Intelligent research and development method of network pharmacology and molecular docking of traditional Chinese medicine monomers based on key genes of copper death in renal clear cell carcinoma

Abstract

The application discloses an intelligent research and development method of traditional Chinese medicine monomer network pharmacology and molecular docking based on a key gene of copper death of renal clear cell carcinoma, and belongs to the technical field of renal clear cell carcinoma treatment.The application first determines that CDKN2A as a key gene of copper death plays a core role in renal clear cell carcinoma, and confirms that high expression of CDKN2A is significantly related to poor prognosis of patients, so that CDKN2A can be used as a biomarker for prognosis evaluation of renal clear cell carcinoma and a molecular target point for drug treatment.A copper death risk score model is constructed by using machine learning based on five genes including CDKN2A, so as to realize accurate prediction of the prognosis of patients.Active ingredients of traditional Chinese medicines, apigenin, kaempferol and quercetin, which are selected by targeting CDKN2A downstream CDK2 / CDK6, can block kinase activity by competitive inhibition, regulate cell cycle G2 / M phase conversion, and play an anti-renal clear cell carcinoma role.

Description

Technical Field

[0001] This invention belongs to the field of renal clear cell carcinoma technology, specifically a smart research and development method based on the network pharmacology and molecular docking of traditional Chinese medicine monomers for key genes of copper death in renal clear cell carcinoma. Background Technology

[0002] Clear cell renal carcinoma (ccRCC) is the most common pathological type of renal cell carcinoma, accounting for 70%-75% of all renal cell carcinoma cases. Its incidence ranks 14th among adult malignancies worldwide, posing a serious threat to human health. Currently, localized clear cell renal carcinoma is primarily treated surgically. However, patients with advanced or metastatic clear cell renal carcinoma have poor prognoses, and existing targeted therapies (such as drugs targeting the VEGF and mTOR signaling pathways) are prone to resistance, while immunotherapy has limited response rates. Therefore, there is an urgent need to develop new therapeutic targets and effective treatments.

[0003] Copper death is a novel programmed cell death mechanism discovered in recent years. Its core characteristics are the accumulation of intracellular copper ions, which triggers the aggregation of mitochondrial lipid proteins and instability of Fe-S cluster proteins, ultimately leading to cell death. Compared with healthy cells, cancer cells have a significantly increased demand for copper ions. Copper ions are involved in regulating signaling pathways related to tumor cell proliferation, invasion, and metastasis. Therefore, copper death-related genes have become potential targets for cancer therapy.

[0004] The CDKN2A gene is an important tumor suppressor gene, encoding a protein that participates in cell cycle regulation and can inhibit abnormal cell proliferation. Existing research indicates that mutations, deletions, or methylation silencing of CDKN2A are widespread in various cancers, but its specific role as a key gene for copper death in clear cell renal cell carcinoma and its targeted applications remain unclear. Our research team, through multi-omics analysis and machine learning validation, discovered that CDKN2A is the key gene for copper death with the highest risk association in clear cell renal cell carcinoma. Abnormal expression of CDKN2A is closely related to patient clinical stage and prognosis, demonstrating its core potential as a therapeutic target.

[0005] Traditional Chinese medicine (TCM) possesses unique advantages in cancer treatment, including multiple targets and low toxicity, but its mechanisms of action remain largely unclear, limiting its clinical translation and application. Network pharmacology and molecular docking technology provide effective means to uncover the targeted effects of active ingredients in TCM, enabling a precise integration of traditional Chinese medicine with modern molecular biology. Based on this, this invention systematically elucidates the therapeutic target value of CDKN2A and screens out active TCM ingredients targeting this gene, providing a new strategy for the precision treatment of clear cell renal cell carcinoma. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent research and development method for the network pharmacology and molecular docking of Chinese medicine monomers based on the copper death key gene in clear cell renal cell carcinoma, to explore a new application of the copper death key gene CDKN2A in the treatment of clear cell renal cell carcinoma, to clarify its value as a therapeutic target, and to provide a CDKN2A-targeted drug for the treatment of clear cell renal cell carcinoma, thereby solving the problems of lack of therapeutic targets and drug resistance in existing clear cell renal cell carcinoma treatments.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0008] Application of CDKN2A gene as a prognostic biomarker or drug target for clear cell renal cell carcinoma.

[0009] The application is in the construction of a prognostic assessment model for clear cell renal cell carcinoma, which uses the expression levels of the CDKN2A gene and copper death-related genes FDX1, PDHA1, PDHB, and DBT as input features.

[0010] The application is in the screening or preparation of therapeutic drugs for clear cell renal cell carcinoma that target the CDKN2A downstream signaling pathway.

[0011] A prognostic assessment system for clear cell renal cell carcinoma includes: a data input module for acquiring expression data of CDKN2A, FDX1, PDHA1, PDHB, and DBT genes in clear cell renal cell carcinoma tissue of a subject; a risk calculation module for inputting the expression data into a pre-trained copper mortality risk scoring model to calculate a risk score; and a result output module for outputting a prognostic assessment result based on the risk score.

[0012] The copper mortality risk scoring model uses the expression levels of CDKN2A and copper mortality-related genes FDX1, PDHA1, PDHB, and DBT as input features and is constructed using the "StepCox [both]+survival-SVM" machine learning algorithm.

[0013] The system further includes a sample processing module containing reagents for detecting CDKN2A gene expression levels. These reagents include specific primer pairs with the following sequences:

[0014] Forward primer: 5'-GGGTTTTCGTGGTTCACATCC-3'

[0015] Reverse primer: 5'-CTAGACGCTGGCTCCTCAGTA-3'.

[0016] A computer-readable storage medium for prognostic assessment of clear cell renal cell carcinoma, having thereon stored a computer program that, when executed by a processor, performs the following steps:

[0017] Expression data of CDKN2A, FDX1, PDHA1, PDHB and DBT genes in clear cell renal cell carcinoma tissue of subjects were obtained; the expression data were input into a pre-trained copper mortality risk scoring model to calculate the risk score; and the prognostic assessment result of clear cell renal cell carcinoma was output based on the risk score.

