Integrated single-cell and bulk RNA-seq analysis reveals a novel T-cell signature for prognosis and treatment response in colorectal cancer

Integrated single-cell and bulk RNA-seq analysis reveals a novel T-cell signature for prognosis and treatment response in colorectal cancer | 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 Integrated single-cell and bulk RNA-seq analysis reveals a novel T-cell signature for prognosis and treatment response in colorectal cancer Peng Cui, Haibo Wang, Zhigang Bai This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4156323/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 30 Aug, 2024 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Colorectal cancer (CRC) is a major contributor to global morbidity and mortality, necessitating more effective therapeutic approaches. T cells, prominent in the tumor microenvironment, exert a crucial role in modulating immunotherapeutic responses and clinical outcomes in CRC. This study introduces a pioneering method for characterizing the CRC immune microenvironment using single-cell sequencing data. Unlike previous approaches, which focused on individual T-cell signature genes, we utilized overall infiltration levels of colorectal cancer signature T-cells. Through weighted gene co-expression network analysis (WGCNA), Lasso regression, and StepCox analysis, we developed a prognostic risk model, TRGS, based on six T cell-related genes. Multivariate Cox analysis identified TRGS as an independent prognostic factor for CRC, showcasing its superior predictive efficacy compared to existing immune-related prognostic models. Immunoreactivity analysis revealed higher Immune Prognostic Score (IPS) and lower Tumor Immune Dysfunction and Exclusion (TIDE) scores in the low-risk group, indicating potential responsiveness to immune checkpoint inhibitor (ICI) therapy. Additionally, patients in the low-risk group demonstrated heightened sensitivity to 5-Fu-based chemotherapy regimens. In summary, TRGS emerges as a standalone prognostic biomarker for CRC, offering insights to optimize patient responses to immunotherapy and chemotherapy, thereby laying the groundwork for personalized tumor management strategies. Biological sciences/Cancer/Cancer therapy/Cancer immunotherapy Biological sciences/Cancer/Cancer therapy/Chemotherapy Biological sciences/Cancer/Cancer genomics Biological sciences/Cancer/Cancer microenvironment Biological sciences/Cancer/Tumour biomarkers Biological sciences/Cancer Biological sciences/Cancer/Gastrointestinal cancer/Colorectal cancer Colorectal cancer T cell Immune checkpoint inhibitors prognostic risk model chemotherapy Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Colorectal cancer (CRC) stands as the third most prevalent malignancy globally, ranking second in mortality 1 , 2 , thereby posing substantial therapeutic challenges and an economic burden. Despite advancements in early CRC screening leading to improved 5-year survival rates, a subset of patients still presents at stage IV, resulting in a dismal prognosis 3 . Immune checkpoint inhibitors (ICIs) have demonstrated remarkable therapeutic efficacy in diverse solid tumors, establishing them as the primary treatment for microsatellite instability-high (MSI-H) metastatic CRC 4 . However, given that this patient subset constitutes less than 5% of all CRC cases 5 , there is a critical need to identify more efficient biomarkers to guide the application of ICIs in colorectal cancer. The tumor microenvironment (TME) plays an important role in tumors, mainly in terms of tumor progression, angiogenesis, tumor metastasis, and influencing treatment outcome 6 – 9 . TME encompasses diverse cellular components such as immune cells, stromal cells, and non-cellular components including extracellular matrix molecules (ECM) and soluble signaling molecules 10 . Among these, T cells, including CD4 + and CD8 + T cells, represent the primary immune cells exerting anti-tumor activity 11 . After antigen-presenting cells present tumor cell antigens to T cells, activating them, the activated T cells infiltrate the tumor vicinity to exert cytotoxic effects on tumor cells 12 . However, tumor cells can evade immune surveillance through various mechanisms, such as HLA-I loss 13 , expression of immune checkpoint molecules 14 , or inhibitory effects from other cells in the microenvironment 15 , 16 . Although more studies are focusing on a specific single T cell, few studies have been reported on overall T cells infiltration. A comprehensive understanding of the alterations in T cells within the tumor microenvironment is crucial for devising future personalized therapeutic strategies. This study utilized bioinformatics methodologies to integrate scRNA-seq data and transcriptome sequencing data from CRC samples. We performed unsupervised consensus clustering of CRC patients based on the overall infiltration levels of characteristic T cells in the tumor microenvironment instead of focusing on a single T cell’s signature genes. Through this approach, a novel T-cell-related prognostic risk model, termed TRGS, was developed and subsequently validated. This model holds significant promise for elucidating the intricate role of T cells within the tumor immune microenvironment of CRC. Notably, when compared with existing prognostic models, TRGS demonstrates superior efficacy in prognostic prediction. Moreover, TRGS exhibits the capacity to predict treatment outcomes, encompassing responses to both immunotherapy and chemotherapy among CRC patients. Additionally, it may identify novel therapeutic targets to improve the effectiveness of ICIs. Methods Data download and processing. For the TCGA-CRC cohort, we retrieved normalized RNA sequencing data (expressed as transcripts per million, TPM) for 549 colorectal cancer (CRC) patients from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). We specifically included patients with a survival time of ≥30 days who did not undergo neoadjuvant therapy. The normalized matrix files for the GSE39582 and GSE38832 cohorts from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/) were utilized as independent external cohorts for the validation of our risk model. Similarly, normalized matrix files for the GSE28702, GSE19860, and GSE45404 cohorts from the GEO database served as independent external cohorts for predicting chemotherapy outcomes. Additionally, single-cell sequencing data were obtained from the GSE132465 dataset, including 23 primary colorectal cancer and 10 matched normal mucosa samples. scRNA-Seq data analysis. The “Seurat” R package 17 was used to perform unsupervised clustering of single cells using the read count matrix as input. We annotated the cells according to the original annotation 18 . Cell populations originally annotated as T-cells were extracted and normalized using the SCTransform method 19 with resolution choice 1.2 for re-clustering. To find marker genes in various clusters, the “Find All Markers” tool was used with |Log2FC| and min. pct cutoff values set to 0.25 and 0.25, respectively. Identification of immune class by consensus clustering. The “ConsensusClusterPlus” R package 20 was used to perform the unsupervised consensus clustering analysis. We used 80% of the items for subsampling and divided each subsample into groups by the k-means algorithm, repeating this clustering process 1000 times. Then, the cumulative distribution function (CDF) curve and consensus matrix are used to identify the optimal k-value. Based on the optimal k-value, patients were clustered into sub-clusters. Tumor immune microenvironment evaluation. TheMCP-counter algorithm 21 based on the “MCPcounter” R package was used to assess the 26 kinds of T cell clusters identified from the scRNA-seq analysis. The ESTIMATE algorithm 22 tool was used to calculate stromal, immune, and estimated scores. Single sample gene set enrichment analysis (ssGSEA) 23 based on the “GSVA” R package was used to quantify 28 kinds of immune cell infiltration scores of the TCGA cohort. Weighted correlation network analysis (WGCNA). Coexpression protein-coding RNA networks of TCGA-CRC were generated using the WGCNA package 24 . An appropriate soft threshold β was calculated to meet the criteria for the scale-free network. Then, the adjacency was transformed into a topological overlap matrix (TOM), and genes with similar expression patterns were divided into the same module. Finally, genes from modules that had high correlation coefficients with T-cell relative immune status were identified for subsequent analysis. Screening and functional annotation of T cell-related genes. DEGs were screened by the limma R package. Adj. P 0.6 were set as the selection conditions for screening the downregulated and upregulated DEGs. T cell-related genes were obtained by overlapping DEGs and WGCNA. Enrichment analysis in Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) were based on the “Clusterprofiler” R package 25 , where “adjusted p < 0.05” was considered significant. Construction and validation of T cell‑related genes prognosis signature (TRGS). The TRGS was developed based on the TCGA-CRC cohort. First, univariate Cox regression analysis was utilized to screen T cell-related genes that were associated with OS using the “survival” R package (P < 0.05). Next, we performed machine learning algorithms including the least absolute shrinkage and selection operator (LASSO) regression analysis and stepwise Cox algorithm to improve the accuracy and reliability of prognostic signature using the “glmnet” and “StepReg” R packages, respectively. the prognostic signature was established as follows: where Coef referred to the coefficient of gene i and Ni represented the expression value of gene i. To further validate the predictive power of the prognostic signature, we performed Kaplan–Meier survival analysis in the TCGA-CRC, GSE39582, and GSE38832 cohorts, where the cutoff value was set to the median TRGS. Time-dependent receiver operating characteristic curves (ROC) were plotted using the “timeROC” R package to predict OS at 1, 3, and 5 years in the training dataset (TCGA-CRC); 1, 2, and 3 years in GSE38832 cohort, and 3, 7, 10 years in GSE39582 cohort. Finally, we further analyzed the relationship between the prognostic signature and clinical characteristics (including age, gender, and stage) to clarify the independent prognostic value of TRGS using multivariate Cox regression analyses. Development and evaluation of the nomogram. We performed the univariate and multivariate Cox regression analyses for clinical parameters and TRGS. In the multivariate Cox model, variables with P < 0.05 were included in the construction of the nomogram by using the “rms” R package. The calibration analysis was used to assess the prognostic accuracy of the nomogram model. Decision curve analysis (DCA) curves were used to evaluate the net benefit of the nomogram. Immune checkpoints inhibitors sensitivity prediction. We utilized Immunophenoscore (IPS) obtained from The Cancer Immunome Atlas (TCIA) website (tcia.at/home) 26 to predict the response to ICIs. We also utilized the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm through an online website (http://tide.dfci.harvard.edu) 27 to predict patients’ ICIs responses. Higher TIDE prediction scores were generally associated with worse ICIs responses. Anticancer drug sensitivity prediction. The half-maximal inhibitory concentration (IC50) of common chemotherapeutic and targeted therapeutic drugs was estimated by using the "pRRophetic" R package 28 . GSE28702, GSE19860, and GSE45404 were transcriptome sequencing data from patients who received adjuvant chemotherapy (ACT), which were utilized as external validation cohorts for predicting chemotherapy drug sensitivity. Statistical analysis. All statistical analyses were performed using R software (version 4.3.1). Wilcoxon test was used to compare the differences between groups. The log-rank test was used to compare Kaplan–Meier survival curves. Univariate and multivariate Cox analyses were performed to establish independent prognostic factors. All P values were two-sided and less than 0.05 were considered statistically significant. Results ScRNA-Seq analysis of CRC samples. The flowchart outlining our study is presented in Supplementary Fig. 1. We obtained 10x single-cell RNA sequencing (scRNA-seq) data