional steps such as tissue lysis, denaturation, reduction, alkylation, digestion, heating, precipitation, and tandem mass tag (TMT) labeling) than existing peptidomics / proteomics methods. In addition, it can be applied to mammalian tissue samples using unbiased mass spectrometry tools. The exact cause of salivary gland tumors is often unknown, and diagnosing these tumors is impossible without a thorough evaluation by medical professionals, including physical examination and imaging tests (ultrasound, computed tomography (CT) scans, magnetic resonance imaging / MRI, etc.). Therefore, the use of extracellular tissue peptidomics for early biomarker screening and diagnosis is essential for developing effective treatments for salivary gland tumors and various disease models. Additional comparisons of previously reported tissue peptidomics are: It is contained in JPEG2026525254000001.jpg694. The principle of extra-tissue peptidomics is unknown.

[0009] Table 1. Comparison of raw materials, sample preparation procedures, and qualitative / quantitative peptide counts for the inventive method (extratissue peptide mix) and the publicly available method targeting intracellular peptides. "Not applicable (N / A)" means the method was not developed for quantitative peptide analysis.

[0010] [Table 1] JPEG2026525254000003.jpg20155

[0011] This invention relates to antibody-free methods for sample preparation, qualitative analysis, and quantitative analysis of extracellular peptidomics and major histocompatibility complex (MHC) class I immunopeptidomics from solid tumors. The invention also relates to extracellular peptidomics and immunopeptidomics from solid tumors based on amino acid sequencing using tandem mass spectrometry. This invention offers a simple, cost-effective, rapid, and comprehensive approach to solid tumor peptidomics profiling. Furthermore, these methods enable the discovery of complex extracellular peptidomics that can be used for biomarker discovery, health status assessment, disease diagnosis, and prognosis prediction. MHC class I immunopeptidomics are used for novel antigen discovery and vaccine development / immunotherapy design. Therefore, this invention provides valuable insights for understanding challenges in basic and applied biology.

[0012] [Advantages of the extracellular peptide approach] Extracellular peptidochemistry is a new field, and tissue peptidochemistry is still in its developmental stages; therefore, the inventors have identified several areas where significant benefits can be gained from this development.

[0013] 1. The nature and complexity of the extracellular peptidophysics are not yet fully understood. Reliable information on the overall abundance of peptides and cell surface secreted peptides in various disease tissue samples, particularly salivary gland tumors, is lacking. Qualitative and quantitative yields can vary significantly in experiments using different extraction methods. To our knowledge, no experiments have precisely addressed this issue. Furthermore, the extracellular peptidophysics is complex and, like the proteome, exhibits a wide concentration dynamic range. This, as with proteomic analysis, presents new analytical challenges, often requiring fractionation or targeted analysis to identify and quantify target (poly)peptides or surface peptides.

[0014] 2. As mentioned above, there is no consensus on the expected yield of extracellular peptides isolated from tissues. This suggests that significant sample loss occurs in reported methods. The inventors believe that comprehensive yields and biomarker discovery from the extracellular peptidomethod of tissues can be achieved with minimal sample preparation procedures. Generally, it is believed that optimal detection of peptides cannot be obtained when cell lysis is performed using only one denaturing agent

[19] . Therefore, the present invention does not involve destroying tissue cells with cell lysing agents or denaturing agents before attempting size fractionation.

[0015] 3. The main difference is that it focuses on the extracellular peptidomethic area of ​​tissue cells, rather than proteins.

[0016] 4. Because the present invention was developed without requiring dissolution, denaturation, heating, removal, or precipitation steps, it is possible to yield more extracellular tissue peptides and isolate peptides that may be bound to tissue / cell surface peptides.

[0017] 5. The sample preparation method of the present invention is faster and more cost-effective than existing peptidomics / proteomics methods (it does not require additional steps such as heating or tandem mass tag (TMT) labeling). Furthermore, by using unbiased mass spectrometry tools, it can be applied to mammalian cells and primary patient cells.

[0018] 6. The present invention does not require the use of additional purification or desalting kits (Ettan 2D clean)

[20]

[21] . All of these procedures are time-consuming and costly.

[0019] 7. For increased sensitivity, known methods

[21] have used fluorescent dyes (Cy3,5) for protein labeling. Reading and imaging proteome profiles from dye-based gels requires specialized equipment (GE Healthcare's Typhoon variable-mode imaging system, which produces digital images of radioactive, fluorescent, or chemiluminescent samples), which also requires an investment of time and money. Therefore, overall, it is a complex sample preparation procedure, and performing this type of analysis requires experienced personnel.

[0020] 8. The present invention requires no additional procedures such as reduction, alkylation, or enzymatic (trypsin) digestion. Furthermore, it makes it possible to identify natural extracellular peptides of tissue cells (which can function as unique markers) without modification.

[0021] 9. This method can concentrate peptides of various sizes depending on the molecular weight filter used to collect extracellular peptides from tissue cells. Therefore, the permeability of the filter is not a limitation of this method. The present invention was tested using a 3 kilodalton (kDa) filter membrane and significantly concentrated peptides between 0.9 and 1.8 kDa.

[0022] 10. This method reveals several important characteristics of extracellular tissue peptidomethylmidic. For example, a) peptide length distribution. By using a 3 kilodalton (kDa) cutoff membrane, peptides with a length of 9 to 18 amino acids are concentrated. b) effective value distribution at the precursor level of quantified peptides. In this invention, most peptides were quantified with an effective value of 5 to 8. This demonstrates that this invention has high reliability in the quantitative quality of extracellular tissue peptidomethylmidic.

[0023] 11. This method is simple and the sample preparation only involves two steps. a) Incubate the tissue sample in an extraction buffer to dissociate protein-peptide interactions and concentrate the extracellular peptides in the tissue and the peptides released from the complex from the tissue surface. b) To purify the separated extracellular peptides in the tissue, use a hydrophilic-lipophilic balance column to remove a large amount of proteins while eluting the extracellular peptides.

[0024] 12. The conventional methods seem to focus only on the identification of proteins / peptides. However, the present invention is based on both qualitative and quantitative performance. The present invention can comprehensively perform both qualitative and quantitative analysis of extracellular peptide mixtures in tissues.

[0025] 13. For the device (the most advanced mass spectrometry model), conventionally, for the separation and identification of peptides, two-dimensional differential gel electrophoresis (2D DIGE), surface-enhanced laser desorption ionization (SELDI), or improved liquid chromatography matrix-assisted laser desorption ionization mass spectrometry (LC-MALDI), matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF / MS)-based platforms have been used for various biological samples. These platforms are less sensitive and comprehensive than the Exploris 480 Orbitrap-based mass spectrometer. That is, the inventors have developed a protein labeling method using fluorescent dyes, gel electrophoresis, desalting, etc., and a direct qualitative, quantitative, and comprehensive extracellular peptide mixture method in tissues.

[0026] 14. In the discovery experiment, the invention of this methodology of extracellular peptide mixture can be complementarily utilized in a wide range of basic research and applied research. Extracellular peptide mixture has been applied to various disease models as a method for identifying biologically relevant novel polypeptides and discovering biomarkers, and has been further developed.

[0027] [Advantages of the Antibody-Free Tissue MHC Class I Immune Peptide Mixture Approach] The inventor has first revealed the immunopeptidome of major histocompatibility (MHC) class I molecules derived from solid tumors using an antibody-free approach. This is a simple and economical parallel approach that can be used in combination with immunoprecipitation (IP). Therefore, the inventor has identified several fields that can benefit greatly from the development of the immunopeptidome of solid tumors using an antibody-free approach.

[0028] 1. The potential of the present invention is to identify and quantify MHC class I patient- and disease-specific immunopeptidomes.

[0029] 2. The antibody-free approach is unbiased with respect to antibody epitope specificity for MHC class I alleles. Therefore, the present invention makes it possible to answer the following questions. a) To what extent is the variability and specificity of the neoantigen peptidome between the control group and each patient-derived tissue sample? b) Since there is a neoantigen landscape unique to each patient, the possibility of developing personalized cancer treatment can be expected.

[0030] 3. Screening and discovery of neoantigens common to various cancer types are suitable for the development of global immunotherapies for various cancer patients. This will establish new practices in the entire medical field, such as routine screening of neoantigens.

[0031] 4. The approach of the present invention (antibody-free approach) is applicable even when the starting tissue material is limited (less than 0.5 grams). On the other hand, in the immunoprecipitation (IP) method, it is recommended to prepare a base sample using at least 1 gram of tissue in order to obtain an optimal yield of MHC class I peptides.

[0032] 5. The approach of the present invention enables cell regeneration after p2M dissociation, that is, after the first isolation of the MHC class I immunopeptidome, so reanalysis is possible after subsequent perturbations. Therefore, the dissociation dynamics of the MHC class I immunopeptidome can be easily analyzed.

[0033] 6. The antibody-free approach is simple, rapid, and highly reproducible. In particular, this concerns qualitative and quantitative tissue sample preparation for major histocompatibility (MHC) class I immunopeptidomics analysis using mammalian tissue samples with an antibody-free approach. This method is characterized by only two steps of sample preparation, based on amino acid sequencing and tandem mass spectrometry to measure the signal intensity of peptides: a) extraction of tissue MHC class I immunopeptidomics using an antibody-free approach, and b) purification of the tissue MHC class I immunopeptidomics.

[0034] 7. The present invention's approach (antibody-free approach) is the only option for the scientific community, particularly for performing immunopeptidomic analysis in non-human tissue samples (such as wild boar, cheetah, and leopard), when antibodies against MHC I molecules are unavailable.

[0035] 8. The present invention requires fewer purification steps and does not require surfactants during sample preparation.

[0036] 9. The present invention does not introduce bias associated with preferential recovery loss of low-affinity / medium-affinity / high-affinity peptides.

[0037] 10. The inventive approach has great potential to open up numerous fields for various diseases based on immune responses and to contribute to the advancement of academia, medicine, biotechnology, and industry.