[0018] A method for prognostic assessment of clear cell renal cell carcinoma includes the following steps:

[0019] (1) Obtain expression data of CDKN2A, FDX1, PDHA1, PDHB and DBT genes in the clear cell renal cell carcinoma tissue of the subjects;

[0020] (2) Input the expressed data into a pre-trained copper mortality risk scoring model and calculate the risk score. The model is constructed using the “StepCox [both]+survival-SVM” machine learning algorithm.

[0021] (3) Assess the prognostic risk level of the subject based on the risk score.

[0022] A therapeutic drug for clear cell renal cell carcinoma targeting the CDKN2A downstream signaling pathway comprises an effective amount of a traditional Chinese medicine active ingredient, wherein the active ingredient is selected from apigenin, kaempferol, quercetin, or their pharmaceutically acceptable salts, solvates, or prodrugs; the active ingredient exerts its anti-clear cell renal cell carcinoma effect by targeting and inhibiting CDK2 and / or CDK6 kinase activity and regulating the G2 / M phase transition of the cell cycle.

[0023] Application of apigenin, kaempferol, or quercetin as CDK2 / CDK6 inhibitors in the preparation of therapeutic drugs for clear cell renal cell carcinoma.

[0024] The drug also includes a pharmaceutically acceptable carrier or excipient, and the dosage form of the drug is selected from tablets, capsules, granules, injections, or lyophilized powder for injection.

[0025] The present invention achieves the following technical effects:

[0026] 1. This invention is the first to clearly define the core role of CDKN2A as a key gene for copper death in clear cell renal cell carcinoma, and confirms that its high expression is significantly associated with the clinical stage and poor prognosis of patients. It has the specificity and effectiveness as a therapeutic target, filling the gap in the research on copper death-related targets in clear cell renal cell carcinoma and providing a new molecular target for the precision treatment of clear cell renal cell carcinoma.

[0027] 2. This invention integrates 10 machine learning algorithms to construct and validate a robust copper mortality risk scoring model. The model has a C-index greater than 0.6 in both the training and validation sets, and can independently identify high-risk patient subgroups with poor prognosis based on traditional clinical factors (age, gender, stage, grade), thus demonstrating good clinical application value.

[0028] 3. The three active ingredients of traditional Chinese medicine screened in this invention—apigenin, kaempferol, and quercetin—can specifically bind to and inhibit the function of CDK2 / CDK6 proteins, exerting anti-cancer effects by regulating the copper death pathway and cell cycle. Compared with traditional chemotherapy drugs, they have the advantages of fewer toxic side effects and less likelihood of inducing drug resistance.

[0029] 4. The therapeutic drug preparation process provided by this invention is simple, the raw materials are widely available, the cost is controllable, and it has good prospects for clinical translation. Attached Figure Description

[0030] Figure 1 This study screened genes related to copper death prognosis in clear cell renal cell carcinoma (PCD). A represents essential regulatory genes for 18 PCD patterns. B represents 12,788 genes differentially expressed between normal and tumor tissues in the TCGA-KIRC cohort. C represents the intersection genes between the 18 PCD patterns and the 12,788 differentially expressed genes in the TCGA-KIRC cohort. D represents the univariate COX gene, the intersection gene between ccRCC and copper death.

[0031] Figure 2 This section describes the construction and predictive performance evaluation of a copper mortality prognostic model. Specifically, A represents the C-index calculation results for 101 copper mortality types in the TCGA-KIRC and GSE22541 datasets. B represents the ROC curve calculation results of the copper mortality prognostic model. C represents the ROC curve calculation results based on clinical information. D represents the survival prognostic analysis results from TCGA and GSE22541.

[0032] Figure 3 This section describes the association of clinical features, prognostic independence, and nomogram construction of a copper mortality risk model. Specifically, AB represents univariate and multivariate Cox regression analyses of clinical features. C shows the TCGA hazard curve and scatter plot. D represents the time-dependent C-index analysis. E shows the predictive performance of the nomogram model for copper mortality. F shows the predictive performance of the nomogram model for progression-free survival. G is the ROC curve of the nomogram model for progression-free survival.

[0033] Figure 4 This section validates the association and stratified stability between the copper mortality risk model and clinical characteristics. A represents a heatmap of clinical information distribution related to copper mortality. BE represents the survival prognostic analysis results for age, sex, grade, and stage of copper mortality.

[0034] Figure 5 This section describes the prognostic value of copper death risk scores in relation to tumor immune microenvironment characteristics and immune cell infiltration. AL represents the survival analysis results for the following immune cell types in the TCGA-KIRC cohort: T cells regulatory (Tregs), T cells follicularhelper (Tfh), T cells CD8, T cells CD4 memory resting, T cells CD4 memory activated, Plasma cells, Monocytes, Mast cells resting, Mast cells activated, Macrophages M2, Macrophages M0, and Dendritic cells resting. M represents the inflammatory immune cell subtype responsible for copper death.

[0035] Figure 6 This involves the tumor immune microenvironment characteristics and multi-algorithm validation of copper mortality risk grouping. Among these, Figure 6 A represents the analysis of immune cell infiltration in the high- and low-risk groups for copper death. Figure 6 B represents different software algorithms used to analyze the infiltration of immune cells that cause copper death. Figure 6 C represents the enrichment analysis of immune cell function in the high- and low-risk groups for copper death.

[0036] Figure 7 This is an association analysis between copper mortality risk scores and immune checkpoint expression profiles. In the figure, A represents the expression levels of genes related to copper mortality immune checkpoints. BM represents the analysis of immune checkpoints associated with high and low risk groups for copper mortality.

[0037] Figure 8 This is an analysis of differentially expressed gene function enrichment and tumor microenvironment characteristics in copper death risk groups. In the figure, AB represents the GO analysis plots for the high- and low-risk copper death groups, respectively. CD represents the KEGG analysis plots for the high- and low-risk copper death groups, respectively. E represents the ratio of immune matrix components in the tumor microenvironment for the high- and low-risk copper death groups.

