Folate Metabolism in Colorectal Cancer Reveals Links Between Clinical and Immune Traits, Identifying CYP26A1 as a Target

Folate Metabolism in Colorectal Cancer Reveals Links Between Clinical and Immune Traits, Identifying CYP26A1 as a Target | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Folate Metabolism in Colorectal Cancer Reveals Links Between Clinical and Immune Traits, Identifying CYP26A1 as a Target Jian Zhang, Yifei Zhu, Teng Zhou, Doudou Li, Yao Zheng, Yanxi Yao, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5647525/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 03 Jul, 2025 Read the published version in Genes & Immunity → Version 1 posted 10 You are reading this latest preprint version Abstract Background: Folic acid plays a key role in cellular regulation and metabolism, commonly found in dietary supplements. However, its complex role in colorectal cancer (CRC), particularly in metabolism and immune evasion, remains unclear. Methods: We developed the FMRG_score system using machine learning algorithms based on TCGA and GEO databases to assess modification patterns influencing CRC patients' clinical and immune characteristics. The system’s reliability was validated using multiple external clinical cohorts receiving immunotherapy. We further explored the relationships between FMRGs-related features and clinical traits, mutation profiles, biological functions, immune infiltration, therapy response, and drug sensitivity. Results: By combining in vitro experiments and bioinformatics analysis, we established a 9-gene risk model associated with folate metabolism to predict CRC prognosis. Notably, CYP26A1, a key component of the model, was upregulated in CRC tissues, promoting cell proliferation, migration, and invasion. Significant differences in clinical traits, immune cell infiltration, immune checkpoint expression, therapy response, and drug sensitivity were observed between risk groups. Conclusion: The folate scoring system can assess CRC prognosis, tumor microenvironment, and immune therapy response. This is the first study proposing CYP26A1 as an oncogene in CRC. Folate metabolism Colorectal cancer CYP26A1 Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Introduction Colorectal cancer (CRC) is the third most common malignancy. It is also the second deadliest cancer worldwide 1 . CRC is 1.5 times more common in men than in women, with most cases occurring after the age of 50 2 . However, CRC incidence has been increasing in countries with a rising human development index, especially in those under age 50 3 . By 2035, the number of deaths from colon cancer and rectal cancer is projected to rise by 60.0% and 71.5%, respectively, due to population growth and aging 4 . Research definitively shows that early diagnosis and precise treatment of CRC significantly improve patient survival rates 5, 6 . Recent advances in molecular biology have facilitated the delineation of critical genetic pathways that play a role in colorectal carcinogenesis 7, 8 . These pathways include the adenomatous polyposis coli (APC) pathway 9 , the microsatellite instability (MSI) pathway 10 , and the CpG island methylator phenotype (CIMP) pathway 11 . Understanding the intricate mechanisms of these pathways can provide valuable insights for developing targeted therapeutic strategies to combat colorectal cancer. In recent years, the rapid advancement of immune checkpoint inhibitor (ICI)-based immunotherapy has brought about a promising new era in anticancer treatment. ICIs enhance the antitumor immune response by disrupting the signaling of key immunosuppressive proteins like programmed cell death 1 (PD-1), programmed cell death-1 ligand 1 (PD-L1), and cytotoxic T lymphocyte antigen 4 (CTLA-4) 12-14 . Some patients with melanoma, non-small-cell lung cancer (NSCLC), and other cancers show sustained responses 15-18 . Immunotherapy using ICI in CRC patients has shown promising safety and efficacy outcomes. Approved immunotherapeutic agents for advanced CRC treatment include pembrolizumab, nivolumab, and ipilimumab 19 . In 2015, Le DT et al. discovered that the anti-PD-1 drug pembrolizumab exhibited a notably higher response rate in patients with the DNA mismatch repair–deficient (dMMR)/microsatellite instability-high (MSI-H) molecular subtype of CRC 20 . This finding implies the potential benefits of anti-PD-1 therapy for patients with both CRC and non-CRC dMMR tumors. ICIs are effective in treating dMMR/microsatellite-stable (MSS) CRC due to the elevated mutational load, abundance of neoantigens, and increased immune cell presence within this tumor subtype 21, 22 . Significantly, MSS-pMMR patients make up a substantial portion of CRC patients 23 . Hence, it is essential to actively explore new biomarkers and innovative approaches for early detection and treatment to understand the underlying mechanisms of varied treatment responses in current colorectal cancer research. The relationship between folate metabolism and CRC has been extensively studied due to folate's crucial role in DNA methylation, repair, and synthesis processes 24-28 . Folate, a- vitamin B naturally present in foods, and its synthetic form folic acid, commonly utilized in supplements and food fortification, play a vital role in cell production and upkeep, especially during stages of accelerated cell division and growth, such as infancy and pregnancy 29, 30 . Based on these pivotal functions, it has been hypothesized by researchers that sufficient folate intake has the potential to mitigate DNA alterations that might serve as precursors to cancer development, particularly in the colon and rectum 31 . Recent studies and reviews suggest a complex relationship between folate intake, folate status, and CRC risk. A systematic review and meta-analysis examining the impact of folic acid supplement intake and total folate intake on CRC risk found no significant effect of folic acid supplements in randomized controlled trials. Despite these findings, the effect of folate status, as measured by red blood cell folate content, on CRC risk was not significant, indicating that the relationship may depend on the form of folate consumed and other individual factors 32 . A study found that high folic acid intake accelerates methionine cycling in cancerous tissues in vivo, potentially contributing to the development of hepatocellular carcinoma 33 . Recent evidence suggests that the relationship between high plasma levels of folate and CRC risk may not be straightforward 34-36 . For example, one study observed a significant decrease in CRC risk among women in the highest quartile of plasma folate levels 37 . In contrast, another study identified an increased risk of CRC associated with high plasma folate levels over a follow-up period 38 . These inconsistencies underscore the challenges inherent in accurately assessing the impact of folate on CRC risk. The relevance of genetic polymorphisms in folate metabolism genes, particularly in genes such as MTHFR that code for enzymes in folate metabolism, should be noted 39 . However, the impact of folate metabolism on TME characteristics and clinical outcomes of CRC patients is still uncertain. We systematically analyzed the underlying effects of folate metabolism in CRC in this study. Through our analysis, we constructed a folate metabolism-related signature to predict the prognostic outcomes, TME characteristics, and immunotherapy response in CRC patients. Our findings not only revealed the paramount role of folate metabolism in the complex heterogeneity of CRC but also indicated its potential to enhance individualized management strategies for CRC patients. This study also marked the first instance of proposing the potential anticancer function of Cytochrome P450 Family 26 Subfamily A Member 1 (CYP26A1) in CRC, laying the groundwork for further exploration of this promising molecular target in CRC. Materials and methods Data collection The training cohort included the transcriptome data and clinical information of 585 patients with colorectal adenocarcinoma (COAD and READ) were retrieved from the Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/repository). Validation sets were obtained from GSE39582 (579 patients), GSE17536 (177 patients), GSE38832 (122 patients) GSE17537 (55 patients) expression profiles from the Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo). Single-cell RNA sequencing (scRNA-seq) data for colorectal adenocarcinomas were obtained from the GSE161277 and GSE205506 databases, which include 13 and 19 CRC samples, respectively. 410 folate metabolism-related genes (FMRGs) were retrieved from GeneCards (https://www.genecards.org/). The complete gene details are displayed in Table S1. The RNA-seq sample expression levels were converted from fragments per kilobase of transcript per million mapped reads (FPKM) to transcripts per million (TPM), and then log2(TPM + 1) was calculated. Somatic mutation data, copy number variation (CNV) files, and tumor mutation burden (TMB) data of CRC patients were retrieved from the TCGA database. Construction and validation of FMRGs risk signature First, we analyzed the genes related to folate metabolism and the differential genes from TCGA, leading to the identification of 138 genes. Subsequently, we conducted univariate analysis on these 138 genes to pinpoint those significantly correlated with patient survival (p<0.05). Following this, a combination of the least absolute shrinkage and selection operator (LASSO) technique with multivariate regression analysis was employed to refine the gene selection process and ascertain risk coefficients strongly linked to prognosis. In order to identify high-impact genes, tenfold cross-validation was used to determine the optimal regularization parameter. Genes with non-zero coefficients were then deemed potential prognostic markers. Candidate genes were chosen through multivariate Cox analysis to create a prognostic FMRG_score in the training dataset. The FMRG_score was calculated as follows: In the training set, 585 patients were stratified into low-risk and high-risk groups based on the median risk score, which was calculated using the risk coefficient (Coefi) and expression of each gene (Expi). Subsequently, Kaplan-Meier survival analysis was conducted on the two groups to assess their survival outcomes. The predictive performance of the signature was evaluated using receiver operating characteristic (ROC) curves. Kaplan–Meier survival curves were plotted, and log-rank tests were performed to assess the statistical significance of the observed differences in survival between the two risk groups. The effectiveness of the prediction model was further validated in three independent GEO datasets (GSE38832, GSE39582, GSE17537) through survival analysis and calculation of the area under the curve (AUC) in ROC analysis. Relationship between risk groupings and clinical characteristics The study analyzed clinical factors such as age, gender, and TNM stage. Differences in prognostic outcomes were assessed using Kaplan-Meier analysis in R software with the “survival” and “survminer” packages 40 . Development and validation of a nomogram scoring system To determine the independent prognostic values and the predictive efficacy of the FMRG_score in predicting the survival of CRC patients, we conducted univariable and multivariate Cox regression analyses for better clinical practice. Furthermore, we investigated the relationship between the FMRG_score and various clinical characteristics. Additionally, in order to improve the prognostic accuracy of our model, a nomogram was developed utilizing the risk score, T stage, and N stage as independent prognostic factors to estimate the probability of overall survival (OS) at 2-, 4-, and 6-years. Subsequently, the predictive efficacy was compared between the FMRG_score and different clinical pathological factors based on AUC. Functional and pathway enrichment analysis To identify specific biological pathways enriched between the high- and low-risk groups, we performed a comprehensive analysis using various techniques. Initially, we conducted Gene Ontology (GO) analysis (Supplementary Table 8) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis (Supplementary Table 7) to understand the functions of the screened candidate genes and related pathways. Following this, we utilized the "ClustProfiler" and "GSVA" packages for further elucidation 41, 42 . Additionally, we employed Gene Set Variation Analysis (GSVA) (Supplementary Table 10) to identify variations in gene sets. Subsequently, Gene Set Enrichment Analysis (GSEA) was utilized to identify enriched pathways or gene sets based on differential expression results between the high-risk and low-risk groups (Supplementary Table 9). Moreover, GSVA (Supplementary Table 11) and correlation analysis between Hallmark pathway activities and the FMRG risk score were conducted to explore potential pathways associated with the identified signature. Exploration of the immune landscape in distinct risk groupings The immune characteristics of 585 CRC samples were assessed by evaluating the scores of tumor microenvironment (TME) cells using the single sample gene set enrichment analysis (ssGSEA) algorithm 43 . To analyze the relative proportion of 22 immune cells within the CRC samples, CIBERSORT (https://cibersort.stanford.edu) was employed. The CIBERSORT algorithm was executed using R software. By leveraging the 585 samples gene expression matrix and the provided gene expression feature set of the 22 immune cell subtypes from the official website, simulation calculations were iterated 1000 times to derive the relative composition ratio of the 22 immune cells in each sample. Subsequently, the immune score and ESTIMATE score of each patient were assessed utilizing the R package of estimate 44 . Prediction of response to immunotherapy Initially, the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm was utilized to evaluate the potential disparities in treatment responses between high- and low-risk groups 45 . A higher TIDE score was found to be associated with reduced treatment efficacy, thereby highlighting a negative correlation between the TIDE score and treatment effectiveness. Furthermore, data on immunotherapy was gathered from various datasets including the IMvigor210 dataset for urothelial carcinoma (UC), the TCGA dataset, GSE17536 dataset, GSE39582 dataset, and GSE38832 dataset for CRC. Subsequently, within each dataset, the FMRG_score was calculated to predict responses to immunotherapy. Mutation and drug susceptibility analysis To explore the differences in therapeutic effects of chemotherapeutic drugs between high- and low-risk groups of CRC patients, we first generated the mutation annotation format (MAF) from the TCGA database using the "maftools" R package. Subsequently, we calculated the TMB score for each patient within these groups. Additionally, we determined the semi-inhibitory concentration (IC50) values of commonly used chemotherapeutic drugs for treating CRC by utilizing the "pRRophetic" package. Single-cell RNA sequencing analysis The scRNA-seq analysis was conducted using the "Seurat" R package for data processing, including quality control steps where cells expressing 200-7000 genes were retained, and cells with more than 20% mitochondrial gene expression were excluded. The cells were classified into eight primary types using t-SNE dimensionality reduction. The "inferCNV" package was used to infer copy number variations (CNVs) from single-cell RNA sequencing data by comparing gene expression profiles to a reference dataset. This approach allowed identification of regions with CNV alterations. Based on these CNV patterns, the malignancy score was calculated to classify cells as either malignant or non-malignant, reflecting tumor characteristics. The "AUCell" algorithm was employed to assess gene set activity at the single-cell level. Cell-to-cell communication networks were analyzed using "CellChat," focusing on receptor-ligand interactions. Additionally, the "Monocle 2" algorithm was used to construct pseudotime trajectories to track the dynamic functional changes in FDRG-related signatures. Cell culture and reagents The human colon carcinoma cell line HCT116 and RKO were obtained from the Cell Bank of Chinese Academy of Sciences (Shanghai, China). All cell lines were authenticated by monitoring cell vitality, mycoplasma contamination, and short tandem repeat profiling. Cells were maintained in DMEM medium supplemented with 10% FBS and 1% penicillin/streptomycin. Culture media and supplements were obtained from BasalMedia (Shanghai, China). DNA constructs, transfection, and viral transduction CYP26A1 cDNA in penter vector with C terminal Flag and His tag was purchased from Vigene Bioscience. Short hairpin RNAs (shRNAs) targeting human CYP26A1 (shCYP26A1) in GIPZ lentiviral vector and corresponding control constructs were purchased from Dharmacon. The detailed information of expression vectors for molecular cloning is provided in Supplementary Table 14. shRNAs in lentiviral expression vectors were transfected into HEK293T together with packaging plasmid mix using Neofect DNA transfection reagent. Supernatants were collected after 48 h of transfection and used for infecting cells in the presence of 8 mg/mL of polybrene. After 24 h of infection, cells were selected with 2 mg/mL of puromycin (Cayman, Ann Arbor, USA) for 1-2 week. Cell viability and colony formation assays Cell proliferation assays were performed using Cell Counting Kit-8 (CCK-8) (Dojindo, Shanghai, China) according to the manufacturer’s instructions. The absorbance was measured at a wavelength of 450nm (A450). For colony-formation assays, cells were grown onto 6-well plate at a density of 2000 cells/well for 14 days with media replacement every 3 days. Cells were stained with 1% crystal violet and the number of survival colonies was counted. Western blotting Proteins from cell lines and tissues were extracted using RIPA lysis buffer (Thermo Fisher Scientific, USA) on ice. After centrifugation (15000g, 10 min), protein concentration was determined with a BCA protein assay kit (Thermo Fisher Scientific, USA). Protein samples were separated on PAGE gels (Epizyme Biomedical Technology, Shanghai, China) and transferred to 0.22 µm Immobilon PVDF membranes (Millipore Sigma, USA). Membranes were blocked with 5% milk and incubated with primary antibodies overnight at 4°C. Secondary antibodies (anti-Rabbit IgG) were applied at room temperature for 1 h, and immunoreactivity was detected using an ECL system (Share-bio, Shanghai, China). Primary antibodies used: pan-kla (#PTM-1401RM, PTMBio, Hangzhou, China) at 1:1000. Secondary antibody dilution was 1:5000. Transwell migration assays and Matrigel invasion assays Cell migration and invasion were assessed using transwell chambers (Corning Biocoat, Tewksbury, USA). For invasion assays, the Matrigel (BD Biosciences) was diluted 1:8 (dilution ratio determined based on the MMP expression levels of the cells) and applied to coat the upper surface of the chamber membrane. The coated chambers were incubated at 37°C for 30 minutes to allow the Matrigel to polymerize and hydrated prior to use. For both assays, cells were resuspended in DMEM containing 1% FBS and loaded at a density of 5 ×10^4 cells per well onto the upper well of chambers, while growth medium containing 10% FBS was placed in the lower chamber as a chemoattractant. After 24h incubation, cells on the upper surface were gently removed with a cotton swab, and cells that had migrated or invaded to the lower surface were fixed with methanol and stained with 1% crystal violet solution. The number of cells was counted