Construction method of colorectal cancer prognosis model based on M7G modification

Abstract

The invention belongs to the technical field of gene technology and biomedicine, and provides a construction method of a colorectal cancer prognosis model based on M7G modification, which comprises the following steps: firstly, acquiring and downloading the miRNA expression quantity of METTL1 and cofactors WDR4 thereof of 539 tumor samples and 9 normal samples from a TCGA database, and then analyzing differential genes between the tumor samples and the normal samples by using an edgeR packet, then selecting 20 miRNAs before the top according to a difference multiple, and drawing a heat map and a volcano map by using a pheatmap and a gplo2 packet so as to present different expression modes of the 20 genes; the colorectal cancer prognosis model consisting of 13 M7G modification related miRNAs is established, can be used as a prognosis evaluation index, and has potential values of helping patients to layer and formulate personalized treatment strategies; and meanwhile, the M7G modification has potential influence on immune cell subset infiltration and immune function in the tumor microenvironment, which is possibly beneficial to deepening the understanding of people on the colorectal cancer tumor microenvironment and brings new hope for the development of immunotherapy intervention in the future.

Description

Technical Field

[0001] This invention belongs to the fields of gene technology and biomedical technology, specifically a method for constructing a prognostic model for colorectal cancer based on M7G modification. Background Technology

[0002] Colorectal cancer (CRC) is one of the most common cancers worldwide, with 1 to 2 million new cases diagnosed each year, making it the third most common cancer and the fourth leading cause of cancer-related deaths, causing 70,000 deaths annually, after lung, liver, and stomach cancer. By sex, colorectal cancer is the second most common cancer in women (9.2%) and the third most common in men (10%). Therefore, identifying new biomarkers for effective prognosis of colorectal cancer is crucial.

[0003] Colorectal cancer is caused by mutations in oncogenes, tumor suppressor genes, and genes related to DNA repair mechanisms. Abnormal methylation can lead to disease and cancer, and RNA methylation plays a crucial role in regulating gene expression. Among these, N1-methyladenosine (m1A), N6-methyladenosine (m6A), 5-methylcytosine (m5C), and 7-methylguanosine (M7G) RNA methylations possess various biological characteristics. M7G is the most common modification at the 5' end of miRNAs, catalyzed by the Trm8 / Trm82 complex in yeast and the METTL1 and WDR4 complex in humans under the action of methyltransferases. However, Its detection in low-abundance microRNAs (miRNAs) is hampered by a lack of sensitive detection strategies; the most characteristic enzyme mediating internal M7G methylation is the TRMT8 yeast enzyme homolog METTL1 (methyltransferase-like 1), which, together with its cofactor WDR4 (WD repeat domain 4), catalyzes M7G at specific miRNA sites; METTL1 methyltransferase mediates M7G methylation within miRNAs, and this enzyme regulates cell migration through its catalytic activity; meanwhile, increasing evidence suggests that M7G modification can regulate mRNA transcription, nuclear processing, tRNA stability, miRNA biological function, and 18S... rRNA maturation plays a crucial role in the translation of oncogenic miRNAs and cancer development. Existing research indicates that M7G modification is a common type of post-transcriptional RNA modification. Its function in the tumor microenvironment (TME) has attracted widespread attention in recent years. M7G modification can affect the proliferation of immune cells such as dendritic cells, M1 macrophages, and T cells, while adipocytes, fibroblasts, immune cells, and tumor-associated macrophages are mainly found in the tumor immune microenvironment. The tumor microenvironment plays a vital role in cancer development; for example, the main function of M1 macrophages is to promote antigen presentation and secretion. M7G secretes immune-activating factors and exerts anti-tumor effects; fibroblasts can inhibit cancer immunity, promote cancer progression, and cause patients to develop resistance to immunotherapy, which is associated with poor prognosis; therefore, it is inferred that M7G modification has an impact on the tumor immune microenvironment; in domestic studies, a prognostic risk model for COAD patients composed of M7G-related lncRNAs was created, and it was found that the tumor immune infiltration microenvironment was closely related to the risk model; however, to date, few studies have explored the potential role of METTL1 and WDR4-mediated M7G in the tumor immune microenvironment (TIME) and the prognosis of CRC patients.

[0004] To address the problems raised in the background art, those skilled in the art have proposed a method for constructing a prognostic model for colorectal cancer based on M7G modification. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method for constructing a prognostic model for colorectal cancer based on M7G modification, thereby resolving the issues existing in the prior art.