[0038] 11. This invention holds great potential for translational research and commercialization aimed at high-throughput screening of MHC class I peptides, and ultimately, for the discovery of neoantigens from various clinical tumor samples.

[0039] Therefore, the patient-derived tissue samples and their extracellular peptidomethyl and MHC class I immunopeptidomethyl data of the present invention complement other omics technologies such as proteomics. The present invention is readily applicable worldwide based on clinical samples such as tissues. Furthermore, the approach of the present invention is relatively rapid (in terms of time) and inexpensive for directly discovering important biological extracellular tissue peptide biomarkers and potential neoantigens from MHC class I immunopeptidomethyl data screened from patient-derived tissue samples using unbiased mass spectrometry tools.

[0040] The present invention provides a method for directly screening and quantifying extracellular peptidomethyl and neoantigens from tissues for the purpose of evaluating health status, diagnosing diseases, and discovering neoantigens.

[0041] The method of the present invention is applicable to all types of tissue clinical samples from both human and non-human samples. The present invention includes a unique sample preparation method for qualitative and quantitative analysis of extracellular tissue and MHC class I-related immunopeptidomics using an antibody-free approach. The workflow consists of two steps in sample preparation: a. extraction and b. purification of tissue peptides. The present invention is largely unchanged, with particular emphasis on optimizing the first extraction step, namely 1. pre-washing, 2. incubation time, 3. extraction buffer pH, and 4. extraction buffer volume, to achieve enrichment of the tissue extracellular peptidomic and tissue MHC class I immunopeptidomics. The second important step is purification using an Oasis column (hydrophilic-lipophilic balance) to remove P2 microglobulin (B2M) / excess protein. This method is simple yet effectively removes interference, enabling the detection of the tissue extracellular peptidomic and tissue MHC class I immunopeptidomics without compromising quantitative reproducibility.

[0042] [I. Extraction of extracellular tissue peptides:] Prior to extraction, tissue samples are washed 2-3 times with phosphate-buffered saline (PBS) (without calcium, magnesium, or phenol red) at pH 7.2-7.5, and then centrifuged at 4°C and 500g for 5 minutes. Next, the test tissue samples are incubated with extraction buffer (citric acid-phosphate buffer) at pH 3.1-3.6, prepared with liquid chromatography-mass spectrometry (LC-MS) grade ultrapure water (purity 99.9% or higher). The amount of buffer is 7-8 mL per 0.05-1 g of tissue sample. The incubation step is performed at 4-10°C with stirring for 2-3 minutes to dissociate peptides from the high-concentration tissue membrane surface, and the extracellular tissue peptidomethicone is recovered by centrifugation at 4°C and 500g for 5 minutes. Therefore, this method is characterized by the unexpected use of a low pH buffer for the concentration of extracellular tissue peptides and the release of complexes from the tissue surface.

[0043] [II. Extraction of tissue MHC class I immunopeptidomics using antibody-free methods:] Prior to extraction, the tissue sample is washed 7-10 times with phosphate-buffered saline (PBS) (without calcium, magnesium, or phenol red) at pH 7.2-7.5, and then centrifuged at 4°C at 500g for 5 minutes. Next, the test tissue sample is incubated with an extraction buffer (citric acid-phosphate buffer) at pH 3.3, prepared with ultrapure water of liquid chromatography-mass spectrometry (LC-MS) grade with a purity of over 99.9%. The amount of buffer is 5-6 ml for every 0.5-1 g of tissue sample. The incubation step involves stirring at 4-10°C for 4-5 minutes to dissociate the MHC class I peptides on the tissue membrane surface. The MHC class I peptides on the tissue membrane surface are recovered by centrifuging at 4°C at 500g for 5 minutes.

[0044] [III. Purification of the extracellular tissue peptideme and MHC class I immunopeptideme] The supernatants collected from both sample sets were passed through an Oasis cartridge (30 mg, Waters, hydrophilic-lipophilic balance column) for reverse-phase purification of polypeptides. All steps were performed at room temperature using gravity flow only. The detailed steps are as follows. i) Cartridge condition: The cartridges were conditioned twice with 0.2% formic acid / methanol (volume / volume (v / v)) (the final percentage of formic acid in methanol in this solution was 0.2%). The volume of each washing solution was 1 ml / cartridge. ii) Cartridge equilibration: The cartridges were equilibrated with 0.2% formic acid / water (volume / volume (v / v)) (the final percentage of formic acid in the water in this solution was 0.2%). The volume of each wash was 1 ml per cartridge. iii) Sample loading: In this step, 6-8 ml of the sample, which is the supernatant collected after centrifugation, was loaded into a cartridge (hydrophilic-lipophilic balanced column). iv) Washing step: The cartridges were washed three times with water containing 5% methanol / 0.2% formic acid (volume / volume (v / v)) (the final concentrations of methanol and formic acid in the water were 5% and 0.2%, respectively). The volume of each wash was 1 ml per cartridge. v) Elution: The bound substance (peptide) was eluted with 1 ml of water containing 80% methanol / 0.2% formic acid (volume / volume (v / v)) (the final concentrations of methanol and formic acid in this solution were 80% and 0.2%, respectively). Subsequently, it was diluted with water / 40% methanol / 0.2% formic acid (volume / volume (v / v)) (the final concentrations of methanol and formic acid in this solution were 40% and 0.2%, respectively).

[0045] For extracellular tissue peptidophores, conditioning and filtration of all filtration units were performed immediately after elution. According to the instructions, ultrafiltration was performed using a selected molecular weight cutoff filter (3-30 kilodaltons (kDa), preferably 3, 10, or 30 kDa) to remove proteins greater than 15-30 kDa from the peptide concentrate and simultaneously elute tissue extracellular peptides. After further separation of tissue extracellular peptides, the sample was centrifuged at 4000 g, 4-10°C for 120-135 minutes to remove high molecular weight proteins greater than 30 kDa and other cellular compounds. Finally, the eluted and diluted peptide solution (2 ml) was injected into the device using a pipette, taking care to prevent the pipette tip from touching the membrane. The filtrate (extracellular tissue peptides) was collected and evaporated to dryness. All dried peptide samples were stored at -80°C until mass spectrometry. In tissue MHC class I immunopeptide mix, after peptide concentration, extracellular material greater than 3 kDa (P2 microglobulin (B2M) or extracellular material greater than 3 kDa) was removed by centrifugation at 4000 g at 4-10°C for 120-135 minutes using a selected molecular weight cutoff filter with a 3 kDa molecular weight filter, and MHC class I peptides less than 3 kDa were recovered. Finally, the eluted and diluted peptide solution (2 ml) was injected into the device using a pipette. Care was taken to prevent the pipette tip from touching the membrane. The filtrate (MHC class I peptides) was collected and evaporated to dryness. All dried peptide samples were stored at -80°C until mass spectrometry.

[0046] The present invention is accurately described in the examples and drawings, which provide the following information. [Brief explanation of the drawing]

[0047] [Figure 1] A workflow for screening and quantifying the tissue extracellular peptideome and tissue immunopeptideome using an antibody-free approach is presented. [Figure 2]A Venn diagram of quantitative analysis comparing quantitative values ​​of extracellular peptides in salivary gland tumor tissue (n=4) and healthy tissue samples (n=4). [Figure 3A-3B] Figure 3A shows the peptide length distribution of extracellular tissue peptides identified by DIA (Data Independent Acquisition) spectroscopy in salivary gland tumor tissue and healthy tissue peptidometer samples. Figure 3B shows the distribution of effective values ​​at the precursor level obtained from quantified extracellular tissue peptides. [Figure 4] Quality control for tissue extracellular peptide dome extracts differs significantly from that for MHC class I peptides. [Figure 5] The tissue extracellular peptide volcano plot shows abnormal extracellular peptides (as black dots) with log2 multiplicity changes > 2 and -logp values ​​> 2 in comparison between salivary gland tumor tissue and healthy tissue samples. [Figure 6] This shows a string network analysis of specific sources of peptides that exhibit statistically different abundances in salivary gland tumor tissue and healthy tissue samples. [Figure 7] Analysis of the distribution of extracellular peptide sources in tissue cells and changes in differentiation peptide sources. [Figure 8] Detailed insights into tissue MHC class I immunopeptide dome peptide identification using an antibody-free approach are presented. [Figure 9] Peptide length distribution of tissue MHC class I immunopeptide dome eluates from healthy tissue and NSCLC tissue samples. [Figure 10] Analysis of tissue MHC class I immunopeptide prediction. [Figure 11] Gene set enrichment analysis plot quantitatively comparing the immunopeptidomes of healthy tissue and NSCLC tissue samples. [Figure 12] Evaluation of MHC class I binding of a subset of significantly altered peptides associated with lymphocyte activation, obtained from tissue immunopeptidemic analysis comparing healthy tissue and NSCLC tissue. [Figure 13]A volcano plot showing abnormal protein sources and corresponding MHC class I peptides, obtained from tissue immunopeptidemic analysis comparing healthy tissue and NSCLC tissue samples. [Modes for carrying out the invention]

[0048] [Detailed description of the drawing] [Figure 1] Workflow for screening and quantification of tissue extracellular peptides and tissue immunopeptides using an antibody-free approach. This workflow requires tissue as the biological sample, and sample preparation consists of two steps: a) extraction and b) purification of tissue peptides. The present invention, with minimal changes in the first step, i.e., the extraction step (pre-washing step, incubation time, extraction buffer pH, etc.), achieves enrichment of the tissue extracellular peptideme and tissue MHC class I immunopeptideme. The second crucial step is purification using an Oasis column (hydrophilic-lipophilic balance) to remove β2 microglobulin (B2M) / excess protein while simultaneously eluting the tissue peptides of interest. The peptides are further separated by ultrafiltration using a 3 kilodalton (3 kDa) molecular weight cutoff (MWCO) filter, removing high molecular weight proteins and other intracellular compounds. This method is simple yet removes sufficient interference to enable detection of the tissue extracellular peptideme and tissue MHC class I immunopeptideme without compromising quantitative reproducibility.