under a light microscope with a magnification of 100. All assays were conducted in triplicate and repeated at least three times. RNA isolation and quantitative reverse transcription-PCR Total RNA was isolated from cell lines and tissue samples using Trizol reagent (Invitrogen), Reverse transcription was performed using a PrimeScript RT reagent Kit (TaKara). The resultant cDNA was subjected to quantitative real-time PCR (qPCR) using SYBR Premix Ex Taq (Tli RNaseH Plus) (Takara) on an Eppendorf Mastercycler ep realplex4 instrument. Primer sequences for qPCR are listed in Supplementary Table 15. All reactions were performed in triplicate. The data is present as mean ± SD. In vitro T cell activation and tumor cells-T cells coculture model PBMCs from healthy donors were treated with RBC lysis buffer and sorted using CD8 A antibody by flow cytometry. After activation with anti-CD3 (BioLegend, 5 μg/ml) and anti-CD28 (BioLegend, 2.5 μg/ml) at the indicated concentrations and cultured in media containing IL-2 (PeproTech, 10 ng/ml), the cells were used in co-culture models. T cells (2 x 10^5 cells/well) were seeded in the top chamber of a transwell (pore size: 0.4 µm), while tumor cells, with or without CYP26A1 knockdown, were seeded in the bottom chamber for indirect co-culture. In the direct co-culture model, T cells and tumor cells were mixed directly at a 1:4 ratio without using a chamber. Following co-culture, T cells from both models were subjected to surface marker and intracellular cytokine staining and analyzed by FACS. Statistical analysis All experiments were replicated at least three times and the results are presented as mean ± standard deviation (SD). For comparisons between two groups, unpaired Student's t-tests were used for normally distributed variables, while Mann-Whitney U tests were employed for non-normally distributed variables. For multiple group comparisons, one-way ANOVA (parametric) or Kruskal-Wallis tests (nonparametric) were performed as appropriate. Pearson's or Spearman's correlation analyses were conducted to evaluate linear relationships between variables. Survival curves were generated using the Kaplan-Meier method, and differences between groups were assessed using the log-rank test. ROC curves were generated and area under the AUC values were calculated to evaluate predictive performance. Statistical analyses were performed using SPSS version 23.0, R software (version 4.3.2), and GraphPad Prism 9. P-values < 0.05 were considered statistically significant. Results Construction and validation of the prognostic FMRG_score The flowchart for this article was shown in Figure 1. Oue To construct the FMRG_score, we identified 138 target genes through differential analysis of CRC data from TCGA, intersecting with genes associated with folate metabolism (Fig. 2A, Supplementary Table 1-2). We chose to conduct additional analysis on the TCGA cohort to extract a prognostic signature for FMRG. Initially, a univariable Cox regression analysis revealed 19 candidate genes that were found to be associated with CRC, with a p-value < 0.05 (Supplementary Table 3-4). To delve deeper into the characteristics of folate metabolism in CRC patients, we utilized LASSO Cox regression to determine the optimal λ value, leading to the identification of 9 key genes (Supplementary Table 5-6). The mRNA expressions of the 9 key genes in tumor and normal specimens were assessed using the TCGA-CRC dataset. Notably, AHCY, PLK1, WNT5A, CYP26A1, BDNF and DRD4 exhibited upregulation, while NAT1, CD36 and GSTM1displayed downregulation in tumor (Fig. 2D). Five of the risk genes (NAT1, WNT5A, CYP26A1, BDNF and DRD4) were identified as independent predictive factors (Table 1). The locations of these 9 risk genes on chromosomes were visualized in Figure 2I. We displayed the relationship between these nine risk genes and patient prognosis through the TCGA CRC dataset (Fig. S1). This comprehensive model integrates the predictive power of individual genes to provide a more precise tool for assessing CRC prognosis. The risk-score model was determined using the following equation: FMRG_score = (-0.168)*AHCY + (-0.157)*PLK1 + (-0.26)*NAT1 + (0.136)*CD36 + (-0.194)*WNT5A + (0.181)*CYP26A1 + (0.156)*BDNF + (0.235)*DRD4 + (0.155)*GSTM1 Patients were divided into two groups based on their FMRG_score: those with a score lower than the median were categorized as low-risk (n = 292), while those with a score higher than the median were classified as high-risk (n = 293). The distribution plot showed that as FMRG_score increased, survival times decreased and recurrence rates increased (Fig. 2F-G). The Kaplan–Meier survival curves indicated that patients with low scores had a significantly better overall survival compared to those with high scores (log-rank test, p < 0.001; Fig. 2E). Furthermore, the 2-, 4-, and 6-year survival rates predicted by the FMRG_score were reflected in the AUC values of 0.7026, 0.7027, and 0.6927, respectively (Fig. 2C). Clustering heatmaps revealed that CYP26A1, DRD4, GSTM1, CD36, and BDNF were more prominent in the high-risk group, while AHCY, PLK1, NAT1, and WNT5A were more prevalent in the low-risk group (Fig. 2H). To validate the prognostic performance of the FMRG_score, we calculated FMRG_scores across three external validation groups (GSE38832, GSE39582, GSE17537) (Fig. S2). Patients were also stratified into low- or high-risk groups based on the formula utilized for the training set. Survival analysis was conducted, revealing a significantly better prognosis in the low-risk group compared to the high-risk group (log rank; p < 0.05). Assessment of the 2-, 4-, and 6-year prognostic prediction classification efficiencies demonstrated that the FMRG_score maintained relatively high AUC values. These results indicate that the FMRG_score exhibits excellent ability to predict the survival outcomes of CRC patients.We combined clinical characteristics with the FMRG_score to create a nomogram predicting the 2-, 4-, and 6-year OS of CRC patients based on the strong relationship between the FMRG_score system and patient prognosis (Figure S3A). The calibration plot showed excellent agreement between the predicted overall OS using the nomogram and the actual OS of CRC patients at 2, 4, and 6 years (Fig. 3A). Correlation between Clinical Characteristics and the FMRGs-related Predictive Signature Univariable and multivariate Cox regression was used to explore the relationship between the FMRG_score and clinical characteristics (Fig. 3B and Fig. S3B), and the results showed that FMRG_score and M stage could be used as independent prognostic factors (HR = 2.177, 95% CI = 1.628, 2.912, P < 0.001; HR = 2.455, 95% CI = 1.508, 3.998, P < 0.001, respectively). Kaplan-Meier analyses were conducted across subgroups stratified by distinct clinical factors to comprehensively validate the predictive reliability of the FMRG_score. Our findings revealed that patients with a low FMRG_score tended to have a lower TNM stage, whereas those with a higher FMRG_score were more likely to present with an advanced TNM stage (Fig. 3C). Specifically, the high-risk group had a smaller proportion of patients with stages I and II relative to the low-risk group, and a higher proportion of patients with stages III and IV. Our results showed that the FMRG_score increased with advancing stages and was significant in all components except for patients at stages I and II, where no significant difference was observed (Fig. 3D-G). This highlights the important role of the FMRG_score in predicting clinical stage. Stratified survival analyses showed that the prognostic outcomes for colorectal cancer patients varied significantly between the high and low FMRG-Groups, particularly in advanced TNM stages (Fig. 3H, Fig. S3C). At the same time, patients with high FMRG_score had a worse prognosis in both young and elderly patients (Fig. 3I-J), indicating the generalization of the FMRG_score. To gain insights into the biological processes linked to the poor survival in the high-risk group and explore the molecular implications of transcriptomic and genetic variances between high- and low-risk groups, we delved into the genomic heterogeneity of the FMRGs model within the TCGA cohort. Our examination of the mutation landscape of FMRGs aimed to uncover the specific genetic alterations associated with the high-risk group and further elucidate the underlying mechanisms contributing to the differential survival outcomes observed. Our study revealed that the high-risk group had more mutations in the genes APC , TP53 , TTN , and K-RAS , which may have enhanced their cancer-promoting characteristics (Fig. 3K). Landscape of biological characteristics of FMRGs-related signature Functional enrichment analyses were conducted to gain valuable insights into biological processes, molecular functions, and cellular components (Fig. 4). KEGG analysis showed that the high-risk group were considerably enriched in pro-tumor pathways like “Wnt signaling pathway” and “PPAR signaling pathway” (Fig. 4A). In recent years, the role of the tumor microenvironment in cancer development has garnered increasing attention. Newer studies have highlighted the emerging significance of the nervous system as a key factor in promoting tumor growth. Innervation plays a crucial role in the growth of various tumors, as neurons can establish tumor-nerve synapses with tumor cells. Through these signaling mechanisms, typical oncogenic signaling pathways are often activated, consequently fostering tumor growth 46 . GO analysis explained several neural-related sets that may be key to tumor progression (Fig. 4B). The results of the GSEA analysis highlighted that the high-risk group was significantly enriched in sets that promote tumor progression and metastasis such as “ECM RECEPTOR INTERACTION”, while the low-risk group was enriched in “CITRATE CYCLE TCA CYCLE” and “PYRUVATE METABOLISM” (Fig. 4C-D). The results of GSVA analysis also suggested that the high-risk group was enriched in cancer-promoting pathways, while the low-risk group was enriched in cancer-suppressor-related pathways (Fig. 4E-F) scRNA‑seq data processing and analysis of FMRGs-related signature The scRNA-seq data was analyzed with the "Seurat" R program, which included quality control measures such as screening for cells exhibiting low expression levels in the range of 200-7000 genes, and ensuring that mitochondrial genes accounted for no more than 20% of the total gene count. To define TME cell populations of CRC, we identified and visualized 8 main cell types using the T-distributed Stochastic Neighbor Embedding (tSNE) (Fig. 5A). We visualized the annotated cell clusters and found a significant decrease in T and B cells and a slight increase in plasma cells in the tumor compared to normal tissue (Fig. 5B). The "AUCell" R program, which evaluates gene set activity, was used to assign FMRGs-related signatures to individual cells (Fig. 5C). Our study revealed that epithelial cells and macrophages exhibited higher FMRG scores compared to other cell types. Specifically, epithelial cells showed a greater abundance of FMRG scores (Fig. 5D). As a result, we classified epithelial cells into two groups based on their FMRG scores - high and low - for further analysis (Fig. S4A). We analyzed the communication network between the nine cells, and results showed that T cells and the FMRG scores - high group had the most receptor-ligand communication with other cells (Fig. 5E). To explore the differential signaling pathways in the FMRG scores - high group and FMRG scores - low group, we further analyzed the signaling patterns of these nine cells (Fig. 5F).Our results suggest that epithelial cells with high expression of FMRG score are significantly enriched in some oncogenic pathways, such as "TGF", "WNT" and "EGF". To uncover dynamic functional changes of FMRGs-related signature, we adopted the Monocle 2 algorithm to chronologically order FMRGs-related signature in pseudotime and indicate their trajectories (Fig. S4B-C). We next investigated the transcriptional changes of FMRGs-related signature associated with trajectory and found that NAT1 was significantly enriched early in development, BDNF, DRD4, CD36 and WNT5A were progressively enriched as development progressed, and CYP26A1, AHCY and PLK1 were significantly enriched at the terminal end of development (Fig. S4D). The FMRG_score reshape the immune cell infiltration landscape Recent research has highlighted the critical role of the inflammatory environment in the development of CRC, particularly focusing on the activation status and interactions of immune and stromal cells with tumor cells 47 . The varying composition of the TME impacts the effectiveness of immunotherapy. We examined the immune profiles of CRC patients across different FMRG groups to better understand this. We analyzed 22 different types of immune cells in CRC patients with high and low FMRG scores. Significantly heightened infiltrations were observed in the high FMRG group for central memory CD4 T cell, CD56dim natural killer cell, Macrophages, Myeloid-derived suppressor cells (MDSC), natural killer cell, plasmacytoid dendritic cell, and T follicular helper cell. Conversely, diminished infiltrations were noted for activated CD4 T cell and type 2 T helper cell within the same group (Fig. 6A). A correlative analysis between the abundance of immune cell types and FMRG scores was further depicted in Figure S5A. A high FMRG_score was closely associated with a high stromal score, as well as a high estimate score (Fig. 6B). We examined how the nine genes in the proposed model were related to the abundance of immune cells. Our findings showed significant correlations between most immune cells and seven of the genes (Fig. S5B). Using the TIDE score, we found a notable rise in the Dysfunction score and TIDE score in the high-risk group (Fig. S5C-E), indicating an increased risk of immune escape in high-risk patients, which may reduce the efficacy of ICI therapy. We also predicted immunotherapy efficacy using TIDE, which was validated using the IMvigor cohort, TCGA CRC group, GSE17536, GSE39582, and GSE38832, and we found that the higher the FMRG score, the more insensitive the patient may be to immunotherapy leading to ineffectiveness of immunotherapy (Fig. 6C-G). The sensitivity of patients in the low-risk and high-risk groups to chemotherapeutic agents currently used for treating colorectal cancer or undergoing clinical trials was evaluated (Fig. 7). The IC50 values of cisplatin and gemcitabine were notably lower in patients with low FMRG scores, potentially offering more precise clinical treatment guidance. These findings collectively indicate an association between FMRG and drug sensitivity. CYP26A1 plays a carcinogenic role in CRC We validated the FMRGs-related signature by our own sequencing data (Fig. 8A). A protein-protein interaction (PPI) network among the FMRGs-related signature was constructed from the STRING database, in which CYP26A1 interacts most with other proteins (Fig. 8B). We performed a follow-up analysis of CYP26A1, which we verified in the GSE161277 dataset to be highly expressed in tumor tissues (Fig. 8C). The CYP26A1 gene encodes a member of the cytochrome P450 superfamily of enzymes crucial for various biological processes, such as drug metabolism, the synthesis of cholesterol, steroids, and other lipids 48 . In the context of cancer, CYP26A1 plays a significant role in the metabolism of retinoic acid (RA), a key metabolite of vitamin A that governs cell proliferation, differentiation, and apoptosis 49, 50 . Aberrant expression of CYP26A1 has been linked to the pathogenesis and progression of various tumors by affecting RA levels and modulating RA signaling pathways 51 . Studies have shown that increased CYP26A1 expression can diminish RA levels in cells, thereby promoting tumorigenesis by facilitating cell proliferation and maintaining a less differentiated cell state 52 . Conversely, the downregulation of CYP26A1 in specific cancers can elevate RA concentrations, potentially inhibiting tumor growth and inducing apoptosis 53 . Interestingly, in our sequencing data of tumor primary and metastatic foci from 8 pairs of patients, we found that CYP26A1 was highly expressed in metastatic foci. Additionally, analysis of TCGA CRC data indicated an association between elevated CYP26A1 expression and poor prognosis in stage III and IV patients (Fig. 8D-E). We performed GSEA analysis of CYP26A1 using our own data, and we found that some pro-oncogenic pathways were significantly enriched when CYP26A1 was highly expressed, whereas pathways related to immunity and autophagy were enriched when CYP26A1 was expressed, also suggesting that CYP26A1 functions as an oncogene (Fig. 8F). To further explore the relationship between CYP26A1 and the TME of CRC patients, single-cell RNA sequencing data were analyzed from 19 patients with d-MMR/MSI-H CRC treated with the neoadjuvant PD-1 blockade 54 (Fig. 9A). Significant reduction of malignant cells accompanied by an increase in CD8+ T cells and B cells in the CYP26A1 low-expression group (Fig. 9B, C) and CYP26A1 expression negatively correlated with the number of CD8+ T cells (Fig. 9D). Based on classical markers 55, 56 , CD8+ T cells were reclassified into five subpopulations: exhausted T cells (Tex), effector T cells (Teff), tissue resident memory T cells (Trm), naive T cells, and memory T cells (Fig. 9E). The proportion of Teffs was dramatically upregulated in CRCs with low CYP26A1 expression, while the relative ratio of Texs was lower (Fig. 9F, G). In addition, we observed that CYP26A1 expression was higher in non-pathological complete response (pCR) patients, suggesting that CYP26A1 were associated with immunotherapy non-response (Fig. 9H). CYP26A1 is upregulated in colorectal tumor, promotes colon cancer cell proliferation, enhances colorectal cancer cell migratory and invasive potential. We investigate the expression of CYP26A1 in CRC and its impact on immune cell function, revealing the relationship between CYP26A1 and CRC prognosis. Immunohistochemical staining on tissue microarrays demonstrated a significant increase in the overall CYP26A1 expression in CRC tissues compared to normal tissues (Fig. 10A). Additionally, our findings show that higher expression of CYP26A1 are associated with poor prognosis and the high CYP26A1 expression group exhibited a higher recurrence rate (Fig. 10B). These results underscore the potential clinical significance of CYP26A1 as a prognostic marker for CRC, highlighting its relationship with disease progression and outcomes. Further exploration of the molecular mechanisms of CYP26A1 in CRC may provide new therapeutic avenues for this challenging malignancy. To further investigate the role of CYP26A1 in colorectal cancer, we first analyzed CYP26A1 mRNA levels in the Fudan University Cancer Center database via qPCR. The results showed that CYP26A1 mRNA levels were elevated in primary colon tumors compared to normal colon tissues (Fig. 10C). Furthermore, to explore the biological function of CYP26A1 in colon cancer, we stably depleted endogenous CYP26A1 in HCT116 and RKO cells (Fig. 10D) by infecting them with lentiviral vectors expressing shCYP26A1. The expression status of CYP26A1 in these stable cell lines was validated by Western blot (WB) and qPCR. CCK-8 assays showed that knockdown of CYP26A1 suppressed cell proliferation in HCT116 and RKO cells (Fig. 10E). Given the invasive and metastatic nature of colon cancer cells, we next investigated the effect of CYP26A1 on the invasive and metastatic phenotype of these cells. Transwell migration and invasion assays revealed that shCYP26A1-infected HCT116 and