A comprehensive analysis of TRP-related gene signature, and immune infiltration in patients with colorectal cancer

Abstract

Background

Transient receptor potential (TRP) channels are involved in the development and progression of tumors. However, their role in colorectal cancer (CRC) remains unclear, and this study aims to investigate the role of TRP-related genes in CRC.

Methods

Data was obtained from The Cancer Genome Atlas (TCGA) database, and analyses were conducted on the GSE14333 and GSE38832 datasets to assess the prognosis and mark TRP-related genes (TRGs). Subsequently, clustering analysis and immune infiltration analysis were performed to explore the relevant TRGs. In vitro validation of key TRGs’ gene and protein expression was conducted using human colon cancer cells.

Results

Compared to normal tissues, 8 TRGs were significantly upregulated in CRC, while 11 were downregulated. TRPA1 was identified as a protective prognostic factor, whereas TRPM5 (HR = 1.349), TRPV4 (HR = 1.289), and TRPV3 (HR = 1.442) were identified as prognostic risk factors. Receiver operating characteristic (ROC) curves and Kaplan-Meier (KM) analyses yielded similar results. Additionally, lower expression of TRPA1 and higher expression of TRPV4 and TRPM5 were negatively correlated with patient prognosis, and experimental validation confirmed the underexpression of TRPA1 and overexpression of TRPV4 and TRPM5 in CRC cell lines.

Conclusion

This study identifies a TRP channel-related prognosis in CRC, providing a novel approach to stratifying CRC prognosis.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-024-01227-2.

Introduction

As the second biggest cause of cancer-related deaths globally and the third most frequent malignancy, colorectal cancer (CRC) claims around 900,000 lives annually, adding to the burden of cancer in a number of ways [1, 2]. CRC continues to be underdiagnosed despite the fact that diagnostics, prognoses, and therapies have all improved. As a result, the 5-year overall survival rate for CRC continues to be rather poor [3]. The progression of CRC may be attributed to a number of different factors, including genetics and the environment. The earlier the disease is diagnosed, the better the prognosis. Thus, it is crucial to have a solid understanding of the molecular processes driving the development of clinical signs.

Activation of a transient receptor potential (TRP) channel causes a temporary increase in intracellular calcium levels, making TRP channels an important part of the calcium signaling pathway [4, 5]. Cell propagation, aggressiveness, and survival are influenced by Ca²⁺ channels. Consequently, TRP channels are involved in tumor development [6–9]. TRP channels are multifunctional signaling molecules, and their surface localization makes them amenable to drug targeting [10]. Therefore, accurate determination of TRP-related genes (TRGs) would be immensely valuable.

Previous studies have revealed that the TRPs family is upregulated in the gastrointestinal tract [11]. Additionally, Ca²⁺ channels, as well as KCa channels, might be critical in the development and progression of CRC [12]. Due to scientific and technological constraints, most cancer studies to date have only looked at one or two TRGs. However, antitumor effects are defined by the highly coordinated interactions of a large number of genes, including TRPV1, TRPC1, TRPC6, TRPM4, and TRPM5. The published work [13] identified the subtypes of TRPC and constructed a prognostic signature without including direct TRP-related genes and further discussion. Nonetheless, a thorough investigation of TRGs in the early identification and prognosis of CRC patients is necessary.

In the current study, TRG expression in CRC was comprehensively evaluated, and intertumoral immune landscape through computational algorithms was performed. Moreover, clustering subtypes and prognostic signatures associated with TRP channels are crucial for optimizing clinical risk stratification. Thus, based on the above-mentioned evidence, the prognostic value of TRGs and their effect on immune infiltration were explored. Furthermore, we performed cells analysis to confirm the mRNA and histological expression of TRGs, which can identify potential therapeutic targets for future study.

Materials and methods

CRC metadata resources

The Cancer Genome Atlas (TCGA) database was utilized for RNA sequencing data and clinical information about CRC. The GTEx database (https://gtexportal.org/home/datasets) was used to collect all data for normal tissue samples. We collected the normalized expression profiles of GSE14333 and GSE38832 from the GEO database to validate the reliability of the prediction models constructed by TRGs. All RNA-seq data were TPM-normalized for further research. Using “transient receptor potential channel” as the search term.