[0038] Figure 9 This is an association analysis between copper mortality risk grouping and targeted drug sensitivity. A and J represent the drug sensitivity of the high- and low-risk copper mortality groups to Axitinib, Cediranib, Rapamycin, Carmustine, Dihydroorotate, ML323, Ibrutinib, RO-3306, Tovasertib, and Ruxolitinib, respectively.

[0039] Figure 10This represents the mRNA expression level of CDKN2A in clear cell renal cell carcinoma and adjacent normal tissues.

[0040] Figure 11 This section describes the expression of CDKN2A protein in clear cell renal cell carcinoma and adjacent normal tissues. AC represents the expression level of CDKN2A in clear cell renal cell carcinoma (100×, 200×, 400×); DF represents the expression level of CDKN2A in adjacent normal tissues (100×, 200×, 400×); and GI represents the protein expression and subcellular localization verification of CDKN2A in human cell lines A-431, U-251MG, and U2OS.

[0041] Figure 12 This is a roadmap for screening active ingredients in traditional Chinese medicine that target the key gene CDKN2A.

[0042] Figure 13 It is a GO enrichment analysis of the core active ingredient targets of traditional Chinese medicine.

[0043] Figure 14 It is a specific GO entry and signaling pathway enrichment analysis of the core active ingredient targets of traditional Chinese medicine.

[0044] Figure 15 It is a disease enrichment analysis of the core active ingredient targets of traditional Chinese medicine (DisGeNET).

[0045] Figure 16 This is a diagram illustrating the docking pattern between CDK2, CDK6, and the core active monomer molecules of traditional Chinese medicine. Detailed Implementation

[0046] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. The technical route of the present invention is as follows: Figure 1 As shown.

[0047] Example 1

[0048] This embodiment aims to use bioinformatics to screen key genes related to the prognosis of clear cell renal cell carcinoma (ccRCC) from copper death-related genes, and to construct and validate a robust prognostic risk scoring model based on various machine learning algorithms.

[0049] 1. Data Collection

[0050] The TCGA-KIRC cohort data was downloaded from the Clear Cell Renal Cell Carcinoma Project section of the Cancer Genome Atlas (TCGA) database (https: / / portal.gdc.cancer.gov / projects / TCGA-KIRC), yielding RNA sequencing data and corresponding clinicopathological information for 614 ccRCC samples. The ccRCC cohort GSE22541 dataset was obtained from the GEO database (GEO Accession viewer (nih.gov)), containing clinical parameter information and microarray expression analysis data for 68 samples from 27 ccRCC patients.

[0051] Nineteen core regulatory genes of copper death reported in the literature were collected and integrated. These genes include: FDX1, LIAS, LIPT1, DLD, DLAT, PDHA1, PDHB, MTF1, GLS, CDKN2A, DBT, GCSH, DLST, ATP7A, ATP7B, SLC31A1, COX17, CCS, and ATOX1.

[0052] 2. Screening of differentially expressed genes and prognostic genes: Differential expression analysis of TCGA-KIRC cohort data was performed using the R language "edgeR" package (screening criteria: adjusted P value < 0.05 and |log2FC| > 1) to obtain differentially expressed genes. A total of 12,788 genes (DEGs) that were significantly differentially expressed between tumor and normal tissues were screened.

[0053] The intersection of differentially expressed genes and copper death genes yielded six intersection genes: NLRP3, FDX1, PDHA1, PDHB, CDKN2A, and DBT.

[0054] Using ccRCC patients with complete survival information in the TCGA-KIRC cohort as the research subjects, we used the R language "survival" package to perform univariate Cox proportional hazards regression analysis on the relationship between the expression levels of the above 6 intersection genes and the patients' overall survival (OS).

[0055] The results showed that FDX1, PDHA1, PDHB, and DBT were protective factors (HR<1), while CDKN2A was the only independent risk factor (HR=1.362, P<0.001), suggesting that high expression of CDKN2A is significantly associated with poor prognosis in ccRCC, while NLRP3 expression is not significantly associated with prognosis.

[0056] 3. Construction of a machine learning-based prognostic risk scoring model:

[0057] To obtain the most robust prediction model, this study integrated 10 mainstream machine learning algorithms and constructed 101 different algorithm combinations through pairwise pairwise combinations. The algorithms used included: Random Survival Forest (RSF), Elastic Network (Enet), Lasso (Minimum Absolute Shrinkage and Selection Operator), Ridge Regression, Stepwise Cox Regression (with both / backward / forward directions), CoxBoost, Partial Least Squares Regression Cox (plsRcox), Supervised Principal Component Analysis (SuperPC), Generalized Augmented Regression Model (GBM), Survival Support Vector Machine (survival-SVM), and various combinations thereof.

[0058] The expression levels of FDX1, PDHA1, PDHB, CDKN2A, and DBT were used as input features. In the TCGA-KIRC training set, 101 algorithm combinations were trained and their hyperparameters tuned using 10-fold cross-validation. The consistency index (C-index) was used as the evaluation metric for model prediction performance. The average C-index of each model in the training set was calculated, and its performance was further evaluated in the GSE22541 external validation set. Selection criteria: Models with a C-index greater than 0.6 in both the training and validation sets were considered stable models with excellent prediction performance.

[0059] The results showed that several algorithm combinations met the selection criteria. Among them, the model constructed by the "StepCox [both] + survival-SVM" combination demonstrated stable and superior predictive efficacy in both the training set (C-index=0.692) and the validation set (C-index=0.651), and was identified as the optimal prognostic model of this invention. This model calculates a copper mortality risk score for each patient using a weighted integral method. The scoring formula is a linear combination of the expression level of each gene and its corresponding coefficient.

[0060] Analysis of the risk score distribution characteristics showed that as the Risk Score increased, the risk of patient death gradually increased. Deaths in the high-risk group (Risk Score > 0) (red) were mainly distributed in the later stages of follow-up, while patients in the low-risk group (green) had good survival, exhibiting a clear risk stratification effect. A nomogram model was further constructed to enhance clinical applicability. The nomogram, integrating Risk Score with clinical characteristics (age, sex, grade, stage), can intuitively predict 1-year, 3-year, and 5-year overall survival rates, with a C-index of 0.764 (95% CI: 0.729-0.800), significantly improving predictive performance compared to the single Risk Score model (AUC = 0.684). The calibration curve showed that the predicted survival rate closely matched the actual observed values ​​at all time points, indicating good model accuracy. Similarly, the progression-free survival (PFS) nomogram also demonstrated excellent predictive efficacy. Kaplan-Meier analysis confirmed that PFS was significantly shorter in the high-risk group than in the low-risk group (P<0.001), consistent with the overall survival trend.