[0049] [Figure 2] Venn diagram of quantitative analysis comparing quantitative values ​​of extracellular tissue peptides in salivary gland tumor tissue (n=4) and healthy tissue samples (n=4) under two conditions. The present invention comprehensively quantifies at least 9329 types of extracellular tissue peptides.

[0050] [Figure 3] A. Peptide length distribution of extracellular tissue peptides identified by DIA (data-independent acquisition) spectroscopy in salivary gland tumor tissue and healthy tissue peptideome samples. Samples were filtered through a 3 kilodalton (kDa) cutoff membrane that concentrates peptides with lengths of 8–16 amino acids. B. Distribution of effective precursor levels obtained from quantified extracellular tissue peptides.

[0051] [Figure 4] Quality control of extracellular tissue peptidomethic extracts differs significantly from that of MHC class I peptides. The yield of extracellular tissue peptides, distinct from MHC peptides, was confirmed by comparing the detected peptides with random / artificial peptides. NetMHCpan prediction of peptide binding (to any MHC molecule with a known sequence) revealed that the extracellular tissue peptidomethic extract had lower MHC binding affinity than predicted by the random / artificial peptide set.

[0052] [Figure 5] The tissue extracellular peptide volcano plot shows abnormal extracellular peptides (as black dots) with log2 multiplicity changes > 2 and -logp values ​​> 2 in comparison between salivary gland tumor tissue and healthy tissue samples.

[0053] [Figure 6] This shows a string network analysis of specific sources of peptides that exhibit statistically different abundances in salivary gland tumor tissue and healthy tissue samples. The string analysis of the tissue extracellular peptidophysis consists of the majority of the detected peptides from extracellular origins and shows high functional connectivity between the obtained hits.

[0054] [Figure 7] Analysis of the distribution of extracellular peptide sources in tissue cells and changes in differentiated peptide sources. A. Amino acid positions and ploidy changes in the sequences of detected peptides, considering their position in full-length proteins. Grayscale bars indicate Welch's test p-value labels for differences between sample groups for each peptide. B. Peptide regulation profiles in comparison of salivary gland tumor tissue and healthy tissue samples (black - significantly regulated peptides, gray - other peptides that may originate from the same protein).

[0055] [Figure 8] Detailed findings on tissue MHC class I immunopeptidomic peptide identification using an antibody-free approach are shown. A. The bar graph shows the number of tissue MHC class I immunopeptidomic peptides (15671) identified in healthy tissue (n=5) and NSCLC tissue (n=5) samples. B. The Venn diagram shows the overlap of peptides identified in healthy tissue and NSCLC tissue samples. This figure highlights that the NSCLC tissue group has a significantly higher number of peptides than the healthy tissue (control tissue). C. The bar graph shows that peptides were identified for each tissue using two search engines, COMET (CMT) and MS FRAGGER (FRG), and two types of mass spectrometry data acquisition methods (data-dependent acquisition (DDA) and data-independent acquisition (DIA)). The bar graph shows that while the number of peptides identified by DDA data increased, the COMET and MS FRAGGER search engines demonstrated comparable efficiency in peptide identification (P=patient, Series A=healthy tissue, Series B=NSCLC tissue sample; the numbers 6, 8, 11, 13, and 18 indicate different patients identified).

[0056] [Figure 9] Peptide length distribution of tissue MHC class I immunopeptide eluates from healthy tissue samples (A) and NSCLC tissue samples (B). The histogram showing the peptide length distribution indicates that the majority of peptides are in the 7-16 amino acid residue range, with a greater proportion of 9-nucleotide long peptides. This is consistent with the characteristic lengths of the MHC class I immunopeptide range.

[0057] [Figure 10] Analysis of tissue MHC class I immunopeptide predictions. Computational algorithms and prediction models, such as the MHCfhirry prediction model, are used for A) antigen treatment prediction, B) and C) allele-independent prediction, D) antigenicity / immunogenicity prediction of MHC class I peptides from the Immunoepitope Database (IEDB), and E) and F) allele-specific prediction. By integrating these predictions, the inventors of A-F provide a comprehensive understanding of the MHC class I immunopeptide composition of tissue immunopeptidic isolates by an antibody-free approach and further provide insights into the dynamics of immunorecognition and peptide-MHC class I interactions in NSCLC tissue samples (P = patient, numbers 6, 8, 11, 13, and 18 identify different patients, Series A = healthy tissue, Series B = NSCLC tissue sample, Control = artificially generated peptide).

[0058] [Figure 11] Gene set enrichment analysis (GSEA) plot quantitatively comparing the immunopeptideomes of healthy tissue and NSCLC tissue samples. The GSEA plot shows enrichment of immunopeptide products of genes related to lymphocyte activation in an analysis dataset quantitatively comparing the MHC class I immunopeptideome of healthy tissue with that of NSCLC tissue samples. This enrichment pattern suggests that the decreased activity of lymphocyte-related processes in NSCLC tissue is manifested by changes in the immunopeptideome profile.

[0059] [Figure 12] Evaluation of MHC class I binding of a subset of significantly altered peptides associated with lymphocyte activation, obtained from tissue immunopeptidomic analysis comparing healthy tissue and NSCLC tissue. This figure shows that MHC class I binding factors attributable to lymphocyte activation are upregulated more frequently in healthy tissue than in NSCLC tissue samples.

[0060] [Figure 13] Volcano plot showing abnormal protein sources and corresponding MHC class I peptides obtained from tissue immunopeptidemic analysis comparing healthy tissue and NSCLC tissue samples. Selected labels indicate the regulation of peptides resulting from protein precursors linked by lymphocyte activators.

[0061] The following examples demonstrate detailed sample preparation and analysis methods.

[0062] [Example 1] The method of the present invention is novel and simple, enabling screening and quantification of tissue extracellular peptides and tissue MHC class I immunopeptides using an antibody-free approach in only two sample preparation steps and at least four mutation steps (Figure 1). Before commencing any sample preparation step of the tissue peptidomics (extracellular peptidomics and MHC class I immunopeptomics) method, it is strongly recommended to prepare all solutions using LC-MS (liquid chromatography with mass spectrometry) grade water, solvents, reagents, and ultra-clean glass and plastic equipment used in the method. Furthermore, all solutions and buffers should be prepared immediately before use and discarded after being stored for one week. The first step is to prepare all buffers required on the day of tissue peptidomics sample preparation.

[0063] Preparation of buffer solution: Primary buffer solution: Prepare a citrate-phosphate buffer solution (0.131M citrate / 0.066M Na2HPO4, 150mM NaCl) with a pH of 3.1-3.6, and adjust the pH to 3.1-3.6 with 1M NaOH prepared with LC-MS (liquid chromatography / mass spectrometry) grade water.

[0064] As a secondary solution, 0.2% formic acid / methanol (volume / volume (v / v)) was prepared (the final concentration of formic acid in methanol was 0.2%).

[0065] As a tertiary solution, 0.2% formic acid / water (volume / volume (v / v)) was prepared (the final concentration of formic acid in water was 0.2%).

[0066] As the fourth solution, a washing buffer (water / 5% methanol / 0.2% formic acid (volume / volume (v / v))) was prepared (the final ratios of methanol and formic acid in water were 5% and 0.2%, respectively).

[0067] As the fifth solution, an elution buffer (water / 80% methanol / 0.2% formic acid (volume / volume (v / v))) was prepared (the final ratios of methanol and formic acid in water were 80% and 0.2%, respectively).

[0068] We strived to develop a highly reproducible and effective method for obtaining a sufficient amount of sample from a minimal amount of tissue material.

[0069] The workflow consists of two steps: sample preparation (Figure 1) and tissue peptide purification (a). The present invention, which makes minimal changes to the first step, namely the extraction step (pre-washing, incubation time, pH of the extraction buffer, volume of the extraction buffer, etc.), enables the enrichment of the tissue extracellular peptidomethic and tissue MHC class I immunopeptidomethic. The second crucial step is purification using a hydrophilic-lipophilic OASIS column to remove β2-microglobulin (B2M) and excess proteins. This method is simple yet removes sufficient interference to enable the detection of the extracellular peptidomethic and tissue MHC class I immunopeptidomethic without compromising quantitative reproducibility.

[0070] Invention of a sample preparation method for qualitative and quantitative analysis of extracellular tissue peptidomes and MHC class I immunopeptidomes in mammalian tissue samples using an antibody-free approach. This method is based on amino acid sequence analysis using tandem mass spectrometry with peptide signal intensity measurement and is characterized by only two steps of sample preparation.

[0071] [1. Extraction of extracellular tissue peptides:] Before extraction, the tissue sample is washed 2-3 times with phosphate-buffered saline (PBS) (without calcium, magnesium, or phenol red) at pH 7.2-7.5, and then centrifuged at 4°C at 500g for 5 minutes. Next, the test tissue sample is incubated with extraction buffer (citric acid-phosphate buffer) at pH 3.1-3.6, prepared with ultrapure water of liquid chromatography-mass spectrometry (LC-MS) grade with a purity exceeding 99.9%. The amount of buffer is 7-8 ml per 0.05-1 g of tissue sample. The incubation step is performed at 4-10°C with stirring for 2-3 minutes to dissociate peptides from the high-concentration tissue membrane surface, and the extracellular tissue peptidomethicone is recovered by centrifugation at 4°C at 500g for 5 minutes. Therefore, this method is characterized by the unexpected use of a low pH buffer for the concentration of extracellular tissue peptides and the release of complexes from the tissue surface.

[0072] [2. Extraction of tissue MHC class I immunopeptidomics using antibody-free methods:] Prior to extraction, the tissue sample is washed 7-10 times with phosphate-buffered saline (PBS) (without calcium, magnesium, or phenol red) at pH 7.2-7.5, and then centrifuged at 4°C at 500g for 5 minutes. Next, the test tissue sample is incubated with an extraction buffer (citric acid-phosphate buffer) at pH 3.3, prepared with ultrapure water of liquid chromatography-mass spectrometry (LC-MS) grade with a purity of over 99.9%. The amount of this buffer is 5-6 ml for every 0.5-1 g of tissue sample. The incubation step involves stirring at 4-10°C for 4-5 minutes to dissociate the MHC class I peptides on the tissue membrane surface. The MHC class I immunopeptides on the tissue membrane surface are recovered by centrifugation at 4°C at 500g for 5 minutes.