RKO cells displayed reduced migratory and invasive abilities compared to shNC-infected cells (Fig. 10F). Additionally, the study found that CYP26A1 plays a crucial role in regulating CD8+ T cell function. Knocking down CYP26A1 resulted in increased secretion of granzyme B, IFN-γ, and TNF-α by CD8+ T cells in co-culture conditions compared to the control group (shNC). Specifically, both shCYP26A1-1 and shCYP26A1-2 groups showed significantly higher expression of granzyme B and cytokines, with statistically significant differences compared to the control group (P < 0.01, P < 0.001) (Fig. 10G). These results suggest that CYP26A1 regulates immune effector molecules, potentially influencing the tumor immune microenvironment. Our findings provide new evidence for CYP26A1 as an immune therapy target, with potential clinical applications. In conclusion, CYP26A1 plays a key role in immune regulation and tumor progression in CRC. Further exploration of its molecular mechanisms and potential application in immunotherapy will provide new directions and strategies for the treatment of colorectal cancer. Discussion In this study, we developed and validated an FMRG_score to predict prognosis, TME characteristics, and response to immunotherapy in CRC. Our integrative analysis, which combines large-scale datasets from TCGA and GEO with in vitro experiments, underscores the critical role of folate metabolism in CRC heterogeneity. It identifies key molecular pathways associated with immune evasion and treatment resistance. Notably, this study is the first to propose CYP26A1 as a potential oncogene in CRC, highlighting its significance as a biomarker and therapeutic target. The FMRG_score effectively stratified patients into high- and low-risk groups, revealing significant survival differences. Patients with a high FMRG_score exhibited a poor prognosis, which correlated with increased activation of pro-tumor pathways and immunosuppressive cell infiltration in the TME, including a higher prevalence of MDSCs and exhausted T cells. These findings suggest that folate metabolism plays a pivotal role in shaping the immune landscape of CRC, potentially mediating immune escape mechanisms that reduce the efficacy of ICIs. This aligns with existing literature on the complex interactions between metabolic reprogramming and immune regulation in cancer progression. One of the most striking findings of this study is the identification of CYP26A1 as a novel oncogenic driver in CRC. CYP26A1, a member of the cytochrome P450 family, is known for its role in retinoic acid metabolism, which regulates cellular proliferation, differentiation, and apoptosis. Our analysis revealed that high CYP26A1 expression was significantly associated with advanced tumor stage, poor overall survival, and an immunosuppressive TME. These results are consistent with previous reports linking aberrant retinoic acid metabolism to tumorigenesis. Furthermore, our single-cell RNA sequencing data demonstrated a correlation between high CYP26A1 expression and reduced infiltration of effector CD8+ T cells, reinforcing its potential role in immune evasion. These findings open new avenues for targeting CYP26A1 in CRC, either through direct inhibition or by modulating retinoic acid signaling to enhance immune response. However, it is not without limitations. A primary limitation is that the majority of our findings were derived from bioinformatic analyses. While these analyses are powerful, they necessitate further validation in in vivo systems to confirm their clinical relevance. Although we successfully demonstrated the tumor-suppressive function of CYP26A1 in vitro, it is imperative that future studies prioritize the use of animal models and clinical samples to verify these results in more physiologically relevant environments. Despite utilizing comprehensive public datasets, the clinical utility of the FMRG_score in predicting treatment outcomes requires validation through prospective clinical trials. Additionally, further mechanistic studies are essential to elucidate the precise role of CYP26A1 in CRC pathogenesis and to assess its potential as a therapeutic target. Future research should also investigate how folate metabolism interacts with other metabolic pathways, influencing cancer progression and therapy resistance. Abbreviations CRC（colorectal cancer）, APC（adenomatous polyposis coli）, MSI（microsatellite instability）, dMMR（DNA mismatch repair–deficient）, MSS（microsatellite-stable）, CIMP（CpG island methylator phenotype）, ICI（immune checkpoint inhibitor）, PD-1（programmed cell death 1）, PD-L1（programmed cell death-1 ligand 1）, CTLA-4（cytotoxic T lymphocyte antigen 4）, NSCLC（non-small-cell lung cancer）, MSI-H（microsatellite instability-high）, CYP26A1（Cytochrome P450 Family 26 Subfamily A Member 1）, TCGA（the Cancer Genome Atlas）, GEO（Gene Expression Omnibus）, scRNA-seq（single-cell RNA sequencing）, FMRGs（folate metabolism-related genes）, FPKM（per kilobase of transcript per million mapped reads）, TPM（transcripts per million）, CNV（copy number variation）, TMB（tumor mutation burden）, LASSO（least absolute shrinkage and selection operator）, Coefᵢ（coefficient）, Expᵢ（expression of each gene）, ROC（receiver operating characteristic）, AUC（area under the curve）, OS（overall survival）, GO（Gene Ontology）, KEGG（Kyoto Encyclopedia of Genes and Genomes）, GSVA（Gene Set Variation Analysis）, GSEA（Gene Set Enrichment Analysis）, TME（tumor microenvironment）, ssGSEA（single sample gene set enrichment analysis）, TIDE（Tumor Immune Dysfunction and Exclusion）, UC（urothelial carcinoma）, MAF（mutation annotation format）, IC50（semi-inhibitory concentration）, tSNE（T-distributed Stochastic Neighbor Embedding）, MDSCs（Myeloid-derived suppressor cells）, PPI（Protein-protein interaction）, RA（Retinoic acid）, Tₑₓ（exhausted T cells）, Tₑff（effector T cells）, Tᵣₘ（tissue resident memory T cells）, pCR（pathological complete response） Declarations Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. The data that support the results of current study is available on TCGA (https://portal.gdc.cancer.gov/) and GEO websites (http://www.ncbi.nlm.nih.gov/geo). Acknowledgements Thanks for the help of the CRC tissues form Fudan university Shanghai Cancer Center and thanks for the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) Database for sharing a large amount of data. The graphic figure was created with BioRender.com. Disclosure Funding Information The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. This study was supported by the National Key Research and Development Program of China (2023YFF1205000)；National Natural Science Foundation of China (grant no. 82072915 and 82373359); Project of Shanghai Municipal Health Commission (grant no. 202140397); CSCO-ROCHE Cancer Research Fund 2019 (grant no. Y-2019Roche-17 1); and Chinese Young Breast Experts Research project (grant no. CYBER-2021-001). Beijing Science and Technology Innovation Medical Development Foundation Key Project (grant no. KC2022-ZZ-0091-6). Conflict of Interest The authors declare no conflict of interest. Ethics Statement N/A Approval of the research protocol by an Institutional Reviewer Board N/A Informed Consent All authors have read and approved the final version of the manuscript and agree to its publication. Informed consent was obtained from all participants involved in the study. Registry and the Registration No. of the study/trial N/A Animal Studies N/A Author contributions Y.F. and T.Z designed the studies, conducted most of the experiments and data analysis, and wrote the draft manuscript. Y.X., M.X., C.Z. and Y.F conducted some of the molecular experiments. D.D. and J.Z. were involved in the overall design of the study and the revision of the final manuscript. All authors contributed to the article and approved the submitted version. All authors have made significant contributions to the conception, design, analysis, and/or interpretation of the work presented in this manuscript. Furthermore, all authors have thoroughly reviewed and approved the final version of the manuscript. They fully agree with its content and the decision to submit it for consideration. References Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A et al. . Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: a cancer journal for clinicians 2021; 71 (3) : 209-249. Murphy N, Ward HA, Jenab M, Rothwell JA, Boutron-Ruault M, Carbonnel F et al. . Heterogeneity of Colorectal Cancer Risk Factors by Anatomical Subsite in 10 European Countries: A Multinational Cohort Study. Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association 2019; 17 (7) : 1323-1331. Siegel RL, Miller KD, Goding Sauer A, Fedewa SA, Butterly LF, Anderson JC et al. . Colorectal cancer statistics, 2020. CA: a cancer journal for clinicians 2020; 70 (3) : 145-164. Araghi M, Soerjomataram I, Jenkins M, Brierley J, Morris E, Bray F et al. . Global trends in colorectal cancer mortality: projections to the year 2035. INT J CANCER 2019; 144 (12) : 2992-3000. Aguiar Junior S, Oliveira MMD, Silva DRME, Mello CALD, Calsavara VF, Curado MP. SURVIVAL OF PATIENTS WITH COLORECTAL CANCER IN A CANCER CENTER. Arquivos de gastroenterologia 2020; 57 (2) : 172-177. Vega P, Valentín F, Cubiella J. Colorectal cancer diagnosis: Pitfalls and opportunities. WORLD J GASTRO ONCOL 2015; 7 (12) : 422-433. Keum N, Giovannucci E. Global burden of colorectal cancer: emerging trends, risk factors and prevention strategies. Nature reviews. Gastroenterology & hepatology 2019; 16 (12) : 713-732. Fearon ER. Molecular genetics of colorectal cancer. Annual review of pathology 2011; 6: 479-507. Fodde R, Smits R, Clevers H. APC, signal transduction and genetic instability in colorectal cancer. Nature reviews. Cancer 2001; 1 (1) : 55-67. Vilar E, Gruber SB. Microsatellite instability in colorectal cancer-the stable evidence. Nature reviews. Clinical oncology 2010; 7 (3) : 153-162. Wu J, Zhang L, Kuchi A, Otohinoyi D, Hicks C. CpG Site-Based Signature Predicts Survival of Colorectal Cancer. BIOMEDICINES 2022; 10 (12). Wei SC, Duffy CR, Allison JP. Fundamental Mechanisms of Immune Checkpoint Blockade Therapy. CANCER DISCOV 2018; 8 (9) : 1069-1086. Weiner LM. Cancer immunology for the clinician. Clinical advances in hematology & oncology : H&O 2015; 13 (5) : 299-306. Qin S, Xu L, Yi M, Yu S, Wu K, Luo S. Novel immune checkpoint targets: moving beyond PD-1 and CTLA-4. MOL CANCER 2019; 18 (1) : 155. O'Brien M, Paz-Ares L, Marreaud S, Dafni U, Oselin K, Havel L et al. . Pembrolizumab versus placebo as adjuvant therapy for completely resected stage IB-IIIA non-small-cell lung cancer (PEARLS/KEYNOTE-091): an interim analysis of a randomised, triple-blind, phase 3 trial. The Lancet. Oncology 2022; 23 (10) : 1274-1286. Garon EB, Rizvi NA, Hui R, Leighl N, Balmanoukian AS, Eder JP et al. . Pembrolizumab for the treatment of non-small-cell lung cancer. The New England journal of medicine 2015; 372 (21) : 2018-2028. Larkin J, Chiarion-Sileni V, Gonzalez R, Grob JJ, Cowey CL, Lao CD et al. . Combined Nivolumab and Ipilimumab or Monotherapy in Untreated Melanoma. The New England journal of medicine 2015; 373 (1) : 23-34. Homet Moreno B, Ribas A. Anti-programmed cell death protein-1/ligand-1 therapy in different cancers. BRIT J CANCER 2015; 112 (9) : 1421-1427. Zhao W, Jin L, Chen P, Li D, Gao W, Dong G. Colorectal cancer immunotherapy-Recent progress and future directions. CANCER LETT 2022; 545: 215816. Le DT, Uram JN, Wang H, Bartlett BR, Kemberling H, Eyring AD et al. . PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. The New England journal of medicine 2015; 372 (26) : 2509-2520. Guidoboni M, Gafà R, Viel A, Doglioni C, Russo A, Santini A et al. . Microsatellite instability and high content of activated cytotoxic lymphocytes identify colon cancer patients with a favorable prognosis. The American journal of pathology 2001; 159 (1) : 297-304. Overman MJ, Lonardi S, Wong KYM, Lenz H, Gelsomino F, Aglietta M et al. . Durable Clinical Benefit With Nivolumab Plus Ipilimumab in DNA Mismatch Repair-Deficient/Microsatellite Instability-High Metastatic Colorectal Cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2018; 36 (8) : 773-779. Ganesh K. Optimizing immunotherapy for colorectal cancer. Nature reviews. Gastroenterology & hepatology 2022; 19 (2) : 93-94. Kim JW, Jeon YJ, Jang MJ, Kim JO, Chong SY, Ko KH et al. . Association between folate metabolism-related polymorphisms and colorectal cancer risk. MOL CLIN ONCOL 2015; 3 (3) : 639-648. Du W, Li W, Lu R, Fang J. Folate and fiber in the prevention of colorectal cancer: between shadows and the light. WORLD J GASTROENTERO 2010; 16 (8) : 921-926. Kherbek H, Daoud R, Soueycatt T, Soueycatt Y, Ali Z, Ehsan J et al. . The relationship between folic acid and colorectal cancer; a literature review. Annals of medicine and surgery (2012) 2022; 80: 104170. Jang MJ, Kim JW, Jeon YJ, Chong SY, Hong SP, Hwang SG et al. . Polymorphisms of folate metabolism-related genes and survival of patients with colorectal cancer in the Korean population. GENE 2014; 533 (2) : 558-564. Bouras E, Kim AE, Lin Y, Morrison J, Du M, Albanes D et al. . Genome-wide interaction analysis of folate for colorectal cancer risk. The American journal of clinical nutrition 2023; 118 (5) : 881-891. McNulty H, Pentieva K. Folate bioavailability. The Proceedings of the Nutrition Society 2004; 63 (4) : 529-536. Iyer R, Tomar SK. Folate: a functional food constituent. J FOOD SCI 2009; 74 (9) : R114-R122. Hubner RA, Houlston RS. Folate and colorectal cancer prevention. BRIT J CANCER 2009; 100 (2) : 233-239. Moazzen S, Dolatkhah R, Tabrizi JS, Shaarbafi J, Alizadeh BZ, de Bock GH et al. . Folic acid intake and folate status and colorectal cancer risk: A systematic review and meta-analysis. Clinical nutrition (Edinburgh, Scotland) 2018; 37 (6 Pt A) : 1926-1934. Li J, Yang H, Lei M, Zhu W, Su Y, Li K et al. . Dietary folate drives methionine metabolism to promote cancer development by stabilizing MAT IIA. SIGNAL TRANSDUCT TAR 2022; 7 (1) : 192. Ma J, Stampfer MJ, Giovannucci E, Artigas C, Hunter DJ, Fuchs C et al. . Methylenetetrahydrofolate reductase polymorphism, dietary interactions, and risk of colorectal cancer. CANCER RES 1997; 57 (6) : 1098-1102. Van Guelpen B, Hultdin J, Johansson I, Hallmans G, Stenling R, Riboli E et al. . Low folate levels may protect against colorectal cancer. GUT 2006; 55 (10) : 1461-1466. Otani T, Iwasaki M, Sasazuki S, Inoue M, Tsugane S. Plasma folate and risk of colorectal cancer in a nested case-control study: the Japan Public Health Center-based prospective study. Cancer causes & control : CCC 2008; 19 (1) : 67-74. Kato I, Dnistrian AM, Schwartz M, Toniolo P, Koenig K, Shore RE et al. . Serum folate, homocysteine and colorectal cancer risk in women: a nested case-control study. BRIT J CANCER 1999; 79 (11-12) : 1917-1922. Van Guelpen B, Hultdin J, Johansson I, Hallmans G, Stenling R, Riboli E et al. . Low folate levels may protect against colorectal cancer. GUT 2006; 55 (10) : 1461-1466. Weile J, Kishore N, Sun S, Maaieh R, Verby M, Li R et al. . Shifting landscapes of human MTHFR missense-variant effects. AM J HUM GENET 2021; 108 (7) : 1283-1300. Rich JT, Neely JG, Paniello RC, Voelker CCJ, Nussenbaum B, Wang EW. A practical guide to understanding Kaplan-Meier curves. Otolaryngology--head and neck surgery : official journal of American Academy of Otolaryngology-Head and Neck Surgery 2010; 143 (3) : 331-336. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC BIOINFORMATICS 2013; 14: 7. Yu G, Wang L, Han Y, He Q. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics : a journal of integrative biology 2012; 16 (5) : 284-287. Wang W, Chen Z, Hua Y. Bioinformatics Prediction and Experimental Validation Identify a Novel Cuproptosis-Related Gene Signature in Human Synovial Inflammation during Osteoarthritis Progression. BIOMOLECULES 2023; 13 (1). Chen B, Khodadoust MS, Liu CL, Newman AM, Alizadeh AA. Profiling Tumor Infiltrating Immune Cells with CIBERSORT. Methods in molecular biology (Clifton, N.J.) 2018; 1711: 243-259. Jiang P, Gu S, Pan D, Fu J, Sahu A, Hu X et al. . Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. NAT MED 2018; 24 (10) : 1550-1558. Shi DD, Guo JA, Hoffman HI, Su J, Mino-Kenudson M, Barth JL et al. . Therapeutic avenues for cancer neuroscience: translational frontiers and clinical opportunities. The Lancet. Oncology 2022; 23 (2) : e62-e74. Schmitt M, Greten FR. The inflammatory pathogenesis of colorectal cancer. Nature reviews. Immunology 2021; 21 (10) : 653-667. Thatcher JE, Isoherranen N. The role of CYP26 enzymes in retinoic acid clearance. EXPERT OPIN DRUG MET 2009; 5 (8) : 875-886. He W, Wang X, Chen M, Li C, Chen W, Pan L et al. . Metformin reduces hepatocarcinogenesis by inducing downregulation of Cyp26a1 and CD8(+) T cells. CLIN TRANSL MED 2023; 13 (11) : e1465. Osanai M, Sawada N, Lee G. Oncogenic and cell survival properties of the retinoic acid metabolizing enzyme, CYP26A1. ONCOGENE 2010; 29 (8) : 1135-1144. Alshafie GA, Walker JR, Curley RWJ, Clagett-Dame M, Highland MA, Nieves NJ et al. . Inhibition of mammary tumor growth by a novel nontoxic retinoid: chemotherapeutic evaluation of a C-linked analog of 4-HPR-glucuronide. ANTICANCER RES 2005; 25 (3c) : 2391-2398. Osanai M, Lee G. Increased expression of the retinoic acid-metabolizing enzyme CYP26A1 during the progression of cervical squamous neoplasia and head and neck cancer. BMC RES NOTES 2014; 7: 697. Lavudi K, Nuguri SM, Olverson Z, Dhanabalan AK, Patnaik S, Kokkanti RR. Targeting the retinoic acid signaling pathway as a modern precision therapy against cancers. FRONT CELL DEV BIOL 2023; 11: 1254612. Li J, Wu C, Hu H, Qin G, Wu X, Bai F et al. . Remodeling of the immune and stromal cell compartment by PD-1 blockade in mismatch repair-deficient colorectal cancer. CANCER CELL 2023; 41 (6) : 1152-1169. Zhang L, Yu X, Zheng L, Zhang Y, Li Y, Fang Q et al. . Lineage tracking reveals dynamic relationships of T cells in colorectal cancer. NATURE 2018; 564 (7735) : 268-272. Liu C, Yu H, Huang R, Lei T, Li X, Liu M et al. . Radioimmunotherapy-induced intratumoral changes in cervical squamous cell carcinoma at single-cell resolution. Vol 42.; 2022:1407-1411. Table 1 Table 1. Multivariate Cox regression analysis of 5 FMRGs associated with FMRG_score in CRC patients. id coef HR lower upper pvalue NAT1 -0.26 0.770873 0.623926 0.95243 0.015881 WNT5A -0.194 0.823309 0.689291 0.983383 0.031967 CYP26A1 0.181 1.197879 1.050177 1.366354 0.007163 BDNF 0.156 1.168938 1.004565 1.360205 0.0435 DRD4 0.235 1.264762 1.035357 1.544995 0.021435 Additional Declarations Yes there is potential conflict of interest. The authors declare no conflict of interest. Supplementary Files SupplementaryFigureandTableslegends.docx SupplementaryTableS1.xlsx SupplementaryTableS2.xlsx SupplementaryTableS3.xlsx SupplementaryTableS4.xlsx SupplementaryTableS5.xlsx SupplementaryTableS6.xlsx SupplementaryTableS7.xlsx SupplementaryTableS8.xlsx SupplementaryTableS9.xlsx SupplementaryTableS10.xlsx SupplementaryTableS11.xlsx SupplementaryTableS12.xlsx SupplementaryTableS13.xlsx SupplementaryTableS14.xlsx SupplementaryTableS15.xlsx SupplementaryfigS1.jpg SupplementaryfigS2.jpg SupplementaryfigS3.jpg SupplementaryfigS4.jpg SupplementaryfigS5.jpg Cite Share Download PDF Status: Published Journal Publication published 03 Jul, 2025 Read the published version in Genes & Immunity → Version 1 posted Editorial decision: revise 18 Feb, 2025 Review # 3 received at journal 04 Feb, 2025 Reviewer # 3 agreed at journal 01 Feb, 2025 Review # 1 received at journal 01 Feb, 2025 Reviewer # 2 agreed at journal 30 Jan, 2025 Reviewer # 1 agreed at journal 25 Jan, 2025 Reviewers invited by journal 24 Jan, 2025 Submission checks completed at