[0006] A method for constructing a prognostic model for colorectal cancer based on M7G modification includes the following steps:

[0007] S1. First, the miRNA expression levels of METTL1 and its cofactor WDR4 in 539 tumor samples and 9 normal samples were obtained and downloaded from the TCGA database;

[0008] S2. Next, the edgeR package is used to analyze differentially expressed genes between tumor and normal samples. Then, the top 20 miRNAs are selected based on the fold change. Finally, the pheatmap and ggplot2 packages are used to draw heatmaps and volcano plots to show the different expression patterns of these 20 genes.

[0009] S3. Then, univariate Cox regression was used to identify prognostic genes associated with M7G from 111 differentially expressed miRNAs; then Lasso regression analysis was used to analyze 17 key genes to obtain the best fit results.

[0010] S4. Finally, construct a prognostic model based on the best fit results.

[0011] Preferably, in step S3, the analysis of 17 key genes using Lasso regression analysis includes: firstly, performing Lasso regression analysis on these 17 miRNAs to screen out 13 miRNAs that are closely related to the overall survival of colorectal cancer patients, and then using these genes to construct a prognostic model.

[0012] Application of M7G in a prognostic model of colorectal cancer.

[0013] Validation of the preferred M7G-related gene prognostic model.

[0014] Preferably, the verification method includes gene function enrichment analysis, immune correlation analysis, and immune infiltration analysis.

[0015] Preferred gene function enrichment analysis: The ClusterProfiler package in R was used to perform GO (Gene Ontology), DO (Disease Ontology), and KEGG (Kyoto Encyclopedia of Genes and Genomes) enrichment analyses on differentially expressed immune-related genes. The results were visualized using barplot and bubble, and P<0.05 was considered significant enrichment.

[0016] Preferred immune correlation analysis: Using the CIBERSORT and TIMER algorithms, the cellular composition of complex tissues was determined based on the standardized gene expression profile, thereby assessing the infiltration of 16 immune cells and the expression of 13 immune functions in each sample. The correlations between the 16 immune cells and the 13 immune functions were assessed using corrplot, and the immune cell infiltration and immune function expression of each sample were displayed using heatmap.

[0017] Preferably, immune infiltration analysis: The R package "GSVA" was used to quantify the infiltration of each immune cell in high- and low-risk colorectal cancer tissues and test related immune functions using the single-sample gene enrichment analysis (ssGSEA) algorithm. The differences between the high- and low-risk groups were displayed using box plots. Univariate Cox regression analysis was used to screen for genes with prognostic significance from 960 candidate genes (P<0.05). Correlation analysis was further used to screen for 7 immune-related genes for ssGSEA analysis, and the cor R package was used for correlation analysis.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] This invention establishes a colorectal cancer prognostic model composed of 13 M7G-modified miRNAs, which can serve as a prognostic assessment indicator and has the potential value of helping patients stratify and develop personalized treatment strategies. Furthermore, we found that M7G modification has a potential impact on the infiltration of immune cell subsets and immune function in the tumor microenvironment, which may help deepen our understanding of the colorectal cancer tumor microenvironment and bring new hope for the development of future immunotherapeutic interventions. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the prognostic model of the present invention;

[0021] Figure 2 This is a schematic diagram of the prognostic model verification analysis of the present invention;

[0022] Figure 3 This is a schematic diagram of the results of multifactor Cox regression analysis according to the present invention;

[0023] Figure 4 This is a schematic diagram illustrating the enrichment of differentially expressed genes in this invention.

[0024] Figure 5 This is a schematic diagram illustrating the correlation between the 13 immune functions of the present invention;

[0025] Figure 6 This is a schematic diagram illustrating the relationship between immune cells and immune function according to the present invention. Detailed Implementation

[0026] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0027] like Figure 1 As shown:

[0028] Example: This invention provides a method for constructing a prognostic model for colorectal cancer based on M7G modification, comprising the following steps:

[0029] S1. First, the miRNA expression levels of METTL1 and its cofactor WDR4 in 539 tumor samples and 9 normal samples were obtained and downloaded from the TCGA database;

[0030] S2. Next, the edgeR package is used to analyze differentially expressed genes between tumor and normal samples. Then, the top 20 miRNAs are selected based on the fold change. Finally, the pheatmap and ggplot2 packages are used to draw heatmaps and volcano plots to show the different expression patterns of these 20 genes.