[0073] As shown in Figure 1, the extracellular tissue peptideome and MHC class I immunopeptideome were purified.

[0074] The supernatants collected from both sample sets were passed through an Oasis cartridge (30 mg, Waters, hydrophilic-lipophilic balance column) for reverse-phase purification of polypeptides. All steps were performed at room temperature using gravity flow only. The detailed steps are as follows: i) Cartridge condition: The cartridges were conditioned twice with 0.2% formic acid / methanol (volume / volume (v / v)) (the final percentage of formic acid in methanol in this solution was 0.2%). The volume of each washing solution was 1 ml / cartridge. ii) Cartridge equilibration: The cartridges were equilibrated with 0.2% formic acid / water (the final concentration of formic acid in the water in this solution was 0.2%). The volume of each wash was 1 ml per cartridge. iii) Sample Loading: In this step, 6-8 ml of the sample, which is the supernatant collected after centrifugation, was loaded into a cartridge (hydrophilic-lipophilic balanced column). iv) Washing step: The cartridges were washed three times with water containing 5% methanol / 0.2% formic acid (volume / volume (v / v)) (the final concentrations of methanol and formic acid in the water were 5% and 0.2%, respectively). The volume of each wash was 1 ml per cartridge. v) Elution: The bound substance (peptide) was eluted with 1 ml of water containing 80% methanol / 0.2% formic acid (volume / volume (v / v)) (the final concentrations of methanol and formic acid in this solution were 80% and 0.2%, respectively). Subsequently, it was diluted with water / 40% methanol / 0.2% formic acid (volume / volume (v / v)) (the final concentrations of methanol and formic acid in this solution were 40% and 0.2%, respectively).

[0075] For extracellular tissue peptidophores, conditioning and filtration of all filtration units were performed immediately after elution. According to the instructions, to remove proteins greater than 15-30 kDa from the peptide concentrate and simultaneously elute extracellular tissue peptides, further separation was performed by ultrafiltration using a selected molecular weight cutoff filter (3-30 kilodaltons (kDa), preferably 3, 10, or 30 kDa), followed by centrifugation at 4000 g, 4-10°C for 120-135 minutes to remove high molecular weight proteins greater than 30 kDa and other cellular compounds. Finally, the eluted and diluted peptide solution (2 ml) was injected into the device using a pipette, taking care to prevent the pipette tip from touching the membrane. The filtrate (extracellular tissue peptidophore) was collected and evaporated to dryness. All dried peptide samples were stored at -80°C until mass spectrometry.

[0076] In tissue MHC class I immunopeptide mix, after peptide concentration, ultrafiltration is performed using a selected molecular weight cutoff filter with a 3kDa molecular weight filter to remove extracellular material greater than 3kDa (β2-microglobulin (B2M) or other extracellular material greater than 3kDa). The mixture is then centrifuged at 4000g at 4-10°C for 120-135 minutes to recover MHC class I peptides less than 3kDa. Finally, the eluted and diluted peptide solution (2 ml) is injected into the device using a pipette tip. Care is taken to prevent the pipette tip from touching the membrane. The filtrate (MHC class I peptides) is collected and evaporated to dryness. All dried peptide samples were stored at -80°C until mass spectrometry.

[0077] [Example 2 (Extracellular Peptidomics)] In extratissue cell peptide mix, the next crucial step in this invention is measurement based on mass spectrometry, such as DDA (Data-Dependent Acquisition) mode or DIA (Data-Independent Acquisition) mode. Mass spectrometry combined with chromatographic separation is essential for identifying thousands of peptides. Because extratissue cell peptides are nonspecific, successful identification requires significantly higher data quality, particularly mass precision, than standard trypsin-based proteome samples. Only recently have advancements in mass spectrometry technology made it possible to collect data of acceptable quality for such analysis. DIA (Data-Independent Acquisition) methods enable the collection of quantitative data for all peptides present in a sample while maintaining excellent resistance to matrix interference.

[0078] [Extratissue cell peptide mix LC-MS / MS (liquid chromatography and tandem mass spectrometry) analysis] The LC conditions for the DDA and DIA experiments were identical. Samples were separated using a Thermo Scientific RSLC3000nano LC system connected to a Thermo Scientific Orbitrap Exploris 480 mass spectrometer. Chromatography involved concentrating peptides at a flow rate of 5 μl / min for 10 minutes using a 0.3 × 5 mm Cl 8 PepMap trap column (Thermo Scientific), followed by separation using a 0.075 mm × 250 mm 2 μm C18 PepMap RSLC column (Thermo Scientific). The loading buffer was the same as the sample lysis buffer; mobile phase A was 2.5% acetonitrile and 0.1% formic acid aqueous solution, and mobile phase B was 80% acetonitrile and 0.1% formic acid aqueous solution. Subsequently, the mobile phase B concentration was increased from 2.5% to 35% over 80 minutes, then further increased to 60% over 15 minutes, washed with 99% B, and then re-equilibrated to the initial state to separate the peptides.

[0079] DDA data were acquired using a full scan with a resolution of 120k, mass range of 350–1650Th, AGC (Automatic Gain Control) target of 300%, and automatic setting of the maximum implantation time. The full scan was saved as a profile spectrum. For fragmentation, a minimum intensity of 3000 was selected, and dynamic exclusion was set to 20 seconds within a range of ±10 ppm from the fragmented precursor and its isotopes. Charge states 1–6 were allowed for fragmentation. The ion separation window was set to 1.6Th, resolution of 60k, and normalized HCD (High Energy Collision Dissociation) fragmentation energy to 30%. The AGC target was set to standard, and the maximum ion implantation time was set to automatic. The fragment mass range was set to "Define initial mass," and the starting m / z was 110Th. The spectrum was saved in profile mode.

[0080] DIA data were acquired with a full scan at a resolution of 60k, mass range of 350–1450Th, AGC target of 300%, and a maximum injection time of 100ms. The DIA used 62 12Th windows with 1Th overlap, covering the mass range of 350–1100Th. Resolution was set to 30k, AGC to 1000%, HCD fragmentation energy to 30%, and the default ionic charge state was 3+. Spectra were saved in profile mode.

[0081] [Identification of extracellular peptides in tissue cells] The data were primarily analyzed for the quantitative characterization of the intracellular peptidophysis. Raw data files (DDA and DIA runs) were converted to .mzml files using MSConvert, which has centroidization capabilities.

[0082] The converted MS data were searched against the complete Homo sapiens Uniprot database (SwissProt+Trembl, all isoforms), concatenated with index retention time (iRT) peptide sequences, decoy reverse sequences, and contaminants, using MSFragger 3.4, integrated into the Fragpipe suite. The search database was constructed using the Fragpipe v. 15.0 suite. Additional DDA-like data were extracted from DIA (data-independent acquisition) files using the DIA-Umpire signal extraction engine integrated into Fragpipe. The MSFragger search settings were configured with a precursor mass tolerance of + / - 8 ppm and automatic optimization of fragment mass tolerance. Enzyme digestion was set to non-specific, and peptide length to 7-45 amino acids. The peptide mass range was set to 500-5000 daltons. Variable modifications were set to methionine oxidation and protein N-terminal acetylation. The data were mass-recalibrated, and automatic parameter optimization settings were used to adjust fragment mass tolerance. The search results were automatically filtered and validated by Philosopher. The spectral library for use in DIA-based quantitative analysis was created using EasyPQP.

[0083] [Quantification of extracellular peptides in tissue cells] DIA data processing was performed using DIA-NN v.1.8. All .mzml files were converted to .dia files. The MBR option was selected. Precursor FDR was set to 1%. The Robust LC (high precision) option was selected. Mass precision for MS2 and MSI, and the scan window were automatically adjusted for each run. Retention time (RT)-dependent cross-run normalization was applied to the data. Other settings in the tabs were left at their default values. Peptide quantification data were created using the R library diann-r. The data were preprocessed in Python, and the mean of the technical replicates was calculated. Statistical analysis was performed using Perseus v.1.6.15. Peptide quantification data containing missing values ​​were excluded. Regulatory peptides were identified using Student's test with a p-value cutoff of 0.01 and a log2 change cutoff of 2.

[0084] The extracellular peptide mix method of the present invention demonstrates several important characteristics of extracellular peptide mixes. For example, a) Salivary gland tumor tissue (n=4) and healthy tissue samples (n=4) were compared, and at least 9329 extracellular peptides were comprehensively quantified (Figure 2). The Venn diagram (Figure 2) shows that an appropriate mass spectrometry method, the extracellular tissue peptide isolation protocol, and mass spectrometry-based peptide mix acquisition methods (DIA (data-independent acquisition) and DDA (data-dependent acquisition)) were employed. b) Peptide length distribution when a 3 kilodalton (kDa) cutoff membrane was used to concentrate peptides with a length of 9-18 amino acids (Figure 3A). Extracellular peptides can be separated by ultrafiltration using a molecular weight cutoff filter of 3-30 kilodaltons (kDa). Therefore, filter permeability is not a limitation of this method. Figure 3B shows the distribution of effective values ​​at the precursor level of the quantified extracellular peptides. Despite the high variability of clinical samples such as tissues, 37% of the peptides were detected in at least 7 out of 8 measurements, and were deemed suitable for quantification. In this invention, most peptides were quantified with an effective value of 5-8, demonstrating the highly reliable quantitative quality of this invention (Figure 3B).