journal 16 Dec, 2024 First submitted to journal 15 Dec, 2024 Editor assigned by journal 15 Dec, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5647525","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":406858491,"identity":"8cc4fe12-da36-4425-bf5f-9de536b7b0bd","order_by":0,"name":"Jian Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAy0lEQVRIiWNgGAWjYBACxmYQWSHHwHAAxGAjWssZYxK0QPS1kaKFuZ352cOf8wwS+46fPcDwoewwA//sBkIOYzM3kNxmkDjzTF4C44xzhxkk7hwgpIXBTMJw25/EDQdyDJh52w4zGEgkENLC/k0icY5B4obzbwyY/xKnhcdM4mADUMsNoC2MRGopk2w4ZmA888Ybg4M959J5JG4Q0GLYf3yb5I8aA9m+8zmGD36UWcvxzyCkpQGJcwCIefCrBwJ5gipGwSgYBaNgFAAAGCND6uiqujoAAAAASUVORK5CYII=","orcid":"","institution":"Fudan university","correspondingAuthor":true,"prefix":"","firstName":"Jian","middleName":"","lastName":"Zhang","suffix":""},{"id":406858492,"identity":"9ea727b2-7c4b-456b-be56-380cd5ccbff9","order_by":1,"name":"Yifei Zhu","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yifei","middleName":"","lastName":"Zhu","suffix":""},{"id":406858493,"identity":"36a02ec9-c9f7-4b0d-90ae-b02e8175a2eb","order_by":2,"name":"Teng Zhou","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Teng","middleName":"","lastName":"Zhou","suffix":""},{"id":406858494,"identity":"b01843d5-c996-4643-bc70-1b407e400e0d","order_by":3,"name":"Doudou Li","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Doudou","middleName":"","lastName":"Li","suffix":""},{"id":406858495,"identity":"d6e67e3f-fa14-49d5-8a68-a74c929aff9a","order_by":4,"name":"Yao Zheng","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yao","middleName":"","lastName":"Zheng","suffix":""},{"id":406858496,"identity":"fdb384b5-5dd0-4a4a-af03-19bebfeb2c2d","order_by":5,"name":"Yanxi Yao","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yanxi","middleName":"","lastName":"Yao","suffix":""},{"id":406858497,"identity":"00d69560-01fe-4ee6-a0f3-0b8e9d73a855","order_by":6,"name":"Minxi Lin","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Minxi","middleName":"","lastName":"Lin","suffix":""},{"id":406858498,"identity":"1f5befdf-34da-42e0-a8b1-bf87cbe35a4e","order_by":7,"name":"Cheng Zeng","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Cheng","middleName":"","lastName":"Zeng","suffix":""},{"id":406858499,"identity":"73d7fd95-e324-42d3-8061-ee1797c39b37","order_by":8,"name":"Yuxin Yan","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yuxin","middleName":"","lastName":"Yan","suffix":""},{"id":406858500,"identity":"0485dd37-8306-4d36-ab9f-07dafde70b16","order_by":9,"name":"Yifei Zhou","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yifei","middleName":"","lastName":"Zhou","suffix":""}],"badges":[],"createdAt":"2024-12-15 13:05:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5647525/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5647525/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41435-025-00342-6","type":"published","date":"2025-07-03T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":74960479,"identity":"1dedcc56-4376-45d4-91bc-00fcdfc0778c","added_by":"auto","created_at":"2025-01-28 18:51:53","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1291745,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWorkflow of the study.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/4cca36e044f2211a606604dd.jpg"},{"id":74961907,"identity":"c93b36f6-8683-4837-a177-21b141d396ab","added_by":"auto","created_at":"2025-01-28 19:07:56","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":966573,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDevelopment and validation of a prognostic model based on Folate metabolism-related genes (FMRGs) model in CRC\u003cbr\u003e\n\u003c/strong\u003e(A) Veen plots of folate metabolism-related genes and DEGs in TCGA-CRC.\u003c/p\u003e\n\u003cp\u003e(B) LASSO coefficient profiles and partial likelihood deviance plot for gene selection.\u003cbr\u003e\n(C) Time-dependent ROC curves at 2, 4, and 6 years, illustrating the prognostic performance of the FDRG risk model in the TCGA dataset.\u003cbr\u003e\n(D) Boxplot comparing the expression levels of key genes in normal vs. tumor tissues.\u003cbr\u003e\n(E) Kaplan-Meier survival analysis comparing high- and low-risk groups (p = 4.49e-07).\u003cbr\u003e\n(F-G) Risk score distribution, survival status, and scatterplot of patients based on risk score.\u003c/p\u003e\n\u003cp\u003e(H) Heatmap showing the expression levels of selected genes in high- and low-risk groups.\u003c/p\u003e\n\u003cp\u003e(I) Circos plot visualizing the chromosomal locations of genes in the FMRG signature.\u003c/p\u003e","description":"","filename":"fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/365686a3bde74d879e55f7ed.jpg"},{"id":74960482,"identity":"f08752e6-dfed-478d-ab92-238dba2d9fb7","added_by":"auto","created_at":"2025-01-28 18:51:54","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":784410,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation between Clinical Characteristics and the FMRGs-related Predictive Signature\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Calibration curves for the nomogram showing good agreement between predicted and observed 2-, 4-, and 6-year OS.\u003c/p\u003e\n\u003cp\u003e(B) Forest plots from univariate Cox regression analyses demonstrating that the risk score is an independent predictor of OS.\u003c/p\u003e\n\u003cp\u003e(C-G) Boxplots comparing the distribution of risk scores across stages (I-IV), T stage (T1-T4), N stage (N0-N2) and M stage(M0-N1) between low- and high-risk groups.\u003c/p\u003e\n\u003cp\u003e(H-J) Kaplan-Meier survival curves comparing OS in different groups stratified by stage and age. Patients in the high-risk group exhibit worse survival in all comparisons.\u003c/p\u003e\n\u003cp\u003e(K) Oncoprints showing the distribution of key mutations in high-risk.\u003c/p\u003e","description":"","filename":"fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/fe9118e1d854d73d24e701e4.jpg"},{"id":74960483,"identity":"02158d0a-8f7d-4cd9-adf2-0061506098d9","added_by":"auto","created_at":"2025-01-28 18:51:54","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1729142,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLandscape of biological characteristics of FMRGs-related signature\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Bar plot showing enriched Gene Ontology (GO) terms in the high- and low-risk groups.\u003c/p\u003e\n\u003cp\u003e(B) Bar plot showing enriched KEGG terms in the high- and low-risk groups.\u003cbr\u003e\n(C-D) Gene Set Enrichment Analysis (GSEA) plots for key biological processes.\u003cbr\u003e\n(E) Heatmap displaying expression of genes involved in the significantly enriched pathways across the two groups.\u003cbr\u003e\n(F) Bar plot showing the top enriched pathways ranked by GSVA score, highlighting metabolic and signaling pathways.\u003c/p\u003e","description":"","filename":"fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/3f3f4d7a3a32f665192e2015.jpg"},{"id":74960488,"identity":"56b5847f-b412-4a80-97b3-1e7c306621b2","added_by":"auto","created_at":"2025-01-28 18:51:54","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1319942,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSingle-cell Analysis and Cell-Cell Communication Networks in Tumor Microenvironment\u003cbr\u003e\n\u003c/strong\u003e(A) t-SNE plot visualizing distinct cell populations within the tumor microenvironment, identifying eight main cell types, including epithelial cells, T cells, macrophages, and fibroblasts.\u003c/p\u003e\n\u003cp\u003e(B) Histogram of the proportions of different subpopulations of cells.\u003c/p\u003e\n\u003cp\u003e(C) Score distribution of FDRG expression in various cells, showing clear separation of high and low scoring groups.\u003c/p\u003e\n\u003cp\u003e(D) Violin plot illustrating the distribution of FDRG scores across different cell types.\u003c/p\u003e\n\u003cp\u003e(E) Cell-cell communication network analysis, highlighting interaction patterns between different cell types.\u003c/p\u003e\n\u003cp\u003e(F) Heatmap showing incoming signaling pathways among cell types, with TGF, WNT, and EGF signaling predominantly activated in certain cell groups.\u003c/p\u003e","description":"","filename":"fig5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/84dcd89e13b36026fa772d11.jpg"},{"id":74960503,"identity":"2a138e0e-b48f-4fff-8239-af0f51eb49fb","added_by":"auto","created_at":"2025-01-28 18:51:55","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":753216,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe FMRG_score reshape the immune cell infiltration landscape\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Abundance of immune cell types in different FMRG Groups.\u003c/p\u003e\n\u003cp\u003e(B) Differences in the immune, stromal, and ESTIMATE scores between the two FMRG Groups.\u003c/p\u003e\n\u003cp\u003e(C-G) Boxplot showing a significant difference in risk scores between responders and non-responders.\u003c/p\u003e","description":"","filename":"fig6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/9099e835df4387da78426dcb.jpg"},{"id":74960500,"identity":"93c1b8a6-3829-48b5-b3f9-ce7e0a5dd07d","added_by":"auto","created_at":"2025-01-28 18:51:55","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":348026,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChemotherapy and Immunotherapy Response Prediction in High- and Low-Risk Groups\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-I) Boxplots showing the estimated IC50 values for multiple chemotherapeutic drugs across low- and high-risk groups. High-risk groups display significantly higher IC50 values for drugs like Camptothecin (C), Cisplatin (D), Cytarabine (E), Etoposide (F), Gemcitabine (G), Methotrexate (H) and Sorafenib (I), suggesting lower drug sensitivity in the high-risk group.\u003c/p\u003e","description":"","filename":"fig7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/143d2349384b591560f6e997.jpg"},{"id":74960510,"identity":"24fb5895-527a-449e-be08-cc56ee5da6d0","added_by":"auto","created_at":"2025-01-28 18:51:55","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":840762,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of key model gene-CYP26A1 and analysis of its association with the immune microenvironment at the single-cell level\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Boxplot comparing the expression levels of key genes in normal vs. tumor tissues.\u003c/p\u003e\n\u003cp\u003e(B) Protein-protein interaction networks functional enrichment analysis.\u003c/p\u003e\n\u003cp\u003e(C) Dot plot displaying CYP26A1 expression in normal versus tumor tissues from GSE161277, highlighting elevated CYP26A1 expression in tumor tissues.\u003c/p\u003e\n\u003cp\u003e(D) Boxplot of CYP26A1 expression in primary and metastatic foci.\u003c/p\u003e\n\u003cp\u003e(E) Kaplan-Meier survival analysis comparing high- and low-CYP26A1 groups in stage III-IV patients (p = 0.035).\u003c/p\u003e\n\u003cp\u003e(F) Gene Set Enrichment Analysis (GSEA) plots demonstrate enrichment of key pathways in CYP26A1 high and low groups.\u003c/p\u003e","description":"","filename":"fig8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/dabfe41e95b2c5d629e59b00.jpg"},{"id":74960490,"identity":"4786b770-c2a4-41f9-9f02-7a7d4c83602d","added_by":"auto","created_at":"2025-01-28 18:51:54","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":816827,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCYP26A1 is associatied with the immune microenvironment at the single-cell level\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-C) UMAP visualization of immune cell clustering based on CYP26A1 expression. Clusters of immune cells, including CD8+ T cells, macrophages, and fibroblasts, are colored to reflect high and low CYP26A1 expression, indicating distinct immune cell subpopulations based on CYP26A1 levels.\u003c/p\u003e\n\u003cp\u003e(D) Box plot shows T cell ratio according to different CYP26A1 expression.\u003c/p\u003e\n\u003cp\u003e(E) UMAP embedding of CD8+ T cell subtypes stratified.\u003c/p\u003e\n\u003cp\u003e(F-G) Proportional bar plots and Re-clustering analysis of T cell subpopulations based on CYP26A1 levels reveals functional differences, including a shift in activation status, with CYP26A1_high groups exhibiting higher CD8+ effector T cells.\u003c/p\u003e\n\u003cp\u003e(H). Dot plot of CYP26A1 expression in CRC tumors from individuals treated with the neoadjuvant PD-1 blockade. Mean expression is shown as color and is standard scaled (binarized), whereas dot size represents the fraction of samples with expression (pCR n = 15; non-pCR n = 4).\u003c/p\u003e","description":"","filename":"fig9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/49bb193519cdaea34c65e18f.jpg"},{"id":74961540,"identity":"063235dc-658e-40ff-b7f7-a2e14189522b","added_by":"auto","created_at":"2025-01-28 18:59:55","extension":"jpg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":2577232,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCYP26A1 promotes cell proliferation, migration, and invasion in colon cancer cell lines.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Immunohistochemical staining demonstrates significantly higher CYP26A1 expression in colorectal cancer (CRC) tissues compared to normal tissues. The bar graph on the right shows the percentage of positive areas for CYP26A1 in normal and tumor tissues. (B) Survival analysis reveals that the high CYP26A1 expression group has a significantly higher recurrence rate compared to the low expression group. The bar chart on the right shows the proportion of recurrence in each group.(C) qPCR analysis from the FUSCC database shows a significant increase in CYP26A1 mRNA levels in colon tumor tissues compared to normal tissue. (D) Western blot and qPCR validation of stable CYP26A1 knockdown in HCT116 and RKO cells. Both shCYP26A1-1 and shCYP26A1-2 groups showed significantly reduced CYP26A1 expression compared to the shNC group. (E) CCK-8 assays indicate that knockdown of CYP26A1 suppressed cell proliferation in HCT116 and RKO cells. (F) Transwell migration and invasion assays show that shCYP26A1-infected HCT116 and RKO cells exhibit reduced migratory and invasive capabilities compared to shNC-infected cell. (G) Co-culture assays of immune cells show increased secretion of granzyme B, IFN-γ, and TNF-α by CD8+ T cells upon CYP26A1 knockdown compared to the control (shNC) group. Flow cytometry analysis shows a significant increase in granzyme B, IFN-γ, and TNF-α-positive CD8+ T cells in the shCYP26A1 group. Data are shown as mean ± SD. Statistical significance: p \u0026lt; 0.05 (*), p \u0026lt; 0.001 (**), p \u0026lt; 0.001 (***), p \u0026lt; 0.0001 (****).\u003c/p\u003e","description":"","filename":"fig10.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/e1cf034c102a3e52410d2aca.jpg"},{"id":85917876,"identity":"0a5be746-9f6b-4279-bfd6-2e245f01057f","added_by":"auto","created_at":"2025-07-03 07:14:05","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":13305470,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/3e9b5d75-8d06-4538-82a2-b8a4e8cfeefe.pdf"},{"id":74960480,"identity":"c64b76cc-a1fc-490c-842b-2624b9891302","added_by":"auto","created_at":"2025-01-28 18:51:53","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":54929,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigureandTableslegends.docx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/d56186f7ff31f7fd7d9387ea.docx"},{"id":74960485,"identity":"5d4e9b89-dbef-4527-a1a0-fd8b9f0dbd02","added_by":"auto","created_at":"2025-01-28 18:51:54","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":40695,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/fa31d4f579f51791f8c21df5.xlsx"},{"id":74960505,"identity":"5d6770e5-c60d-4997-bfda-3303233fe4b1","added_by":"auto","created_at":"2025-01-28 18:51:55","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":1718073,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/72f71174d6694a9024a19518.xlsx"},{"id":74960513,"identity":"7246bd26-3138-4740-9a50-515f651a0e22","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":22220,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/e7f0a12c85a71f1dfeeead6e.xlsx"},{"id":74960486,"identity":"ad66f233-84da-4578-8e36-5a4c7e4715a7","added_by":"auto","created_at":"2025-01-28 18:51:54","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":10948,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/69731e25a76b97c9458cc9a0.xlsx"},{"id":74960543,"identity":"e75b5176-828d-4ed3-b426-42793d5a4f5b","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":10776,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/678c4ed652df737972938e17.xlsx"},{"id":74960489,"identity":"3e80bb6f-e54b-4b47-bb14-3075c1e219f1","added_by":"auto","created_at":"2025-01-28 18:51:54","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":10982,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/a6880f40f4a0d15f9787b64f.xlsx"},{"id":74960540,"identity":"d3a9873d-63a9-42de-aeca-aaa2a2fd91a5","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":43159,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS7.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/4e8acb459a29ccbe1f6eb502.xlsx"},{"id":74960542,"identity":"db894b7d-2368-4c83-9347-d45c3dba89e0","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"xlsx","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":13827,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS8.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/b354eb5071af635fd169466d.xlsx"},{"id":74961538,"identity":"5c88e3eb-1a98-44c8-8cce-e13577589504","added_by":"auto","created_at":"2025-01-28 18:59:55","extension":"xlsx","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":22160,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS9.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/7280f0b06a6aff71ef04ff38.xlsx"},{"id":74960547,"identity":"46d0f9e0-485d-4b11-b686-a3c07fe27585","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"xlsx","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":1562021,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS10.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/0191f7126c6b0342f1ec122c.xlsx"},{"id":74960545,"identity":"8cb067fc-367c-4c10-806c-0e14245d9f30","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"xlsx","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":10317,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS11.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/a610d9ed066f0515f275eb98.xlsx"},{"id":74960496,"identity":"2756b69a-9e2b-4fb7-aebb-e71cb098e67f","added_by":"auto","created_at":"2025-01-28 18:51:55","extension":"xlsx","order_by":13,"title":"","display":"","copyAsset":false,"role":"supplement","size":15004,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS12.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/6cb3fae42e1a905de7a614f9.xlsx"},{"id":74961541,"identity":"e53e03f0-e68c-4252-999d-2fbd7290a6fd","added_by":"auto","created_at":"2025-01-28 18:59:55","extension":"xlsx","order_by":14,"title":"","display":"","copyAsset":false,"role":"supplement","size":17855,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS13.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/a2568344703c9b15b68a94b8.xlsx"},{"id":74960521,"identity":"f1f3d356-a1dd-435f-8a73-599f8e060e2a","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"xlsx","order_by":15,"title":"","display":"","copyAsset":false,"role":"supplement","size":8745,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS14.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/19687321bd0cab13108a6bac.xlsx"},{"id":74961539,"identity":"72e0910f-9909-4594-9af9-747cdb6bdb69","added_by":"auto","created_at":"2025-01-28 18:59:55","extension":"xlsx","order_by":16,"title":"","display":"","copyAsset":false,"role":"supplement","size":8676,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS15.