[0031] S3. Then, univariate Cox regression was used to identify prognostic genes related to M7G from 111 differentially expressed miRNAs; then Lasso regression analysis was used to analyze 17 key genes, from which 13 miRNAs closely related to the overall survival of colorectal cancer patients were screened out, and the best fit results were obtained.

[0032] S4. Finally, construct a prognostic model based on the best fit results.

[0033] To further understand the relevant miRNAs, we performed multivariate Cox regression analysis.

[0034] In multivariate regression analysis, 13 key genes were identified:

[0035] (hsa-miR-194-3p, hsa-miR-21-3p, hsa-miR-32-5p, hsa-miR-149-5p, hsa-miR-200c-5p, hsa-miR-221-3p, hsa-miR-129-2-3p, hsa-miR-9-5p, hsa-miR-216a-5p, hsa-miR-5683, hsa-miR-3613-3p, hsa-miR-6807-5p, hsa-miR-378d) were associated with overall survival, among which the three genes (hsa-miR-3613-3p, hsa-miR-6807-5p, hsa-miR-378d) showed the strongest association (Figure 1F).

[0036] As shown above, by establishing a colorectal cancer prognostic model composed of 13 M7G modification-related miRNAs, it can serve as a prognostic assessment indicator, which has the potential value of helping patients stratify and develop personalized treatment strategies, and has high accuracy.

[0037] Specifically, the validation of the prognostic model for M7G-related genes.

[0038] We downloaded 496 samples from the TCGA database as a validation set and analyzed them at the miRNA level. Based on median risk scores, the samples were divided into high-risk and low-risk groups. Principal component analysis (PCA) and T-SNE (t-distributed random neighbor embedding) were then used to analyze the differences between the high- and low-risk groups in a two-dimensional space. Receiver operating characteristic (ROC) curve analysis was used to assess the accuracy of the risk model in the validation set, and Kaplan-Meier curve analysis was used to show the overall survival of patients in the high-risk and low-risk groups. Based on median risk scores, the survival time and survival status of the two groups were analyzed.

[0039] We downloaded clinical data from 533 colorectal cancer patients from the Cancer Genome Atlas (TCGA) database to calibrate survival curves. To validate the independence of the prognostic model, we performed univariate and multivariate Cox regression analyses. Additionally, we combined clinical case characteristics and risk scores, and used the RMS package in R software to plot nomograms to predict the accuracy of colorectal cancer patient survival. Kaplan-Meier curves were used to independently analyze the predictive effect of miRNA expression on patient survival. Then, we used the ESTIMATE package to score the latter's immune function, distinguishing between high and low scores and screening for differentially expressed genes. To obtain common differentially expressed genes, we used the venn.diagram function in the Venn Diagram package to plot Venn diagrams.

[0040] like Figure 2 As shown, risk scores for the validation set were obtained using lasso regression analysis, and the groups were divided into high-risk and low-risk groups. PCA and tSNE methods were used to compare the differences between the two groups. The figure shows that PCA (… Figure 2 A) The difference is not significant, but tSNE ( Figure 2 B) Significant differences were observed between the high- and low-risk groups.

[0041] In the validation set, the area under the curve (AUC) showed that the 1-year, 3-year, and 5-year overall survival rates were 0.714, 0.731, and 0.695, respectively, with the 3-year OS rate being the most accurate, indicating that the model has high accuracy in predicting the prognosis of colorectal cancer patients. Figure 2C); Kaplan-Meier survival analysis showed that the overall survival rate in the high-risk group was significantly lower than that in the low-risk group (P<0.01). Figure 2 D) proves the accuracy of the risk prognosis model.

[0042] The validation set was divided into high-risk and low-risk groups based on the median risk score; blue represents the low-risk group and red represents the high-risk group. The results showed that the higher the patient's risk score, the shorter the patient's survival time. Figure 2 E, F);

[0043] To verify the independence of the prognostic model, we performed univariate Cox regression analysis and multivariate Cox regression analysis on age, gender, grading, and risk score. Figure 2 The results showed that grade and risk score could be independent factors affecting survival; based on the risk score and other clinical indicators of patient prognostic characteristics, we constructed a nomogram to more comprehensively predict patient survival. Figure 2 I), and using the correction curve ( Figure 2 G) Evaluating its performance, the results showed that the predicted total score for patients was 230 points, and the accuracy rates for predicting overall survival at 1 year, 3 years, and 5 years were 0.943, 0.871, and 0.759, respectively, indicating that our model has good practicality.

[0044] As can be seen from the above, the AUC values ​​for predicting overall survival at 1 year, 3 years, and 5 years are all greater than 0.65, which makes the risk prognosis model of this invention have good sensitivity and specificity.