[0085] The next step in the invention of the tissue extracellular peptidophysics focuses on the discovery of extracellular biomarker peptides for early diagnosis of salivary gland tumors, assessment of health status, and discovery of novel antigens. However, prior to biomarker discovery, the inventors conducted quality control check experiments to confirm that the MHC class I peptidophysics and the tissue extracellular peptidophysics do not interfere with each other, as both are present on the tissue / cell surface. As an initial quality control, the inventors examined tissue extracellular peptidophysics extracts. These extracts were found to be significantly different from MHC ligand peptides (Figure 4). The tissue extracellular peptides were fed into the NetMHCpan server, and an artificial neural network (ANN) was used to predict the binding of the peptides to any MHC molecule of known sequence. These peptides were classified into whole peptides, binding peptides, and strongly binding peptides for representatives of MHC supertypes. In Figure 4, the yield of tissue extracellular peptides different from MHC class I peptides was confirmed by comparing the detected peptides with random / artificial peptides. The tissue peptidophysics extract contained very few MHC ligand peptides. In the analyzed peptide samples, only 671 out of 9321 peptides (7.2%) showed potential for MHC binding, which is even lower than the result for the random peptide set (799 predicted conjugates (8.5%)). Thus, peptide binding prediction by NetMHCpan revealed that the tissue extracellular peptidophysics has a lower proportion of MHC binding than the MHC conjugates predicted for the random / artificial peptide set (Figure 4). Furthermore, the length distribution shown in Figure 3A does not represent typical MHC class I peptide enrichment of only 9-mer peptides. This strongly suggests that the detected peptides represent a broad population of the tissue extracellular peptidophysics rather than a subpopulation of MHC class I peptides. For further analysis, only tissue extracellular peptides fully quantified in salivary gland tumor tissue (n=4) and healthy tissue (n=4) samples were considered as inventions of tissue extracellular peptides. This shows identification with excellent reproducibility (Figure 2). Venn diagram comparing peptide hits between salivary gland tumor tissue (n=4) and healthy tissue (n=4) (Figure 2).Statistical analysis was performed using Perseus v.1.6.15. Peptide quantifications containing missing values ​​were excluded. Quantitative data obtained with DIA-NN were post-processed in Perseus. Significantly regulated extracellular tissue peptides were selected using a Student's test p-value cutoff of 0.01, a log2-fold change cutoff of 2, and a permutation-based FDR (false detection rate) correction of 1% FDR. The results were visualized as a volcano plot shown in Figure 5. The extracellular tissue peptide volcano plot shows abnormal extracellular peptides (as black dots) with log2-fold changes > 2 and -logp values ​​> 2 or greater when comparing salivary gland tumor tissue (n=4) and healthy tissue (n=4) samples. The extracellular tissue peptide quantification experiment provides a list of peptides that significantly stratify between comparison groups. Quantitative analysis of extracellular tissue peptides sequenced using tandem mass spectrometry can be performed when the signal intensity of the target peptide is 4 times or greater (≧4) than the signal intensity of the corresponding control sample.

[0086] The extracellular tissue peptide mix patterns of salivary gland tumors are a list of at least 52 quantitative peptides for health assessment, disease diagnosis, and novel antigen discovery in human salivary gland tumors. The peptide sequences exhibiting higher multiplicative changes in the extracellular peptide mix patterns in salivary gland tumors, along with the sources of the protein identifiers for each peptide, are listed in Tables 2, 3, and 4, respectively. Multiplicative changes refer only to tumor extracellular quantitative peptides. This indicates how many times higher the signal intensity of the salivary gland tumor sample was compared to a healthy tissue sample.

[0087] [Table 2]

[0088] Table 2 shows the extracellular tissue peptide sequences of each peptide and the origin of their protein identifiers. In the tissues of salivary gland tumor patients, two extracellular tissue peptides were increased by more than 100 times (***) compared with healthy tissue from the same patients.

[0089] [Table 3]

[0090] Table 3 shows the extracellular tissue peptide sequences and the origin of the protein identifiers for each peptide. Fourteen extracellular tissue peptides showed expression levels more than 20 times higher (**) in the tissues of salivary gland tumor patients compared to healthy tissues of the same patients.

[0091] [Table 4] JPEG2026525254000007.jpg119158

[0092] Table 4 shows the extracellular tissue peptide sequences and the origin of the protein identifiers for each peptide. In the tissues of salivary gland tumor patients, 36 extracellular tissue peptides were increased by more than four times (*) compared with healthy tissues from the same patients.

[0093] To further understand the biological significance inferred from quantitative extracellular peptidomethic data of salivary gland tumor patients, we used the Interactor Gene / Protein Search Tool (STRING) (Figure 6). First, we used the interactome map generated by the Interactor Gene / Protein Search Tool (STRING) to determine characteristic nodes abundant in the tissues of salivary gland tumor patients compared to healthy tissues of the same patients. STRING aims to specifically distinguish tumor tissue from healthy samples. We performed string analysis of protein identifiers significantly abnormal in the extracellular matrix, including extracellular matrix proteins or collagen. A filtered subset of cellular component terms describing extracellular matrix-specific protein identifiers corresponding to significantly abnormal peptides (Student's test p-value < 0.01 and fold change > 2) obtained from the comparison of tumor and healthy samples of salivary gland tissue. The results of the analysis revealed interesting STRING nodes indicating the origin of peptides showing statistically different abundances in salivary gland tumor tissue and healthy tissue samples. STRING analysis of the tissue extracellular peptidomethicome showed that the majority of detected peptides consisted of extracellular origins, and there was high functional connectivity between the obtained hits (Figure 6). Thus, interaction networks provide another powerful method for identifying key enriched protein nodes for determining the predictive roles of extracellular peptides in tissues.

[0094] The extracellular peptidochemistry approach enables screening of peptide expression at various tissue conditions. This screening facilitates a precise understanding of biological mechanisms and the discovery of extracellular peptide biomarkers for various diseases. Consequently, the final extracellular peptidochemistry analysis revealed differences in the abundance profiles of multiple peptides between salivary gland tumor patients and healthy tissue samples. The quantitative data of this invention demonstrate that even a single protein can be the source of various extracellular peptides that may have different abundance profiles. Quantitative analysis of intracellular peptides sequenced by tandem mass spectrometry can be performed when the signal intensity of the peptide under study is at 1% FDR (false detection rate) after t-testing and permutation-based false detection rate (FDR) correction between salivary gland tumor and healthy tissue samples. The inventors can quantify that these interesting examples, namely vimentin and latent transformation growth factor β-binding protein 2 (LTBP2) protein, are the source of a subset of peptides present in different amounts in salivary gland tumor tissue compared to healthy tissue samples (Figure 7). We analyze the origin distribution of extracellular peptides in tissues and the ploidy of the origin of differentiating peptides, particularly the amino acid positions and ploidy changes in the sequences of detected peptides in the context of their position in full-length proteins (Figure 7). Furthermore, we show the differences between sample groups for each peptide and the peptide regulation profiles in comparison between salivary gland tumor tissue and healthy tissue samples (Figure 7).

[0095] The inventors demonstrated the potential application of tissue extracellular peptidomethylaminucleochemistry to the discovery of tissue-derived biomarkers, using quantitative analysis and the following two peptides as examples. The analytical data revealed two peptides: one derived from vimentin, a typical intermediate filament protein found in mesenchymal cells, and the other from latent transformation growth factor β-binding protein 2 (LTBP2), an extracellular matrix protein and growth factor-binding protein. The vimentin peptide originates from the N-terminal region of the protein (17-34 amino acids) and is elevated in salivary gland tumor tissue samples. However, upon closer examination of the data, the inventors discovered that almost all vimentin-derived peptides exhibited a similar trend (Figure 7). These peptides are non-randomly distributed across the protein sequence, concentrated in the N-terminus (1-140 amino acids), C-terminus (410-466 amino acids), and the central region of the sequence (260-320 amino acids). These peptides may simply represent patterns of protein degradation and signaling regulation. In the case of LTBP2, the peptide in question showed significantly elevated expression in salivary gland tumor tissue samples. The inventors observed that other peptides derived from this protein showed a similar trend. They all originated from a relatively distinct region located near the N-terminus, approximately 70–140 amino acids. Therefore, the overexpression of vimentin and LTBP2 in salivary gland tumor tissue samples suggests involvement in the progression of salivary gland tumors and makes them potential biomarkers for health assessment, disease diagnosis, and novel antigen discovery. Furthermore, the extratissue peptidomix approach clearly distinguishes between peptides that undergo different and unregulated processes, even if they share the same protein origin (Figure 7). This suggests that in this cell model, different isoform degradation, partial degradation, or other more complex processes exist with respect to these proteins. Therefore, these correlations between different peptides with potentially identical origins may be useful for the biological interpretation of results in salivary gland tumor samples. Thus, this novel approach for extratissue peptidomix sample preparation is comprehensive, qualitative, quantitative, and provides insights into biomarker discovery.Therefore, it will revolutionize the fields of health status assessment, disease diagnosis, and novel antigen discovery, and will provide a roadmap for point-of-care diagnostics (POCD) development. Ultimately, it will improve quality of life and be applied worldwide. Thus, this invention will bring a breakthrough to point-of-care diagnostics (POCD) platforms and revolutionize the medical field. In the future, it will be applicable to all types of (human / non-human) control serum / disease serum samples.

[0096] [Example 3 (Tissue MHC Class I Immunopeptome by Antibody-Free Approach)] The antibody-free approach to tissue MHC class I immunopeptidomethic analysis, when combined with immunoprecipitation (IP), is a simple and economical parallel approach that can be used for novel antigen discovery. Following the isolation and purification of tissue-derived MHC class I immunopeptidomethic analysis using the antibody-free approach, the next crucial step in this invention is measurement based on mass spectrometry, such as DDA (data-dependent acquisition) mode or DIA (data-independent acquisition) mode. Mass spectrometry combined with chromatographic separation is necessary to identify thousands of peptides. Because MHC class I immunopeptidomethic analysis is nonspecific, successful identification requires significantly higher data quality, particularly mass precision, than standard trypsin-based proteome samples. Technological advancements in mass spectrometry have only recently made it possible to collect data of acceptable quality for such analysis. DIA (data-independent acquisition) methods allow for the collection of quantitative data for all peptides present in the sample while maintaining excellent resistance to matrix interference.