from the GSE132465 dataset, which comprised 33 colorectal cancer (CRC) samples, including 23 primary colorectal cancer and 10 matched normal mucosa samples. The 23,115 T cells extracted from the original dataset were re-clustered. Based on the clustree results (Supplementary Fig. 2A), a resolution of 1.2 was chosen, leading to the identification of 26 distinct cell subgroups through hierarchical clustering utilizing UMAP analysis (Fig. 1A). Annotation of the 26 clusters of T cells was performed using literature reports 29, 30 , CellMarker 2.0 31 , and T cellmap 32 . We presented the characteristic genes (Fig. 1B) as well as functional annotations for each cell cluster (Fig. 1C, D). Differential expression genes (DEGs) of each cluster were provided in Supplementary Table S1 and shown in Supplementary Fig. 2B. Furthermore, we explored the differences between these cells in cancer and adjacent tissues (Fig. 1E). Subsequently, the characteristic genes of each cell cluster were employed to estimate the differences in cell populations of TCGA patients using the MCP-counter algorithm. We compared normal and tumor groups to illustrate the validity of the MCP-counter algorithm (Fig. 1F). Identification of T cell infiltration subgroups. Based on the assessment of T cell infiltration using the MCP-counter package, we conducted a consensus cluster analysis, initially dividing all CRC samples into k clusters (k = 2–9). The CDF curves of the consensus score matrix revealed that the optimal number of clusters was achieved when k = 2 (Fig. 2A, B). The two consensus clusters (C1 and C2) exhibited significant differences in the distribution of 26 T cell clusters, with C1 displaying substantially higher T cell infiltration abundance compared to C2 (Fig. 2C). Consequently, we categorized C1 as "immune-hot" tumors and C2 as "immune-cold" tumors. To ensure the robustness of the two consensus clusters and mitigate potential algorithmic biases, ssGSEA was employed to validate the disparity in immune cell infiltration between C1 and C2, revealing a higher overall immune cell infiltration abundance in C1 (Fig. 2D). Additionally, the ESTIMATE algorithm was utilized to illustrate that immune-hot tumors exhibited a higher immune score compared to immune-cold tumors (Fig. 2E). Identification of candidate T cell-related genes. To identify T cell-related genes associated with immune cell infiltration, we initially conducted WGCNA. In this procedure, we set the soft threshold β to 6, resulting in a scale-free topology fit index of 0.85 (Supplementary Fig. 3A). Subsequently, we utilized the “merged dynamics” algorithm to derive fourteen modules (Fig. 3A). The heatmap depicted the eigengene adjacency of these modules (Fig. 3B). Through assessment of the correlation coefficient and P-value, we observed that the blue module exhibited the strongest correlation with the previously identified immune status (Fig. 3C), achieving a correlation coefficient of 0.78 between gene significance (GS) and module membership (MM) (Fig. 3D). Concurrently, we conducted differential expression analysis of the TCGA-CRC cohort using the “limma” package, identifying a total of 686 differentially expressed genes (DEGs). By intersecting these results with those from WGCNA, we extracted 369 overlapping genes deemed as T cell-related genes for subsequent analyses (Fig. 3E). GO and KEGG enrichment analyses were conducted to provide a comprehensive functional overview of the overlapping genes (Fig. 3F, G and Supplementary Table S2). The results revealed predominant enrichment in biological processes (BP) such as leukocyte-mediated immunity, leukocyte cell-cell adhesion, and regulation of T-cell activation. In terms of cellular composition (CC), enrichment was observed in areas including the external side of the plasma membrane, MHC protein complex, and MHC class II protein complex. Additionally, molecular function (MF) analysis highlighted immune receptor activity, chemokine receptor binding, and MHC protein complex binding as significantly enriched categories. Furthermore, KEGG pathway analysis indicated significant enrichment in pathways such as cytokine-cytokine receptor interaction, cell adhesion molecules, and chemokine signaling, all of which were closely associated with anti-cancer immunity. T cell signature establishment and external validation To evaluate the prognostic impact of the identified genes on CRC, we initially identified 12 genes associated with overall survival (OS) through univariate Cox analysis (Supplementary Fig. 3B). Subsequently, to prevent overfitting and exclude co-expressed T cell-related genes (Supplementary Fig. 3C), we employed lasso regression analysis to construct a predictive prognostic model comprising T cell-related genes. The optimal λ value was determined based on the minimum partial likelihood deviance achieved via the leave-one-out cross-validation (LOOCV) framework (Fig. 4A). Ten genes with nonzero Lasso coefficients were further subjected to stepwise Cox proportional hazards regression, resulting in the identification of a final set of 6 genes (Fig. 4B), and we assessed their impact on prognosis at the same time (Supplementary Fig. 3D-I). A linear prediction model was then developed using the weighted regression coefficients of the 6 T cell-related genes, represented as TRGS = (0.325 x ADAM8 exp) + (0.282 x ZNF385A exp) + (0.201 x G0S2 exp) + (-0.195 x TAP1 exp) + (-0.216 x MMP1 exp) + (-0.347 x CCL22 exp). Based on the median of TRGS, all patients were stratified into high- and low-risk groups. Notably, patients in the high-risk group exhibited significantly poorer OS compared to those in the low-risk group (Fig. 4C, E, G). The discriminatory ability of TRGS was assessed using receiver operating characteristic (ROC) analysis, with 1-, 3-, and 5-year areas under the curve (AUCs) of 0.725, 0.710, and 0.700 in TCGA-CRC (Fig. 4D); 3-, 7-, and 10-year AUCs of 0.620, 0.639, and 0.684 in the GSE39582 cohort (Fig. 4F); and 1-, 2-, 3-year AUCs of 0.656, 0.696, and 0.616 in GSE38832 cohort (Fig. 4H). Furthermore, to evaluate the prognostic capabilities of TRGS, we obtained 5 previously published prognostic models developed using immune-related genes 33-37 . Subsequently, we conducted a comparative analysis to assess the predictive performance of TRGS in terms of prognosis. Remarkably, our findings revealed that TRGS exhibited superior performance, as evidenced by higher AUC values in time-dependent ROC analysis when compared to the 5 existing prognostic models in TCGA, GSE39582, and GSE38832 cohorts. (Supplementary Fig. 4A-O). Finally, we developed a nomogram incorporating TRGS and clinicopathological features to assess the clinical significance of the predictive model (Fig. 4I). Calibration curves demonstrated good agreement between the predicted and actual probabilities of survival at 1, 2, and 3 years (Fig. 4J). Decision curve analysis (DCA) revealed that the nomogram model provided a greater net benefit in predicting 5-year OS (Fig. 4K). Meanwhile, multifactorial Cox regression analysis confirmed TRGS as an independent prognostic risk factor for CRC (Fig. 4L and Supplementary Fig. 4P). TRGS evaluation for immunotherapeutic response prediction. To evaluate the predictive capacity of TRGS as a biomarker for ICIs response, we initially investigated the relationship between low- and high-risk groups and the immune-hot and immune-cold phenotypes as defined in Fig. 2. Patients classified in the immune-cold group belonged to the high-risk category, whereas those in the low-risk group exhibited an immune-hot phenotype (Fig. 5A). Subsequently, we examined the association between TRGS and immune checkpoint blocker (ICB)-related pathways, as well as the seven steps in the Cancer-Immunity cycle. Our findings revealed a negative correlation between TRGS and ICB-related pathways and steps in the Cancer-Immunity cycle, suggesting that low-risk patients may possess a richer tumor immune microenvironment, potentially leading to a higher response rate to ICIs (Fig. 5B, C). Furthermore, we analyzed the distribution of IPS and TIDE scores across different risk groups. Our results indicated that the low-risk group exhibited higher IPS scores (Fig. 5D-G), whereas the high-risk group demonstrated elevated TIDE scores (Fig. 5H), Exclusion scores (Fig. 5I), and Dysfunction scores (Fig. 5J). These results were consistent with previous findings, suggesting that low-risk patients may exhibit increased sensitivity to ICIs. Predictive value of fluorouracil-based ACT benefits. We expanded our investigation to assess the predictive capability of TRGS in quantifying the benefits of fluorouracil-based adjuvant chemotherapy (ACT). Survival analysis among patients stratified into low- and high-risk groups, both receiving and not receiving ACT, revealed that individuals in the low-risk category who underwent therapy exhibited the most favorable survival outcomes. Conversely, among patients categorized as high-risk, no substantial disparity in survival was observed between those who received ACT and those who did not (Fig. 6A, B). This suggests that low TRGS may serve as a predictor for a more favorable response to ACT. Subsequently, we calculated the IC50 values of common chemotherapy and targeted drugs for CRC to further explore the drug sensitivity between low- and high-risk groups. Comparison of the IC50 values between the two risk groups indicated that the low-risk group exhibited higher sensitivity to fluorouracil, gemcitabine, paclitaxel, and vincristine (Fig. 6C). However, patients in the high and low-risk groups showed no significant difference in the sensitivity to targeted therapy (Fig. 6D). These findings were validated in three datasets treated with fluorouracil-based ACT. Responders demonstrated a significantly lower TRGS score than non-responders in TCGA, GSE28702, GSE19860, and GSE45404 (Fig. 6E-H). Discussion Currently, conventional treatment strategies for colorectal cancer primarily involve surgery combined with adjuvant chemotherapy or targeted therapy, yet these approaches have not yielded significant efficacy 38 . Moreover, stage IV patients exhibit a high likelihood of chemotherapy resistance 30 , potentially leading to situations of overtreatment or undertreatment within the current staging-based treatment paradigm. Finding biomarkers to predict chemotherapy outcomes is of utmost urgency. Furthermore, immunotherapy, as a novel and effective treatment modality, remains underutilized in colorectal cancer, necessitating more effective biomarkers to guide the application of ICIs in colorectal cancer 39 . T cells, being the most abundant immune cells infiltrating the tumor microenvironment, play a crucial role in anti-tumor immunity. A comprehensive understanding of the T cell population's status within the colorectal cancer tumor microenvironment contributes to the identification of more effective immunotherapeutic strategies. In this study, we first identified colorectal cancer-specific T cell signature genes through scRNA-seq data. Subsequently, we evaluated the levels of T-cell infiltration in the TCGA-CRC cohort using these genes. Next, we constructed an immune classification of colorectal cancer based on the infiltration levels of 26 types of T cells instead of single T cell signature genes and identified T cell-related genes associated with the immunogenic tumor phenotype through WGCNA. We then conducted univariate COX regression to identify prognosis-related genes, followed by the selection of the six most relevant genes for constructing a subsequent prognostic risk model using the Lasso and StepCox machine learning algorithms. Among these six genes, high expression of ADAM8, ZNF385A, and G0S2 was associated with poor prognosis, while high expression of TAP1, MMP1, and CCL22 was associated with favorable prognosis. Interestingly, these six genes have previously been found to be associated with immune cell infiltration in earlier studies, consistent with our findings. Studies have reported that downregulating ADAM8 in colorectal cancer cells enhances the cytotoxic effect of tumor-infiltrating lymphocytes 40 . ZNF385A has been identified as a gene associated with lymphatic vessel formation in CRC, influencing immune cell infiltration, and is also associated with an immunosuppressive phenotype in HCC 41 , 42 . G0S2 is linked to