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/a9805f7513f0ffb95798294a.xlsx"},{"id":74960553,"identity":"a47f0726-cb6c-4390-8fd1-7be4c9dfbedc","added_by":"auto","created_at":"2025-01-28 18:51:57","extension":"jpg","order_by":17,"title":"","display":"","copyAsset":false,"role":"supplement","size":491694,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfigS1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/258834a1b6a3082fec042097.jpg"},{"id":74961533,"identity":"41ecd469-1e4f-4c7d-a238-8ea2fdab23f1","added_by":"auto","created_at":"2025-01-28 18:59:54","extension":"jpg","order_by":18,"title":"","display":"","copyAsset":false,"role":"supplement","size":422386,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfigS2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/98e7bae2fd931741834fea17.jpg"},{"id":74960549,"identity":"e9a984e6-ae4f-4b8c-bdc2-fd175aba06e3","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"jpg","order_by":19,"title":"","display":"","copyAsset":false,"role":"supplement","size":292124,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfigS3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/100545b7fde44c34e6c0dae3.jpg"},{"id":74960512,"identity":"0b9e68fa-2f54-4775-9ba6-36ba3c5a916b","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"jpg","order_by":20,"title":"","display":"","copyAsset":false,"role":"supplement","size":629658,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfigS4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/efe06f8a853960275bc19133.jpg"},{"id":74960544,"identity":"1c083401-74ee-49fc-9b38-d475f95e913b","added_by":"auto","created_at":"2025-01-28 18:51:56","extension":"jpg","order_by":21,"title":"","display":"","copyAsset":false,"role":"supplement","size":1063483,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfigS5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5647525/v1/30dc109203dbcb68cfe34d9a.jpg"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential conflict of interest.\nThe authors declare no conflict of interest.","formattedTitle":"Folate Metabolism in Colorectal Cancer Reveals Links Between Clinical and Immune Traits, Identifying CYP26A1 as a Target","fulltext":[{"header":"Introduction","content":"\u003cp\u003eColorectal cancer (CRC) is the third most common malignancy. It is also the second deadliest cancer\u0026nbsp;worldwide\u003csup\u003e1\u003c/sup\u003e. \u0026nbsp;CRC is 1.5 times more common in men than in women, with most cases occurring after the age of 50\u003csup\u003e2\u003c/sup\u003e. However, CRC incidence has been increasing in countries with a rising human development index, especially in those under age 50\u003csup\u003e3\u003c/sup\u003e. By 2035, the number of deaths from colon cancer and rectal cancer is projected to rise by 60.0% and 71.5%, respectively, due to population growth and aging\u003csup\u003e4\u003c/sup\u003e. Research definitively shows that early diagnosis and precise treatment of CRC significantly improve patient survival rates\u003csup\u003e5, 6\u003c/sup\u003e. Recent advances in molecular biology have facilitated the delineation of critical genetic pathways that play a role in colorectal carcinogenesis\u003csup\u003e7, 8\u003c/sup\u003e. These pathways include the adenomatous polyposis coli (APC) pathway\u003csup\u003e9\u003c/sup\u003e, the microsatellite instability (MSI) pathway\u003csup\u003e10\u003c/sup\u003e, and the CpG island methylator phenotype (CIMP) pathway\u003csup\u003e11\u003c/sup\u003e. Understanding the intricate mechanisms of these pathways can provide valuable insights for developing targeted therapeutic strategies to combat colorectal cancer.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn recent years, the rapid advancement of immune checkpoint inhibitor (ICI)-based immunotherapy has brought about a promising new era in anticancer treatment. ICIs enhance the antitumor immune response by disrupting the signaling of key immunosuppressive proteins like programmed cell death 1 (PD-1), programmed cell death-1 ligand 1 (PD-L1), and cytotoxic T lymphocyte antigen 4 (CTLA-4)\u003csup\u003e12-14\u003c/sup\u003e. Some patients with melanoma, non-small-cell lung cancer (NSCLC), and other cancers show sustained responses\u003csup\u003e15-18\u003c/sup\u003e.\u0026nbsp;Immunotherapy using ICI in CRC patients has shown promising safety and efficacy outcomes. Approved immunotherapeutic agents for advanced CRC treatment include pembrolizumab, nivolumab, and ipilimumab\u003csup\u003e19\u003c/sup\u003e. In 2015, Le DT et al. discovered that the anti-PD-1 drug pembrolizumab exhibited a notably higher response rate in patients with the DNA mismatch repair\u0026ndash;deficient (dMMR)/microsatellite instability-high (MSI-H) molecular subtype of CRC\u003csup\u003e20\u003c/sup\u003e. This finding implies the potential benefits of anti-PD-1 therapy for patients with both CRC and non-CRC dMMR tumors. ICIs are effective in treating dMMR/microsatellite-stable (MSS) CRC due to the elevated mutational load, abundance of neoantigens, and increased immune cell presence within this tumor subtype\u003csup\u003e21, 22\u003c/sup\u003e. Significantly, MSS-pMMR patients make up a substantial portion of CRC patients\u003csup\u003e23\u003c/sup\u003e. Hence, it is essential to actively explore new biomarkers and innovative approaches for early detection and treatment to understand the underlying mechanisms of varied treatment responses in current colorectal cancer research.\u003c/p\u003e\n\u003cp\u003eThe relationship between folate metabolism and CRC has been extensively studied due to folate\u0026apos;s crucial role in DNA methylation, repair, and synthesis processes\u003csup\u003e24-28\u003c/sup\u003e. Folate, a- vitamin B naturally present in foods, and its synthetic form folic acid, commonly utilized in supplements and food fortification, play a vital role in cell production and upkeep, especially during stages of accelerated cell division and growth, such as infancy and pregnancy\u003csup\u003e29, 30\u003c/sup\u003e. Based on these pivotal functions, it has been hypothesized by researchers that sufficient folate intake has the potential to mitigate DNA alterations that might serve as precursors to cancer development, particularly in the colon and rectum\u003csup\u003e31\u003c/sup\u003e. Recent studies and reviews suggest a complex relationship between folate intake, folate status, and CRC risk. A systematic review and meta-analysis examining the impact of folic acid supplement intake and total folate intake on CRC risk found no significant effect of folic acid supplements in randomized controlled trials. Despite these findings, the effect of folate status, as measured by red blood cell folate content, on CRC risk was not significant, indicating that the relationship may depend on the form of folate consumed and other individual factors\u003csup\u003e32\u003c/sup\u003e. A study found that high folic acid intake accelerates methionine cycling in cancerous tissues in vivo, potentially contributing to the development of hepatocellular carcinoma\u003csup\u003e33\u003c/sup\u003e. Recent evidence suggests that the relationship between high plasma levels of folate and CRC risk may not be straightforward\u003csup\u003e34-36\u003c/sup\u003e. For example, one study observed a significant decrease in CRC risk among women in the highest quartile of plasma folate levels\u003csup\u003e37\u003c/sup\u003e. In contrast, another study identified an increased risk of CRC associated with high plasma folate levels over a follow-up period\u003csup\u003e38\u003c/sup\u003e. These inconsistencies underscore the challenges inherent in accurately assessing the impact of folate on CRC risk. The relevance of genetic polymorphisms in folate metabolism genes, particularly in genes such as MTHFR that code for enzymes in folate metabolism, should be noted\u003csup\u003e39\u003c/sup\u003e. However, the impact of folate metabolism on TME characteristics and clinical outcomes of CRC patients is still uncertain.\u003c/p\u003e\n\u003cp\u003eWe systematically analyzed the underlying effects of folate metabolism in CRC in this study. Through our analysis, we constructed a folate metabolism-related signature to predict the prognostic outcomes, TME characteristics, and immunotherapy response in CRC patients. Our findings not only revealed the paramount role of folate metabolism in the complex heterogeneity of CRC but also indicated its potential to enhance individualized management strategies for CRC patients. This study also marked the first instance of proposing the potential anticancer function of Cytochrome P450 Family 26 Subfamily A Member 1 (CYP26A1) in CRC, laying the groundwork for further exploration of this promising molecular target in CRC.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003e\u003cstrong\u003eData\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ecollection\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe training cohort included the transcriptome data and clinical information of 585 patients with colorectal adenocarcinoma (COAD and READ) were retrieved from the Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/repository). Validation sets were obtained from GSE39582 (579 patients), GSE17536 (177 patients), GSE38832 (122 patients) GSE17537 (55 patients) expression profiles from the Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo). Single-cell RNA sequencing (scRNA-seq) data for colorectal adenocarcinomas were obtained from the GSE161277 and GSE205506 databases, which include 13 and 19 CRC samples, respectively. 410 folate metabolism-related genes (FMRGs) were retrieved from GeneCards (https://www.genecards.org/). The complete gene details are displayed in Table S1. The RNA-seq sample expression levels were converted from fragments per kilobase of transcript per million mapped reads (FPKM) to transcripts per million (TPM), and then log2(TPM + 1) was calculated. Somatic mutation data, copy number variation (CNV) files, and tumor mutation burden (TMB) data of CRC patients were retrieved from the TCGA database.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConstruction and\u0026nbsp;validation of\u0026nbsp;FMRGs risk signature\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFirst, we analyzed the genes related to folate metabolism and the differential genes from TCGA, leading to the identification of 138 genes. Subsequently, we conducted univariate analysis on these 138 genes to pinpoint those significantly correlated with patient survival (p\u0026lt;0.05). Following this, a combination of the least absolute shrinkage and selection operator (LASSO) technique with multivariate regression analysis was employed to refine the gene selection process and ascertain risk coefficients strongly linked to prognosis. In order to identify high-impact genes, tenfold cross-validation was used to determine the optimal regularization parameter. Genes with non-zero coefficients were then deemed potential prognostic markers. Candidate genes were chosen through multivariate Cox analysis to create a prognostic FMRG_score in the training dataset. The FMRG_score was calculated as follows:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cimg src=\"https://myfiles.space/user_files/122228_c8a1650c59388082/122228_custom_files/img1738087807.png\"\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003c/strong\u003eIn the training set, 585 patients were stratified into low-risk and high-risk groups based on the median risk score, which was calculated using the risk coefficient (Coefi) and expression of each gene (Expi). Subsequently, Kaplan-Meier survival analysis was conducted on the two groups to assess their survival outcomes. The predictive performance of the signature was evaluated using receiver operating characteristic (ROC) curves. Kaplan\u0026ndash;Meier survival curves were plotted, and log-rank tests were performed to assess the statistical significance of the observed differences in survival between the two risk groups. The effectiveness of the prediction model was further validated in three independent GEO datasets (GSE38832, GSE39582, GSE17537) through survival analysis and calculation of the area under the curve (AUC) in ROC analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRelationship between risk groupings and clinical characteristics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study analyzed clinical factors such as age, gender, and TNM stage. Differences in prognostic outcomes were assessed using Kaplan-Meier analysis in R software with the \u0026ldquo;survival\u0026rdquo; and \u0026ldquo;survminer\u0026rdquo; packages\u003csup\u003e40\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDevelopment and validation of a nomogram scoring system\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine the independent prognostic values and the predictive efficacy of the FMRG_score in predicting the survival of CRC patients, we conducted univariable and multivariate Cox regression analyses for better clinical practice. Furthermore, we investigated the relationship between the FMRG_score and various clinical characteristics. Additionally, in order to improve the prognostic accuracy of our model, a nomogram was developed utilizing the risk score, T stage, and N stage as independent prognostic factors to estimate the probability of overall survival (OS) at 2-, 4-, and 6-years. Subsequently, the predictive efficacy was compared between the FMRG_score and different clinical pathological factors based on AUC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunctional and pathway enrichment analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify specific biological pathways enriched between the high- and low-risk groups, we performed a comprehensive analysis using various techniques. Initially, we conducted Gene Ontology (GO) analysis (Supplementary Table 8) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis (Supplementary Table 7) to understand the functions of the screened candidate genes and related pathways. Following this, we utilized the \u0026quot;ClustProfiler\u0026quot; and \u0026quot;GSVA\u0026quot; packages for further elucidation\u003csup\u003e41, 42\u003c/sup\u003e. Additionally, we employed Gene Set Variation Analysis (GSVA) (Supplementary Table 10) to identify variations in gene sets. Subsequently, Gene Set Enrichment Analysis (GSEA) was utilized to identify enriched pathways or gene sets based on differential expression results between the high-risk and low-risk groups (Supplementary Table 9). Moreover, GSVA (Supplementary Table 11) and correlation analysis between Hallmark pathway activities and the FMRG risk score were conducted to explore potential pathways associated with the identified signature.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eExploration of the immune landscape in distinct risk groupings\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe immune characteristics of 585 CRC samples were assessed by evaluating the scores of tumor microenvironment (TME) cells using the single sample gene set enrichment analysis (ssGSEA) algorithm\u003csup\u003e43\u003c/sup\u003e. To analyze the relative proportion of 22 immune cells within the CRC samples, CIBERSORT (https://cibersort.stanford.edu) was employed. The CIBERSORT algorithm was executed using R software. By leveraging the 585 samples gene expression matrix and the provided gene expression feature set of the 22 immune cell subtypes from the official website, simulation calculations were iterated 1000 times to derive the relative composition ratio of the 22 immune cells in each sample. Subsequently, the immune score and ESTIMATE score of each patient were assessed utilizing the R package of estimate\u003csup\u003e44\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePrediction of response to immunotherapy\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInitially, the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm was utilized to evaluate the potential disparities in treatment responses between high- and low-risk groups\u003csup\u003e45\u003c/sup\u003e. A higher TIDE score was found to be associated with reduced treatment efficacy, thereby highlighting a negative correlation between the TIDE score and treatment effectiveness. Furthermore, data on immunotherapy was gathered from various datasets including the IMvigor210 dataset for urothelial carcinoma (UC), the TCGA dataset, GSE17536 dataset, GSE39582 dataset, and GSE38832 dataset for CRC. Subsequently, within each dataset, the FMRG_score was calculated to predict responses to immunotherapy.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMutation and drug susceptibility analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore the differences in therapeutic effects of chemotherapeutic drugs between high- and low-risk groups of CRC patients, we first generated the mutation annotation format (MAF) from the TCGA database using the \u0026quot;maftools\u0026quot; R package. Subsequently, we calculated the TMB score for each patient within these groups. Additionally, we determined the semi-inhibitory concentration (IC50) values of commonly used chemotherapeutic drugs for treating CRC by utilizing the \u0026quot;pRRophetic\u0026quot; package.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSingle-cell RNA sequencing analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe scRNA-seq analysis was conducted using the \u0026quot;Seurat\u0026quot; R package for data processing, including quality control steps where cells expressing 200-7000 genes were retained, and cells with more than 20% mitochondrial gene expression were excluded. The cells were classified into eight primary types using t-SNE dimensionality reduction. The \u0026quot;inferCNV\u0026quot; package was used to infer copy number variations (CNVs) from single-cell RNA sequencing data by comparing gene expression profiles to a reference dataset. This approach allowed identification of regions with CNV alterations. Based on these CNV patterns, the malignancy score was calculated to classify cells as either malignant or non-malignant, reflecting tumor characteristics. The \u0026quot;AUCell\u0026quot; algorithm was employed to assess gene set activity at the single-cell level. Cell-to-cell communication networks were analyzed using \u0026quot;CellChat,\u0026quot; focusing on receptor-ligand interactions. Additionally, the \u0026quot;Monocle 2\u0026quot; algorithm was used to construct pseudotime trajectories to track the dynamic functional changes in FDRG-related signatures.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell culture and reagents\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe human colon carcinoma cell line HCT116 and RKO were obtained from the Cell Bank of Chinese Academy of Sciences (Shanghai, China). All cell lines were authenticated by monitoring cell vitality, mycoplasma contamination, and short tandem repeat profiling.\u0026nbsp;Cells\u0026nbsp;were maintained in DMEM medium supplemented with 10% FBS and 1% penicillin/streptomycin.\u0026nbsp;Culture media and supplements were obtained from BasalMedia (Shanghai, China).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDNA constructs, transfection, and viral transduction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCYP26A1 cDNA in penter vector with C terminal Flag and His tag was purchased from Vigene Bioscience. Short hairpin RNAs (shRNAs) targeting human CYP26A1 (shCYP26A1) in GIPZ lentiviral vector and corresponding control constructs were purchased from Dharmacon.