Differential expressions and correlations of DEGs in TRGs

R-4.1.3-win and R packages listed below. The differential expressions and correlation of TRGs were displayed by the R software package software packages ggplot2 [14], ggstatsplot [15], and pheatmap. Our correlation analysis was conducted using Spearman’s correlation analysis without a normal distribution, while differential expressions was conpared through Wilcox test.

Survival analysis of TRGs in CRC

A comparison of the survival differences between CRC and normal tissue data was performed by using the Kaplan-Meier (KM) to explore the TRG expression data and survival information. A p-value was calculated for KM curves by log-rank testing. The cutoff points were chosen as the median value of TRGs.

A predictive nomogram of TRG construction

We conducted Cox regression analyses, including univariate and multivariate analyses, to identify the risk variables in clinical data and TRGs of CRC patients for building a nomogram. We constructed a random forest plot utilizing the R package ‘forestplot’ to graphically present P values, HRs, and 95% CIs. We developed a nomogram to forecast cumulative recurrence across certain intervals (1-, 2-, 3-, and 5-year). In the nomogram, the variables were graphically depicted. We utilized the ‘rms’ R package to give weights to several risk indicators to determine a patient’s probability of recurrence.

TRGs-associated prognostic signature development and validation

Log-rank tests were employed to analyze the disparity in survival rates between the aforementioned colorectal cancer (CRC) and normal groups, with the aim of facilitating comparison and investigating the potential for developing predictive models based on TRGs. Receiver operating characteristic (ROC) (version 0.4) and least absolute shrinkage and selection operator (LASSO) analyses were also performed [16]. The TRGs signature was validated by GSE14333 and GSE38832 through ROC and KM analyses.

Molecular classification of TRGs using consensus clustering analysis

ConsensusClusterPlus (v1.54.0) [17] is an R package used for cluster analysis, which was used to randomly choose 100 samples from 80% of the entire sample and assign them to the cluster with the greatest number (clusterAlg = “hc”, innerLinkage = “ward.D2”).

Immune cell infiltration influenced by TRGs

Using CIBERSORT, our latest research employed predictive analysis to estimate the proportions of immune cells that had infiltrated tumors in patients diagnosed with CRC within two subtypes identified by TRGs. Spearman’s correlations were used for the investigation of TRGs between non-normally distributed quantitative components.

Relationship between immune checkpoints and TRGs

The expression of 8 checkpoints was selected in comparison to different subtypes derived by TRGs to represent the evaluation of response to immune therapy to further analyze the contribution of different TRGs subtypes in the analysis of immune checkpoints.

Experimental verification

The CRC cell lines (SW480 and HCT116) and the normal human epithelial cell line of the large intestine (FHC) were purchased from American Type Culture Collection (ATCC). Meanwhile, all of the cell cultures described above were incubated in a CO₂ incubator (Panasonic, model: MCO-20AIC) at 37 °C with 5% CO₂ to achieve standard cell culture condition.

10 expression verification of TRGs

With TRIzol™Plus RNA (Invitrogen, United States), we isolated total RNA under the experimental procedures. All-In-One 5X RT MasterMix (Beijing Huaruikang Technology Co., Ltd, Beijing) was employed to synthesize first-strand cDNA from 2 µg total RNA. Follow the instructions, activation of the polymerase at 95 °C for 30 s was followed by 40 cycles of 3 s at 95 °C, followed by 30 s at 60 °C. GAPDH was selected as a reference gene for the whole procedure. We calculated the expression of selected TRGs by using the relative quantification method of 2^−ΔΔCT. A list of all primer sequences can be found in Table 1.

Table 1.

