Metabolic gene characteristic-based squamous cell carcinoma marker as well as typing and application thereof

Abstract

According to the application of the metabolic gene marker combination in construction of the squamous cell carcinoma typing model, the gene markers comprise 16 types such as CAD, CYP27B1, SLC44A1, GGT5, GLS, PPT1, GFPT2, NT5E, PADI3, PLA2G7, LPCAT1, SLC16A1, SLC7A7, GMPS, VKORC1 and SLC3A2, and the gene markers are divided into metabolic gene low expression types and metabolic gene high expression types. Through verification, the marker combination can be suitable for squamous cell carcinoma of four different organs, has high general applicability, effectively reduces the research and development cost of independently developing a typing system for SCC cancer species of various organs, provides a uniform molecular typing platform for precise medical treatment of the generic cancer species, not only reduces the detection cost, but also is low in technical threshold, and has wide application prospects. The detection period is shortened, and the clinical application efficiency is improved.

Description

Technical Field

[0001] The present invention relates to a new use of a known compound, in particular to a detection marker for squamous cell carcinoma screening, and the use of the marker in diagnostic models and the preparation of diagnostic kits. Background Art

[0002] Currently, squamous cell carcinoma (SCC), a tumor type that is widely present in multiple organs (such as the head and neck, cervix, esophagus, and lungs), has complex molecular characteristics and strong heterogeneity. Traditional tumor classification methods mostly rely on histological morphology, gene mutations, or single transcriptome analysis methods, but fail to fully integrate metabolomics information, making it difficult to reflect the metabolic reprogramming state within the tumor and its association with prognosis and immune environment.

[0003] The main existing molecular typing methods include:

[0004] Typing based on gene expression profiles, such as PAM50 and Oncotype DX, lacks metabolic dimension information.

[0005] Typing based on mutation spectrum, such as TMB and MSI typing, cannot reflect functional metabolic status.

[0006] Typing based on immune signatures, such as the immune scoring system, ignores metabolic-immune interactions.

[0007] Based on the analysis of single metabolites, it lacks systematic and cross-cancer conservation validation.

[0008] In summary, the main defects of the existing technology are as follows:

[0009] 1. Single information dimension: Most existing classification methods are based on a single omics data set (transcriptome or metabolome), which cannot fully reflect the biological complexity of tumors, especially the synergistic mechanism of metabolic reprogramming and gene regulatory networks.

[0010] 2. Insufficient conservatism across cancer types: Existing classification methods are mostly developed for specific cancer types and lack universal verification across squamous cell carcinomas of different organs, limiting their application value in pan-cancer precision medicine.

[0011] An in-depth analysis of existing squamous cell carcinoma classification technologies revealed that they often rely on single-omics data (e.g., gene expression or mutation data), making it difficult to accurately reflect the metabolic state and immune escape characteristics of tumors. Significant deficiencies in metabolic-transcriptome integration, cross-cancer conservation validation, and clinical translational applications limit their clinical application in prognostic assessment and treatment decision-making. Existing technologies are unable to effectively capture key regulatory genes for tumor metabolic reprogramming, and lack a systematic understanding of the relationship between metabolic state and the immune microenvironment, prognosis, and treatment response.

[0012] To overcome the above-mentioned defects, there is an urgent need for a typing method that integrates metabolomics and transcriptomics to identify functional subtypes that are conserved across cancer types and to achieve a more guiding molecular typing strategy, which will be beneficial for its application in prognosis prediction and treatment guidance. Summary of the Invention

[0013] One purpose of the present invention is to provide a combination of squamous cell carcinoma detection markers, obtain molecular subtypes based on multi-omics data, and establish a prediction model closely related to squamous cell carcinoma (SCC) metabolic genes to predict the patient's 5-year survival rate.

[0014] Another object of the present invention is to provide a squamous cell carcinoma detection marker combination to improve the versatility of predicting squamous cell carcinoma in various organs.

[0015] Another object of the present invention is to provide a combination of squamous cell carcinoma detection markers to provide a reference basis for formulating clinical treatment plans for patients and improve the level of personalized medical treatment.