[0045] Combination Figure 3 As shown, based on the results of the multivariate Cox regression analysis, we further analyzed hsa-miR-21-3p separately. Figure 3 A), hsa-miR-149-5p ( Figure 3 B), hsa-miR-194-3p Figure 3 C), hsa-miR-200c-5p ( Figure 3 D), hsa-miR-378d Figure 3 We used the expression of hsa-miR-21-3p, hsa-miR-194-3p, and hsa-miR-200c-5p to validate patient survival rates; we found that patients with high expression of hsa-miR-21-3p, hsa-miR-194-3p, and hsa-miR-200c-5p had higher survival rates than those with low expression, and patients with low expression of hsa-miR-149-5p and hsa-miR-378d had higher survival rates than those with high expression; after intersecting 2032 immunologically differentially expressed genes with 1878 risk and prognostic differentially expressed genes, we obtained 960 prognostic-related immunologically co-expressed genes (E). Figure 3 F).

[0046] More specifically, the verification methods include gene function enrichment analysis, immune correlation analysis, and immune infiltration analysis.

[0047] Gene function enrichment analysis: The ClusterProfiler package in R was used to perform GO (Gene Ontology), DO (Disease Ontology), and KEGG (Kyoto Encyclopedia of Genes and Genomes) enrichment analyses on differentially expressed immune-related genes. The results were visualized using barplot and bubble. P < 0.05 was considered significant enrichment.

[0048] To further understand gene functional enrichment analysis, combined with Figure 4 As shown, we performed GO and KEGG analyses on the aforementioned 960 differentially expressed genes to observe the enrichment of these genes; GO analysis ( Figure 4 (A, B) indicates that, from a biological process perspective, these genes are mainly enriched in nucleosome assembly, chromatin assembly or disassembly, DNA packaging, protein-DNA complex assembly, gene expression regulation, and epigenetics; from a cellular localization perspective, these genes are significantly enriched in nucleosomes, DNA packaging complexes, protein-DNA complexes, and spliceosome snRNP complexes; and at the molecular functional level, they are significantly enriched in protein heterodimerization, receptor ligand activity, and signal receptor activators; while KEGG analysis ( Figure 4 C, D) showed that these differentially expressed genes were mainly enriched in the cAMP signaling pathway, NF-kappa B signaling pathway, cytokine receptor interaction, and neuroactive ligand-receptor interaction pathways; in addition, our DO analysis ( Figure 4 E, F) indicates that four of these immune-related genes are enriched in symbiotic bacterial infectious diseases.

[0049] Immune correlation analysis: Using the CIBERSORT and TIMER algorithms, the cellular composition of complex tissues was determined based on the standardized gene expression profile, thereby assessing the infiltration of 16 immune cells and the expression of 13 immune functions in each sample. The correlations between the 16 immune cells and the 13 immune functions were evaluated using corrplot, and the immune cell infiltration and immune function expression of each sample were displayed using heatmap.

[0050] To further understand immune-related analysis, combined with Figure 5As shown, based on immune scores, we assessed the correlation between the proportions of different subsets of tumor-infiltrating immune cells, visualized as a heatmap. The analysis revealed that the degree of infiltration among the 16 immune cell subsets was generally moderate to highly correlated, with particularly significant correlations between Treg and TIL, TIL and T-helper-cells, TIL and TH1-cells, and TIL and B-cells. Figure 5 A); We also assessed the correlations among 13 immune functions. Heatmap visualization results showed that these immune functions generally exhibited moderate to high correlations, with particularly significant correlations among Check-point and CCR, APC_co_stimulation, Inflammation−promoting, T_cell_co−inhibition, and T_cell_co−stimulation. Figure 5 B); We present the expression of immune cells and immune function in each sample in a heatmap ( Figure 5 C).

[0051] Immune infiltration analysis: The R package "GSVA" was used to quantify the infiltration of each immune cell type in high- and low-risk colorectal cancer tissues and test related immune functions using the single-sample gene enrichment analysis (ssGSEA) algorithm. The differences between the high- and low-risk groups were displayed using box plots. Univariate Cox regression analysis was used to screen for genes with prognostic significance from 960 candidate genes (P<0.05). Correlation analysis was further used to screen for 7 immune-related genes for ssGSEA analysis, and the cor R package was used for correlation analysis.