\u0026nbsp;The detailed information of expression vectors for molecular cloning is provided in Supplementary Table 14. shRNAs in lentiviral expression vectors were transfected into HEK293T together with packaging plasmid mix using Neofect DNA transfection reagent. Supernatants were collected after 48 h of transfection and used for infecting cells in the presence of 8 mg/mL of polybrene. After 24 h of infection, cells were selected with 2 mg/mL of puromycin (Cayman, Ann Arbor, USA) for 1-2 week.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell viability and colony formation assays\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCell proliferation assays were performed using Cell Counting Kit-8 (CCK-8) (Dojindo, Shanghai, China) according to the manufacturer\u0026rsquo;s instructions.\u0026nbsp;The absorbance was measured at a wavelength of 450nm (A450). For colony-formation assays, cells were grown onto 6-well plate at a density of 2000 cells/well for 14 days with media replacement every 3 days.\u0026nbsp;Cells were stained with 1% crystal violet and the number of survival colonies was counted.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWestern blotting\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eProteins from cell lines and tissues were extracted using RIPA lysis buffer (Thermo Fisher Scientific, USA) on ice. After centrifugation (15000g, 10 min), protein concentration was determined with a BCA protein assay kit (Thermo Fisher Scientific, USA). Protein samples were separated on PAGE gels (Epizyme Biomedical Technology, Shanghai, China) and transferred to 0.22 \u0026micro;m Immobilon PVDF membranes (Millipore Sigma, USA). Membranes were blocked with 5% milk and incubated with primary antibodies overnight at 4\u0026deg;C. Secondary antibodies (anti-Rabbit IgG) were applied at room temperature for 1 h, and immunoreactivity was detected using an ECL system (Share-bio, Shanghai, China). Primary antibodies used: pan-kla (#PTM-1401RM, PTMBio, Hangzhou, China) at 1:1000. Secondary antibody dilution was 1:5000.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTranswell migration assays and Matrigel invasion assays\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCell migration and invasion were assessed using transwell chambers (Corning Biocoat, Tewksbury, USA). For invasion assays, the Matrigel (BD Biosciences) was diluted 1:8 (dilution ratio determined based on the MMP expression levels of the cells) and applied to coat the upper surface of the chamber membrane. The coated chambers were incubated at 37\u0026deg;C for 30 minutes to allow the Matrigel to polymerize and hydrated prior to use. For both assays, cells were resuspended in DMEM containing 1% FBS and loaded at a density of 5 \u0026times;10^4 cells per well onto the upper well of chambers, while growth medium containing 10% FBS was placed in the lower chamber as a chemoattractant. After 24h incubation, cells on the upper surface were gently removed with a cotton swab, and cells that had migrated or invaded to the lower surface were fixed with methanol and stained with 1% crystal violet solution. The number of cells was counted under a light microscope with a magnification of 100. All assays were conducted in triplicate and repeated at least three times.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRNA isolation and quantitative reverse transcription-PCR\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTotal RNA was isolated from cell lines and tissue samples using Trizol reagent (Invitrogen), Reverse transcription was performed using a PrimeScript RT reagent Kit (TaKara). The resultant cDNA was subjected to quantitative real-time PCR (qPCR) using SYBR Premix Ex Taq (Tli RNaseH Plus) (Takara) on an Eppendorf Mastercycler ep realplex4 instrument. Primer sequences for qPCR are listed in Supplementary Table 15. All reactions were performed in triplicate. The data is present as mean \u0026plusmn; SD.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIn vitro T cell activation and tumor cells-T cells coculture model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePBMCs from healthy donors were treated with RBC lysis buffer and sorted using CD8 A antibody by flow cytometry. After activation with anti-CD3 (BioLegend, 5 \u0026mu;g/ml) and anti-CD28 (BioLegend, 2.5 \u0026mu;g/ml) at the indicated concentrations and cultured in media containing IL-2 (PeproTech, 10 ng/ml), the cells were used in co-culture models. T cells (2 x 10^5 cells/well) were seeded in the top chamber of a transwell (pore size: 0.4 \u0026micro;m), while tumor cells, with or without CYP26A1 knockdown, were seeded in the bottom chamber for indirect co-culture. In the direct co-culture model, T cells and tumor cells were mixed directly at a 1:4 ratio without using a chamber. Following co-culture, T cells from both models were subjected to surface marker and intracellular cytokine staining and analyzed by FACS.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll experiments were replicated at least three times and the results are presented as mean \u0026plusmn; standard deviation (SD). For comparisons between two groups, unpaired Student\u0026apos;s t-tests were used for normally distributed variables, while Mann-Whitney U tests were employed for non-normally distributed variables. For multiple group comparisons, one-way ANOVA (parametric) or Kruskal-Wallis tests (nonparametric) were performed as appropriate. Pearson\u0026apos;s or Spearman\u0026apos;s correlation analyses were conducted to evaluate linear relationships between variables. Survival curves were generated using the Kaplan-Meier method, and differences between groups were assessed using the log-rank test. ROC curves were generated and area under the AUC values were calculated to evaluate predictive performance. Statistical analyses were performed using SPSS version 23.0, R software (version 4.3.2), and GraphPad Prism 9. P-values \u0026lt; 0.05 were considered statistically significant.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eConstruction and validation of the prognostic FMRG_score\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe flowchart for this article was shown in Figure 1. Oue To construct the FMRG_score, we identified 138 target genes through differential analysis of CRC data from TCGA, intersecting with genes associated with folate metabolism (Fig. 2A, Supplementary Table 1-2). We chose to conduct additional analysis on the TCGA cohort to extract a prognostic signature for FMRG. Initially, a univariable Cox regression analysis revealed 19 candidate genes that were found to be associated with CRC, with a p-value \u0026lt; 0.05 (Supplementary Table 3-4). To delve deeper into the characteristics of folate metabolism in CRC patients, we utilized LASSO Cox regression to determine the optimal λ value, leading to the identification of 9 key genes (Supplementary Table 5-6). The mRNA expressions of the 9 key genes in tumor and normal specimens were assessed using the TCGA-CRC dataset. Notably, AHCY, PLK1, WNT5A, CYP26A1, BDNF and DRD4 exhibited upregulation, while NAT1, CD36 and GSTM1displayed downregulation in tumor (Fig. 2D). Five of the risk genes (NAT1, WNT5A, CYP26A1, BDNF and DRD4) were identified as independent predictive factors (Table 1). The locations of these 9 risk genes on chromosomes were visualized in Figure 2I. We displayed the relationship between these nine risk genes and patient prognosis through the TCGA CRC dataset (Fig. S1). This comprehensive model integrates the predictive power of individual genes to provide a more precise tool for assessing CRC prognosis. The risk-score model was determined using the following equation:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFMRG_score = (-0.168)*AHCY + (-0.157)*PLK1\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e+ (-0.26)*NAT1 + (0.136)*CD36\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e+ (-0.194)*WNT5A + (0.181)*CYP26A1\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e+ (0.156)*BDNF + (0.235)*DRD4\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e+ (0.155)*GSTM1\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePatients were divided into two groups based on their FMRG_score: those with a score lower than the median were categorized as low-risk (n = 292), while those with a score higher than the median were classified as high-risk (n = 293). The distribution plot showed that as FMRG_score increased, survival times decreased and recurrence rates increased (Fig. 2F-G). The Kaplan–Meier survival curves indicated that patients with low scores had a significantly better overall survival compared to those with high scores (log-rank test, p \u0026lt; 0.001; Fig. 2E). Furthermore, the 2-, 4-, and 6-year survival rates predicted by the FMRG_score were reflected in the AUC values of 0.7026, 0.7027, and 0.6927, respectively (Fig. 2C).\u0026nbsp;Clustering heatmaps revealed that CYP26A1, DRD4, GSTM1, CD36, and BDNF were more prominent in the high-risk group, while AHCY, PLK1, NAT1, and WNT5A were more prevalent in the low-risk group (Fig. 2H). To validate the prognostic performance of the FMRG_score, we calculated FMRG_scores across three external validation groups (GSE38832, GSE39582, GSE17537) (Fig. S2). Patients were also stratified into low- or high-risk groups based on the formula utilized for the training set. Survival analysis was conducted, revealing a significantly better prognosis in the low-risk group compared to the high-risk group (log rank; p \u0026lt; 0.05). Assessment of the 2-, 4-, and 6-year prognostic prediction classification efficiencies demonstrated that the FMRG_score maintained relatively high AUC values. These results indicate that the FMRG_score exhibits excellent ability to predict the survival outcomes of CRC patients.We combined clinical characteristics with the FMRG_score to create a nomogram predicting the 2-, 4-, and 6-year OS of CRC patients based on the strong relationship between the FMRG_score system and patient prognosis (Figure S3A). The calibration plot showed excellent agreement between the predicted overall OS using the nomogram and the actual OS of CRC patients at 2, 4, and 6 years (Fig. 3A).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorrelation between Clinical Characteristics and the FMRGs-related Predictive Signature\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUnivariable and multivariate Cox regression was used to explore the relationship between the FMRG_score and clinical characteristics (Fig. 3B and Fig. S3B), and the results showed that FMRG_score and M stage could be used as independent prognostic factors (HR = 2.177, 95% CI = 1.628, 2.912, P \u0026lt; 0.001; HR = 2.455, 95% CI = 1.508, 3.998, P \u0026lt; 0.001, respectively). Kaplan-Meier analyses were conducted across subgroups stratified by distinct clinical factors to comprehensively validate the predictive reliability of the FMRG_score. Our findings revealed that patients with a low FMRG_score tended to have a lower TNM stage, whereas those with a higher FMRG_score were more likely to present with an advanced TNM stage (Fig. 3C). Specifically, the high-risk group had a smaller proportion of patients with stages I and II relative to the low-risk group, and a higher proportion of patients with stages III and IV. Our results showed that the FMRG_score increased with advancing stages and was significant in all components except for patients at stages I and II, where no significant difference was observed (Fig. 3D-G). This highlights the important role of the FMRG_score in predicting clinical stage. Stratified survival analyses showed that the prognostic outcomes for colorectal cancer patients varied significantly between the high and low FMRG-Groups, particularly in advanced TNM stages (Fig. 3H, Fig. S3C). At the same time, patients with high FMRG_score had a worse prognosis in both young and elderly patients (Fig. 3I-J), indicating the generalization of the FMRG_score. To gain insights into the biological processes linked to the poor survival in the high-risk group and explore the molecular implications of transcriptomic and genetic variances between high- and low-risk groups, we delved into the genomic heterogeneity of the FMRGs model within the TCGA cohort. Our examination of the mutation landscape of FMRGs aimed to uncover the specific genetic alterations associated with the high-risk group and further elucidate the underlying mechanisms contributing to the differential survival outcomes observed. Our study revealed that the high-risk group had more mutations in the genes \u003cem\u003eAPC\u003c/em\u003e, \u003cem\u003eTP53\u003c/em\u003e, \u003cem\u003eTTN\u003c/em\u003e, and \u003cem\u003eK-RAS\u003c/em\u003e, which may have enhanced their cancer-promoting characteristics (Fig. 3K).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLandscape of biological characteristics of FMRGs-related signature\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFunctional enrichment analyses were conducted to gain valuable insights into biological processes, molecular functions, and cellular components (Fig. 4). KEGG analysis showed that the high-risk group were considerably enriched in pro-tumor pathways like “Wnt signaling pathway” and “PPAR signaling pathway” (Fig. 4A). In recent years, the role of the tumor microenvironment in cancer development has garnered increasing attention. Newer studies have highlighted the emerging significance of the nervous system as a key factor in promoting tumor growth. Innervation plays a crucial role in the growth of various tumors, as neurons can establish tumor-nerve synapses with tumor cells. Through these signaling mechanisms, typical oncogenic signaling pathways are often activated, consequently fostering tumor growth\u003csup\u003e46\u003c/sup\u003e. GO analysis explained several neural-related sets that may be key to tumor progression (Fig. 4B). The results of the GSEA analysis highlighted that the high-risk group was significantly enriched in sets that promote tumor progression and metastasis such as “ECM RECEPTOR INTERACTION”, while the low-risk group was enriched in “CITRATE CYCLE TCA CYCLE” and “PYRUVATE METABOLISM” (Fig. 4C-D). The results of GSVA analysis also suggested that the high-risk group was enriched in cancer-promoting pathways, while the low-risk group was enriched in cancer-suppressor-related pathways (Fig. 4E-F)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003escRNA‑seq data processing and analysis of FMRGs-related signature\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe scRNA-seq data was analyzed with the \"Seurat\" R program, which included quality control measures such as screening for cells exhibiting low expression levels in the range of 200-7000 genes, and ensuring that mitochondrial genes accounted for no more than 20% of the total gene count. To define TME cell populations of CRC, we identified and visualized 8 main cell types using the T-distributed Stochastic Neighbor Embedding (tSNE) (Fig. 5A). We visualized the annotated cell clusters and found a significant decrease in T and B cells and a slight increase in plasma cells in the tumor compared to normal tissue (Fig. 5B). The \"AUCell\" R program, which evaluates gene set activity, was used to assign FMRGs-related signatures to individual cells (Fig. 5C). Our study revealed that epithelial cells and macrophages exhibited higher FMRG scores compared to other cell types. Specifically, epithelial cells showed a greater abundance of FMRG scores (Fig. 5D). As a result, we classified epithelial cells into two groups based on their FMRG scores - high and low - for further analysis (Fig. S4A). We analyzed the communication network between the nine cells, and results showed that T cells and the FMRG scores - high group had the most receptor-ligand communication with other cells (Fig. 5E). To explore the differential signaling pathways in the FMRG scores - high group and FMRG scores - low group, we further analyzed the signaling patterns of these nine cells (Fig. 5F).Our results suggest that epithelial cells with high expression of FMRG score are significantly enriched in some oncogenic pathways, such as \"TGF\", \"WNT\" and \"EGF\". To uncover dynamic functional changes of FMRGs-related signature, we adopted the Monocle 2 algorithm to chronologically order FMRGs-related signature in pseudotime and indicate their trajectories (Fig. S4B-C). We next investigated the transcriptional changes of FMRGs-related signature associated with trajectory and found that NAT1 was significantly enriched early in development, BDNF, DRD4, CD36 and WNT5A were progressively enriched as development progressed, and CYP26A1, AHCY and PLK1 were significantly enriched at the terminal end of development (Fig. S4D).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe FMRG_score reshape the immune cell infiltration landscape\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRecent research has highlighted the critical role of the inflammatory environment in the development of CRC, particularly focusing on the activation status and interactions of immune and stromal cells with tumor cells\u003csup\u003e47\u003c/sup\u003e. The varying composition of the TME impacts the effectiveness of immunotherapy. We examined the immune profiles of CRC patients across different FMRG groups to better understand this. We analyzed 22 different types of immune cells in CRC patients with high and low FMRG scores. Significantly heightened infiltrations were observed in the high FMRG group for central memory CD4 T cell, CD56dim natural killer cell, Macrophages, Myeloid-derived suppressor cells (MDSC), natural killer cell, plasmacytoid dendritic cell, and T follicular helper cell. Conversely, diminished infiltrations were noted for activated CD4 T cell and type 2 T helper cell within the same group (Fig. 6A). A correlative analysis between the abundance of immune cell types and FMRG scores was further depicted in Figure S5A. A high FMRG_score was closely associated with a high stromal score, as well as a high estimate score (Fig. 6B). We examined how the nine genes in the proposed model were related to the abundance of immune cells. Our findings showed significant correlations between most immune cells and seven of the genes (Fig. S5B). Using the TIDE score, we found a notable rise in the Dysfunction score and TIDE score in the high-risk group (Fig. S5C-E), indicating an increased risk of immune escape in high-risk patients, which may reduce the efficacy of ICI therapy. We also predicted immunotherapy efficacy using TIDE, which was validated using the IMvigor cohort, TCGA CRC group, GSE17536, GSE39582, and GSE38832, and we found that the higher the FMRG score, the more insensitive the patient may be to immunotherapy leading to ineffectiveness of immunotherapy (Fig. 6C-G). The sensitivity of patients in the low-risk and high-risk groups to chemotherapeutic agents currently used for treating colorectal cancer or undergoing clinical trials was evaluated (Fig. 7). The IC50 values of cisplatin and gemcitabine were notably lower in patients with low FMRG scores, potentially offering more precise clinical treatment guidance. These findings collectively indicate an association between FMRG and drug sensitivity.