[0016] Another object of the present invention is to provide a squamous cell carcinoma detection marker combination and its application in constructing a squamous cell carcinoma diagnostic model.

[0017] The fifth object of the present invention is to provide a squamous cell carcinoma detection kit to improve the convenience of detection and typing of squamous cell carcinoma patients, shorten the detection cycle, and improve clinical application efficiency.

[0018] The present invention systematically collects metabolite data in patient serum based on a large sample size, and combines it with broad-target metabolomics technology to screen out key metabolite combinations with statistical significance and disease relevance from hundreds of metabolites, thereby establishing a highly sensitive and specific HNSCC metabolic diagnostic model suitable for early clinical screening and auxiliary diagnosis.

[0019] Using a systems biology strategy, metabolomics analysis was performed on serum samples from HNSCC (Head and Neck Squamous Cell Carcinoma, SCC) patients to identify differential metabolites. Human metabolome database (HMDB) annotations and TCGA transcriptome data were integrated to identify core metabolic regulatory gene sets. A molecular classification system for squamous cell carcinoma based on 16 conserved metabolic genes was established, and its clinical value and cross-cancer applicability were verified through multidimensional biological feature analysis.

[0020] The levels of dihydroorotase (CAD), 25-hydroxyvitamin D3-1α-hydroxylase (CYP27B1), solute carrier family 44 member A1 (SLC44A1), gamma-glutamyltransferase 5 (GGT5), glutaminase (GLS), palmitoyl protein thioesterase 1 (PPT1), glutamine-fructose-6-phosphate transaminase 2 (GFPT2), 5'-nucleotidase (NT5E), peptidyl arginine deiminase 3 (PADI3), and phospholipase A group 7 (PLA2G7) were detected. The expression results of 16 genes, including lysophosphatidylcholine acyltransferase 1 (LPCAT1), solute carrier family 16 member A1 (SLC16A1), solute carrier family 7 member A7 (SLC7A7), guanylate synthase (GMPS), vitamin K epoxide reductase complex subunit 1 (VKORC1) and solute carrier family 3 member A2 (SLC3A2), were divided into: metabolic gene low expression type, recorded as: C1 subtype, and metabolic gene high expression subtype, recorded as: C2 subtype.

[0021] The high expression or low expression referred to in the present invention is defined by comparing the difference in the median expression between the two groups. It is a relative term rather than a clear numerical value to define "high" or "low".

[0022] Application of a metabolic gene marker combination in constructing a squamous cell carcinoma classification model.

[0023] Application of another metabolic gene marker combination in constructing a classification model for head and neck squamous cell carcinoma.

[0024] Subtype C1 has been shown to exhibit "immune-hot" tumor characteristics, while subtype C2 exhibits "immune-cold" tumor characteristics, with enhanced immune evasion and poorer response to immune checkpoint inhibitors. Compared with subtype C1, subtype C2 has a higher rate of new tumor events, a higher rate of mortality, and a higher frequency of TP53 mutations.

[0025] The C2 subtype was also found to be more sensitive to cisplatin, docetaxel, and gemcitabine, but less sensitive to lapatinib and paclitaxel. Therefore, clinically, when SCC patients have high expression of the 16 genes and are classified as C2 subtype, cisplatin, docetaxel, and gemcitabine should be prioritized over lapatinib and paclitaxel. This can help guide clinical practice in developing targeted medication regimens and achieve personalized, targeted treatment.

[0026] Based on metabolic-transcriptional integrated analysis, the application of the squamous cell carcinoma detection marker combination of the present invention in constructing a cancer typing model is beneficial for providing an objective molecular basis for cancer prognosis prediction and treatment guidance.

[0027] The application of the squamous cell carcinoma detection marker combination of the present invention in constructing a squamous cell carcinoma prognosis prediction model is beneficial for providing a basis for formulating a prognosis diagnosis and treatment plan for patients.

[0028] The application of the squamous cell carcinoma detection marker combination of the present invention in the drug administration prediction model for squamous cell carcinoma patients provides a basis for the targeted and accurate medication of patients and effectively avoids the improper use of drugs.