[0061] To further analyze the relationship between risk scores and clinicopathological features and to validate the independent prognostic stratification value of the copper mortality prognostic model, this study systematically analyzed the distribution of clinical characteristics and stratified prognosis in high- and low-risk groups. Clinical characteristic heatmaps showed that the high-risk group was significantly correlated with advanced age (Age≥65), advanced pathological grade (Grade III-IV), high stage (Stage III-IV), tumor invasion depth (T3-4), and distant metastasis (M1) (all P<0.001), while gender and lymph node metastasis (N) were relatively evenly distributed between the two groups. Stratified survival analysis showed that the high-risk group exhibited significantly worse overall survival in all demographic and pathological subgroups (P<0.001). These results confirm that the copper mortality risk score has prognostic stratification value independent of traditional clinical factors; even within the same clinical stage or pathological grade, the Risk Score can further identify high-risk patient subgroups with poor prognosis.

[0062] Performance validation and evaluation of prognostic models

[0063] 1. Risk grouping and survival analysis

[0064] Patients were divided into high-risk and low-risk groups using the median RiskScore in the TCGA training set as the cutoff value. Survival analysis showed that in the TCGA training set, the overall survival of patients in the high-risk group was significantly shorter than that in the low-risk group (p<0.001); in the GSE22541 validation set, the high-risk group also showed a worse survival prognosis (p=0.034), confirming the reliable prognostic stratification ability of the RiskScore model in different cohorts.

[0065] 2. Predictive Performance Evaluation

[0066] Further evaluation of the model's diagnostic efficacy was conducted by calculating the area under the receiver operating characteristic (ROC) curve. Compared to traditional clinicopathological features, the 5-year AUC of the copper mortality Risk Score (0.684) was significantly better than that of demographic features (Age: 0.592, Gender: 0.489), comparable to pathological grading (Grade: 0.679), and slightly lower than complete TNM staging (Stage: 0.721), but the difference was small (ΔAUC = 0.037). Notably, the model exhibited time-dependent characteristics: the 1-year AUC was 0.611, the 3-year AUC was 0.630, and the 5-year AUC reached 0.684, suggesting better predictive value for long-term prognosis. Survival analysis showed that in the TCGA training set, the overall survival of patients in the high-risk group was significantly shorter than that in the low-risk group (p<0.001); in the GSE22541 validation set, the high-risk group also showed a worse survival prognosis (p=0.034), confirming the reliable prognostic stratification ability of the Risk Score model in different cohorts.

[0067] Compared with traditional clinicopathological features, the 5-year AUC of the copper mortality Risk Score (0.684) was significantly better than that of age (0.592) and sex (0.489), and comparable to that of pathological grade (0.679), suggesting its potential as an independent prognostic indicator.

[0068] 3. Independent prognostic analysis

[0069] To further analyze the relationship between risk scores and clinicopathological features, and to verify the independent prognostic stratification value of the copper mortality prognostic model, this study systematically analyzed the distribution of clinical characteristics and stratified prognosis in high- and low-risk groups. Figure 4 Clinical characteristic heatmaps showed that the high-risk group was significantly associated with advanced age (Age≥65), advanced pathological grade (Grade III-IV), high stage (Stage III-IV), tumor invasion depth (T3-4), and distant metastasis (M1) (all P<0.001), while gender and lymph node metastasis (N) were relatively evenly distributed between the two groups. Stratified survival analysis ( Figure 4The results (BE) showed that, regardless of the demographic or pathological subgroup, the high-risk group exhibited significantly worse overall survival (P<0.001). These results confirm that the copper mortality risk score has prognostic stratification value independent of traditional clinical factors, and that even within the same clinical stage or pathological grade, the Risk Score can further identify high-risk patient subgroups with poor prognosis.

[0070] 4. Nodal Chart Analysis

[0071] To further improve the clinical applicability of the model, RiskScore was integrated with independent clinical prognostic factors (age, gender, grade, stage), and a visual nomogram model was constructed using the "rms" package in R language.

[0072] This nomogram can visually predict the survival probability of ccRCC patients at 1, 3, and 5 years post-surgery. Calibration curves show good agreement between the survival rates predicted by the nomogram and actual observed values. The nomogram model has a C-index as high as 0.764 (95% CI: 0.729–0.800), demonstrating a significant improvement in predictive performance compared to the single RiskScore model.

[0073] In summary, this embodiment successfully constructed and validated a risk scoring model based on copper death-related genes that can accurately predict the prognosis of ccRCC patients, and established the core position of CDKN2A in the model, providing a key basis for subsequent drug target screening.

[0074] Example 2: Validation of CDKN2A expression in clear cell renal cell carcinoma tissue

[0075] This embodiment aims to verify the expression level of CDKN2A in ccRCC tissue using clinical samples, in order to confirm the reliability of the bioinformatics analysis results.

[0076] A retrospective study was conducted on 36 patients with clear cell renal cell carcinoma admitted to the Second Affiliated Hospital of Guangxi Medical University between October 2020 and December 2025. All patients were diagnosed with clear cell renal cell carcinoma by senior pathologists after surgery, with clinical stages I-IV (AJCC 8th edition).

[0077] Inclusion criteria: (1) Age 18-80 years, gender not limited; (2) Postoperative diagnosis of clear cell renal cell carcinoma; (3) Initial diagnosis, no prior radiotherapy, chemotherapy, or biological therapy; (4) No history of other types of tumors. Exclusion criteria: (1) Prior radiotherapy, chemotherapy, or biological therapy; (2) Incomplete clinical data; (3) Concomitant other types of tumors. Paraffin-embedded specimens of cancerous tissue and paired adjacent normal tissue were collected. This study was approved by the hospital ethics committee, and all patients signed informed consent forms.