[0097] [MHC Class I Immunopeptide LC-MS / MS (Liquid Chromatography using Tandem Mass Spectrometry) Analysis] All dried immunopeptide samples were resuspended in 30 μL of loading buffer (containing water with 0.08% trifluoroacetic acid (TFA) and 2.5% acetonitrile (ACN)). Each sample was supplemented with retention time standard iRT peptide (Biognosys, Switzerland) according to the manufacturer's guidelines. Next, 6 μL of the dissolved sample was added to Thermo Scientific, UltiMate TM The peptide was injected into a 3000 RSLC nano System (Thermo Scientific, MA, USA) liquid chromatography system. The peptide was then processed using a 5 μL / min loading buffer flow to obtain Acclaim. TM PepMap TM The peptide was loaded into a 100Cl8, 5μM, ID=1mm, length=5mm (Thermo Scientific, MA, USA) trap column. The peptide was then processed using PepMap. TM Analytical separation was performed using a 100Cl8, 5μM, ID=1mm, length=5mm, particle size=2μM (catalog number: 164534) (Thermo Scientific, MA, USA) analytical column. The separation was performed by nonlinearly increasing mobile phase B (0.1% formic acid (FA) acetonitrile solution (v / v)) in mobile phase A (0.1% formic acid (FA) aqueous solution (v / v)). The nonlinear gradient started at a B concentration of 2.5%, increased linearly to 35% over 180 minutes, and then linearly increased to 60% at a constant flow rate of 300 nl / min over 15 minutes. Immunopeptides eluted from the column were ionized using a nanoelectrospray ion source (NSI) and connected online to a liquid chromatograph using an Orbitrap Exploris. TM It was introduced to the 480 mass spectrometer (Thermo Scientific, MA, USA).

[0098] Data-dependent acquisition (DDA), abbreviated as Exploris 480, was used to acquire data in data-dependent peptide mode. Full scans were performed in profile mode with a resolution of 120,000, and the precursor MS scan range was set from m / z 350Th to m / z 1650Th. The normalized AGC target was set to 300%, and the maximum injection time was 100 milliseconds. After each MS scan, the top 20 precursor ions with the highest intensity were fragmented, and their MS / MS spectra were acquired. Dynamic mass exclusion was set 20 seconds after the initial precursor ion fragmentation. Precursor isotopes were excluded, and the precursor separation mass tolerance was set to 10 ppm. The minimum precursor ion intensity was set to 3.0e3, and only precursor charge states from +1 to +5 were included in the experiment. The precursor separation window was set to 1.6Th. Normalized collision energy type and fixed collision energy mode were selected. The collision energy was set to 30%. The orbitrap resolution was set to 60000. The normalized AGC target was set to 100%. The maximum injection time was set automatically, and the data type was set to centroid.

[0099] Data-independent acquisition (DIA) was used to acquire quantitative tissue immunopeptidomics data. LC separation parameters were the same as for DDA acquisition. The Orbitrap Exploris 480 mass spectrometer was operated in positive polarity data-independent mode (DIA) and a full scan was performed at a resolution of 60000. The full scan mass range was set from m / z 350Th to m / z 1450Th, the normalized AGC target was set to 300%, and the maximum injection time was 100 milliseconds. The DIA method was cycled with a window width of 12Th and an overlap of 1Th over the mass range from m / z 350Th to m / z 1100Th. This method performed 62 precursor window / scan events per cycle. Normalized collision energy type and fixed collision energy mode were selected. The collision energy was set to 30%, and the Orbitrap resolution was 30000. The normalized AGC target was set to 1000%, and the maximum injection time was automatically set. The data type was a profile.

[0100] [Data analysis of MHC class I immunopeptides] Qualitative tissue MHC class I immunopeptide elution analysis and spectral library preparation. A multi-search engine strategy was employed to identify immunopeptides in both DDA and DIA MS data. Raw DDA files were centroidized using MSconvert version 3.0.19094 and converted to mzML and mzXML formats. Raw DIA files were processed directly with DIA Umpire SE, integrated into FragPipe (v.15). Extracted pseudo-DDA data were converted to centroidized mzML and mzXML formats. mzML files were searched using MSFragger 3.4, and mzXML files using the Comet (release 2020.01, revision 2) engine in the Homo sapiens SwissProt+UniProt and iRT (Biognosys, Switzerland) search databases (December 2022). Target search databases were linked with reverse decoy databases containing the target's reverse target sequences and common contaminating protein sequences. The following search settings were used in MSFragger: The precursor mass tolerance was set to ±8 ppm, and the fragment mass tolerance to 10 ppm. Enzyme digestion was set nonspecifically, and the peptide length was set to 7–45 amino acids. The peptide mass range was set to 200–5000 Da. Variable modifications were set to methionine oxidation and protein N-terminal acetylation. The data were mass recalibrated, and the fragment mass tolerance was adjusted using the automatic parameter optimization setting. The output file format was set to pep.XML. The comet precursor mass tolerance was set to 8 ppm, and the fragment mass tolerance was set to 8 ppm. Other settings were the same as MSFragger. The search results (pep.XML) file was processed with PeptideProphet and iProphet as part of TPP5.2.0. Peptide identification in the obtained pep.XML file was further processed in R for visualization and visualized with the ggplot2 and eulerr6.1.1.R packages. The pep.XML file, containing the recalculated peptide probabilities, was further processed against the spectral library using Skyline-daily (64-bit, 20.1.9.234).Before loading the pep.XML files, I set up an empty document in Skyline-daily. In "Peptide Settings," I selected the "Library" tab. I selected the "Build" option, set the "Cutoff Score" to 0.99 in the newly popped-up window, and selected "Biognosys-11(iRT-C18)" as the iRT standard peptide. All pep.XML files were loaded. Next, in the "Digestion" tab of the "Background Proteome" section, I selected the "Add" option. In the newly popped-up window, the background proteome was created from the .FASTA file that I had previously used as the search library. Next, in the "Library" section of the "Library" tab, I selected the newly created spectral library and clicked the "Explore" button to open the library. In the library window, I checked the "Protein Association" option and clicked the "Add All" button. Based on the retention time values ​​of the iRT peptides observed in the library pep.XML files, a retention time calculator was automatically generated. The Skyline file thus created was saved to the PC's hard drive.

[0101] [Data-Independent Acquisition (DIA) Data Extraction] DIA data extraction was performed using Skyline-daily (64-bit, 20.1.9.234) with the spectral library and documentation generated in the previous step. In the "Peptide Settings" tab, "Maximum number of missing cleavages" was set to 0. Next, in the "Filter" tab, the maximum peptide length was set from 4 to 200 amino acids. The "Exclude N-terminal amino acids" option was set to 0. The "Automatically select all matching peptides" function was selected. In the "Modification" tab, no modifications were selected. In the "Library" window of the "Library" tab, the spectral library created in the previous step was selected. Other tab settings were left at their defaults. In the transition settings, the "Prediction" tab was left at its default. In the "Filter" tab, peptide precursor charges of +1, +2, +3, +5, and +6 were specified. Ionic charges were set to +1 and +2. Ionic types were set to y and b. Product ion selection was set as follows: In the "From:" window, the "Ion 4" option was selected, and in the "To:" window, the "Last Ion" option was selected. The "Automatically select all matching transitions" option at the bottom of the tab was selected, and the "N-terminal to Proline" option was selected in the "Special ions:" menu. The "Ion match tolerance" window in the "Library" tab was set to 0.05 m / z, and only peptides with at least four product ions were retained for analysis. Furthermore, if more product ions were available for a peptide, the six strongest ones were selected from the filtered product ions. The function "From filtered ion charges and types" was selected. In the "Instrument" tab, the inventor included product ions from m / z 350 to m / z 1100. The "Method match tolerance m / z" window was set to a mass tolerance of 0.055 m / z.In the Full Scan tab, MSI filtering was set to none, the "DIA" option was selected in the "Acquisition method" window of the MS / MS filtering section, and the "Orbitrap" option was selected in "Product mass analyzer". The "Add" option was selected for "Isolation scheme:", and the "Prespecified isolation windows" option was selected in the "Edit isolation scheme" pop-up window. Next, the option to "Import" isolation windows from the DIA.raw file was selected, and the isolation scheme was read from the DIAraw file. In the "Resolving power:" section of the "Full-Scan" tab, the resolution was set to 30000 at 200m / z. In the retention time filtering section, it was specified to "Use only scans of MS / MS IDs within 5 minutes". In the "Purification" tab in the "Document" section of the "Advanced Settings" window, the "Minimum number of transitions per precursor" option was set to 4. Empty proteins were removed from the document. The same number of reverse sequence decoys were added using the "Add decoy" function in the "Purification" tab. In the "Add Decoy Peptide" pop-up window, the reverse sequence decoy generation method for creating the decoy peptide was selected. Next, the DIA.raw file was imported. The mProphet model to reintegrate the peptide boundary was trained as follows: The "Reintegrate" function was selected in the "Purify" tab. In the "Reintegrate" pop-up window for "Peak Scoring Model," the "Add" option was selected. Next, in "Edit Peak Scoring Model," the mProphet model was activated. The mProphet model was trained using the target and decoy. The mProphet model was retrained by deleting the "Score Rows" where the "Weights" and / or "Contribution Rate" (highlighted in red) were negative values. The new mProphet peak scoring model was selected in the "Edit Peak Scoring Model" window and applied to the reintegration of the new peak boundary.I created a quantitative report for downstream analysis using the report export function, which exports an overview of all dependencies required for MSstats 4.0.1.