[0029] A detection device, such as a sequencing device, synchronously or asynchronously detects and obtains information on the squamous cell carcinoma detection marker combination of the present invention, and types this information to quickly provide objective information for clinical diagnosis and treatment.

[0030] A system for squamous cell carcinoma classification and prediction, comprising the following units:

[0031] a data aggregation unit that stores the detection data of 16 gene expressions in patients with squamous cell carcinoma;

[0032] an analysis unit that uses the results detected by the data aggregation unit as input items for analysis; and

[0033] The evaluation unit outputs the metabolic genotyping of the individual corresponding to the sample, as well as recommendations for prognosis prediction and treatment guidance.

[0034] The system of the present invention further includes a detection unit, wherein information of the detection unit is obtained by detection using a kit or a sequencing instrument. The kit or the sequencing instrument synchronously or asynchronously detects and obtains expression information of the 16 genes.

[0035] Compared with the existing technology in this field, the squamous cell carcinoma detection marker combination of the present invention has clinical prognostic prediction value, provides an objective basis for the patient's 5-year survival rate, and is conducive to clinical follow-up and the formulation of diagnosis and treatment plans.

[0036] The squamous cell carcinoma detection marker combination of the present invention provides a correlation between patient response rates to immune checkpoint inhibitors, providing an objective molecular basis for clinical drug use and improving the accuracy of medical decision-making. For example, the C2 subtype is sensitive to cisplatin, docetaxel, and gemcitabine, and these drugs should be given priority for such patients. The C1 subtype is sensitive to lapatinib and paclitaxel, and these drugs should be given priority for patients with the C1 subtype.

[0037] The squamous cell carcinoma detection marker combination of the present invention has been verified to be applicable to squamous cell carcinomas of four different organs, showing high universal applicability. It effectively reduces the R&D cost of developing separate classification systems for SCC cancers of various organs, and provides a unified molecular classification platform for pan-cancer precision medicine.

[0038] Compared with the whole transcriptome detection method, only the 16 squamous cell carcinoma detection markers of the present invention need to be detected to provide objective information for clinical diagnosis and treatment. Not only is the detection cost reduced, the technical threshold is also low. It can be implemented on qPCR or small-scale sequencing platforms, the detection cycle is shortened, and the efficiency of clinical application is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Provide a technical flowchart for integrated metabolomics-transcriptome analysis;

[0040] Figure 2 Schematic diagram of the screening strategy for core metabolic regulatory genes;

[0041] Figure 3 Figure 2 is the clustering result of HNSCC molecular subtypes; A is the result of principal component analysis (PCA), and B is the result of heat map analysis, which uses color gradient to show the expression levels of 16 core metabolic regulatory genes in different patient samples, with blue representing low expression and red representing high expression;

[0042] Figure 4 The following are multi-dimensional feature comparisons of various subtypes; A is a survival analysis result curve, B is a clinical feature comparison statistical chart, C is a differential gene enrichment analysis result chart, D is a TP53 mutation frequency analysis heat map, and E is a statistical chart of m6A-related gene expression differences;

[0043] Figure 5 This is a diagram analyzing the characteristics of the subtype-specific immune microenvironment in HNSCC;

[0044] Figure 6Figure 1 is a cross-cancer validation result diagram; A is the TCGA-LUSC (lung squamous cell carcinoma, n=501) cohort validation result diagram, from left to right, including PCA clustering results, survival analysis (Log-rank P=0.025, HR(C2)=1.369) results, and gene expression results; B is the TCGA-CSCC (cervical squamous cell carcinoma, n=253) cohort validation result diagram, from left to right, including PCA clustering results, survival analysis (Log-rank P=0.015, HR(C2)=2.017) results, and gene expression results; C is the TCGA-ESCC (esophageal squamous cell carcinoma, n=95) cohort validation result diagram, from left to right, including PCA clustering results, survival analysis (Log-rank P=0.0428, HR(C2)=0.728) results, and gene expression results;

[0045] Figure 7 These are treatment prediction analysis charts based on different classifications; A is a statistical chart comparing tumor immune dysfunction and rejection (TIDE) scores, and B is a statistical chart showing the IC50 prediction results of common chemotherapy drugs;

[0046] Figure 8 Statistical graph of 16 gene expression results for C1 and C2 types. DETAILED DESCRIPTION

[0047] The technical solution of the present invention is described in detail below with reference to the accompanying drawings. The embodiments of the present invention are intended only to illustrate the technical solution of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the invention may be modified or replaced with equivalents without departing from the spirit and scope of the technical solution of the present invention, and all such modifications or equivalents should be included in the scope of the claims of the present invention.