[0078] 1. RT-qPCR method to verify CDKN2A mRNA expression level

[0079] Main instruments: refrigerated centrifuge, spectrophotometer, real-time quantitative PCR instrument, ice box, semi-automatic slide drying machine, drying oven, pressure cooker, humidification chamber, microscope. Main reagents: TRIZOL reagent, chloroform, isopropanol, reverse transcription kit (including reverse transcription buffer, dNTPs, MMLV), real-time quantitative PCR kit (including SYBR Green I dye, Taq enzyme, dNTPs), rabbit anti-human CDKN2A / p16INK4a monoclonal antibody (clone number R02-5N7, Zenbio, 380963), Tris-EDTA antigen retrieval solution (ShareBio, SB-MY001), MaxVision™ HRP polymer method universal secondary antibody and DAB colorimetric kit (MXB Biotechnologies, Kit-0014), antibody dilution buffer (Zhongshan Jinqiao, ZLI-9030), and specific primers for CDKN2A and the internal reference gene GAPDH were synthesized by Sangon Biotech (Shanghai) Co., Ltd.

[0080] Primer sequences:

[0081] CDKN2A:Forward:5'-GGGTTTTCGTGGTTCACATCC-3';Reverse:5'-CTAGACGCTGGCTCCTCAGTA-3'.

[0082] GAPDH:Forward:5'-CTGGGCTACACTGAGCACC-3';Reverse:5'-AAGTGGTCGTTGAGGGCAATG-3'.

[0083] Total RNA extraction from paraffin tissue

[0084] (1) Take 5-8 paraffin-embedded tissue sections (5-10 μm thick) from patients with clear cell renal cell carcinoma and place them in RNase-free centrifuge tubes. (2) Add 1000 μL of xylene for dewaxing, vortex to dissolve the paraffin in the sections; then centrifuge at 12000 rpm for 3 min, discard the supernatant, and repeat 3 times. (3) Add 1000 μL of anhydrous ethanol to the precipitate, vortex, centrifuge at 12000 rpm for 3 min, discard the supernatant, and repeat washing 2-3 times to remove residual xylene. Open the tube cap and let it stand at room temperature for 8-10 minutes to allow the ethanol to evaporate completely and the tissue precipitate to become translucent. (4) Add 180 μl of Buffer AT and 20 μl of proteinase K stock solution to the dried tissue precipitate; vortex to mix. Place in a metal bath and incubate at 56 °C for 15 min, then incubate at 80 °C for 15 min (to reverse formaldehyde crosslinking). (5) Add 200 μl of Buffer SL to the digestion solution and centrifuge at 12,000 rpm for 5 min. (6) Transfer the supernatant to a new RNase-free centrifuge tube and add 660 μL of anhydrous ethanol. Transfer the entire mixture to the matching nucleic acid purification column, centrifuge at 12,000 rpm for 30 s, and discard the waste liquid. (7) Add 600 μL of Buffer WBR to the purification column, centrifuge at 12,000 rpm for 30 s, discard the waste liquid, and repeat twice. (8) Place the purification column back into the collection tube and centrifuge at 14,000 rpm for 1 min. (9) Place the purification column in a new 1.5 mL RNase-free centrifuge tube, add 100 μL of preheated RNase-free water (at 56 °C) to the center of the column membrane, let stand at room temperature for 1 min, and centrifuge at 12,000 rpm for 30 s to elute the RNA. (10) Immediately store the eluted RNA solution in a -80 °C ultra-low temperature freezer for later use.

[0085] Quality testing and DNA synthesis

[0086] RNA concentration and purity were determined using a NanoDrop One micro-volume spectrophotometer. The A260 / A280 ratio should be between 1.8 and 2.0, and the A260 / A230 ratio should be >1.5.

[0087] Reverse transcription was performed using the PrimeScript™ RT reagent Kit with gDNA Eraser: (1) Prepare a 10 μL genomic DNA removal reaction system (containing 2 μL 5×gDNA Eraser Buffer, 1 μL gDNA Eraser, ≤500 ng total RNA, and RNase-free water) on ice and react at 42°C for 2 min. (2) Add 10 μL of reverse transcription reaction solution (containing 4 μL 5×PrimeScript Buffer 2, 1 μL RT Primer Mix, 1 μL PrimeScript RTEnzyme Mix I, and 4 μL RNase-free water) directly to the previous step reaction solution, mix well, and then perform the reverse transcription reaction: 37°C for 15 min, followed by 85°C for 5 s. The resulting cDNA was stored at -20°C.

[0088] Real-time quantitative PCR

[0089] Prepare the reaction system (20 μL) on ice: containing 10 μL 2X SYBR Green qPCR MasterMix II, 0.8 μL CDKN2A or GAPDH forward primer (10 μM), 0.8 μL corresponding reverse primer (10 μM), 2 μL cDNA template, and 6.4 μL sterile water. Experimental controls: Each sample was tested in 3 replicates, and template-free and reverse transcription-free controls were also included. Reaction program: (1) Pre-denaturation: 95℃ for 30 s. (2) Cyclic reaction (40 cycles): 95℃ for 5 s, 60℃ for 30 s. (3) Melting curve analysis: 95℃ for 15 s, 60℃ for 1 min, 95℃ for 15 s.

[0090] Data Analysis

[0091] After the reaction, a single peak in the melting curve and a Tm value within the 80-90℃ range indicate good amplification specificity. If double peaks or abnormal Tm values ​​appear, the data are considered invalid and discarded. Record the Ct value using a 2-1Tt method. (-ΔΔCt) The relative expression level is calculated using the following method. The specific calculation formula is: ΔCt = Ct_ CDKN2A -Ct_ GAPDH ;ΔΔCt=ΔCt_ 癌组织 -ΔCt_ 配对癌旁组织 Paired adjacent normal tissues from each patient were used as controls, with CDKN2A mRNA expression levels calibrated to 1, and the relative fold increase in expression within the cancerous tissues was calculated. Statistical analysis was performed using paired-samples t-tests or Wilcoxon signed-rank tests, with P < 0.05 considered statistically significant.