[0052] To further understand immune infiltration analysis, combined with Figure 6 As shown, the infiltration of the immune system is related to cancer progression and patient prognosis. We assessed the correlation between risk scores and the level of immune infiltration. We calculated patient scores based on the constructed risk score model to divide patients into high-risk and low-risk groups. Then, we performed differential analysis of immune cells and immune function in the high-risk and low-risk groups. The results were visualized using box plots. The immune cell analysis results showed significant differences in the content of B cells, macrophages, pDCs, and Th2 cells between the high-risk and low-risk groups. Among them, the content of B cells, macrophages, and pDCs was higher in the high-risk group, while the content of Th2 cells was higher in the low-risk group. Figure 6 A) Immune function analysis results showed significant differences in the expression of MHC_class_I and Type_II_IFN_Response between high-risk and low-risk groups, with MHC_class_I...

[0053] The expression of MHC_class_I and Type_II_IFN_Response showed significant differences, with MHC_class_I being expressed at a higher level in the low-risk group, while Type_II_IFN_Response was expressed at a higher level in the high-risk group. Figure 6 B).

[0054] Finally, we calculated the expression levels of 960 genes in the overlapping area of ​​the aforementioned Venn diagram across all samples. Based on univariate Cox regression analysis, we screened out 35 genes associated with patient prognosis. We then performed correlation analysis on these genes with immune cells and immune function, ultimately selecting seven significantly associated (P<0.05) prognostic genes—CCL19, IGHV2-70D, FABP4, LINC01419, CALML5, ISM2, and HAND1—for ssGSEA analysis to assess their correlation with immune infiltration. The results showed that genes such as FABP4, IGHV2-70D, and CCL19 were positively correlated with the aforementioned 16 immune cell subsets and 13 immune functions, with CCL19 showing a particularly significant correlation. Conversely, genes such as ISM2 and LINC01419 were negatively correlated with these immune cells and immune functions. Figure 6 C).

[0055] As shown above, using a validation set to validate the model can provide a more accurate and intuitive understanding of the independent prognostic value of the risk model and the results of systematic analysis of immune-related gene function, immune correlation, and tumor immune cell infiltration.

[0056] In summary, by constructing a prognostic assessment model composed of the expression of 13 M7G-related genes, Kaplan-Meier curves and receiver operating characteristic (ROC) curves confirmed that the risk characteristics of the 13 M7G-related genes have accurate predictive value for the prognosis of CRC patients; and the prognostic assessment capability of the model was validated in the test dataset. The immune correlation analysis of tumor-infiltrating immune cells showed that the degree of infiltration among 16 immune cell subsets and 13 immune functions were generally moderate to high. The correlation analysis between immune-related prognostic genes and immune infiltration showed that genes such as FABP4, IGHV2-70D, and CCL19 were positively correlated with the aforementioned 16 immune cell subsets and 13 immune functions, with CCL19 showing a particularly significant correlation, while genes such as ISM2 and LINC01419 were negatively correlated with these immune cells and immune functions.

[0057] The embodiments of the present invention are given for the purposes of illustration and description. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for constructing a prognosis model of colorectal cancer based on M7G modification, characterized by comprising the following steps: Comprising the following steps: ​ S1, first obtain and download the miRNA expression of METTL1 and its cofactor WDR4 from 539 tumor samples and 9 normal samples in TCGA database; S2, then analyze the differential genes between tumor and normal samples using edgeR R package, then select the top 20 miRNAs according to the difference fold, and then use pheatmap and ggplot2 packages to draw heat map and volcano plot to present the different expression patterns of the 20 genes; S3, then use single factor cox regression to identify the prognosis genes related to M7G from 111 miRNAs differential genes; then use Lasso regression analysis to analyze 17 key genes to obtain the best fitting result; S4, finally construct the prognosis model according to the best fitting result.

2. The method for constructing a prognosis model of colorectal cancer based on M7G modification according to claim 1, wherein: In step S3, the Lasso regression analysis of 17 key genes includes: first, lasso regression analysis is performed on the 17 miRNAs, and 13 miRNAs closely related to the overall survival rate of patients with colorectal cancer are screened out, and then a prognosis model is constructed using these genes.

3. Application of M7G in a prognosis model for colorectal cancer.

4. Use of M7G in a prognostic model for colorectal cancer according to claim 3, characterized in that, Comprising: Verification of the M7G related gene prognosis model.

5. Use of M7G in a prognostic model for colorectal cancer according to claim 4, characterized in that: The verification method comprises gene function enrichment analysis, immune correlation analysis and immune infiltration analysis.

Images