[0048] To elucidate the molecular mechanisms underlying the metabolic changes observed in HNSCC, this example established a multi-omics integrated analysis framework based on systems biology. Figure 1This is a flowchart of the integrated metabolomics-transcriptome analysis technology. It illustrates the complete technical process from multi-omics data acquisition to molecular subtype identification, including: data acquisition and preprocessing (targeted metabolomics analysis of 248 HNSCC and 250 healthy individuals' serum, using the TCGA database); screening of core metabolic regulatory genes (355 candidate genes were annotated using the HMDB database, 1,504 upregulated genes were identified using the TCGA-HNSCC cohort, 27 overlapping genes were obtained through intersection analysis, and 16 core genes were ultimately identified through biological validation); consensus clustering and subtype identification (C1 subtype: low metabolic gene expression, C2 subtype: high metabolic gene expression); multidimensional subtype feature analysis (covering seven dimensions: clinicopathological characteristics, survival prognosis, genomic variation, transcriptome functional enrichment, epigenetics, immune microenvironment, and drug sensitivity); and cross-cancer validation (verification of the conserved features of the subtype in the TCGA-LUSC, TCGA-ESCC, and TCGA-CSCC cohorts).

[0049] The biological screening criteria and process from 355 candidate genes to 16 core genes can be found in Figure 2 The figure shows the candidate gene annotation strategy on the left (based on the HMDB database metabolic gene annotation, screening of 355 metabolic-related candidate genes with biosynthesis, degradation, and transport functions); the differential expression analysis method on the right (using the DESeq2 algorithm to analyze the TCGA-HNSC cohort, with the screening conditions of FDR < 0.05 and |log2FC| > 1, identifying 1,504 up-regulated genes); the biological validation screening process is shown below (27 genes were obtained through intersection analysis of 355∩1,504, and 16 core genes were finally determined through biological consistency verification). These 16 core metabolic regulatory genes are clearly listed: CAD, CYP27B1, SLC44A1, GGT5, GLS, PPT1, GFPT2, NT5E, PADI3, PLA2G7, LPCAT1, SLC16A1, SLC7A7, GMPS, VKORC1, and SLC3A2.

[0050] Example 1 Molecular subtype identification and feature analysis

[0051] Step 1.1: Consistency cluster analysis

[0052] The expression data of 16 core genes in the TCGA-HNSC cohort were extracted and a gene expression matrix was constructed. ConsensusClusterPlus algorithm was used for consensus cluster analysis with the following parameters:

[0053] Number of clusters: k = 2

[0054] Resampling ratio: 80%

[0055] Resampling times: 1000 times

[0056] Distance metric: Euclidean distance

[0057] Clustering algorithm: Hierarchical clustering (complete linkage)

[0058] Figure 3 A shows the results of principal component analysis (PCA), which clearly shows the obvious separation of the two subtypes C1 (blue-black dots) and C2 (orange dots) in two-dimensional space, proving that the typing method based on 16 metabolic genes has good discrimination. Figure 3 Figure B shows the results of the heat map analysis, which uses a color gradient to display the expression levels of 16 core metabolic regulatory genes in different patient samples, with blue representing low expression and red representing high expression, clearly demonstrating the significant difference in gene expression patterns between the C1 and C2 subtypes.

[0059] Cluster analysis classified HNSCC patients into two stable molecular subtypes:

[0060] C1 subtype: metabolic gene low expression subtype (n=229)

[0061] C2 subtype: metabolic gene high expression subtype (n = 275)

[0062] According to statistics, the results are as follows Figure 8 As shown in Figure 3, the median expression of 16 genes in the two subtypes was compared, and it was found that the median expression of the 16 genes in the C1 subtype was lower than that in the C2 subtype.