[0092] RT-qPCR results

[0093] The expression level of CDKN2A mRNA in 12 paired paraffin-embedded tissues was detected by RT-qPCR. The results showed that the expression level of CDKN2A mRNA was significantly upregulated in clear cell renal cell carcinoma tissues compared with adjacent normal tissues. Figure 10 Quantitative analysis showed that the relative expression level of CDKN2A in cancerous tissue was 4.12±0.85 (mean±SD), which was significantly higher than that in paired adjacent normal tissue (normalized to 1.00, P<0.05, Wilcoxon signed-rank test).

[0094] 2. Immunohistochemistry was used to verify the expression level of CDKN2A protein.

[0095] (1) Section preparation: Place the paraffin blocks of cancerous tissue and adjacent tissue from KIRC patients on the freezing stage for 5-10 minutes, and use a semi-automatic microtome to section them. The section thickness is 4 μm. Place the slides into the slide rack and place them in the drying oven at 65°C for 1-2 hours.

[0096] (2) Dewaxing and rehydration: Place the slices in an environmentally friendly dewaxing solution for 5 minutes in a 3-cylinder solution while the wax is still hot and melted; place them in an anhydrous alcohol solution for 2 minutes in a 2-cylinder solution; place them in 95% alcohol, 85% alcohol and 75% alcohol solutions for 1 minute each in a 1-cylinder solution; rinse with running water for 5 minutes; and soak them in distilled water until ready for use.

[0097] (3) Antigen retrieval: Place the slides into a retrieval box containing Tris-EDTA antigen retrieval solution, and place the retrieval box in a preheated and boiling pressure cooker. After covering the pressure cooker, adjust the heating power to 1200W. After the pressure cooker releases steam, adjust the power to 800W and time for 2-3 minutes. After the pressure cooker is cooled to room temperature in a cold water bath, remove the slides and soak them in distilled water for later use.

[0098] (4) Block endogenous peroxidase: Soak in 3% H2O2 at room temperature for 13 minutes, then rinse with distilled water for 3×2 minutes.

[0099] (5) Antigen-antibody reaction: Add rabbit anti-human CDKN2A / p16INK4a antibody (1:400), place in a humidified chamber, cover the chamber, and incubate at 4°C for 12 hours. After removal, wash thoroughly with PBS for 3×3 minutes. Add MaxVision™ HRP polymer method universal secondary antibody, incubate at room temperature for 30 minutes, and then wash thoroughly with PBS for 3×3 minutes.

[0100] (6) Color development: Add DAB color developer, develop color at room temperature for 2-3 minutes, then rinse with tap water to stop the display.

[0101] (7) Counterstaining: Place the slices in hematoxylin staining solution for 2 minutes, differentiate for 2 seconds, then rinse with tap water for several minutes, then return to blue for 2-3 seconds, and finally rinse with tap water.

[0102] (8) Dehydration and clearing: Dehydration with anhydrous alcohol for 3×2 minutes, xylene for 3×2 minutes, and then sealing with neutral resin.

[0103] Immunohistochemical analysis of CDKN2A was performed independently by two senior pathologists under double-blind conditions. CDKN2A protein expression in the cell nucleus was defined as positive. The results were semi-quantitatively evaluated using the H-Score (histochemical scoring) system, which is derived by multiplying staining intensity by the number of positive cells; the theoretical range is 0-300 points.

[0104] The calculation formula is: H-Score = (0 × percentage of unstained cells %) + (1 × percentage of weakly positive cells %) + (2 × percentage of moderately positive cells %) + (3 × percentage of strongly positive cells %). Based on the final score, CDKN2A expression levels are defined as follows: 0-100 points for negative; 101-150 points for weakly positive; 151-200 points for moderately positive; and 201-300 points for strongly positive. For statistical analysis, negative (-) and weakly positive (+) were combined into the "low expression group" (≤150 points), and moderately positive (++) and strongly positive (+++) were combined into the "high expression group" (150 points).

[0105] Immunohistochemical staining showed that the positive signal of CDKN2A protein was mainly located in the cell nucleus. In 36 cases of adjacent normal kidney tissue, CDKN2A was mostly negative or weakly positive, with 35 cases (97.2%) being negative, 1 case (2.8%) being weakly positive, and no moderately or strongly positive expression. However, in 36 cases of renal cell carcinoma tissue, the expression level of CDKN2A was significantly increased, with 19 cases (52.8%) being negative, 3 cases (8.3%) being weakly positive, 4 cases (11.1%) being moderately positive, and 10 cases (27.8%) being strongly positive (as shown in Table 1).

[0106] Table 1. Immunohistochemical expression results of CDKN2A in renal cell carcinoma and adjacent normal tissues.

[0107]

[0108] The results of the intergroup difference analysis showed (Table 2) that the high expression rate of CDKN2A in renal cell carcinoma tissue was 38.9% (14 / 36), which was significantly higher than that in adjacent normal tissue (0 / 36). The difference between the two groups was statistically significant (χ²=17.51, P<0.001).

[0109] Table 2 Immunohistochemical expression of CDKN2A in renal cell carcinoma and adjacent normal tissues.

[0110]

[0111] The above results suggest that CDKN2A is significantly upregulated in clear cell renal cell carcinoma, consistent with the results of TCGA database analysis, further confirming the potential role of CDKN2A in the development and progression of renal cell carcinoma.

[0112] To further validate the expression level and subcellular localization characteristics of CDKN2A protein, this study retrieved immunofluorescence data of CDKN2A in normal human cell lines from The Human Protein Atlas (HPA) database (https: / / www.proteinatlas.org / ). CDKN2A in human cell line A-4... (Green fluorescence), and is mainly located in the cell nucleus ( Figure 11 The nuclear localization features observed in the immunohistochemical assay (GI) were completely consistent with those observed in the cellular assay. This data further confirms at the cellular level the stable expression pattern of CDKN2A as a nuclear protein, supporting the reliability of the high expression of CDKN2A observed in ccRCC tissues in this study.

[0113] Example 3

[0114] This embodiment aims to screen potential active ingredients of traditional Chinese medicine that can target CDK2 / CDK6, a key downstream molecule of CDKN2A, using network pharmacology methods and molecular docking technology.