evasion from cytotoxic T lymphocytes and knocking down G0S2 significantly inhibits proliferation and migration of CRC cell lines 43 . TAP1, a transporter associated with antigen processing protein, plays a crucial role in antigen presentation 44 , and its downregulation in colorectal cancer is associated with low infiltration of various lymphocytes 45 . MMP1, one of the markers of epithelial-mesenchymal transition, is associated with metastasis and angiogenesis in colorectal cancer 46 , 47 . In refractory HCC treated with Transcatheter arterial chemoembolization (TACE), a prognostic model incorporating MMP1 is related to immune cell infiltration 48 . Additionally, MMP1 may be one of the targets to enhance the effectiveness of immunotherapy in PDAC 49 . CCL22 is a chemokine that not only recruits monocytes but also regulates the infiltration of Th1 and Treg cells in colorectal cancer 50 , 51 . Subsequently, we constructed a risk model of T cell-related genes (TRGS) using these six genes, and patients were stratified into high-risk and low-risk groups based on the median TRGS value. Compared to patients in the high-risk group, those in the low-risk group exhibited significantly prolonged overall survival. Furthermore, TRGS demonstrated superior predictive efficacy in prognosis compared to other existing immune-related models. Meanwhile, the analysis results indicated that TRGS could serve as an independent prognostic factor for colorectal cancer. Additionally, we utilized IPS and TIDE scores to predict the response of patients in the high- and low-risk groups to ICIs treatment. The results revealed a statistically significant increase in IPS scores among patients in the low-risk group, whereas those in the high-risk group demonstrated higher Dysfunction and Exclusion scores, suggesting a greater likelihood of immune evasion in this group. These findings suggest that ICIs therapy may be more effective for patients in the low-risk group. Previous studies have shown that 5-FU treatment can induce a reduction in tumor-infiltrating myeloid-derived suppressor cells (MDSCs) and promote T cell-mediated anti-tumor immunity 52 . Additionally, in metastatic colorectal cancer patients receiving FOLFOX combined with bevacizumab, the accumulation of Th17 cells was associated with poor prognosis 53 . Similarly, in lung cancer, the combination treatment of oxaliplatin and cyclophosphamide led to an increased ratio of CD8 T cells to Treg cells within the tumor, resulting in significant inhibition of tumor growth 54 . These findings collectively indicate the existence of interactions between T cells and chemotherapy within the tumor microenvironment. Therefore, we subsequently compared the IC50 values of common chemotherapy drugs to determine the sensitivity of high- and low-risk groups to chemotherapy drugs. The results showed that the IC50 values of common chemotherapy drugs including fluorouracil, gemcitabine, and paclitaxel were decreased in the low-risk group, indicating higher sensitivity. Furthermore, we used the GSE28702 and GSE19860 datasets to validate patient responses to fluorouracil-based ACT. The results successfully validated our findings, demonstrating that patients responsive to fluorouracil-based ACT exhibited lower risk scores. While our study has been successfully validated in external cohorts, several limitations warrant consideration. Firstly, our study relied on publicly available databases, and both our training cohort (TCGA) and validation cohorts (GEO) were retrospective. Hence, prospective studies are warranted to validate our conclusions rigorously. Secondly, there is still a shortage of sequencing databases containing data from CRC patients treated with ICIs, as well as a deficiency of large clinical cohorts of patients undergoing ICIs treatment. Therefore, our results still need validation in a large clinical cohort in the future. Thirdly, further research is needed to identify hub genes and elucidate the potential regulatory mechanisms to facilitate the discovery of targets that enhance the efficacy of ICIs and chemotherapy. Conclusion In summary, we pioneered a novel approach to categorize the colorectal cancer immune microenvironment utilizing single-cell data. Rather than focusing on individual T-cell signature genes, we employed the infiltration levels of the comprehensive colorectal cancer signature T-cells. This innovative strategy led to the construction of a TRGS comprising ADAM8, ZNF385A, G0S2, TAP1, MMP1, and CCL22. Our findings underscore the superiority of TRGS in prognostic prediction compared to existing immune-related prognostic models. Moreover, TRGS exhibits predictive capabilities not only for ICIs treatment but also for chemotherapy efficacy, presenting novel perspectives for the stratified diagnosis and treatment of colorectal cancer patients. Declarations Acknowledgements The authors thank researchers and contributors of the TCGA and GEO databases for their valuable efforts and contributions. Author contributions P.C. and Z.G.B. conceptualized and designed the study. P.C. was responsible for data collection, analysis, and initial manuscript drafting. H.B.W. reviewed the statistical analysis and performed manuscript editing. Z.G.B. oversaw the project, contributed to manuscript editing, provided critical feedback, and supervised the study. All authors contributed to the critical revision of the manuscript and approved the final version for publication. Data availability statement The raw data utilized in our study were obtained from the TCGA dataset (https://portal.gdc.cancer.gov/) and the GEO dataset (https://www.ncbi.nlm.nih.gov/). Specifically, TCGA accession numbers for our study are TCGA-COAD and TCGA-READ, while GEO accession numbers are GSE39582, GSE38832, GSE132465, GSE28702, GSE19860, and GSE45404. The datasets analyzed and R language codes employed in the current study are accessible upon reasonable request from the corresponding author. Additional information Competing Interests: The authors declare no competing interests. References Xi, Y. & Xu, P. Global colorectal cancer burden in 2020 and projections to 2040. Transl Oncol 14 ,101174 (2021). Sung, H. et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 71 ,209-249 (2021). Siegel, R.L. et al. Colorectal cancer statistics, 2020. CA Cancer J Clin 70 ,145-164 (2020). Andre, T. et al. Pembrolizumab in Microsatellite-Instability-High Advanced Colorectal Cancer. N Engl J Med 383 ,2207-2218 (2020). Koopman, M. et al. Deficient mismatch repair system in patients with sporadic advanced colorectal cancer. Br J Cancer 100 ,266-273 (2009). Quail, D.F. & Joyce, J.A. Microenvironmental regulation of tumor progression and metastasis. Nat Med 19 ,1423-1437 (2013). Angelova, M. et al. Evolution of Metastases in Space and Time under Immune Selection. Cell 175 ,751-765 e716 (2018). Jin, M.Z. & Jin, W.L. The updated landscape of tumor microenvironment and drug repurposing. Signal Transduct Target Ther 5 ,166 (2020). Hernandez-Camarero, P., Lopez-Ruiz, E., Marchal, J.A. & Peran, M. Cancer: a mirrored room between tumor bulk and tumor microenvironment. J Exp Clin Cancer Res 40 ,217 (2021). Belli, C. et al. Targeting Cellular Components of the Tumor Microenvironment in Solid Malignancies. Cancers (Basel) 14 (2022). Pelka, K. et al. Spatially organized multicellular immune hubs in human colorectal cancer. Cell 184 ,4734-4752 e4720 (2021). Mellman, I., Chen, D.S., Powles, T. & Turley, S.J. The cancer-immunity cycle: Indication, genotype, and immunotype. Immunity 56 ,2188-2205 (2023). Anderson, P., Aptsiauri, N., Ruiz-Cabello, F. & Garrido, F. HLA class I loss in colorectal cancer: implications for immune escape and immunotherapy. Cell Mol Immunol 18 ,556-565 (2021). Chen, Q. et al. Demethylase JMJD2D induces PD-L1 expression to promote colorectal cancer immune escape by enhancing IFNGR1-STAT3-IRF1 signaling. Oncogene 41 ,1421-1433 (2022). Xue, R. et al. Liver tumour immune microenvironment subtypes and neutrophil heterogeneity. Nature 612 ,141-147 (2022). Fridman, W.H. et al. B cells and tertiary lymphoid structures as determinants of tumour immune contexture and clinical outcome. Nat Rev Clin Oncol 19 ,441-457 (2022). Hao, Y. et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat Biotechnol 42 ,293-304 (2024). Lee, H.O. et al. Lineage-dependent gene expression programs influence the immune landscape of colorectal cancer. Nat Genet 52 ,594-603 (2020). Lause, J., Berens, P. & Kobak, D. Analytic Pearson residuals for normalization of single-cell RNA-seq UMI data. Genome Biol 22 ,258 (2021). Wilkerson, M.D. & Hayes, D.N. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics 26 ,1572-1573 (2010). Becht, E. et al. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome Biol 17 ,218 (2016). Yoshihara, K. et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat Commun 4 ,2612 (2013). Barbie, D.A. et al. Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature 462 ,108-112 (2009). Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9 ,559 (2008). Yu, G., Wang, L.G., Han, Y. & He, Q.Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16 ,284-287 (2012). Charoentong, P. et al. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep 18 ,248-262 (2017). Jiang, P. et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nat Med 24 ,1550-1558 (2018). Geeleher, P., Cox, N. & Huang, R.S. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS One 9 ,e107468 (2014). Zhang, L. et al. Lineage tracking reveals dynamic relationships of T cells in colorectal cancer. Nature 564 ,268-272 (2018). Zheng, L. et al. Pan-cancer single-cell landscape of tumor-infiltrating T cells. Science 374 ,abe6474 (2021). Hu, C. et al. CellMarker 2.0: an updated database of manually curated cell markers in human/mouse and web tools based on scRNA-seq data. Nucleic Acids Res 51 ,D870-D876 (2023). Chu, Y. et al. Pan-cancer T cell atlas links a cellular stress response state to immunotherapy resistance. Nat Med 29 ,1550-1562 (2023). Tu, W. et al. Elucidating the role of T-cell exhaustion-related genes in colorectal cancer: a single-cell bioinformatics perspective. Funct Integr Genomics 23 ,259 (2023). Chen, J. et al. Comprehensive analysis and experimental verification of the mechanism of action of T cell-mediated tumor-killing related genes in Colon adenocarcinoma. Transl Oncol 43 ,101918 (2024). Liao, K. et al. Identification of signature of tumor-infiltrating CD8 T lymphocytes in prognosis and immunotherapy of colon cancer by machine learning. Clin Immunol 257 ,109811 (2023). Li, Q., Chu, Y., Yao, Y. & Song, Q. A Treg-related riskscore model may improve the prognosis evaluation of colorectal cancer. J Gene Med 26 ,e3668 (2024). Li, X. et al. Identification of an Immune Signature Predicting Prognosis Risk and Lymphocyte Infiltration in Colon Cancer. Front Immunol 11 ,1678 (2020). Bockelman, C., Engelmann, B.E., Kaprio, T., Hansen, T.F. & Glimelius, B. Risk of recurrence in patients with colon cancer stage II and III: a systematic review and meta-analysis of recent literature. Acta Oncol 54 ,5-16 (2015). Ganesh, K. et al. Immunotherapy in colorectal cancer: rationale, challenges and potential. Nat Rev Gastroenterol Hepatol 16 ,361-375 (2019). Liao, R. et al. Identification of biomarkers related to Tumor-Infiltrating Lymphocytes (TILs) infiltration with gene co-expression network in colorectal cancer. Bioengineered 12 ,1676-1688 (2021). Liu, H., Shi, H. & Sun, Y. Identification of a novel lymphangiogenesis signature associated with immune cell infiltration in colorectal cancer based on bioinformatics analysis. BMC Med Genomics 17 ,2 (2024). Peng, Q. et al. ZNF385A and ZNF346 Serve as Prognostic Biomarkers Associated with an Inflamed Immunosuppressive Tumor Microenvironment in Hepatocellular Carcinoma. Int J Mol Sci 24 (2023). Wang, X. et al. Robust machine-learning based prognostic index using cytotoxic T lymphocyte evasion genes highlights potential therapeutic targets in colorectal cancer. Cancer Cell Int 24 ,52 (2024). Blees, A. et al. Structure of the human MHC-I peptide-loading complex. Nature 551 ,525-528 (2017). Ling, A. et al. TAP1 down-regulation elicits immune escape and poor prognosis in colorectal cancer. Oncoimmunology 6 ,e1356143 (2017). Yang, P. et al. TCONS_00012883 promotes proliferation and metastasis via DDX3/YY1/MMP1/PI3K-AKT axis in colorectal