[0063] Step 1.2: Subtype biological characteristics analysis

[0064] A systematic multidimensional biological characterization analysis of the two molecular subtypes was performed. Figure 4 A shows the results of survival analysis. The Kaplan-Meier survival curve shows that the overall survival rate of patients with C2 subtype (red line) is significantly lower than that of patients with C1 subtype (blue line). Log-rank test P = 0.042, hazard ratio HR (C2) = 1.321 (95% CI: 1.011-1.727). Figure 4 B shows the comparison of clinical characteristics. The pie chart comparison shows that the C2 subtype is significantly worse than the C1 subtype in terms of histological grade (Grade, P=0.011), overall survival (OS, P=0.0028), and progression-free survival (PFI, P=0.032); Figure 4 C shows the results of differential gene enrichment analysis. The bubble chart shows the main signaling pathways enriched in the C2 subtype, including ECM-receptor interaction, adherens junction, and PI3K-Akt signaling pathway; Figure 4D shows the analysis of TP53 mutation frequency. The heat map shows that the TP53 mutation frequency of C2 subtype (77.6%) is significantly higher than that of C1 subtype (58.8%). 2 Test statistic = 18.94, P < 0.001; Figure 4 E shows the differences in expression of m6A-related genes. The box plot shows that most m6A regulatory factors (including "writers", "readers" and "erasers") are significantly overexpressed in the C2 subtype (P<0.001), especially the significant upregulation of IGF2BP1 / 2 / 3, suggesting enhanced mRNA stability and translation efficiency.

[0065] Based on these results, the biological characteristics presented by the two subtypes are summarized as follows:

[0066] (1) Comparison of clinicopathological features:

[0067] The C2 subtype had a higher histological grade (P=0.011).

[0068] The mortality rate of C2 subtype was higher (P=0.0028).

[0069] The C2 subtype had a higher incidence of new tumor events (P=0.032).

[0070] There were no significant differences in age and gender distribution.

[0071] (2) Survival prognosis analysis:

[0072] Kaplan-Meier survival analysis and Log-rank test showed that the overall survival rate of patients with C2 subtype was significantly lower than that of patients with C1 subtype (P=0.042, HR=1.32, 95% CI: 1.01-1.73).

[0073] (3) Genomic variation characteristics:

[0074] The frequency of TP53 mutation in C2 subtype was significantly higher than that in C1 subtype (Chi-square test, χ 2 =18.94, P < 0.001), and there was no significant difference in the mutation frequencies of other driver genes.

[0075] (4) Transcriptome functional enrichment:

[0076] The signaling pathways enriched in the C2 subtype include: ECM-receptor interaction pathway, adherens junction pathway, and PI3K-Akt signaling pathway (P<0.001).

[0077] Example 2 Analysis of immune microenvironment characteristics

[0078] Step 2.1: Immune infiltration analysis

[0079] The CIBERSORTx algorithm was used to quantitatively analyze tumor immune cell infiltration. Figure 5 As shown in the figure, there are significant differences in immune cell infiltration between the two subtypes: C1 tumors are enriched in anti-tumor immune populations, including activated NK cells, T cell follicular helper subsets, and CD8 + In contrast, C2 tumors showed a significant infiltration of M2 macrophages, suggesting a tumor immune microenvironment with a more immunosuppressive effect. The characteristics of the two subtypes are summarized as follows:

[0080] (1) Immune characteristics of C1 subtype:

[0081] The infiltration of activated NK cells increased (P<0.01).

[0082] Follicular helper T cell infiltration increased (P<0.01),

[0083] CD8+T cell infiltration increased (P<0.01),

[0084] The overall manifestation is "immune fever" tumor characteristics.

[0085] (2) C2 subtype immune characteristics:

[0086] The infiltration of M2 macrophages increased significantly (P<0.001).

[0087] CD8+ T cell infiltration decreased,

[0088] The overall manifestation is "immune cold" tumor characteristics.