[0115] Given that the p16 protein encoded by CDKN2A regulates the cell cycle by inhibiting CDK4 / 6 activity, and that its downstream CDK2 plays a key role in the G2 / M transition of the cell cycle and resistance to CDK4 / 6 inhibitors, this study identified CDK2 and CDK6 as the core targets for drug screening.

[0116] Using the key gene for copper death, “CDKN2A,” as the search term, potential active ingredients of traditional Chinese medicine (TCM) were retrieved through the TCM Systems Pharmacology Database and Analysis Platform (TCMSP, v2.3, https: / / www.tcmsp-e.com / ). Entering “CDKN2A” (Target ID: TAR04110) in the “Target name” field yielded five TCMs associated with CDKN2A: Bupleurum chinense (Radix Bupleuri), Cyathula chinensis (Radix Cyathulae), Pueraria lobata (Flos Puerariae), Hemerocallis fulva (Radix Hemerocallis), and Pueraria lobata (Radix Puerariae), as well as the only directly associated active monomer, puerarin. All active ingredients from these five TCMs were collected, and a screening threshold of oral bioavailability (OB) ≥ 30% and drug-likeness (DL) ≥ 0.18 was set, resulting in 51 candidate ingredients. Twelve overlapping items from the five TCMs were extracted. Although the OB value of puerarin is less than 30%, it is retained based on the principle of prioritizing direct association.

[0117] Reverse target-component association prediction was performed on the bioinformatics annotation database platform for molecular mechanisms of traditional Chinese medicine (BATMAN-TCM, v2.0, http: / / bionet.ncpsb.org.cn / batman-tcm / ). The "herbs bygenes" module was accessed, "CDKN2A" was entered, and a prediction score threshold of Confidence score > 0.8 was set. Twelve candidate active monomers of traditional Chinese medicine were obtained, including flavonoids and isoflavones such as quercetin, kaempferol, and luteolin.

[0118] A search was conducted using the Herbal Identification High-Throughput Experimental Database (HERB, v2.0, http: / / 47.92.70.12 / ). The search term "CDKN2A" was entered in the "Target" search box to filter for "high-affinity components" validated through biological experiments. The significance threshold was set to p<0.05, resulting in 18 candidate active monomers of traditional Chinese medicine.

[0119] By merging the search results from the three databases mentioned above and removing duplicates based on the standard PubChem CID of the compounds, an initial library of active ingredients of traditional Chinese medicine for CDKN2A was constructed, containing a total of 39 traditional Chinese medicine monomers.

[0120] Drug-likeness screening of all initially screened components was performed using the SwissADME online platform (https: / / swissadme.ch / ). The normalized SMILES sequences of each component were imported in batches, and the screening criteria were set as follows: (1) GI absorption: High (considered OB ≥ 30%); (2) Bioavailability Score ≥ 0.55 (considered DL ≥ 0.18); (3) Number of violations of Lipinski's five rules ≤ 1. Compounds that did not meet any of the criteria were eliminated.

[0121] Through this step, 22 candidate components that meet the requirements for drug development were selected from 39 initial screening components, with a pass rate of 56.4%.

[0122] Target reverse validation was performed on the 22 components screened via ADME. Using the online platform SwissTargetPrediction (https: / / www.swisstargetprediction.ch / ), the SMILES structures of the active components were entered one by one, and the species "Homo sapiens" was selected to predict the target proteins. A screening threshold of Probability > 0.4 was set, and components containing CDK2 and / or CDK6 as predicted targets were selected.

[0123] The results showed that the predicted probability of apigenin targeting CDK6 was 1.00; the predicted probability of kaempferol targeting both CDK2 and CDK6 was 0.48; and the predicted probability of quercetin targeting both CDK2 and CDK6 was 0.50. Therefore, these three flavonoids were identified as the core active ingredients of the traditional Chinese medicine.

[0124] All predicted targets (Probability>0.2) of the three core components were merged and deduplicated to obtain 109 target genes, which were then submitted to the Metascape platform for multidimensional functional enrichment analysis.

[0125] like Figure 13-14As shown, the target genes were most significantly enriched at the molecular functional level in "phosphotransferase activity," suggesting that the core component targets are widely involved in kinase-related phosphorylation signal transduction. In biological processes, the targets were significantly enriched in "G2 / M Transition of Mitotic Cell Cycle," highly consistent with previous predictions that the components target core cell cycle kinases such as CDK2 / 6. Furthermore, key biological processes such as "cellular response to hormone stimulus," "positive regulation of cell migration," "cell activation," and "carboxylic acid metabolism" were also significantly enriched. Notably, "carboxylic acid metabolism" is closely related to the tricarboxylic acid cycle and is a key potential node connecting the "copper death" mechanism.

[0126] KEGG pathway analysis showed that ( Figure 14 The target genes were significantly enriched in pathways closely related to immune and infectious responses, such as tuberculosis and epithelial cell signaling in Helicobacter pylori infection. These pathways involve macrophage activation, inflammatory responses, and regulation of epithelial cell invasion, suggesting that the core components may exert anti-tumor effects by regulating tumor-associated macrophage (TAM) function and inhibiting epithelial-mesenchymal transition (EMT) in renal cancer cells. Enrichment results from the Reactome pathway database further confirmed that the target is involved in a broad and important signaling network, including receptor tyrosine kinase signaling, platelet activation, and TP53 activity regulation. The TP53 activity regulation pathway is particularly relevant to the pathological mechanisms of clear cell renal cell carcinoma (ccRCC).

[0127] Disease enrichment analysis using the DisGeNET database ( Figure 15The results indicate that these target genes are significantly associated with a range of malignant tumors, including precursor T-cell lymphoblastic leukemia-Lymphoma, mesothelioma, rhabdomyosarcoma, and Ewing sarcoma. Although clear cell renal cell carcinoma (ccRCC) was not among the top-ranked tumors, the strong association between the targets and broad categories such as "hematopoietic system tumors" and "refractory cancers" supports the broad anti-tumor potential of the core active ingredient at the disease level.