cancer. Clin Transl Med 10 ,e211 (2020). Ng, L. et al. CD26 Induces Colorectal Cancer Angiogenesis and Metastasis through CAV1/MMP1 Signaling. Int J Mol Sci 23 (2022). Cheng, X. et al. The role of hypoxia-related genes in TACE-refractory hepatocellular carcinoma: Exploration of prognosis, immunological characteristics and drug resistance based on onco-multi-OMICS approach. Front Pharmacol 13 ,1011033 (2022). Hung, H.C. et al. Effect of chimeric antigen receptor T cells against protease-activated receptor 1 for treating pancreatic cancer. BMC Med 21 ,338 (2023). Cremonesi, E. et al. Gut microbiota modulate T cell trafficking into human colorectal cancer. Gut 67 ,1984-1994 (2018). Dong, S., Guo, X., Han, F., He, Z. & Wang, Y. Emerging role of natural products in cancer immunotherapy. Acta Pharm Sin B 12 ,1163-1185 (2022). Vincent, J. et al. 5-Fluorouracil selectively kills tumor-associated myeloid-derived suppressor cells resulting in enhanced T cell-dependent antitumor immunity. Cancer Res 70 ,3052-3061 (2010). Limagne, E. et al. Accumulation of MDSC and Th17 Cells in Patients with Metastatic Colorectal Cancer Predicts the Efficacy of a FOLFOX-Bevacizumab Drug Treatment Regimen. Cancer Res 76 ,5241-5252 (2016). Pfirschke, C. et al. Immunogenic Chemotherapy Sensitizes Tumors to Checkpoint Blockade Therapy. Immunity 44 ,343-354 (2016). Additional Declarations No competing interests reported. Supplementary Files SupplementaryFigures.pdf SupplementaryTableS1.xlsx SupplementaryTableS2.xlsx SupplementrayFigurelegends.docx Cite Share Download PDF Status: Published Journal Publication published 30 Aug, 2024 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 14 May, 2024 Reviews received at journal 10 May, 2024 Reviewers agreed at journal 30 Apr, 2024 Reviews received at journal 02 Apr, 2024 Reviewers agreed at journal 02 Apr, 2024 Reviewers invited by journal 02 Apr, 2024 Editor assigned by journal 02 Apr, 2024 Editor invited by journal 02 Apr, 2024 Submission checks completed at journal 01 Apr, 2024 First submitted to journal 23 Mar, 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-4156323","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":287026191,"identity":"e29b9696-076a-4fd8-a0e2-c7eb68d694ae","order_by":0,"name":"Peng Cui","email":"","orcid":"","institution":"Capital Medical University, National Clinical Research Center for Digestive Diseases","correspondingAuthor":false,"prefix":"","firstName":"Peng","middleName":"","lastName":"Cui","suffix":""},{"id":287026192,"identity":"1cc44d51-9555-4a31-8a5c-d31a5856f14b","order_by":1,"name":"Haibo Wang","email":"","orcid":"","institution":"Capital Medical University","correspondingAuthor":false,"prefix":"","firstName":"Haibo","middleName":"","lastName":"Wang","suffix":""},{"id":287026193,"identity":"64117a3f-47f3-44a0-9f18-24639dbd5539","order_by":2,"name":"Zhigang Bai","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+ElEQVRIiWNgGAWjYBACAwbGBmYgCQTMB6BiCURrYYMpJagFaDyEyWNAnBZz9sPNnwsKauXN+dd8/PAz5zADP3uOAcPPHbi1WPYkNhjPMDhuuHPG282SvdsOM0j2vDFg7D2Dx2EHEhuSeQyOMW64cXYbMyNQi8GNHANmxjY8Ws4/bDgM1GK/4caZZ2At9gS13EhsbOYxqEnccL6HDWKLBAEtljMeNjPzGBxI3nCDzRjol3QeiTPPCg724tFizp/++DPPnzrbDecPP/zwc5u1HH978sYHP/FogYLDDAwSCWAWD4g4QFADA0MdAwM/MepGwSgYBaNgRAIADCZWUHWO/zUAAAAASUVORK5CYII=","orcid":"","institution":"Capital Medical University, National Clinical Research Center for Digestive Diseases","correspondingAuthor":true,"prefix":"","firstName":"Zhigang","middleName":"","lastName":"Bai","suffix":""}],"badges":[],"createdAt":"2024-03-24 03:14:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4156323/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4156323/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-024-70422-6","type":"published","date":"2024-08-30T15:57:27+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":54162350,"identity":"abe61705-d47c-4363-944e-91f3bf5fa700","added_by":"auto","created_at":"2024-04-05 13:09:14","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":783211,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDifferent T cell clustering in 10X scRNA-seq data of colorectal carcinoma. \u003c/strong\u003e(A) Cluster annotation and cell type identification using UMAP.\u003cstrong\u003e \u003c/strong\u003e(B) Bubble plot displaying marker gene expression across identified clusters.\u003cstrong\u003e \u003c/strong\u003e(C, D) Heatmap depicting the expression of curated gene signatures within CD8+ T and CD4+ T cell clusters.\u003cstrong\u003e \u003c/strong\u003e(E) Distribution of defined T cell clusters between tumor and normal groups in GSE132465.\u003cstrong\u003e \u003c/strong\u003e(F) MCP-counter enrichment scores of defined T cell clusters between normal and tumor groups in the TCGA-CRC cohort.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/e87fe1f995c7e69d38db4cc8.png"},{"id":54162354,"identity":"88262b80-c647-46f0-a4fb-d91d4b176653","added_by":"auto","created_at":"2024-04-05 13:09:16","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1436634,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUnsupervised consensus cluster analysis of CRC patients in the TCGA cohort based on MCP-counter enrichment score.\u003c/strong\u003e (A) The consensus score matrix of all samples when k=2. (B) The CDF curves of the consensus matrix for each k (indicated by colors). (C) The infiltration abundance of defined T cell clusters for two clusters. (D) The infiltration abundance of 28 immune cell subsets evaluated by ssGSEA for two clusters. (E) The distribution of immune score inferred by ESTIMATE algorithm between two clusters in the TCGA-CRC cohort. Statistic test: two-sided unpaired t-test. (*\u003cem\u003eP\u003c/em\u003e \u0026lt;0.05, ** \u003cem\u003eP\u003c/em\u003e \u0026lt;0.01, *** \u003cem\u003eP\u003c/em\u003e \u0026lt;0.001)\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/daac853857e7d3224d11a067.png"},{"id":54162353,"identity":"6fafedbb-1213-47e7-b697-8344e830952b","added_by":"auto","created_at":"2024-04-05 13:09:16","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":829293,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of T cell-related genes. \u003c/strong\u003e(A) The dendrogram of modules. (B) The heatmap revealed the eigengene adjacency of modules. (C) Correlation analysis between module eigengenes and clinical traits. (D) The strong correlation between GS and MM in the blue module. (E) Venn diagram displaying overlapping genes between WGCNA and TRDEGs. (F, G) Go and KEGG function enrichment of the intersected genes.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/0fe10868f74af34e0f2d2f6b.png"},{"id":54162360,"identity":"bdc45428-e49b-4ab1-99a6-353a3959be1e","added_by":"auto","created_at":"2024-04-05 13:09:17","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":823705,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe establishment and external validation of the T cell-related genes signature.\u003c/strong\u003e (A) Lasso regression profiles of TRGS to prevent over-fitting and 10-fold cross-validation of variable selection with Lasso. (B) Coefficients of 6 T cell-related genes finally obtained in stepwise Cox regression. (C, E, G) Kaplan-Meier curves illustrating the prognostic value of TRGS. (D, F, H) Time-dependent ROC curvesdemonstrating the prognostic signature's predictive performance. (I) Nomogram integrating age, stage, and TRGS for prognosis assessment. (J) The calibration curve of the constructed nomogram of 1-year, 2-year, and 3-year survival prediction. (K) Decision curve analysis evaluating the 5-year survival benefit in the TCGA cohort. (L) Univariate and multivariate COX regression analysis of TRGS and various clinical features in TCGA cohort.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/8438a562b143cd4784f32ef5.png"},{"id":54162361,"identity":"86ff2cf4-fbcf-447a-b7df-144ca5b70513","added_by":"auto","created_at":"2024-04-05 13:09:17","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":842225,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTRGS correlation with immune phenotype and ICIs response predictors in CRC.\u003c/strong\u003e (A) Alluvial plot depicting the distribution of Immune hot and Immune cold phenotypes across different risk subtypes and survival statuses. (B, C) Correlation analysis of TRGS with immune checkpoint blockade (ICB) related pathways (B) and individual steps of the cancer-immune cycle (C). (D-J) Distribution of IPS, IPS_PD1_blocker, IPS_CTLA4_blocker, and IPS_CTLA4_PD1_blocker. (H-I) Distribution of TIDE, T cell exclusion score, and T cell dysfunction score. (*\u003cem\u003eP\u003c/em\u003e \u0026lt;0.05, ** \u003cem\u003eP \u003c/em\u003e\u0026lt;0.01, *** \u003cem\u003eP \u003c/em\u003e\u0026lt;0.001)\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/40d95f17fb8a09b91871a7b8.png"},{"id":54162356,"identity":"de9cf53f-c5af-4f4f-aaac-5a61c71045a7","added_by":"auto","created_at":"2024-04-05 13:09:17","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":321522,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTRGS predicts the sensitivity of fluorouracil-based ACT.\u003c/strong\u003e (A, B) Kaplan-Meier curves comparing patients who received ACT with those who did not in TCGA (A) and GSE39582 (B) cohorts. (C, D) The predicted IC50 values of chemotherapy (C) and targeted therapy (D) drugs in different risk groups in the TCGA cohort. (E-H) Distribution of TRGS between responders and non-responders of fluorouracil-based ACT in TCGA (E), GSE28702 (F), GSE19860 (G), and GSE45404 (H) cohorts. (*\u003cem\u003eP\u003c/em\u003e \u0026lt;0.05, ** \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01, *** \u003cem\u003eP\u003c/em\u003e \u0026lt;0.001)\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/075d3bd97c1bfd2a6a2b4922.png"},{"id":63820896,"identity":"0f810f9b-5a4c-4fc0-9b46-dbce943718b2","added_by":"auto","created_at":"2024-09-02 16:09:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6487594,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/194f6695-997d-4a6c-87bb-3648227e4650.pdf"},{"id":54162348,"identity":"e8623a88-c710-410e-b689-9770468ff123","added_by":"auto","created_at":"2024-04-05 13:09:13","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":11340312,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigures.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/efb66e55985daef8b5e38a47.pdf"},{"id":54162359,"identity":"046a2957-2ccb-415a-af9e-a255f96f7a5f","added_by":"auto","created_at":"2024-04-05 13:09:17","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":43867,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/9d080ea2464bedb436d6fe60.xlsx"},{"id":54162349,"identity":"f73783c9-9db0-4a58-b9f2-7fad65790f0d","added_by":"auto","created_at":"2024-04-05 13:09:13","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":153494,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/f7188bcd0b033905767a88e1.xlsx"},{"id":54162355,"identity":"b7c81fed-dcf5-4261-ab88-34948656704a","added_by":"auto","created_at":"2024-04-05 13:09:17","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":13704,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementrayFigurelegends.docx","url":"https://assets-eu.researchsquare.com/files/rs-4156323/v1/052d9ba67a7bba0e4f368109.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Integrated single-cell and bulk RNA-seq analysis reveals a novel T-cell signature for prognosis and treatment response in colorectal cancer","fulltext":[{"header":"Introduction","content":"\u003cp\u003eColorectal cancer (CRC) stands as the third most prevalent malignancy globally, ranking second in mortality\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, thereby posing substantial therapeutic challenges and an economic burden. Despite advancements in early CRC screening leading to improved 5-year survival rates, a subset of patients still presents at stage IV, resulting in a dismal prognosis\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Immune checkpoint inhibitors (ICIs) have demonstrated remarkable therapeutic efficacy in diverse solid tumors, establishing them as the primary treatment for microsatellite instability-high (MSI-H) metastatic CRC\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. However, given that this patient subset constitutes less than 5% of all CRC cases\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, there is a critical need to identify more efficient biomarkers to guide the application of ICIs in colorectal cancer.