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCYP26A1 plays a carcinogenic role in CRC\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe validated the FMRGs-related signature by our own sequencing data (Fig. 8A). A protein-protein interaction (PPI) network among the FMRGs-related signature was constructed from the STRING database, in which CYP26A1 interacts most with other proteins (Fig. 8B). We performed a follow-up analysis of CYP26A1, which we verified in the GSE161277 dataset to be highly expressed in tumor tissues (Fig. 8C). The CYP26A1 gene encodes a member of the cytochrome P450 superfamily of enzymes crucial for various biological processes, such as drug metabolism, the synthesis of cholesterol, steroids, and other lipids\u003csup\u003e48\u003c/sup\u003e. In the context of cancer, CYP26A1 plays a significant role in the metabolism of retinoic acid (RA), a key metabolite of vitamin A that governs cell proliferation, differentiation, and apoptosis\u003csup\u003e49, 50\u003c/sup\u003e. Aberrant expression of CYP26A1 has been linked to the pathogenesis and progression of various tumors by affecting RA levels and modulating RA signaling pathways\u003csup\u003e51\u003c/sup\u003e. Studies have shown that increased CYP26A1 expression can diminish RA levels in cells, thereby promoting tumorigenesis by facilitating cell proliferation and maintaining a less differentiated cell state\u003csup\u003e52\u003c/sup\u003e. Conversely, the downregulation of CYP26A1 in specific cancers can elevate RA concentrations, potentially inhibiting tumor growth and inducing apoptosis\u003csup\u003e53\u003c/sup\u003e. Interestingly, in our sequencing data of tumor primary and metastatic foci from 8 pairs of patients, we found that CYP26A1 was highly expressed in metastatic foci. Additionally, analysis of TCGA CRC data indicated an association between elevated CYP26A1 expression and poor prognosis in stage III and IV patients (Fig. 8D-E). We performed GSEA analysis of CYP26A1 using our own data, and we found that some pro-oncogenic pathways were significantly enriched when CYP26A1 was highly expressed, whereas pathways related to immunity and autophagy were enriched when CYP26A1 was expressed, also suggesting that CYP26A1 functions as an oncogene (Fig. 8F). To further explore the relationship between CYP26A1 and the TME of CRC patients, single-cell RNA sequencing data were analyzed from 19 patients with d-MMR/MSI-H CRC treated with the neoadjuvant PD-1 blockade\u003csup\u003e54\u003c/sup\u003e (Fig. 9A). Significant reduction of malignant cells accompanied by an increase in CD8+ T cells and B cells in the CYP26A1 low-expression group (Fig. 9B, C) and CYP26A1 expression negatively correlated with the number of CD8+ T cells (Fig. 9D). Based on classical markers\u003csup\u003e55, 56\u003c/sup\u003e, CD8+ T cells were reclassified into five subpopulations: exhausted T cells (Tex), effector T cells (Teff), tissue resident memory T cells (Trm), naive T cells, and memory T cells (Fig. 9E). The proportion of Teffs was dramatically upregulated in CRCs with low CYP26A1 expression, while the relative ratio of Texs was lower (Fig. 9F, G). In addition, we observed that CYP26A1 expression was higher in non-pathological complete response (pCR) patients, suggesting that CYP26A1 were associated with immunotherapy non-response (Fig. 9H).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCYP26A1\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;is upregulated in colorectal tumor,\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003epromotes colon cancer cell proliferation,\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eenhances colorectal cancer cell migratory and invasive potential.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe investigate the expression of CYP26A1 in CRC and its impact on immune cell function, revealing the relationship between CYP26A1 and CRC prognosis. Immunohistochemical staining on tissue microarrays demonstrated a significant increase in the overall CYP26A1 expression in CRC tissues compared to normal tissues (Fig. 10A). Additionally, our findings show that higher expression of CYP26A1 are associated with poor prognosis and the high CYP26A1 expression group exhibited a higher recurrence rate (Fig. 10B). These results underscore the potential clinical significance of CYP26A1 as a prognostic marker for CRC, highlighting its relationship with disease progression and outcomes. Further exploration of the molecular mechanisms of CYP26A1 in CRC may provide new therapeutic avenues for this challenging malignancy.\u003c/p\u003e\n\u003cp\u003eTo further investigate the role of CYP26A1 in colorectal cancer, we first analyzed CYP26A1 mRNA levels in the Fudan University Cancer Center database via qPCR. The results showed that CYP26A1 mRNA levels were elevated in primary colon tumors compared to normal colon tissues (Fig. 10C). Furthermore, to explore the biological function of CYP26A1 in colon cancer, we stably depleted endogenous CYP26A1 in HCT116 and RKO cells (Fig. 10D) by infecting them with lentiviral vectors expressing shCYP26A1. The expression status of CYP26A1 in these stable cell lines was validated by Western blot (WB) and qPCR. CCK-8 assays showed that knockdown of CYP26A1 suppressed cell proliferation in HCT116 and RKO cells (Fig. 10E). Given the invasive and metastatic nature of colon cancer cells, we next investigated the effect of CYP26A1 on the invasive and metastatic phenotype of these cells. Transwell migration and invasion assays revealed that shCYP26A1-infected HCT116 and RKO cells displayed reduced migratory and invasive abilities compared to shNC-infected cells (Fig. 10F).\u003c/p\u003e\n\u003cp\u003eAdditionally, the study found that CYP26A1 plays a crucial role in regulating CD8+ T cell function. Knocking down CYP26A1 resulted in increased secretion of granzyme B, IFN-γ, and TNF-α by CD8+ T cells in co-culture conditions compared to the control group (shNC). Specifically, both shCYP26A1-1 and shCYP26A1-2 groups showed significantly higher expression of granzyme B and cytokines, with statistically significant differences compared to the control group (P \u0026lt; 0.01, P \u0026lt; 0.001) (Fig. 10G). These results suggest that CYP26A1 regulates immune effector molecules, potentially influencing the tumor immune microenvironment. Our findings provide new evidence for CYP26A1 as an immune therapy target, with potential clinical applications.\u003c/p\u003e\n\u003cp\u003eIn conclusion, CYP26A1 plays a key role in immune regulation and tumor progression in CRC. Further exploration of its molecular mechanisms and potential application in immunotherapy will provide new directions and strategies for the treatment of colorectal cancer.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we developed and validated an FMRG_score to predict prognosis, TME characteristics, and response to immunotherapy in CRC. Our integrative analysis, which combines large-scale datasets from TCGA and GEO with in vitro experiments, underscores the critical role of folate metabolism in CRC heterogeneity. It identifies key molecular pathways associated with immune evasion and treatment resistance. Notably, this study is the first to propose CYP26A1 as a potential oncogene in CRC, highlighting its significance as a biomarker and therapeutic target.\u003c/p\u003e\n\u003cp\u003eThe FMRG_score effectively stratified patients into high- and low-risk groups, revealing significant survival differences. Patients with a high FMRG_score exhibited a poor prognosis, which correlated with increased activation of pro-tumor pathways and immunosuppressive cell infiltration in the TME, including a higher prevalence of MDSCs and exhausted T cells. These findings suggest that folate metabolism plays a pivotal role in shaping the immune landscape of CRC, potentially mediating immune escape mechanisms that reduce the efficacy of ICIs. This aligns with existing literature on the complex interactions between metabolic reprogramming and immune regulation in cancer progression.\u003c/p\u003e\n\u003cp\u003eOne of the most striking findings of this study is the identification of CYP26A1 as a novel oncogenic driver in CRC. CYP26A1, a member of the cytochrome P450 family, is known for its role in retinoic acid metabolism, which regulates cellular proliferation, differentiation, and apoptosis. Our analysis revealed that high CYP26A1 expression was significantly associated with advanced tumor stage, poor overall survival, and an immunosuppressive TME. These results are consistent with previous reports linking aberrant retinoic acid metabolism to tumorigenesis. Furthermore, our single-cell RNA sequencing data demonstrated a correlation between high CYP26A1 expression and reduced infiltration of effector CD8+ T cells, reinforcing its potential role in immune evasion. These findings open new avenues for targeting CYP26A1 in CRC, either through direct inhibition or by modulating retinoic acid signaling to enhance immune response.\u003c/p\u003e\n\u003cp\u003eHowever, it is not without limitations. A primary limitation is that the majority of our findings were derived from bioinformatic analyses. While these analyses are powerful, they necessitate further validation in in vivo systems to confirm their clinical relevance. Although we successfully demonstrated the tumor-suppressive function of CYP26A1 in vitro, it is imperative that future studies prioritize the use of animal models and clinical samples to verify these results in more physiologically relevant environments. Despite utilizing comprehensive public datasets, the clinical utility of the FMRG_score in predicting treatment outcomes requires validation through prospective clinical trials. Additionally, further mechanistic studies are essential to elucidate the precise role of CYP26A1 in CRC pathogenesis and to assess its potential as a therapeutic target. Future research should also investigate how folate metabolism interacts with other metabolic pathways, influencing cancer progression and therapy resistance.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eCRC（colorectal cancer）, APC（adenomatous polyposis coli）, MSI（microsatellite instability）, dMMR（DNA mismatch repair\u0026ndash;deficient）, MSS（microsatellite-stable）, CIMP（CpG island methylator phenotype）, ICI（immune checkpoint inhibitor）, PD-1（programmed cell death 1）, PD-L1（programmed cell death-1 ligand 1）, CTLA-4（cytotoxic T lymphocyte antigen 4）, NSCLC（non-small-cell lung cancer）, MSI-H（microsatellite instability-high）, CYP26A1（Cytochrome P450 Family 26 Subfamily A Member 1）, TCGA（the Cancer Genome Atlas）, GEO（Gene Expression Omnibus）, scRNA-seq（single-cell RNA sequencing）, FMRGs（folate metabolism-related genes）, FPKM（per kilobase of transcript per million mapped reads）, TPM（transcripts per million）, CNV（copy number variation）, TMB（tumor mutation burden）, LASSO（least absolute shrinkage and selection operator）, Coefᵢ（coefficient）, Expᵢ（expression of each gene）, ROC（receiver operating characteristic）, AUC（area under the curve）, OS（overall survival）, GO（Gene Ontology）, KEGG（Kyoto Encyclopedia of Genes and Genomes）, GSVA（Gene Set Variation Analysis）, GSEA（Gene Set Enrichment Analysis）, TME（tumor microenvironment）, ssGSEA（single sample gene set enrichment analysis）, TIDE（Tumor Immune Dysfunction and Exclusion）, UC（urothelial carcinoma）, MAF（mutation annotation format）, IC50（semi-inhibitory concentration）, tSNE（T-distributed Stochastic Neighbor Embedding）, MDSCs（Myeloid-derived suppressor cells）, PPI（Protein-protein interaction）, RA（Retinoic acid）, Tₑₓ（exhausted T cells）, Tₑff（effector T cells）, Tᵣₘ（tissue resident memory T cells）, pCR（pathological complete response）\u003c/p\u003e\n"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. The data that support the results of current study is available on TCGA (https://portal.gdc.cancer.gov/) and GEO websites (http://www.ncbi.nlm.nih.gov/geo).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThanks for the help of the CRC tissues form Fudan university Shanghai Cancer Center and thanks for the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) Database for sharing a large amount of data. The graphic figure was created with BioRender.com. \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDisclosure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author(s) declare that financial support was received for the research, authorship, and/or publication of this article. This study was supported by the National Key Research and Development Program of China\u0026nbsp;(2023YFF1205000)；National Natural Science Foundation of China (grant no. 82072915 and 82373359); Project of Shanghai Municipal Health Commission (grant no. 202140397); CSCO-ROCHE Cancer Research Fund 2019 (grant no. Y-2019Roche-17 1); and Chinese Young Breast Experts Research project (grant no. CYBER-2021-001). Beijing Science and Technology Innovation Medical Development Foundation Key Project (grant no. KC2022-ZZ-0091-6).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eN/A\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eApproval of the research protocol by an Institutional Reviewer Board\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eN/A\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed Consent\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors have read and approved the final version of the manuscript and agree to its publication. Informed consent was obtained from all participants involved in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRegistry and the Registration No. of the study/trial\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eN/A\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnimal Studies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eN/A\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eY.F. and T.Z designed the studies, conducted most of the experiments and data analysis, and wrote the draft manuscript. Y.X., M.X., C.Z. and Y.F conducted some of the molecular experiments. D.D. and J.Z. were involved in the overall design of the study and the revision of the final manuscript. All authors contributed to the article and approved the submitted version.\u003c/p\u003e\u003cp\u003eAll authors have made significant contributions to the conception, design, analysis, and/or interpretation of the work presented in this manuscript. Furthermore, all authors have thoroughly reviewed and approved the final version of the manuscript. They fully agree with its content and the decision to submit it for consideration.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A\u003cem\u003e et al.\u003c/em\u003e. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. \u003cem\u003eCA: a cancer journal for clinicians\u003c/em\u003e 2021; \u003cstrong\u003e71\u003c/strong\u003e(3)\u003cstrong\u003e:\u003c/strong\u003e 209-249.\u003c/li\u003e\n\u003cli\u003eMurphy N, Ward HA, Jenab M, Rothwell JA, Boutron-Ruault M, Carbonnel F\u003cem\u003e et al.\u003c/em\u003e. Heterogeneity of Colorectal Cancer Risk Factors by Anatomical Subsite in 10 European Countries: A Multinational Cohort Study. \u003cem\u003eClinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association\u003c/em\u003e 2019; \u003cstrong\u003e17\u003c/strong\u003e(7)\u003cstrong\u003e:\u003c/strong\u003e 1323-1331.\u003c/li\u003e\n\u003cli\u003eSiegel RL, Miller KD, Goding Sauer A, Fedewa SA, Butterly LF, Anderson JC\u003cem\u003e et al.\u003c/em\u003e. Colorectal cancer statistics, 2020. \u003cem\u003eCA: a cancer journal for clinicians\u003c/em\u003e 2020; \u003cstrong\u003e70\u003c/strong\u003e(3)\u003cstrong\u003e:\u003c/strong\u003e 145-164.\u003c/li\u003e\n\u003cli\u003eAraghi M, Soerjomataram I, Jenkins M, Brierley J, Morris E, Bray F\u003cem\u003e et al.\u003c/em\u003e. Global trends in colorectal cancer mortality: projections to the year 2035. \u003cem\u003eINT J CANCER\u003c/em\u003e 2019; \u003cstrong\u003e144\u003c/strong\u003e(12)\u003cstrong\u003e:\u003c/strong\u003e 2992-3000.\u003c/li\u003e\n\u003cli\u003eAguiar Junior S, Oliveira MMD, Silva DRME, Mello CALD, Calsavara VF, Curado MP. SURVIVAL OF PATIENTS WITH COLORECTAL CANCER IN A CANCER CENTER. \u003cem\u003eArquivos de gastroenterologia\u003c/em\u003e 2020; \u003cstrong\u003e57\u003c/strong\u003e(2)\u003cstrong\u003e:\u003c/strong\u003e 172-177.\u003c/li\u003e\n\u003cli\u003eVega P, Valent\u0026iacute;n F, Cubiella J. Colorectal cancer diagnosis: Pitfalls and opportunities. \u003cem\u003eWORLD J GASTRO ONCOL\u003c/em\u003e 2015; \u003cstrong\u003e7\u003c/strong\u003e(12)\u003cstrong\u003e:\u003c/strong\u003e 422-433.\u003c/li\u003e\n\u003cli\u003eKeum N, Giovannucci E. Global burden of colorectal cancer: emerging trends, risk factors and prevention strategies. \u003cem\u003eNature reviews. Gastroenterology \u0026amp; hepatology\u003c/em\u003e 2019; \u003cstrong\u003e16\u003c/strong\u003e(12)\u003cstrong\u003e:\u003c/strong\u003e 713-732.\u003c/li\u003e\n\u003cli\u003eFearon ER. Molecular genetics of colorectal cancer. \u003cem\u003eAnnual review of pathology\u003c/em\u003e 2011; \u003cstrong\u003e6:\u003c/strong\u003e 479-507.\u003c/li\u003e\n\u003cli\u003eFodde R, Smits R, Clevers H. APC, signal transduction and genetic instability in colorectal cancer. \u003cem\u003eNature reviews. Cancer\u003c/em\u003e 2001; \u003cstrong\u003e1\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 55-67.\u003c/li\u003e\n\u003cli\u003eVilar E, Gruber SB. Microsatellite instability in colorectal cancer-the stable evidence. \u003cem\u003eNature reviews. Clinical oncology\u003c/em\u003e 2010; \u003cstrong\u003e7\u003c/strong\u003e(3)\u003cstrong\u003e:\u003c/strong\u003e 153-162.\u003c/li\u003e\n\u003cli\u003eWu J, Zhang L, Kuchi A, Otohinoyi D, Hicks C. CpG Site-Based Signature Predicts Survival of Colorectal Cancer. \u003cem\u003eBIOMEDICINES\u003c/em\u003e 2022; \u003cstrong\u003e10\u003c/strong\u003e(12).\u003c/li\u003e\n\u003cli\u003eWei SC, Duffy CR, Allison JP. Fundamental Mechanisms of Immune Checkpoint Blockade Therapy. \u003cem\u003eCANCER DISCOV\u003c/em\u003e 2018; \u003cstrong\u003e8\u003c/strong\u003e(9)\u003cstrong\u003e:\u003c/strong\u003e 1069-1086.\u003c/li\u003e\n\u003cli\u003eWeiner LM. Cancer immunology for the clinician. \u003cem\u003eClinical advances in hematology \u0026amp; oncology : H\u0026amp;O\u003c/em\u003e 2015; \u003cstrong\u003e13\u003c/strong\u003e(5)\u003cstrong\u003e:\u003c/strong\u003e 299-306.\u003c/li\u003e\n\u003cli\u003eQin S, Xu L, Yi M, Yu S, Wu K, Luo S. Novel immune checkpoint targets: moving beyond PD-1 and CTLA-4. \u003cem\u003eMOL CANCER\u003c/em\u003e 2019; \u003cstrong\u003e18\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 155.\u003c/li\u003e\n\u003cli\u003eO\u0026apos;Brien M, Paz-Ares L, Marreaud S, Dafni U, Oselin K, Havel L\u003cem\u003e et al.\u003c/em\u003e. Pembrolizumab versus placebo as adjuvant therapy for completely resected stage IB-IIIA non-small-cell lung cancer (PEARLS/KEYNOTE-091): an interim analysis of a randomised, triple-blind, phase 3 trial. \u003cem\u003eThe Lancet. Oncology\u003c/em\u003e 2022; \u003cstrong\u003e23\u003c/strong\u003e(10)\u003cstrong\u003e:\u003c/strong\u003e 1274-1286.\u003c/li\u003e\n\u003cli\u003eGaron EB, Rizvi NA, Hui R, Leighl N, Balmanoukian AS, Eder JP\u003cem\u003e et al.\u003c/em\u003e. Pembrolizumab for the treatment of non-small-cell lung cancer. \u003cem\u003eThe New England journal of medicine\u003c/em\u003e 2015; \u003cstrong\u003e372\u003c/strong\u003e(21)\u003cstrong\u003e:\u003c/strong\u003e 2018-2028.\u003c/li\u003e\n\u003cli\u003eLarkin J, Chiarion-Sileni V, Gonzalez R, Grob JJ, Cowey CL, Lao CD\u003cem\u003e et al.\u003c/em\u003e. Combined Nivolumab and Ipilimumab or Monotherapy in Untreated Melanoma. \u003cem\u003eThe New England journal of medicine\u003c/em\u003e 2015; \u003cstrong\u003e373\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 23-34.\u003c/li\u003e\n\u003cli\u003eHomet Moreno B, Ribas A. Anti-programmed cell death protein-1/ligand-1 therapy in different cancers. \u003cem\u003eBRIT J CANCER\u003c/em\u003e 2015; \u003cstrong\u003e112\u003c/strong\u003e(9)\u003cstrong\u003e:\u003c/strong\u003e 1421-1427.