[0102] [Data-Independent Data Acquisition (DIA) statistical data analysis in Msstats] Statistical analysis of DIA data extracted from Skyline was performed using the R (version 4.1.2) package MSstats 4.0.1. Analysis at the peptide level was maintained by combining protein and peptide sequence columns. This prevented the summation of peptide intensities originating from a single protein. Extracted peak groups were reduced by filtering with an mProphet q value cutoff of 0.01. The SkylinetoMSstatsFormat function was configured to retain proteins as a single feature and convert Skyline output to MSstats input. Furthermore, peptide intensities were log2 transformed and quantile normalized. Differential quantification of immunopeptides under specified conditions was performed pairwise using a mixed-effects model implemented in the MSstats "groupComparison" function. p-values ​​were adjusted using the Benjamini-Hochberg method, and the output matrix was exported for downstream analysis. Volcano plots were generated with Enhanced Volcano 1.10.0, and Venn diagrams with Eulerr 6.1.1. The bar graphs are created using the plyr1.8.6 package and ggplot2 3.3.5. All were executed on R (version 4.1.2).

[0103] [Tissue MHC Class I Immunopeptide Prediction] To predict the binding affinity of peptide sequences to allele sets with a frequency of 6.4% or higher, the inventors used the MHCflurry package (version 3.1.0) in Python 3.9.13. This package utilizes a pre-trained deep learning model and a large dataset of peptide-MHC binding affinities. Peptides with predicted affinity less than 50 were classified as strongly binding, those with affinity between 50 and 150 as moderately binding, those with affinity between 150 and 500 as weakly binding, and those with affinity above 500 as unbinding.

[0104] Allele-independent MHCflurry prediction utilized deep learning architectures such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to learn complex patterns from peptide sequences and their physicochemical properties. These models were trained on diverse datasets containing peptide-MHC binding affinity measurements for various MHC alleles. By extracting relevant features from input peptide sequences, the models accurately predicted peptide-MHC binding affinity independently of specific alleles. In presentation prediction, peptides with an allele-independent presentation score above 0.8 were deemed likely to be presented, while peptides with an allele-independent presentation score below 0.8 were deemed unlikely to be presented and therefore considered HLA-unbound peptides. The percentile distribution of MHCflurry presentation scores evaluated the relative ranking of peptide presentation predictions across a diverse population of peptides in MHCflurry. Factors such as peptide sequence, length, and potential MHC binding affinity were considered. Peptides with a presentation percentile of less than 1% were considered to have a favorable presentation percentile, while peptides with a presentation percentile greater than 1% were considered to have an unfavorable presentation percentile. Furthermore, the inventors employed a deep learning architecture trained on a real dataset to predict processing by the proteasome. Peptides with a presentation score greater than 0.8 were considered to be likely to be processed, while peptides with a presentation score less than 0.8 were considered to be unlikely to be processed. The MHCflurry results were visualized as bar graphs using the ggplot2 R package.

[0105] To evaluate the antigenicity of MHC class I peptides, the inventors used the IEDB MHC class I peptide antigenicity prediction tool in Python 3.9.13. The prediction script was obtained from the Immunoepitope Database (IEDB) website (https: / / www.iedb.org / ). The IEDB estimated the immunogenicity of peptides using a computational algorithm and prediction model that takes into account factors such as peptide length, amino acid sequence, and MHC molecular mask. Thresholds for determining immunogenic and non-immunogenic immunopeptides were set according to the recommendations of the IEDB database. Peptides with an immunogenicity score of less than 0 were determined to be non-immunogenic, and peptides with a score greater than 0 were determined to be immunogenic. The results of the IEDB MHC class I peptide antigenicity prediction were visualized as bar graphs using the ggplot2R package.

[0106] The inventive antibody-free approach to tissue MHC class I immunopeptideome analysis focuses on the quantitative evaluation of complex peptide mixtures, revealing the diverse properties of the tissue MHC class I immunopeptideome. The workflow of this invention comprehensively identifies at least 15,671 MHC class I peptides (iPROB<0.99) from 2,290 protein precursors in both NSCLC tissue (n=5) and healthy donor (n=5) samples. Of these, at least 8,565 MHC class I peptides were quantified by comparing healthy tissue (n=5) and NSCLC tissue (n=5) samples (mPROPHET q value<0.01). Figure 8 provides detailed insights into tissue MHC class I immunopeptideome peptide identification using an antibody-free approach. Figures 8A-C include additional details and visual representations showing MHC class I peptide identification in individual tissue samples and information regarding overlap in peptide identification between NSCLC tissue and control tissue sample groups. The Venn diagram highlights the significantly higher number of peptides in the NSCLC tissue group compared to healthy tissue (control tissue). Each search engine, such as COMET (CMT) and MS FRAGGER (FRG), incorporates algorithms to consider the comprehensiveness of data analysis. This invention employs two types of mass spectrometry data acquisition—data-dependent acquisition (DDA) and data-independent acquisition (DIA)—and uses multiple search engine strategies to create a spectral library of qualitative and quantitative tissue MHC class I immunopeptidomes, facilitating the distinction between the NSCLC tissue and control tissue sample groups. As a result, a comprehensive and qualitative tissue MHC class I immunopeptidome was identified for the first time (Figure 8). The next step is quality control of the MHC class I immunopeptidome. The peptide length distribution between 7-16 mer amino acids is a key feature of MHC class I peptides. Figure 9 shows the peptide length distribution of all peptides identified in the tissue MHC class I immunopeptidome. Most peptides fall within the 7-16 mer range, with a greater proportion of 9 mer peptides. This is consistent with the characteristic length of the MHC class I immunopeptideome.To date, several platforms are available for predicting the tissue MHC class I immunopeptideome. The present invention, the MHCflurry platform, was implemented to provide an overview of tissue MHC class I immunopeptide enrichment between contaminating peptides in quantitative immunopeptidemix analysis comparing non-small cell lung cancer (NSCLC) tissue with healthy tissue (Figure 10). MHCflurry utilizes deep learning architectures such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to predict peptide-MHC binding affinity in an allele-independent manner. Furthermore, an ensemble model was used to capture allele-specific binding preferences, taking into account the unique characteristics and preferences of each MHC allele, such as peptide anchor location and residue preferences (Figure 10). Antigen processing prediction is also based on machine learning models trained on experimentally observed proteasome cleavage sites (Figure 10). These models identify proteasome cleavage sites by considering sequence motifs, amino acid characteristics, and context-specific information. Furthermore, the predictions incorporated the preference of transporter (TAP) complexes associated with antigen processing, which play a crucial role in peptide transport for MHC loading. Additionally, the IEDB MHC class I peptide binding prediction workflow was implemented to estimate the immunogenicity of isolated peptides (Figure 10). Immunogenicity determination depended on parameters such as peptide length, amino acid sequence, and MHC molecule characteristics. Considering these approaches provided insights into the enrichment of immunopeptides in tissue class I immunopeptide mix samples obtained from isolated NSCLC tissue and healthy control tissue. Integrating these prediction results, a comprehensive understanding of the MHC class I immunopeptide composition of tissue immunopeptide mix isolates obtained through antibody-free approaches was achieved, as shown in Figures 10A–F, providing insights into the dynamics of immunorecognition and peptide-MHC class I interactions in NSCLC tissue samples.

[0107] The present invention's quantitative tissue MHC class I immunopeptide-mix analysis using an antibody-free approach provides insights into the discovery of novel antigens. Following multi-bioinformatics analysis, the present invention's quantitative tissue MHC class I immunopeptide-mix workflow revealed variability in immunopeptide-mix signaling within NSCLC tissue samples compared to a healthy sample group, indicating a decay of lymphocyte-associated immunopeptide-mix in NSCLC tissue samples. This may be a source of novel antigens, such as NSCLC tumor-specific mutations or abnormally expressed proteins, making them attractive targets for cancer immunotherapy. However, further validation is needed to determine the efficacy of immunotherapy. In particular, quantitative tissue MHC class I immunopeptide-mix analysis aimed to identify and characterize differences in peptide repertoire between healthy tissue and non-small cell lung cancer (NSCLC) tissue samples. Comprehensive analysis using gene set enrichment analysis (GSEA) revealed a subset of significantly altered peptides and their corresponding protein precursors belonging to the lymphocyte activation term (G0:0046649), as shown in the GSEA results in Figure 11. The enrichment pattern suggests reduced activity of lymphocyte-related processes in NSCLC tissue, represented by changes in the immunopeptide profile (Figure 11). Furthermore, Figure 12 summarizes the human leukocyte antigen (HLA) system (human major histocompatibility complex [MHC]) HLA binding assessment to a subset of MHC class I peptides that showed significant changes related to the lymphocyte activation term in Figure 11. Figure 12 shows the differential regulation of HLA binders derived from this term in healthy tissue compared to NSCLC tissue samples. In contrast, a population of MHC class I peptides showing opposing regulation was observed, but generally, most of these peptides were predicted not to bind to HLA molecules. This suggests that this peptide population contains many impurities and may not be directly involved in the changes in the presentation pattern indicating lymphocyte inactivation in NSCLC samples.The different regulatory mechanisms observed in HLA-binding peptides associated with lymphocyte activation further support the hypothesis of lymphocyte attenuation in NSCLC tissue samples compared to healthy tissue samples, and we further provide a list of immunopeptide biomarkers / neoantigens (Table 5).

[0108] Finally, the inventors selected several unique MHC class I peptidomes as the origin of neoantigen peptidomes and biomarker candidates for the treatment or diagnosis of NSCLC. Figure 13 shows a volcano plot illustrating abnormal protein sources and their corresponding MHC class I immunopeptide sequences derived from the protein precursor of lymphocyte activation term (GO: 0046649) for diagnosing lymphocyte attenuation in non-small cell lung cancer (NSCLC), distinguishing NSCLC tissue from healthy tissue samples (Figures 11, 12). The volcano plot focuses particularly on the selected targets proposed in the invention and represents immunopeptides derived from the protein precursor linked by lymphocyte activation term (G0:0046649). The volcano plot in Figure 13 shows the multiplicative changes (x-axis) and statistical significance (y-axis) of the expression or regulation of abnormal MHC class I immunopeptides. Each point on the plot represents an immunopeptide, and its position indicates both the magnitude of regulation and the level of statistical significance. The labels selected on the plot highlight the regulation of MHC class I immunopeptides derived from protein precursors associated with the lymphocyte activation term (GO:0046649). These labels provide crucial information regarding the upregulation and downregulation of specific MHC class I immunopeptides, and their potential relevance to lymphocyte activation in NSCLC samples. Thus, all of these at least 25 quantitative immunopeptides (Table 5), HLA binders, and peptide lengths of 7–16 amino acids (mostly 9 nucleotides) constitute the MHC class I immunopeptide pattern of lymphocyte-associated immune responses (lymphocyte attenuation) in NSCLC.