[0089] Step 2.2: Immune escape assessment

[0090] The TIDE (Tumor Immune Dysfunction and Exclusion) algorithm is used to assess immune escape ability:

[0091] The TIDE score of C2 subtype was significantly higher than that of C1 subtype (P<0.001).

[0092] This suggests that the C2 subtype has a stronger ability to escape immune system.

[0093] The C2 subtype is predicted to be less responsive to immune checkpoint inhibitors.

[0094] Example 3 Cross-cancer validation

[0095] Step 3.1: Expand the validation queue

[0096] The effectiveness of the classification method was verified in squamous cell carcinoma cohorts including TCGA-LUSC (lung squamous cell carcinoma, n=501), TCGA-ESCC (esophageal squamous cell carcinoma, n=95), and TCGA-CESC (cervical squamous cell carcinoma, n=253). Figure 6 As shown in the figure, the PCA clustering results, survival analysis and gene expression results of various squamous cell carcinomas all showed that the 16-gene typing method had good consistency and prognostic prediction ability in different squamous cell carcinomas.

[0097] Step 3.2: Consistency Verification Results

[0098] A stable dual-subtype classification pattern, a consistent trend toward worse prognosis for the C2 subtype, and conserved metabolic gene expression differences were observed in all validation cohorts.

[0099] Figure 7 A shows the comparison of tumor immune dysfunction and rejection (TIDE) scores. The violin plot shows that the TIDE score of the C2 subtype is significantly higher than that of the C1 subtype (P<0.0001), indicating that the C2 subtype has a stronger immune escape ability and a lower responsiveness to immunotherapy. Figure 7 B shows the IC50 prediction results of common chemotherapy drugs. The violin plot shows the differences in sensitivity of different subtypes to various chemotherapy drugs: the C2 subtype has higher sensitivity to cisplatin, docetaxel, and gemcitabine (P<0.0001), but relatively lower sensitivity to lapatinib and paclitaxel (P=0.0259), providing an important reference for the selection of individualized chemotherapy regimens in clinical practice.

Claims

1. Use of a metabolic gene marker combination in constructing a squamous cell carcinoma typing model, wherein the gene markers include CAD, CYP27B1, SLC44A1, GGT5, GLS, PPT1, GFPT2, NT5E, PADI3, PLA2G7, LPCAT1, SLC16A1, SLC7A7, GMPS, VKORC1, and SLC3A2.

2. The use according to claim 1, characterized in that Application in constructing a classification model for head and neck squamous cell carcinoma.

3. The use according to claim 1, characterized in that Squamous cell carcinoma is divided into two types: metabolic gene low expression type and metabolic gene high expression type.

4. The use according to claim 3, characterized in that The low expression type of the metabolic gene shows "immune hot" tumor characteristics, and the high expression type of the metabolic gene shows "immune cold" tumor characteristics.

5. The use according to claim 3, characterized in that The low expression type of the metabolic gene shows high sensitivity to lapatinib and paclitaxel, and the high expression type of the metabolic gene shows higher sensitivity to cisplatin, docetaxel and gemcitabine.

6. The use of the squamous cell carcinoma detection marker combination according to claim 1 in constructing a squamous cell carcinoma prognosis prediction model is beneficial for providing a basis for formulating a patient prognosis diagnosis and treatment plan.

7. Use of the squamous cell carcinoma detection marker combination according to claim 1 in a drug administration prediction model for squamous cell carcinoma patients provides a basis for the targeted and accurate medication of patients and effectively avoids improper use of drugs.

8. A system for squamous cell carcinoma classification and prediction, characterized in that: The following units are included: a data aggregation unit that stores the detection data of 16 gene expressions in patients with squamous cell carcinoma; An analysis unit, which uses the results obtained by the data aggregation unit as input items for analysis; as well as The evaluation unit outputs the metabolic genotyping of the individual corresponding to the sample, as well as recommendations for prognosis prediction and treatment guidance.

9. The system according to claim 8, characterized in that It also includes a detection unit, the information of which is obtained from detection using a kit or a sequencing instrument.

10. A detection device, characterized in that: Detect the metabolic gene marker combination according to claim 1.

Images