\u003c/li\u003e\n\u003cli\u003eZhao W, Jin L, Chen P, Li D, Gao W, Dong G. Colorectal cancer immunotherapy-Recent progress and future directions. \u003cem\u003eCANCER LETT\u003c/em\u003e 2022; \u003cstrong\u003e545:\u003c/strong\u003e 215816.\u003c/li\u003e\n\u003cli\u003eLe DT, Uram JN, Wang H, Bartlett BR, Kemberling H, Eyring AD\u003cem\u003e et al.\u003c/em\u003e. PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. \u003cem\u003eThe New England journal of medicine\u003c/em\u003e 2015; \u003cstrong\u003e372\u003c/strong\u003e(26)\u003cstrong\u003e:\u003c/strong\u003e 2509-2520.\u003c/li\u003e\n\u003cli\u003eGuidoboni M, Gaf\u0026agrave; R, Viel A, Doglioni C, Russo A, Santini A\u003cem\u003e et al.\u003c/em\u003e. Microsatellite instability and high content of activated cytotoxic lymphocytes identify colon cancer patients with a favorable prognosis. \u003cem\u003eThe American journal of pathology\u003c/em\u003e 2001; \u003cstrong\u003e159\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 297-304.\u003c/li\u003e\n\u003cli\u003eOverman MJ, Lonardi S, Wong KYM, Lenz H, Gelsomino F, Aglietta M\u003cem\u003e et al.\u003c/em\u003e. Durable Clinical Benefit With Nivolumab Plus Ipilimumab in DNA Mismatch Repair-Deficient/Microsatellite Instability-High Metastatic Colorectal Cancer. \u003cem\u003eJournal of clinical oncology : official journal of the American Society of Clinical Oncology\u003c/em\u003e 2018; \u003cstrong\u003e36\u003c/strong\u003e(8)\u003cstrong\u003e:\u003c/strong\u003e 773-779.\u003c/li\u003e\n\u003cli\u003eGanesh K. Optimizing immunotherapy for colorectal cancer. \u003cem\u003eNature reviews. Gastroenterology \u0026amp; hepatology\u003c/em\u003e 2022; \u003cstrong\u003e19\u003c/strong\u003e(2)\u003cstrong\u003e:\u003c/strong\u003e 93-94.\u003c/li\u003e\n\u003cli\u003eKim JW, Jeon YJ, Jang MJ, Kim JO, Chong SY, Ko KH\u003cem\u003e et al.\u003c/em\u003e. Association between folate metabolism-related polymorphisms and colorectal cancer risk. \u003cem\u003eMOL CLIN ONCOL\u003c/em\u003e 2015; \u003cstrong\u003e3\u003c/strong\u003e(3)\u003cstrong\u003e:\u003c/strong\u003e 639-648.\u003c/li\u003e\n\u003cli\u003eDu W, Li W, Lu R, Fang J. Folate and fiber in the prevention of colorectal cancer: between shadows and the light. \u003cem\u003eWORLD J GASTROENTERO\u003c/em\u003e 2010; \u003cstrong\u003e16\u003c/strong\u003e(8)\u003cstrong\u003e:\u003c/strong\u003e 921-926.\u003c/li\u003e\n\u003cli\u003eKherbek H, Daoud R, Soueycatt T, Soueycatt Y, Ali Z, Ehsan J\u003cem\u003e et al.\u003c/em\u003e. The relationship between folic acid and colorectal cancer; a literature review. \u003cem\u003eAnnals of medicine and surgery (2012)\u003c/em\u003e 2022; \u003cstrong\u003e80:\u003c/strong\u003e 104170.\u003c/li\u003e\n\u003cli\u003eJang MJ, Kim JW, Jeon YJ, Chong SY, Hong SP, Hwang SG\u003cem\u003e et al.\u003c/em\u003e. Polymorphisms of folate metabolism-related genes and survival of patients with colorectal cancer in the Korean population. \u003cem\u003eGENE\u003c/em\u003e 2014; \u003cstrong\u003e533\u003c/strong\u003e(2)\u003cstrong\u003e:\u003c/strong\u003e 558-564.\u003c/li\u003e\n\u003cli\u003eBouras E, Kim AE, Lin Y, Morrison J, Du M, Albanes D\u003cem\u003e et al.\u003c/em\u003e. Genome-wide interaction analysis of folate for colorectal cancer risk. \u003cem\u003eThe American journal of clinical nutrition\u003c/em\u003e 2023; \u003cstrong\u003e118\u003c/strong\u003e(5)\u003cstrong\u003e:\u003c/strong\u003e 881-891.\u003c/li\u003e\n\u003cli\u003eMcNulty H, Pentieva K. Folate bioavailability. \u003cem\u003eThe Proceedings of the Nutrition Society\u003c/em\u003e 2004; \u003cstrong\u003e63\u003c/strong\u003e(4)\u003cstrong\u003e:\u003c/strong\u003e 529-536.\u003c/li\u003e\n\u003cli\u003eIyer R, Tomar SK. Folate: a functional food constituent. \u003cem\u003eJ FOOD SCI\u003c/em\u003e 2009; \u003cstrong\u003e74\u003c/strong\u003e(9)\u003cstrong\u003e:\u003c/strong\u003e R114-R122.\u003c/li\u003e\n\u003cli\u003eHubner RA, Houlston RS. Folate and colorectal cancer prevention. \u003cem\u003eBRIT J CANCER\u003c/em\u003e 2009; \u003cstrong\u003e100\u003c/strong\u003e(2)\u003cstrong\u003e:\u003c/strong\u003e 233-239.\u003c/li\u003e\n\u003cli\u003eMoazzen S, Dolatkhah R, Tabrizi JS, Shaarbafi J, Alizadeh BZ, de Bock GH\u003cem\u003e et al.\u003c/em\u003e. Folic acid intake and folate status and colorectal cancer risk: A systematic review and meta-analysis. \u003cem\u003eClinical nutrition (Edinburgh, Scotland)\u003c/em\u003e 2018; \u003cstrong\u003e37\u003c/strong\u003e(6 Pt A)\u003cstrong\u003e:\u003c/strong\u003e 1926-1934.\u003c/li\u003e\n\u003cli\u003eLi J, Yang H, Lei M, Zhu W, Su Y, Li K\u003cem\u003e et al.\u003c/em\u003e. Dietary folate drives methionine metabolism to promote cancer development by stabilizing MAT IIA. \u003cem\u003eSIGNAL TRANSDUCT TAR\u003c/em\u003e 2022; \u003cstrong\u003e7\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 192.\u003c/li\u003e\n\u003cli\u003eMa J, Stampfer MJ, Giovannucci E, Artigas C, Hunter DJ, Fuchs C\u003cem\u003e et al.\u003c/em\u003e. Methylenetetrahydrofolate reductase polymorphism, dietary interactions, and risk of colorectal cancer. \u003cem\u003eCANCER RES\u003c/em\u003e 1997; \u003cstrong\u003e57\u003c/strong\u003e(6)\u003cstrong\u003e:\u003c/strong\u003e 1098-1102.\u003c/li\u003e\n\u003cli\u003eVan Guelpen B, Hultdin J, Johansson I, Hallmans G, Stenling R, Riboli E\u003cem\u003e et al.\u003c/em\u003e. Low folate levels may protect against colorectal cancer. \u003cem\u003eGUT\u003c/em\u003e 2006; \u003cstrong\u003e55\u003c/strong\u003e(10)\u003cstrong\u003e:\u003c/strong\u003e 1461-1466.\u003c/li\u003e\n\u003cli\u003eOtani T, Iwasaki M, Sasazuki S, Inoue M, Tsugane S. Plasma folate and risk of colorectal cancer in a nested case-control study: the Japan Public Health Center-based prospective study. \u003cem\u003eCancer causes \u0026amp; control : CCC\u003c/em\u003e 2008; \u003cstrong\u003e19\u003c/strong\u003e(1)\u003cstrong\u003e:\u003c/strong\u003e 67-74.\u003c/li\u003e\n\u003cli\u003eKato I, Dnistrian AM, Schwartz M, Toniolo P, Koenig K, Shore RE\u003cem\u003e et al.\u003c/em\u003e. Serum folate, homocysteine and colorectal cancer risk in women: a nested case-control study. \u003cem\u003eBRIT J CANCER\u003c/em\u003e 1999; \u003cstrong\u003e79\u003c/strong\u003e(11-12)\u003cstrong\u003e:\u003c/strong\u003e 1917-1922.\u003c/li\u003e\n\u003cli\u003eVan Guelpen B, Hultdin J, Johansson I, Hallmans G, Stenling R, Riboli E\u003cem\u003e et al.\u003c/em\u003e. Low folate levels may protect against colorectal cancer. \u003cem\u003eGUT\u003c/em\u003e 2006; \u003cstrong\u003e55\u003c/strong\u003e(10)\u003cstrong\u003e:\u003c/strong\u003e 1461-1466.\u003c/li\u003e\n\u003cli\u003eWeile J, Kishore N, Sun S, Maaieh R, Verby M, Li R\u003cem\u003e et al.\u003c/em\u003e. Shifting landscapes of human MTHFR missense-variant effects. \u003cem\u003eAM J HUM GENET\u003c/em\u003e 2021; \u003cstrong\u003e108\u003c/strong\u003e(7)\u003cstrong\u003e:\u003c/strong\u003e 1283-1300.\u003c/li\u003e\n\u003cli\u003eRich JT, Neely JG, Paniello RC, Voelker CCJ, Nussenbaum B, Wang EW. A practical guide to understanding Kaplan-Meier curves. \u003cem\u003eOtolaryngology--head and neck surgery : official journal of American Academy of Otolaryngology-Head and Neck Surgery\u003c/em\u003e 2010; \u003cstrong\u003e143\u003c/strong\u003e(3)\u003cstrong\u003e:\u003c/strong\u003e 331-336.\u003c/li\u003e\n\u003cli\u003eH\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. \u003cem\u003eBMC BIOINFORMATICS\u003c/em\u003e 2013; \u003cstrong\u003e14:\u003c/strong\u003e 7.\u003c/li\u003e\n\u003cli\u003eYu G, Wang L, Han Y, He Q. clusterProfiler: an R package for comparing biological themes among gene clusters. \u003cem\u003eOmics : a journal of integrative biology\u003c/em\u003e 2012; \u003cstrong\u003e16\u003c/strong\u003e(5)\u003cstrong\u003e:\u003c/strong\u003e 284-287.\u003c/li\u003e\n\u003cli\u003eWang W, Chen Z, Hua Y. Bioinformatics Prediction and Experimental Validation Identify a Novel Cuproptosis-Related Gene Signature in Human Synovial Inflammation during Osteoarthritis Progression. \u003cem\u003eBIOMOLECULES\u003c/em\u003e 2023; \u003cstrong\u003e13\u003c/strong\u003e(1).\u003c/li\u003e\n\u003cli\u003eChen B, Khodadoust MS, Liu CL, Newman AM, Alizadeh AA. Profiling Tumor Infiltrating Immune Cells with CIBERSORT. \u003cem\u003eMethods in molecular biology (Clifton, N.J.)\u003c/em\u003e 2018; \u003cstrong\u003e1711:\u003c/strong\u003e 243-259.\u003c/li\u003e\n\u003cli\u003eJiang P, Gu S, Pan D, Fu J, Sahu A, Hu X\u003cem\u003e et al.\u003c/em\u003e. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. \u003cem\u003eNAT MED\u003c/em\u003e 2018; \u003cstrong\u003e24\u003c/strong\u003e(10)\u003cstrong\u003e:\u003c/strong\u003e 1550-1558.\u003c/li\u003e\n\u003cli\u003eShi DD, Guo JA, Hoffman HI, Su J, Mino-Kenudson M, Barth JL\u003cem\u003e et al.\u003c/em\u003e. Therapeutic avenues for cancer neuroscience: translational frontiers and clinical opportunities. \u003cem\u003eThe Lancet. Oncology\u003c/em\u003e 2022; \u003cstrong\u003e23\u003c/strong\u003e(2)\u003cstrong\u003e:\u003c/strong\u003e e62-e74.\u003c/li\u003e\n\u003cli\u003eSchmitt M, Greten FR. The inflammatory pathogenesis of colorectal cancer. \u003cem\u003eNature reviews. Immunology\u003c/em\u003e 2021; \u003cstrong\u003e21\u003c/strong\u003e(10)\u003cstrong\u003e:\u003c/strong\u003e 653-667.\u003c/li\u003e\n\u003cli\u003eThatcher JE, Isoherranen N. The role of CYP26 enzymes in retinoic acid clearance. \u003cem\u003eEXPERT OPIN DRUG MET\u003c/em\u003e 2009; \u003cstrong\u003e5\u003c/strong\u003e(8)\u003cstrong\u003e:\u003c/strong\u003e 875-886.\u003c/li\u003e\n\u003cli\u003eHe W, Wang X, Chen M, Li C, Chen W, Pan L\u003cem\u003e et al.\u003c/em\u003e. Metformin reduces hepatocarcinogenesis by inducing downregulation of Cyp26a1 and CD8(+) T cells. \u003cem\u003eCLIN TRANSL MED\u003c/em\u003e 2023; \u003cstrong\u003e13\u003c/strong\u003e(11)\u003cstrong\u003e:\u003c/strong\u003e e1465.\u003c/li\u003e\n\u003cli\u003eOsanai M, Sawada N, Lee G. Oncogenic and cell survival properties of the retinoic acid metabolizing enzyme, CYP26A1. \u003cem\u003eONCOGENE\u003c/em\u003e 2010; \u003cstrong\u003e29\u003c/strong\u003e(8)\u003cstrong\u003e:\u003c/strong\u003e 1135-1144.\u003c/li\u003e\n\u003cli\u003eAlshafie GA, Walker JR, Curley RWJ, Clagett-Dame M, Highland MA, Nieves NJ\u003cem\u003e et al.\u003c/em\u003e. Inhibition of mammary tumor growth by a novel nontoxic retinoid: chemotherapeutic evaluation of a C-linked analog of 4-HPR-glucuronide. \u003cem\u003eANTICANCER RES\u003c/em\u003e 2005; \u003cstrong\u003e25\u003c/strong\u003e(3c)\u003cstrong\u003e:\u003c/strong\u003e 2391-2398.\u003c/li\u003e\n\u003cli\u003eOsanai M, Lee G. Increased expression of the retinoic acid-metabolizing enzyme CYP26A1 during the progression of cervical squamous neoplasia and head and neck cancer. \u003cem\u003eBMC RES NOTES\u003c/em\u003e 2014; \u003cstrong\u003e7:\u003c/strong\u003e 697.\u003c/li\u003e\n\u003cli\u003eLavudi K, Nuguri SM, Olverson Z, Dhanabalan AK, Patnaik S, Kokkanti RR. Targeting the retinoic acid signaling pathway as a modern precision therapy against cancers. \u003cem\u003eFRONT CELL DEV BIOL\u003c/em\u003e 2023; \u003cstrong\u003e11:\u003c/strong\u003e 1254612.\u003c/li\u003e\n\u003cli\u003eLi J, Wu C, Hu H, Qin G, Wu X, Bai F\u003cem\u003e et al.\u003c/em\u003e. Remodeling of the immune and stromal cell compartment by PD-1 blockade in mismatch repair-deficient colorectal cancer. \u003cem\u003eCANCER CELL\u003c/em\u003e 2023; \u003cstrong\u003e41\u003c/strong\u003e(6)\u003cstrong\u003e:\u003c/strong\u003e 1152-1169.\u003c/li\u003e\n\u003cli\u003eZhang L, Yu X, Zheng L, Zhang Y, Li Y, Fang Q\u003cem\u003e et al.\u003c/em\u003e. Lineage tracking reveals dynamic relationships of T cells in colorectal cancer. \u003cem\u003eNATURE\u003c/em\u003e 2018; \u003cstrong\u003e564\u003c/strong\u003e(7735)\u003cstrong\u003e:\u003c/strong\u003e 268-272.\u003c/li\u003e\n\u003cli\u003eLiu C, Yu H, Huang R, Lei T, Li X, Liu M\u003cem\u003e et al.\u003c/em\u003e. Radioimmunotherapy-induced intratumoral changes in cervical squamous cell carcinoma at single-cell resolution.\u003cem\u003e \u003c/em\u003eVol 42.; 2022:1407-1411.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Table 1","content":"\u003cp\u003e\u003cstrong\u003eTable 1. Multivariate Cox regression analysis of 5 FMRGs associated with FMRG_score in CRC patients.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"593\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 108px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eid\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 93px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ecoef\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHR\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003elower\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eupper\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003epvalue\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 108px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNAT1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 93px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.26\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.770873\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.623926\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.95243\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.015881\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 108px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eWNT5A\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 93px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.194\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.823309\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.689291\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.983383\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.031967\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 108px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCYP26A1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 93px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.181\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.197879\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.050177\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.366354\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.007163\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 108px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBDNF\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 93px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.156\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.168938\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.004565\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.360205\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.0435\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 108px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDRD4\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 93px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.235\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.264762\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.035357\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.544995\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.021435\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"genes-and-immunity","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"genes","sideBox":"Learn more about [Genes \u0026 Immunity](http://www.nature.com/gene/)","snPcode":"41435","submissionUrl":"https://mts-gene.nature.com/cgi-bin/main.plex","title":"Genes \u0026 Immunity","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Folate metabolism, Colorectal cancer, CYP26A1","lastPublishedDoi":"10.21203/rs.3.rs-5647525/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5647525/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003eFolic acid plays a key role in cellular regulation and metabolism, commonly found in dietary supplements. However, its complex role in colorectal cancer (CRC), particularly in metabolism and immune evasion, remains unclear.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003eWe developed the FMRG_score system using machine learning algorithms based on TCGA and GEO databases to assess modification patterns influencing CRC patients' clinical and immune characteristics. The system’s reliability was validated using multiple external clinical cohorts receiving immunotherapy. We further explored the relationships between FMRGs-related features and clinical traits, mutation profiles, biological functions, immune infiltration, therapy response, and drug sensitivity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003eBy combining in vitro experiments and bioinformatics analysis, we established a 9-gene risk model associated with folate metabolism to predict CRC prognosis. Notably, CYP26A1, a key component of the model, was upregulated in CRC tissues, promoting cell proliferation, migration, and invasion. Significant differences in clinical traits, immune cell infiltration, immune checkpoint expression, therapy response, and drug sensitivity were observed between risk groups.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion:\u003c/strong\u003eThe folate scoring system can assess CRC prognosis, tumor microenvironment, and immune therapy response. This is the first study proposing CYP26A1 as an oncogene in CRC.\u003c/p\u003e","manuscriptTitle":"Folate Metabolism in Colorectal Cancer Reveals Links Between Clinical and Immune Traits, Identifying CYP26A1 as a Target","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-28 18:51:48","doi":"10.21203/rs.3.rs-5647525/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"revise","date":"2025-02-18T17:05:06+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-02-04T15:13:16+00:00","index":3,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-02-01T14:23:51+00:00","index":3,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-02-01T07:09:22+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-01-30T16:21:21+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-01-25T10:07:25+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2025-01-24T21:20:30+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-12-16T12:18:44+00:00","index":"","fulltext":""},{"type":"submitted","content":"Genes \u0026 Immunity","date":"2024-12-15T13:00:36+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-12-15T13:00:36+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"genes-and-immunity","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"genes","sideBox":"Learn more about [Genes \u0026 Immunity](http://www.nature.com/gene/)","snPcode":"41435","submissionUrl":"https://mts-gene.nature.com/cgi-bin/main.plex","title":"Genes \u0026 Immunity","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"51398e7a-3f77-4476-a694-f1f2ad0f0bba","owner":[],"postedDate":"January 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-07-03T07:13:55+00:00","versionOfRecord":{"articleIdentity":"rs-5647525","link":"https://doi.org/10.1038/s41435-025-00342-6","journal":{"identity":"genes-and-immunity","isVorOnly":false,"title":"Genes \u0026 Immunity"},"publishedOn":"2025-07-03 04:00:00","publishedOnDateReadable":"July 3rd, 2025"},"versionCreatedAt":"2025-01-28 18:51:48","video":"","vorDoi":"10.1038/s41435-025-00342-6","vorDoiUrl":"https://doi.org/10.1038/s41435-025-00342-6","workflowStages":[]},"version":"v1","identity":"rs-5647525","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5647525","identity":"rs-5647525","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}