\u003c/p\u003e \u003cp\u003eThe tumor microenvironment (TME) plays an important role in tumors, mainly in terms of tumor progression, angiogenesis, tumor metastasis, and influencing treatment outcome \u003csup\u003e\u003cspan additionalcitationids=\"CR7 CR8\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. TME encompasses diverse cellular components such as immune cells, stromal cells, and non-cellular components including extracellular matrix molecules (ECM) and soluble signaling molecules\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Among these, T cells, including CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;T cells, represent the primary immune cells exerting anti-tumor activity\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. After antigen-presenting cells present tumor cell antigens to T cells, activating them, the activated T cells infiltrate the tumor vicinity to exert cytotoxic effects on tumor cells\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. However, tumor cells can evade immune surveillance through various mechanisms, such as HLA-I loss\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, expression of immune checkpoint molecules\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, or inhibitory effects from other cells in the microenvironment\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Although more studies are focusing on a specific single T cell, few studies have been reported on overall T cells infiltration. A comprehensive understanding of the alterations in T cells within the tumor microenvironment is crucial for devising future personalized therapeutic strategies.\u003c/p\u003e \u003cp\u003eThis study utilized bioinformatics methodologies to integrate scRNA-seq data and transcriptome sequencing data from CRC samples. We performed unsupervised consensus clustering of CRC patients based on the overall infiltration levels of characteristic T cells in the tumor microenvironment instead of focusing on a single T cell\u0026rsquo;s signature genes. Through this approach, a novel T-cell-related prognostic risk model, termed TRGS, was developed and subsequently validated. This model holds significant promise for elucidating the intricate role of T cells within the tumor immune microenvironment of CRC. Notably, when compared with existing prognostic models, TRGS demonstrates superior efficacy in prognostic prediction. Moreover, TRGS exhibits the capacity to predict treatment outcomes, encompassing responses to both immunotherapy and chemotherapy among CRC patients. Additionally, it may identify novel therapeutic targets to improve the effectiveness of ICIs.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eData download and processing.\u003c/strong\u003e For the TCGA-CRC cohort, we retrieved normalized RNA sequencing data (expressed as transcripts per million, TPM) for 549 colorectal cancer (CRC) patients from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). We specifically included patients with a survival time of \u0026ge;30 days who did not undergo neoadjuvant therapy. The normalized matrix files for the GSE39582 and GSE38832 cohorts from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/) were utilized as independent external cohorts for the validation of our risk model. Similarly, normalized matrix files for the GSE28702, GSE19860, and GSE45404 cohorts from the GEO database served as independent external cohorts for predicting chemotherapy outcomes. Additionally, single-cell sequencing data were obtained from the GSE132465 dataset, including 23 primary colorectal cancer and 10 matched normal mucosa samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003escRNA-Seq data analysis.\u003c/strong\u003e The \u0026ldquo;Seurat\u0026rdquo; R package\u003csup\u003e17\u003c/sup\u003e was used to perform unsupervised clustering of single cells using the read count matrix as input. We annotated the cells according to the original annotation\u003csup\u003e18\u003c/sup\u003e. Cell populations originally annotated as T-cells were extracted and normalized using the SCTransform method\u003csup\u003e19\u003c/sup\u003e with resolution choice 1.2 for re-clustering. To find marker genes in various clusters, the \u0026ldquo;Find All Markers\u0026rdquo; tool was used with |Log2FC| and min. pct cutoff values set to 0.25 and 0.25, respectively.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of immune class by consensus clustering. \u003c/strong\u003eThe \u0026ldquo;ConsensusClusterPlus\u0026rdquo; R package\u003csup\u003e20\u003c/sup\u003e was used to perform the unsupervised consensus clustering analysis. We used 80% of the items for subsampling and divided each subsample into groups by the k-means algorithm, repeating this clustering process 1000 times. Then, the cumulative distribution function (CDF) curve and consensus matrix are used to identify the optimal k-value. Based on the optimal k-value, patients were clustered into sub-clusters.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTumor immune microenvironment evaluation. \u003c/strong\u003eTheMCP-counter algorithm\u003csup\u003e21\u003c/sup\u003e based on the \u0026ldquo;MCPcounter\u0026rdquo; R package was used to assess the 26 kinds of T cell clusters identified from the scRNA-seq analysis. The ESTIMATE algorithm\u003csup\u003e22\u003c/sup\u003e tool was used to calculate stromal, immune, and estimated scores. Single sample gene set enrichment analysis (ssGSEA) \u003csup\u003e23\u003c/sup\u003e based on the \u0026ldquo;GSVA\u0026rdquo; R package was used to quantify 28 kinds of immune cell infiltration scores of the TCGA cohort.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWeighted correlation network analysis (WGCNA). \u003c/strong\u003eCoexpression protein-coding RNA networks of TCGA-CRC were generated using the WGCNA package\u003csup\u003e24\u003c/sup\u003e. An appropriate soft threshold \u0026beta; was calculated to meet the criteria for the scale-free network. Then, the adjacency was transformed into a topological overlap matrix (TOM), and genes with similar expression patterns were divided into the same module. Finally, genes from modules that had high correlation coefficients with T-cell relative immune status were identified for subsequent analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eScreening and functional annotation of T cell-related genes. \u003c/strong\u003eDEGs were screened by the limma R package. Adj. P \u0026lt; 0.05 and |log\u003csub\u003e2\u003c/sub\u003eFC| \u0026gt; 0.6 were set as the selection conditions for screening the downregulated and upregulated DEGs. T cell-related genes were obtained by overlapping DEGs and WGCNA. Enrichment analysis in Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) were based on the \u0026ldquo;Clusterprofiler\u0026rdquo; R package\u003csup\u003e25\u003c/sup\u003e, where \u0026ldquo;adjusted p \u0026lt; 0.05\u0026rdquo; was considered significant. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConstruction and validation of T cell‑related genes prognosis signature (TRGS). \u003c/strong\u003eThe TRGS was developed based on the TCGA-CRC cohort. First, univariate Cox regression analysis was utilized to screen T cell-related genes that were associated with OS using the \u0026ldquo;survival\u0026rdquo; R package (P \u0026lt; 0.05). Next, we performed machine learning algorithms including the least absolute shrinkage and selection operator (LASSO) regression analysis and stepwise Cox algorithm to improve the accuracy and reliability of prognostic signature using the \u0026ldquo;glmnet\u0026rdquo; and \u0026ldquo;StepReg\u0026rdquo; R packages, respectively. the prognostic signature was established as follows: \u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"332\" height=\"52\"\u003e\u003c/p\u003e\n\u003cp\u003ewhere Coef referred to the coefficient of gene i and Ni represented the expression value of gene i.\u003c/p\u003e\n\u003cp\u003eTo further validate the predictive power of the prognostic signature, we performed Kaplan\u0026ndash;Meier survival analysis in the TCGA-CRC, GSE39582, and GSE38832 cohorts, where the cutoff value was set to the median TRGS. Time-dependent receiver operating characteristic curves (ROC) were plotted using the \u0026ldquo;timeROC\u0026rdquo; R package to predict OS at 1, 3, and 5 years in the training dataset (TCGA-CRC); 1, 2, and 3 years in GSE38832 cohort, and 3, 7, 10 years in GSE39582 cohort. Finally, we further analyzed the relationship between the prognostic signature and clinical characteristics (including age, gender, and stage) to clarify the independent prognostic value of TRGS using multivariate Cox regression analyses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDevelopment and evaluation of the nomogram.\u003c/strong\u003e We performed the univariate and multivariate Cox regression analyses for clinical parameters and TRGS. In the multivariate Cox model, variables with P \u0026lt; 0.05 were included in the construction of the nomogram by using the \u0026ldquo;rms\u0026rdquo; R package. The calibration analysis was used to assess the prognostic accuracy of the nomogram model. Decision curve analysis (DCA) curves were used to evaluate the net benefit of the nomogram.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImmune checkpoints inhibitors sensitivity prediction.\u003c/strong\u003e We utilized Immunophenoscore (IPS) obtained from The Cancer Immunome Atlas (TCIA) website (tcia.at/home)\u003csup\u003e26\u003c/sup\u003e to predict the response to ICIs. We also utilized the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm through an online website (http://tide.dfci.harvard.edu) \u003csup\u003e27\u003c/sup\u003e to predict patients\u0026rsquo; ICIs responses. Higher TIDE prediction scores were generally associated with worse ICIs responses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnticancer drug sensitivity prediction. \u003c/strong\u003eThe half-maximal inhibitory concentration (IC50) of common chemotherapeutic and targeted therapeutic drugs was estimated by using the \u0026quot;pRRophetic\u0026quot; R package\u003csup\u003e28\u003c/sup\u003e. GSE28702, GSE19860, and GSE45404 were transcriptome sequencing data from patients who received adjuvant chemotherapy (ACT), which were utilized as external validation cohorts for predicting chemotherapy drug sensitivity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis. \u003c/strong\u003e\u003cstrong\u003e\u003c/strong\u003eAll statistical analyses were performed using R software (version 4.3.1). Wilcoxon test was used to compare the differences between groups. The log-rank test was used to compare Kaplan\u0026ndash;Meier survival curves. Univariate and multivariate Cox analyses were performed to establish independent prognostic factors. All P values were two-sided and less than 0.05 were considered statistically significant.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eScRNA-Seq analysis of CRC samples.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e The flowchart outlining our study is presented in Supplementary Fig. 1. We obtained 10x single-cell RNA sequencing (scRNA-seq) data from the GSE132465 dataset, which comprised 33 colorectal cancer (CRC) samples, including 23 primary colorectal cancer and 10 matched normal mucosa samples. The 23,115 T cells extracted from the original dataset were re-clustered. Based on the clustree results (Supplementary Fig. 2A), a resolution of 1.2 was chosen, leading to the identification of 26 distinct cell subgroups through hierarchical clustering utilizing UMAP analysis (Fig. 1A). Annotation of the 26 clusters of T cells was performed using literature reports\u003csup\u003e29, 30\u003c/sup\u003e, CellMarker 2.0\u003csup\u003e31\u003c/sup\u003e, and T cellmap\u003csup\u003e32\u003c/sup\u003e. We presented the characteristic genes (Fig. 1B) as well as functional annotations for each cell cluster (Fig. 1C, D). Differential expression genes (DEGs) of each cluster were provided in Supplementary Table S1 and shown in Supplementary Fig. 2B. Furthermore, we explored the differences between these cells in cancer and adjacent tissues (Fig. 1E). Subsequently, the characteristic genes of each cell cluster were employed to estimate the differences in cell populations of TCGA patients using the MCP-counter algorithm. We compared normal and tumor groups to illustrate the validity of the MCP-counter algorithm (Fig. 1F).