[0109] [Table 5]

[0110] Table 5. MHC class I quantitative tissue immunopeptide therapy panel for non-small cell lung cancer (NSCLC). Sequence of therapeutic MHC class I peptides, origin protein identifier, length, HLA binding predictive value for each peptide, and fold change in upregulation (↑) and downregulation (↓) peptides in comparison of healthy tissue and NSCLC tissue samples. This regulation / fold change refers only to MHC class I tissue quantitative peptides. This indicates how many times higher the signaling intensity of healthy tissue samples was compared to NSCLC tissue samples. At least 25 quantitative immunopeptides shown in Table 5 reveal MHC class I immunopeptide sequences derived from protein precursors of lymphocyte activation terms for diagnosing lymphocyte attenuation in NSCLC tissue samples.

[0111] The discovery of potential mechanisms of lymphocyte dysfunction in NSCLCs suggests that these immunopeptides may be useful in assessing health status, diagnosing diseases, discovering new antigens, and developing therapeutic strategies to address lymphocyte-associated immune responses in NSCLCs.

[0112] [Reference] [1] R. Richter et al., “Composition of the peptide fraction in human blood plasma: database of circulating human peptides,” J. Chromatogr. B. Biomed. Sci. App., vol. 726, no. 1, pp. 25-35, Apr. 1999, doi: 10.1016 / S0378-4347(99)00012-2. [2] T. Cardon et al., “Optimized Sample Preparation Workflow for Improved Identification of Ghost Proteins,” Anal. Chem., 2020, doi: 10.1021 / acs.analchem.9b04188. [3] B. Fabre, J.-P. Combier, and S. Plaza, “Recent advances in mass spectrometry-based peptidomics workflows to identify short-open-reading-frame-encoded peptides and explore their functions,” Curr. Opin. Chem. BioL, vol. 60, pp. 122-130, Feb. 2021, doi: 10.1016 / j.cbpa.2020.12.002. [4] A. W. Purcell, S. H. Ramarathinam, and N. Ternette, “Mass spectrometry-based identification of MHC -bound peptides for immunopeptidomics,” Nat. Protoc., vol. 14, no. 6, pp. 1687-1707, Jun. 2019, doi: 10.1038 / s41596-019-0133-y. [5] F.-R. Schumacher et al., “Building proteomic tool boxes to monitor MHC class I and class II peptides,” Proteomics, vol. 17, no. 1-2, Jan. 2017, doi: 10.1002 / pmic.201600061. [6] M. B as sani- Sternberg, S. Pletscher-Frankild, L. J. Jensen, and M. Mann, “Mass spectrometry of human leukocyte antigen class I peptidomes reveals strong effects of protein abundance and turnover on antigen presentation,” Mol. Cell. Proteomics MCP , vol. 14, no. 3, pp. 658-673, Mar. 2015, doi: 10.1074 / mcp.Ml 14.042812. [7] D. Gebreselassie, H. Spiegel, and S. Vukmanovic, “Sampling of major histocompatibility complex class I-associated peptidome suggests relatively looser global association of HLA-B*5101 with peptides,” Hum. Immunol., vol. 67, no. 11, pp. 894-906, Nov. 2006, doi: 10.1016 / j.humimm.2006.08.294. [8] L. Wu et al., “Peptidomic Analysis of Cultured Cardiomyocytes Exposed to Acute Ischemic-Hypoxia,” Cell. Physiol. Biochem., vol. 41, no. 1, pp. 358-368, 2017, doi: 10.1159 / 000456282. [9] Y. Xue et al., “Peptidomic Analysis of Endometrial Tissue from Patients with Ovarian Endometriosis,” Cell. Physiol. Biochem. Int. J. Exp. Cell. Physiol. Biochem. Pharmacol., vol. 47, no. 1, pp. 107-118, 2018, doi: 10.1159 / 000489753.

[10] G. P. Roberts et al., “Comparison of Human and Murine Enteroendocrine Cells by Transcriptomic and Peptidomic Profiling,” Diabetes, vol. 68, no. 5, Art. no. 5, May 2019, doi: 10.2337 / dbl8-0883.

[11] C.-X. Zhou, M. Gao, B. Han, H. Cong, X.-Q. Zhu, and H.-Y. Zhou, “Quantitative Peptidomics of Mouse Brain After Infection With Cyst-Forming Toxoplasma gondii,” Front. Immunol., vol. 12, 2021, Accessed: Feb. 15, 2023. [Online], Available: https: / / www.frontiersin.org / articles / 10.3389 / fimmu.2021.681242

[12] L. Zhang et al., “Peptidomics Analysis Reveals Peptide PDCryabl Inhibits Doxorubicin-Induced Cardiotoxicity,” Oxid. Med. Cell. Longev., vol. 2020, p. 7182428, Oct. 2020, doi: 10.1155 / 2020 / 7182428.

[13] Y. Xue et al., “Peptidomic Analysis of Endometrial Tissue from Patients with Ovarian Endometriosis,” Cell. Physiol. Biochem., vol. 47, no. 1, pp. 107-118, 2018, doi: 10.1159 / 000489753.

[14] J. Hu et al., “Peptidomic analysis on synovial tissue reveals galectin-1 derived peptide as a potential bioactive molecule against rheumatoid arthritis,” Cytokine, vol. 131, p. 155020, Jul. 2020, doi: 10.1016 / j.cyto.2020.155020.

[15] F. Zhang et al., “Peptidome Analysis of Pancreatic Tissue Derived from T1DM Mice: Insights into the Pathogenesis and Clinical Treatments of T1DM,” BioMed Res. Int., vol. 2021, p. 9987042, May 2021, doi: 10.1155 / 2021 / 9987042.

[16] X. Hua etal., “Peptidome analysis of human intrauterine adhesion tissues and the identification of antifibrotic peptide,” J. Biomed. Res., vol. 36, no. 4, pp. 280-296, Jul. 2022, doi: 10.7555 / JBR.36.20220059.

[17] T. Chen, Z. Zhang, Q. Lu, and J. Ma, “Screening and functional analysis of the differential peptides from the placenta of patients with healthy pregnancy and preeclampsia using placental peptidome,” Front. Genet., vol. 13, p. 1014836, 2022, doi: 10.3389 / fgene.2022.1014836.

[18] X. Li et al., “Comparative peptidome profiling reveals critical roles for peptides in the pathology of pancreatic cancer,” Int. J. Biochem. CellBioL, vol. 120, p. 105687, Mar. 2020, doi: 10.1016 / j.biocel.2020.105687.

[19] S. Dasgupta et al., “Proteasome Inhibitors Alter Levels of Intracellular Peptides in HEK293T and SH-SY5Y Cells,” PLOS ONE, vol. 9, no. 7, p. el03604, Jul. 2014, doi: 10.1371 / journal. pone.0103604.

[20] T. GARCIA-BERROCOSO and J. M. Vilallonga, “Biomarkers for the prognosis of ischemic stroke,” W02014064202A1, May 01, 2014 Accessed: Feb. 15, 2023. [Online], Available: https: / / patents. google. com / patent / W02014064202Al / en?oq=W02014%2f064202Al

[21] C. Watson, M. Ledwidge, K. McDonald, and J. Baugh, “Biomarkers of cardiovascular disease including Irg,” US20130084276A1, Apr. 04, 2013 Accessed: Feb. 15, 2023. [Online], Available: https: / / patents. google. com / patent / US20130084276Al / en?oq=US2013%2f0084276Al

Claims

1. A method for screening extracellular tissue peptides in mammalian tissue samples for health status assessment, disease diagnosis, and discovery of new antigens, comprising amino acid sequencing and tandem mass spectrometry, and measurement of peptide signal intensity, characterized in that the two steps of sample preparation are carried out as follows. a) Wash the tissue sample two to three times with washed phosphate-buffered saline (PBS) with a pH of 7.2 to 7.

5. Next, the tissue sample is incubated with an extraction buffer prepared in ultrapure water with citrate phosphate buffer (purity >99.9%) at pH 3.1–3.

6. Here, the amount of buffer is 7–8 ml per 0.05–1 gram of tissue sample, and the incubation step is carried out for 2–3 minutes with mixing at 4–10°C. b) The separated extracellular peptidomethyl phosphate is purified using a hydrophilic-lipophilic balance column at room temperature for 120–180 minutes, then centrifuged at at least 4000 g at 4–10°C for 120–135 minutes using a 3 kDa molecular weight cutoff filter, and separated by ultrafiltration.

2. a) The method according to claim 1, wherein the tissue sample is washed 7 to 10 times with phosphate-buffered saline (PBS) pH 7.2 to 7.5 without additives (calcium / magnesium / phenol red), and then the tissue sample is incubated with an extraction buffer (citrate phosphate buffer) pH 3.3 prepared with liquid chromatography-mass spectrometry (LC-MS) grade ultrapure water of purity >99.9%, the volume of the buffer being 5 to 6 ml per 0.5 to 1 gram of tissue sample, the incubation step taking 4 to 5 minutes with mixing at a temperature of 4 to 10°C, and the purification of the isolated tissue MHC class I immunopeptideome being carried out at room temperature for 120 to 180 minutes using a hydrophilic-lipophilic balance column, and further separated by ultrafiltration by centrifugation at at least 4000 g at 4 to 10°C for 120 to 135 minutes using a selected molecular weight cutoff filter having a molecular weight cutoff of 3 kDa.

Images