[0128] Molecular docking verification

[0129] High-resolution crystal structures of human CDK6 (PDB ID: 3NUP) and CDK2 (PDB ID: 3QQF) were downloaded from the Protein Structure Database (RCSB PDB, https: / / www.rcsb.org / ). Water molecules, protoligand molecules, and irrelevant ions were removed from the crystal structures using PyMOL software, and the processed structures were saved in PDB format.

[0130] The 3D structures of three core active ingredients and a positive control drug were downloaded from the PubChem database: apigenin (CID:5280443), kaempferol (CID:5280863), quercetin (CID:5280343), and palbociclib (CID:5330286, as a positive control). The ligands were subjected to energy minimization using Chem3D software (MM2 force field, convergence criterion: RMSGradient < 0.001 kcal / mol·Å), and the results were saved in SDF format.

[0131] Blind docking was performed on each treated receptor protein using the CB-Dock2 online platform (http: / / 183.56.231.194:8001 / cb-dock2 / index.php) according to receptor-ligand combinations (2 targets × [3 core components + positive control]). Potential binding pockets were automatically identified, and the binding domain with the largest volume and optimal binding energy in the positive control results was selected as the final docking site. This verified whether all components occupied the same catalytic site as the positive control palbociclib, thereby validating the competitive inhibition mechanism.

[0132] The molecular docking results are shown in Table 3. All three flavonoid components exhibited strong affinity for both CDK2 and CDK6 (binding energies ranging from -8.2 to -8.7 kcal / mol), significantly exceeding the strong binding threshold (-7.0 kcal / mol). Apigenin and quercetin showed equivalent inhibition of CDK2 / CDK6 (binding energy difference <0.1 kcal / mol), suggesting broad-spectrum activity in cell cycle regulation. All components bound to the ATP-binding pocket (CDK2: center 70, 34, -44; CDK6: center 24, 36, -1), similar to the binding pattern of palbociclib, confirming their competitive inhibition of kinase activity. Visualization analysis showed that the flavonoid components were embedded in the catalytic center of CDK2 / CDK6, forming hydrophobic interactions and hydrogen bonds with key residues such as Lys33 and Ala41, stably binding to the ATP pocket.

[0133] Table 3. Molecular docking binding energy between core components and CDK2 / CDK6

[0134]

[0135] This embodiment successfully obtained three active ingredients from traditional Chinese medicine that can target CDK2 / CDK6 with high probability—apigenin, kaempferol, and quercetin—through network pharmacology screening and molecular docking verification. They have a strong binding ability to CDK2 / CDK6, and the binding mode is similar to that of the clinical drug palbociclib, suggesting that they may exert anti-ccRCC effects by regulating the cell cycle, laying a solid foundation for subsequent cell and animal experimental verification.

[0136] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. Application of CDKN2A gene as a prognostic biomarker or drug target for clear cell renal cell carcinoma.

2. The application according to claim 1, characterized in that, The application is in the construction of a prognostic assessment model for clear cell renal cell carcinoma, which uses the expression levels of the CDKN2A gene and copper death-related genes FDX1, PDHA1, PDHB, and DBT as input features.

3. The application according to claim 1, characterized in that, The application is in the screening or preparation of therapeutic drugs for clear cell renal cell carcinoma that target the CDKN2A downstream signaling pathway.

4. A prognostic assessment system for clear cell renal cell carcinoma, characterized in that, include: Data input module: used to obtain expression data of CDKN2A, FDX1, PDHA1, PDHB and DBT genes in the clear cell renal cell carcinoma tissue of the subjects; Risk calculation module: used to input the expressed data into a pre-trained copper mortality risk scoring model and calculate the risk score value; Result output module: used to output prognostic assessment results based on the risk score value; The copper mortality risk scoring model uses the expression levels of CDKN2A and copper mortality-related genes FDX1, PDHA1, PDHB, and DBT as input features and is constructed using the "StepCox [both]+survival-SVM" machine learning algorithm.

5. The prognostic assessment system according to claim 4, characterized in that, The system further includes a sample processing module containing reagents for detecting CDKN2A gene expression levels. These reagents include specific primer pairs with the following sequences: Forward primer: 5'-GGGTTTTCGTGGTTCACATCC-3'; Reverse primer: 5'-CTAGACGCTGGCTCCTCAGTA-3'.

6. A computer-readable storage medium for prognostic assessment of clear cell renal cell carcinoma, having stored thereon a computer program, characterized in that, When the computer program is executed by the processor, it performs the following steps: The expression data of CDKN2A, FDX1, PDHA1, PDHB and DBT genes were obtained in the clear cell renal cell carcinoma tissue of the subjects. The expressed data is input into a pre-trained copper mortality risk scoring model to calculate the risk score. Based on the risk score, the prognostic assessment results for clear cell renal cell carcinoma are output.

7. A method for prognostic assessment of clear cell renal cell carcinoma, characterized in that, Includes the following steps: (1) Obtain expression data of CDKN2A, FDX1, PDHA1, PDHB and DBT genes in the clear cell renal cell carcinoma tissue of the subjects; (2) Input the expressed data into a pre-trained copper mortality risk scoring model and calculate the risk score. The model is constructed using the "StepCox [both]+survival-SVM" machine learning algorithm. (3) Assess the prognostic risk level of the subject based on the risk score.

8. A therapeutic drug for clear cell renal cell carcinoma targeting the CDKN2A downstream signaling pathway, characterized in that, It contains an effective amount of active ingredients from traditional Chinese medicine, wherein the active ingredients are selected from apigenin, kaempferol, quercetin or their pharmaceutically acceptable salts, solvates or prodrugs; the active ingredients exert their anti-renal clear cell carcinoma effect by targeting and inhibiting CDK2 and / or CDK6 kinase activity and regulating the G2 / M phase transition of the cell cycle.

9. Application of apigenin, kaempferol or quercetin as CDK2 / CDK6 inhibitors in the preparation of therapeutic drugs for clear cell renal cell carcinoma.

10. The therapeutic drug according to claim 9, characterized in that, The drug also includes a pharmaceutically acceptable carrier or excipient, and the dosage form of the drug is selected from tablets, capsules, granules, injections, or lyophilized powder for injection.

Images