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of T cell infiltration subgroups.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the assessment of T cell infiltration using the MCP-counter package, we conducted a consensus cluster analysis, initially dividing all CRC samples into k clusters (k = 2\u0026ndash;9). The CDF curves of the consensus score matrix revealed that the optimal number of clusters was achieved when k = 2 (Fig. 2A, B). The two consensus clusters (C1 and C2) exhibited significant differences in the distribution of 26 T cell clusters, with C1 displaying substantially higher T cell infiltration abundance compared to C2 (Fig. 2C). Consequently, we categorized C1 as \u0026quot;immune-hot\u0026quot; tumors and C2 as \u0026quot;immune-cold\u0026quot; tumors. To ensure the robustness of the two consensus clusters and mitigate potential algorithmic biases, ssGSEA was employed to validate the disparity in immune cell infiltration between C1 and C2, revealing a higher overall immune cell infiltration abundance in C1 (Fig. 2D). Additionally, the ESTIMATE algorithm was utilized to illustrate that immune-hot tumors exhibited a higher immune score compared to immune-cold tumors (Fig. 2E).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of candidate T cell-related genes.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify T cell-related genes associated with immune cell infiltration, we initially conducted WGCNA. In this procedure, we set the soft threshold \u0026beta; to 6, resulting in a scale-free topology fit index of 0.85 (Supplementary Fig. 3A). Subsequently, we utilized the \u0026ldquo;merged dynamics\u0026rdquo; algorithm to derive fourteen modules (Fig. 3A). The heatmap depicted the eigengene adjacency of these modules (Fig. 3B). Through assessment of the correlation coefficient and P-value, we observed that the blue module exhibited the strongest correlation with the previously identified immune status (Fig. 3C), achieving a correlation coefficient of 0.78 between gene significance (GS) and module membership (MM) (Fig. 3D). Concurrently, we conducted differential expression analysis of the TCGA-CRC cohort using the \u0026ldquo;limma\u0026rdquo; package, identifying a total of 686 differentially expressed genes (DEGs). By intersecting these results with those from WGCNA, we extracted 369 overlapping genes deemed as T cell-related genes for subsequent analyses (Fig. 3E). GO and KEGG enrichment analyses were conducted to provide a comprehensive functional overview of the overlapping genes (Fig. 3F, G and Supplementary Table S2). The results revealed predominant enrichment in biological processes (BP) such as leukocyte-mediated immunity, leukocyte cell-cell adhesion, and regulation of T-cell activation. In terms of cellular composition (CC), enrichment was observed in areas including the external side of the plasma membrane, MHC protein complex, and MHC class II protein complex. Additionally, molecular function (MF) analysis highlighted immune receptor activity, chemokine receptor binding, and MHC protein complex binding as significantly enriched categories. Furthermore, KEGG pathway analysis indicated significant enrichment in pathways such as cytokine-cytokine receptor interaction, cell adhesion molecules, and chemokine signaling, all of which were closely associated with anti-cancer immunity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eT cell signature establishment and external validation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the prognostic impact of the identified genes on CRC, we initially identified 12 genes associated with overall survival (OS) through univariate Cox analysis (Supplementary Fig. 3B). Subsequently, to prevent overfitting and exclude co-expressed T cell-related genes (Supplementary Fig. 3C), we employed lasso regression analysis to construct a predictive prognostic model comprising T cell-related genes. The optimal \u0026lambda; value was determined based on the minimum partial likelihood deviance achieved via the leave-one-out cross-validation (LOOCV) framework (Fig. 4A). Ten genes with nonzero Lasso coefficients were further subjected to stepwise Cox proportional hazards regression, resulting in the identification of a final set of 6 genes (Fig. 4B), and we assessed their impact on prognosis at the same time (Supplementary Fig. 3D-I).\u003c/p\u003e\n\u003cp\u003eA linear prediction model was then developed using the weighted regression coefficients of the 6 T cell-related genes, represented as TRGS = (0.325 x ADAM8 exp) + (0.282 x ZNF385A exp) + (0.201 x G0S2 exp) + (-0.195 x TAP1 exp) + (-0.216 x MMP1 exp) + (-0.347 x CCL22 exp). Based on the median of TRGS, all patients were stratified into high- and low-risk groups. Notably, patients in the high-risk group exhibited significantly poorer OS compared to those in the low-risk group (Fig. 4C, E, G). The discriminatory ability of TRGS was assessed using receiver operating characteristic (ROC) analysis, with 1-, 3-, and 5-year areas under the curve (AUCs) of 0.725, 0.710, and 0.700 in TCGA-CRC (Fig. 4D); 3-, 7-, and 10-year AUCs of 0.620, 0.639, and 0.684 in the GSE39582 cohort (Fig. 4F); and 1-, 2-, 3-year AUCs of 0.656, 0.696, and 0.616 in GSE38832 cohort (Fig. 4H). Furthermore, to evaluate the prognostic capabilities of TRGS, we obtained 5 previously published prognostic models developed using immune-related genes\u003csup\u003e33-37\u003c/sup\u003e. Subsequently, we conducted a comparative analysis to assess the predictive performance of TRGS in terms of prognosis. Remarkably, our findings revealed that TRGS exhibited superior performance, as evidenced by higher AUC values in time-dependent ROC analysis when compared to the 5 existing prognostic models in TCGA, GSE39582, and GSE38832 cohorts. (Supplementary Fig. 4A-O).\u003c/p\u003e\n\u003cp\u003eFinally, we developed a nomogram incorporating TRGS and clinicopathological features to assess the clinical significance of the predictive model (Fig. 4I). Calibration curves demonstrated good agreement between the predicted and actual probabilities of survival at 1, 2, and 3 years (Fig. 4J). Decision curve analysis (DCA) revealed that the nomogram model provided a greater net benefit in predicting 5-year OS (Fig. 4K). Meanwhile, multifactorial Cox regression analysis confirmed TRGS as an independent prognostic risk factor for CRC (Fig. 4L and Supplementary Fig. 4P).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTRGS evaluation for immunotherapeutic response prediction.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the predictive capacity of TRGS as a biomarker for ICIs response, we initially investigated the relationship between low- and high-risk groups and the immune-hot and immune-cold phenotypes as defined in Fig. 2. Patients classified in the immune-cold group belonged to the high-risk category, whereas those in the low-risk group exhibited an immune-hot phenotype (Fig. 5A). Subsequently, we examined the association between TRGS and immune checkpoint blocker (ICB)-related pathways, as well as the seven steps in the Cancer-Immunity cycle. Our findings revealed a negative correlation between TRGS and ICB-related pathways and steps in the Cancer-Immunity cycle, suggesting that low-risk patients may possess a richer tumor immune microenvironment, potentially leading to a higher response rate to ICIs (Fig. 5B, C). Furthermore, we analyzed the distribution of IPS and TIDE scores across different risk groups. Our results indicated that the low-risk group exhibited higher IPS scores (Fig. 5D-G), whereas the high-risk group demonstrated elevated TIDE scores (Fig. 5H), Exclusion scores (Fig. 5I), and Dysfunction scores (Fig. 5J). These results were consistent with previous findings, suggesting that low-risk patients may exhibit increased sensitivity to ICIs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePredictive value of fluorouracil-based ACT benefits.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe expanded our investigation to assess the predictive capability of TRGS in quantifying the benefits of fluorouracil-based adjuvant chemotherapy (ACT). Survival analysis among patients stratified into low- and high-risk groups, both receiving and not receiving ACT, revealed that individuals in the low-risk category who underwent therapy exhibited the most favorable survival outcomes. Conversely, among patients categorized as high-risk, no substantial disparity in survival was observed between those who received ACT and those who did not (Fig. 6A, B). This suggests that low TRGS may serve as a predictor for a more favorable response to ACT.\u003c/p\u003e\n\u003cp\u003eSubsequently, we calculated the IC50 values of common chemotherapy and targeted drugs for CRC to further explore the drug sensitivity between low- and high-risk groups. Comparison of the IC50 values between the two risk groups indicated that the low-risk group exhibited higher sensitivity to fluorouracil, gemcitabine, paclitaxel, and vincristine (Fig. 6C). However, patients in the high and low-risk groups showed no significant difference in the sensitivity to targeted therapy (Fig. 6D). These findings were validated in three datasets treated with fluorouracil-based ACT. Responders demonstrated a significantly lower TRGS score than non-responders in TCGA, GSE28702, GSE19860, and GSE45404 (Fig. 6E-H).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eCurrently, conventional treatment strategies for colorectal cancer primarily involve surgery combined with adjuvant chemotherapy or targeted therapy, yet these approaches have not yielded significant efficacy\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Moreover, stage IV patients exhibit a high likelihood of chemotherapy resistance\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, potentially leading to situations of overtreatment or undertreatment within the current staging-based treatment paradigm. Finding biomarkers to predict chemotherapy outcomes is of utmost urgency. Furthermore, immunotherapy, as a novel and effective treatment modality, remains underutilized in colorectal cancer, necessitating more effective biomarkers to guide the application of ICIs in colorectal cancer\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. T cells, being the most abundant immune cells infiltrating the tumor microenvironment, play a crucial role in anti-tumor immunity. A comprehensive understanding of the T cell population\u0026apos;s status within the colorectal cancer tumor microenvironment contributes to the identification of more effective immunotherapeutic strategies.\u003c/p\u003e\n\u003cp\u003eIn this study, we first identified colorectal cancer-specific T cell signature genes through scRNA-seq data. Subsequently, we evaluated the levels of T-cell infiltration in the TCGA-CRC cohort using these genes. Next, we constructed an immune classification of colorectal cancer based on the infiltration levels of 26 types of T cells instead of single T cell signature genes and identified T cell-related genes associated with the immunogenic tumor phenotype through WGCNA. We then conducted univariate COX regression to identify prognosis-related genes, followed by the selection of the six most relevant genes for constructing a subsequent prognostic risk model using the Lasso and StepCox machine learning algorithms. Among these six genes, high expression of ADAM8, ZNF385A, and G0S2 was associated with poor prognosis, while high expression of TAP1, MMP1, and CCL22 was associated with favorable prognosis. Interestingly, these six genes have previously been found to be associated with immune cell infiltration in earlier studies, consistent with our findings. Studies have reported that downregulating ADAM8 in colorectal cancer cells enhances the cytotoxic effect of tumor-infiltrating lymphocytes\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. ZNF385A has been identified as a gene associated with lymphatic vessel formation in CRC, influencing immune cell infiltration, and is also associated with an immunosuppressive phenotype in HCC\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. G0S2 is linked to evasion from cytotoxic T lymphocytes and knocking down G0S2 significantly inhibits proliferation and migration of CRC cell lines\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. TAP1, a transporter associated with antigen processing protein, plays a crucial role in antigen presentation\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e, and its downregulation in colorectal cancer is associated with low infiltration of various lymphocytes\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. MMP1, one of the markers of epithelial-mesenchymal transition, is associated with metastasis and angiogenesis in colorectal cancer\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. In refractory HCC treated with Transcatheter arterial chemoembolization (TACE), a prognostic model incorporating MMP1 is related to immune cell infiltration\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. Additionally, MMP1 may be one of the targets to enhance the effectiveness of immunotherapy in PDAC\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. CCL22 is a chemokine that not only recruits monocytes but also regulates the infiltration of Th1 and Treg cells in colorectal cancer\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eSubsequently, we constructed a risk model of T cell-related genes (TRGS) using these six genes, and patients were stratified into high-risk and low-risk groups based on the median TRGS value. Compared to patients in the high-risk group, those in the low-risk group exhibited significantly prolonged overall survival. Furthermore, TRGS demonstrated superior predictive efficacy in prognosis compared to other existing immune-related models. Meanwhile, the analysis results indicated that TRGS could serve as an independent prognostic factor for colorectal cancer. Additionally, we utilized IPS and TIDE scores to predict the response of patients in the high- and low-risk groups to ICIs treatment. The results revealed a statistically significant increase in IPS scores among patients in the low-risk group, whereas those in the high-risk group demonstrated higher Dysfunction and Exclusion scores, suggesting a greater likelihood of immune evasion in this group. These findings suggest that ICIs therapy may be more effective for patients in the low-risk group.\u003c/p\u003e\n\u003cp\u003ePrevious studies have shown that 5-FU treatment can induce a reduction in tumor-infiltrating myeloid-derived suppressor cells (MDSCs) and promote T cell-mediated anti-tumor immunity\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Additionally, in metastatic colorectal cancer patients receiving FOLFOX combined with bevacizumab, the accumulation of Th17 cells was associated with poor prognosis\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. Similarly, in lung cancer, the combination treatment of oxaliplatin and cyclophosphamide led to an increased ratio of CD8 T cells to Treg cells within the tumor, resulting in significant inhibition of tumor growth\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. These findings collectively indicate the existence of interactions between T cells and chemotherapy within the tumor microenvironment. Therefore, we subsequently compared the IC50 values of common chemotherapy drugs to determine the sensitivity of high- and low-risk groups to chemotherapy drugs. The results showed that the IC50 values of common chemotherapy drugs including fluorouracil, gemcitabine, and paclitaxel were decreased in the low-risk group, indicating higher sensitivity. Furthermore, we used the GSE28702 and GSE19860 datasets to validate patient responses to fluorouracil-based ACT. The results successfully validated our findings, demonstrating that patients responsive to fluorouracil-based ACT exhibited lower risk scores.\u003c/p\u003e\n\u003cp\u003eWhile our study has been successfully validated in external cohorts, several limitations warrant consideration. Firstly, our study relied on publicly available databases, and both our training cohort (TCGA) and validation cohorts (GEO) were retrospective. Hence, prospective studies are warranted to validate our conclusions rigorously. Secondly, there is still a shortage of sequencing databases containing data from CRC patients treated with ICIs, as well as a deficiency of large clinical cohorts of patients undergoing ICIs treatment. Therefore, our results still need validation in a large clinical cohort in the future. Thirdly, further research is needed to identify hub genes and elucidate the potential regulatory mechanisms to facilitate the discovery of targets that enhance the efficacy of ICIs and chemotherapy.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn summary, we pioneered a novel approach to categorize the colorectal cancer immune microenvironment utilizing single-cell data. Rather than focusing on individual T-cell signature genes, we employed the infiltration levels of the comprehensive colorectal cancer signature T-cells. This innovative strategy led to the construction of a TRGS comprising ADAM8, ZNF385A, G0S2, TAP1, MMP1, and CCL22. Our findings underscore the superiority of TRGS in prognostic prediction compared to existing immune-related prognostic models. Moreover, TRGS exhibits predictive capabilities not only for ICIs treatment but also for chemotherapy efficacy, presenting novel perspectives for the stratified diagnosis and treatment of colorectal cancer patients.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors thank researchers and contributors of the TCGA and GEO databases for their valuable efforts and contributions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eP.C. and Z.G.B. conceptualized and designed the study. P.C. was responsible for data collection, analysis, and initial manuscript drafting. H.B.W. reviewed the statistical analysis and performed manuscript editing. Z.G.B. oversaw the project, contributed to manuscript editing, provided critical feedback, and supervised the study. All authors contributed to the critical revision of the manuscript and approved the final version for publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe raw data utilized in our study were obtained from the TCGA dataset (https://portal.gdc.cancer.gov/) and the GEO dataset (https://www.ncbi.nlm.nih.gov/). Specifically, TCGA accession numbers for our study are TCGA-COAD and TCGA-READ, while GEO accession numbers are GSE39582, GSE38832, GSE132465, GSE28702, GSE19860, and GSE45404. The datasets analyzed and R language codes employed in the current study are accessible upon reasonable request from the corresponding author.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCompeting Interests: The authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eXi, Y. \u0026amp; Xu, P. Global colorectal cancer burden in 2020 and projections to 2040. \u003cem\u003eTransl Oncol\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e,101174 (2021).\u003c/li\u003e\n\u003cli\u003eSung, H.\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 Cancer J Clin\u003c/em\u003e \u003cstrong\u003e71\u003c/strong\u003e,209-249 (2021).\u003c/li\u003e\n\u003cli\u003eSiegel, R.L.\u003cem\u003e et al.\u003c/em\u003e Colorectal cancer statistics, 2020. \u003cem\u003eCA Cancer J Clin\u003c/em\u003e \u003cstrong\u003e70\u003c/strong\u003e,145-164 (2020).\u003c/li\u003e\n\u003cli\u003eAndre, T.\u003cem\u003e et al.\u003c/em\u003e Pembrolizumab in Microsatellite-Instability-High Advanced Colorectal Cancer. \u003cem\u003eN Engl J Med\u003c/em\u003e \u003cstrong\u003e383\u003c/strong\u003e,2207-2218 (2020).\u003c/li\u003e\n\u003cli\u003eKoopman, M.\u003cem\u003e et al.\u003c/em\u003e Deficient mismatch repair system in patients with sporadic advanced colorectal cancer. \u003cem\u003eBr J Cancer\u003c/em\u003e \u003cstrong\u003e100\u003c/strong\u003e,266-273 (2009).\u003c/li\u003e\n\u003cli\u003eQuail, D.F. \u0026amp; Joyce, J.A. Microenvironmental regulation of tumor progression and metastasis. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e,1423-1437 (2013).\u003c/li\u003e\n\u003cli\u003eAngelova, M.\u003cem\u003e et al.\u003c/em\u003e Evolution of Metastases in Space and Time under Immune Selection. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e175\u003c/strong\u003e,751-765 e716 (2018).\u003c/li\u003e\n\u003cli\u003eJin, M.Z. \u0026amp; Jin, W.L. The updated landscape of tumor microenvironment and drug repurposing. \u003cem\u003eSignal Transduct Target Ther\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e,166 (2020).\u003c/li\u003e\n\u003cli\u003eHernandez-Camarero, P., Lopez-Ruiz, E., Marchal, J.A. \u0026amp; Peran, M. Cancer: a mirrored room between tumor bulk and tumor microenvironment. \u003cem\u003eJ Exp Clin Cancer Res\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e,217 (2021).\u003c/li\u003e\n\u003cli\u003eBelli, C.\u003cem\u003e et al.\u003c/em\u003e Targeting Cellular Components of the Tumor Microenvironment in Solid Malignancies. \u003cem\u003eCancers (Basel)\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e (2022).\u003c/li\u003e\n\u003cli\u003ePelka, K.\u003cem\u003e et al.\u003c/em\u003e Spatially organized multicellular immune hubs in human colorectal cancer. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e184\u003c/strong\u003e,4734-4752 e4720 (2021).\u003c/li\u003e\n\u003cli\u003eMellman, I., Chen, D.S., Powles, T. \u0026amp; Turley, S.J. The cancer-immunity cycle: Indication, genotype, and immunotype. \u003cem\u003eImmunity\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e,2188-2205 (2023).\u003c/li\u003e\n\u003cli\u003eAnderson, P., Aptsiauri, N., Ruiz-Cabello, F. \u0026amp; Garrido, F. HLA class I loss in colorectal cancer: implications for immune escape and immunotherapy. \u003cem\u003eCell Mol Immunol\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e,556-565 (2021).\u003c/li\u003e\n\u003cli\u003eChen, Q.\u003cem\u003e et al.\u003c/em\u003e Demethylase JMJD2D induces PD-L1 expression to promote colorectal cancer immune escape by enhancing IFNGR1-STAT3-IRF1 signaling. \u003cem\u003eOncogene\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e,1421-1433 (2022).\u003c/li\u003e\n\u003cli\u003eXue, R.\u003cem\u003e et al.\u003c/em\u003e Liver tumour immune microenvironment subtypes and neutrophil heterogeneity. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e612\u003c/strong\u003e,141-147 (2022).\u003c/li\u003e\n\u003cli\u003eFridman, W.H.\u003cem\u003e et al.\u003c/em\u003e B cells and tertiary lymphoid structures as determinants of tumour immune contexture and clinical outcome. \u003cem\u003eNat Rev Clin Oncol\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e,441-457 (2022).\u003c/li\u003e\n\u003cli\u003eHao, Y.\u003cem\u003e et al.\u003c/em\u003e Dictionary learning for integrative, multimodal and scalable single-cell analysis. \u003cem\u003eNat Biotechnol\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e,293-304 (2024).\u003c/li\u003e\n\u003cli\u003eLee, H.O.\u003cem\u003e et al.\u003c/em\u003e Lineage-dependent gene expression programs influence the immune landscape of colorectal cancer. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e,594-603 (2020).\u003c/li\u003e\n\u003cli\u003eLause, J., Berens, P. \u0026amp; Kobak, D. Analytic Pearson residuals for normalization of single-cell RNA-seq UMI data. \u003cem\u003eGenome Biol\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e,258 (2021).\u003c/li\u003e\n\u003cli\u003eWilkerson, M.D. \u0026amp; Hayes, D.N. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e,1572-1573 (2010).\u003c/li\u003e\n\u003cli\u003eBecht, E.\u003cem\u003e et al.\u003c/em\u003e Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. \u003cem\u003eGenome Biol\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e,218 (2016).\u003c/li\u003e\n\u003cli\u003eYoshihara, K.\u003cem\u003e et al.\u003c/em\u003e Inferring tumour purity and stromal and immune cell admixture from expression data. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e,2612 (2013).\u003c/li\u003e\n\u003cli\u003eBarbie, D.A.\u003cem\u003e et al.\u003c/em\u003e Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e462\u003c/strong\u003e,108-112 (2009).\u003c/li\u003e\n\u003cli\u003eLangfelder, P. \u0026amp; Horvath, S. WGCNA: an R package for weighted correlation network analysis. \u003cem\u003eBMC Bioinformatics\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e,559 (2008).\u003c/li\u003e\n\u003cli\u003eYu, G., Wang, L.G., Han, Y. \u0026amp; He, Q.Y. clusterProfiler: an R package for comparing biological themes among gene clusters. \u003cem\u003eOMICS\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e,284-287 (2012).\u003c/li\u003e\n\u003cli\u003eCharoentong, P.\u003cem\u003e et al.\u003c/em\u003e Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. \u003cem\u003eCell Rep\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e,248-262 (2017).\u003c/li\u003e\n\u003cli\u003eJiang, P.\u003cem\u003e et al.\u003c/em\u003e Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e,1550-1558 (2018).\u003c/li\u003e\n\u003cli\u003eGeeleher, P., Cox, N. \u0026amp; Huang, R.S. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. \u003cem\u003ePLoS One\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e,e107468 (2014).\u003c/li\u003e\n\u003cli\u003eZhang, L.\u003cem\u003e et al.\u003c/em\u003e Lineage tracking reveals dynamic relationships of T cells in colorectal cancer. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e564\u003c/strong\u003e,268-272 (2018).\u003c/li\u003e\n\u003cli\u003eZheng, L.\u003cem\u003e et al.\u003c/em\u003e Pan-cancer single-cell landscape of tumor-infiltrating T cells. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e374\u003c/strong\u003e,abe6474 (2021).\u003c/li\u003e\n\u003cli\u003eHu, C.\u003cem\u003e et al.\u003c/em\u003e CellMarker 2.0: an updated database of manually curated cell markers in human/mouse and web tools based on scRNA-seq data. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e,D870-D876 (2023).\u003c/li\u003e\n\u003cli\u003eChu, Y.\u003cem\u003e et al.\u003c/em\u003e Pan-cancer T cell atlas links a cellular stress response state to immunotherapy resistance. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e,1550-1562 (2023).\u003c/li\u003e\n\u003cli\u003eTu, W.\u003cem\u003e et al.\u003c/em\u003e Elucidating the role of T-cell exhaustion-related genes in colorectal cancer: a single-cell bioinformatics perspective. \u003cem\u003eFunct Integr Genomics\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e,259 (2023).\u003c/li\u003e\n\u003cli\u003eChen, J.\u003cem\u003e et al.\u003c/em\u003e Comprehensive analysis and experimental verification of the mechanism of action of T cell-mediated tumor-killing related genes in Colon adenocarcinoma. \u003cem\u003eTransl Oncol\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e,101918 (2024).\u003c/li\u003e\n\u003cli\u003eLiao, K.\u003cem\u003e et al.\u003c/em\u003e Identification of signature of tumor-infiltrating CD8 T lymphocytes in prognosis and immunotherapy of colon cancer by machine learning. \u003cem\u003eClin Immunol\u003c/em\u003e \u003cstrong\u003e257\u003c/strong\u003e,109811 (2023).\u003c/li\u003e\n\u003cli\u003eLi, Q., Chu, Y., Yao, Y. \u0026amp; Song, Q. A Treg-related riskscore model may improve the prognosis evaluation of colorectal cancer. \u003cem\u003eJ Gene Med\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e,e3668 (2024).\u003c/li\u003e\n\u003cli\u003eLi, X.\u003cem\u003e et al.\u003c/em\u003e Identification of an Immune Signature Predicting Prognosis Risk and Lymphocyte Infiltration in Colon Cancer. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e,1678 (2020).\u003c/li\u003e\n\u003cli\u003eBockelman, C., Engelmann, B.E., Kaprio, T., Hansen, T.F. \u0026amp; Glimelius, B. Risk of recurrence in patients with colon cancer stage II and III: a systematic review and meta-analysis of recent literature. \u003cem\u003eActa Oncol\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e,5-16 (2015).\u003c/li\u003e\n\u003cli\u003eGanesh, K.\u003cem\u003e et al.\u003c/em\u003e Immunotherapy in colorectal cancer: rationale, challenges and potential. \u003cem\u003eNat Rev Gastroenterol Hepatol\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e,361-375 (2019).\u003c/li\u003e\n\u003cli\u003eLiao, R.\u003cem\u003e et al.\u003c/em\u003e Identification of biomarkers related to Tumor-Infiltrating Lymphocytes (TILs) infiltration with gene co-expression network in colorectal cancer. \u003cem\u003eBioengineered\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e,1676-1688 (2021).\u003c/li\u003e\n\u003cli\u003eLiu, H., Shi, H. \u0026amp; Sun, Y. Identification of a novel lymphangiogenesis signature associated with immune cell infiltration in colorectal cancer based on bioinformatics analysis. \u003cem\u003eBMC Med Genomics\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e,2 (2024).\u003c/li\u003e\n\u003cli\u003ePeng, Q.\u003cem\u003e et al.\u003c/em\u003e ZNF385A and ZNF346 Serve as Prognostic Biomarkers Associated with an Inflamed Immunosuppressive Tumor Microenvironment in Hepatocellular Carcinoma. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e (2023).\u003c/li\u003e\n\u003cli\u003eWang, X.\u003cem\u003e et al.\u003c/em\u003e Robust machine-learning based prognostic index using cytotoxic T lymphocyte evasion genes highlights potential therapeutic targets in colorectal cancer. \u003cem\u003eCancer Cell Int\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e,52 (2024).\u003c/li\u003e\n\u003cli\u003eBlees, A.\u003cem\u003e et al.\u003c/em\u003e Structure of the human MHC-I peptide-loading complex. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e551\u003c/strong\u003e,525-528 (2017).\u003c/li\u003e\n\u003cli\u003eLing, A.\u003cem\u003e et al.\u003c/em\u003e TAP1 down-regulation elicits immune escape and poor prognosis in colorectal cancer. \u003cem\u003eOncoimmunology\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e,e1356143 (2017).\u003c/li\u003e\n\u003cli\u003eYang, P.\u003cem\u003e et al.\u003c/em\u003e TCONS_00012883 promotes proliferation and metastasis via DDX3/YY1/MMP1/PI3K-AKT axis in colorectal cancer. \u003cem\u003eClin Transl Med\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e,e211 (2020).\u003c/li\u003e\n\u003cli\u003eNg, L.\u003cem\u003e et al.\u003c/em\u003e CD26 Induces Colorectal Cancer Angiogenesis and Metastasis through CAV1/MMP1 Signaling. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e (2022).\u003c/li\u003e\n\u003cli\u003eCheng, X.\u003cem\u003e et al.\u003c/em\u003e The role of hypoxia-related genes in TACE-refractory hepatocellular carcinoma: Exploration of prognosis, immunological characteristics and drug resistance based on onco-multi-OMICS approach. \u003cem\u003eFront Pharmacol\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e,1011033 (2022).\u003c/li\u003e\n\u003cli\u003eHung, H.C.\u003cem\u003e et al.\u003c/em\u003e Effect of chimeric antigen receptor T cells against protease-activated receptor 1 for treating pancreatic cancer. \u003cem\u003eBMC Med\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e,338 (2023).\u003c/li\u003e\n\u003cli\u003eCremonesi, E.\u003cem\u003e et al.\u003c/em\u003e Gut microbiota modulate T cell trafficking into human colorectal cancer. \u003cem\u003eGut\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e,1984-1994 (2018).\u003c/li\u003e\n\u003cli\u003eDong, S., Guo, X., Han, F., He, Z. \u0026amp; Wang, Y. Emerging role of natural products in cancer immunotherapy. \u003cem\u003eActa Pharm Sin B\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e,1163-1185 (2022).\u003c/li\u003e\n\u003cli\u003eVincent, J.\u003cem\u003e et al.\u003c/em\u003e 5-Fluorouracil selectively kills tumor-associated myeloid-derived suppressor cells resulting in enhanced T cell-dependent antitumor immunity. \u003cem\u003eCancer Res\u003c/em\u003e \u003cstrong\u003e70\u003c/strong\u003e,3052-3061 (2010).\u003c/li\u003e\n\u003cli\u003eLimagne, E.\u003cem\u003e et al.\u003c/em\u003e Accumulation of MDSC and Th17 Cells in Patients with Metastatic Colorectal Cancer Predicts the Efficacy of a FOLFOX-Bevacizumab Drug Treatment Regimen. \u003cem\u003eCancer Res\u003c/em\u003e \u003cstrong\u003e76\u003c/strong\u003e,5241-5252 (2016).\u003c/li\u003e\n\u003cli\u003ePfirschke, C.\u003cem\u003e et al.\u003c/em\u003e Immunogenic Chemotherapy Sensitizes Tumors to Checkpoint Blockade Therapy. \u003cem\u003eImmunity\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e,343-354 (2016).\u003c/li\u003e\n\u003c/ol\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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Colorectal cancer, T cell, Immune checkpoint inhibitors, prognostic risk model, chemotherapy","lastPublishedDoi":"10.21203/rs.3.rs-4156323/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4156323/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eColorectal cancer (CRC) is a major contributor to global morbidity and mortality, necessitating more effective therapeutic approaches. T cells, prominent in the tumor microenvironment, exert a crucial role in modulating immunotherapeutic responses and clinical outcomes in CRC. This study introduces a pioneering method for characterizing the CRC immune microenvironment using single-cell sequencing data. Unlike previous approaches, which focused on individual T-cell signature genes, we utilized overall infiltration levels of colorectal cancer signature T-cells. Through weighted gene co-expression network analysis (WGCNA), Lasso regression, and StepCox analysis, we developed a prognostic risk model, TRGS, based on six T cell-related genes. Multivariate Cox analysis identified TRGS as an independent prognostic factor for CRC, showcasing its superior predictive efficacy compared to existing immune-related prognostic models. Immunoreactivity analysis revealed higher Immune Prognostic Score (IPS) and lower Tumor Immune Dysfunction and Exclusion (TIDE) scores in the low-risk group, indicating potential responsiveness to immune checkpoint inhibitor (ICI) therapy. Additionally, patients in the low-risk group demonstrated heightened sensitivity to 5-Fu-based chemotherapy regimens. In summary, TRGS emerges as a standalone prognostic biomarker for CRC, offering insights to optimize patient responses to immunotherapy and chemotherapy, thereby laying the groundwork for personalized tumor management strategies.\u003c/p\u003e","manuscriptTitle":"Integrated single-cell and bulk RNA-seq analysis reveals a novel T-cell signature for prognosis and treatment response in colorectal cancer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-05 13:09:04","doi":"10.21203/rs.3.rs-4156323/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-05-14T07:06:31+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-10T16:20:09+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"414f7c13-cf4c-4e2a-b50e-ee4c4232b2ce","date":"2024-04-30T14:04:17+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-04-02T15:02:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"aff8e320-dcc2-4f91-9051-adc0c2be00ca","date":"2024-04-02T10:59:47+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-04-02T10:34:48+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-04-02T10:24:42+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-04-02T09:53:16+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-04-01T12:27:50+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-03-24T03:06:16+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"193b85ef-e108-419a-811d-5ffd2b6e4947","owner":[],"postedDate":"April 5th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":30211526,"name":"Biological sciences/Cancer/Cancer therapy/Cancer immunotherapy"},{"id":30211527,"name":"Biological sciences/Cancer/Cancer therapy/Chemotherapy"},{"id":30211528,"name":"Biological sciences/Cancer/Cancer genomics"},{"id":30211529,"name":"Biological sciences/Cancer/Cancer microenvironment"},{"id":30211530,"name":"Biological sciences/Cancer/Tumour biomarkers"},{"id":30211531,"name":"Biological sciences/Cancer"},{"id":30211532,"name":"Biological sciences/Cancer/Gastrointestinal cancer/Colorectal cancer"}],"tags":[],"updatedAt":"2024-09-02T16:01:15+00:00","versionOfRecord":{"articleIdentity":"rs-4156323","link":"https://doi.org/10.1038/s41598-024-70422-6","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2024-08-30 15:57:27","publishedOnDateReadable":"August 30th, 2024"},"versionCreatedAt":"2024-04-05 13:09:04","video":"","vorDoi":"10.1038/s41598-024-70422-6","vorDoiUrl":"https://doi.org/10.1038/s41598-024-70422-6","workflowStages":[]},"version":"v1","identity":"rs-4156323","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4156323","identity":"rs-4156323","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}