Primers used for qRT-PCR

Gene name	Forward primer	Reverse primer
GAPDH	5ʹ-GCACCGTCAAGGCTGAGAAC-3ʹ	5ʹ-GCCTTCTCCATGGTGGTGAA-3ʹ
TRPA1	5ʹ-GGGGCCCTGGTTCTGTAAAT-3ʹ	5ʹ-ATACGCCCATAACTGGCTGC-3ʹ
TRPM5	5ʹ-GAAACCCGCTGTAGAGGGCA-3ʹ	5ʹ-AGCTGAAATCCACCCTGCGTC-3ʹ
TRPV4	5ʹ-AGCAAGATTGAGAACCGCCA-3ʹ	5ʹ-AGATGACCATGGCACACAGG-3ʹ
TRPC1	5ʹ-AAGCTTTTCTTGCTGGCGTG-3ʹ	5ʹ-CCCGACATCTGTCCAAACCA-3ʹ
TRPC2	5ʹ-CGAGAATTCGAGGAGACCCG-3ʹ	5ʹ-ACATTTGTGCGTTCACCTGC-3ʹ
TRPC3	5ʹ-GCAAATGAGAGCTTTGGCCC-3ʹ	5ʹ-TGATAACGTGTTGGCTGATTGA-3ʹ
TRPC4	5ʹ-AACTGAAGAAGGCCTGACCG-3ʹ	5ʹ-GGCAATTGCTGCTGATCTCG-3ʹ
TRPC5	5ʹ-GTGCCACTTTGCTGGTCTTG-3ʹ	5ʹ-AGATGGGCAGGGTAGTTTGC-3ʹ
TRPC6	5ʹ-AGATGGGCAGGGTAGTTTGC-3ʹ	5ʹ-TCCCCAACTCGAGAGAGGTT-3ʹ
TRPC7	5ʹ-ACACAACGCTGAGGGAGAAG-3ʹ	5ʹ-GGTTCGTCCTAGCTTAAATTCAGT-3ʹ
TRPV3	5ʹ-GGCCTGTAAGACGAACAGCA-3ʹ	5ʹ-GTCTCCTTTTCCCCGAAGGG-3ʹ
TRPV4	5ʹ-ATTCAGGAAGCGCGGATCTC-3ʹ	5ʹ-GGTGACGATAGGTGCCGTAG-3ʹ
TRPV5	5ʹ-CACCCTTTGTCCTTTGCTGC-3ʹ	5ʹ-ACGTGTCTCATAGGCCTCCT-3ʹ
TRPV6	5ʹ-AAGCCCAGGACCAATAACCG-3ʹ	5ʹ-ATGGCAAAGGCAGCATAGGT-3ʹ
TRPM2	5ʹ-TGCTTCTTTCCAAGGTGCGA-3ʹ	5ʹ-GCTCCAGGGTGTCTCCCAT-3ʹ
TRPM3	5ʹ-TGCCTGCCGTTTTTCTCTCT-3ʹ	5ʹ-CCTGGAAAGTTACCTACCCGC-3ʹ
TRPM4	5ʹ-ACGGATCCAGCTGCAGTTTA-3ʹ	5ʹ-TGATTCGATCTCGGGCTTCG-3ʹ
TRPM5	5ʹ-GCATTGCTATGGGCAGGTCA-3ʹ	5ʹ-CTCAGAAGCTGACCCCAGTG-3ʹ
TRPM6	5ʹ-TCTCTGGTCTCCACCCAAAG-3ʹ	5ʹ-TGGTACAGGCACACCACATC-3ʹ
TRPM7	5ʹ-CAGAACAGAGCCCAACGGAT-3ʹ	5ʹ-CAGCTCTGCCTGTTCCTTCA-3ʹ
TRPA1	5ʹ-TTCCATGAAGAAAGGTGCCC-3ʹ	5ʹ-TTCGGAGGTTTGGGTTTGCT-3ʹ
TRPV1	5ʹ-GCCTTCAAGCTCACAAGCAC-3ʹ	5ʹ-ATGGTGGTGTTCTGTCCGTC-3ʹ

Results

TRGs as potential biomarkers of CRC

We comprehensively calculated the expressions of 21 TRGs between 620 CRC and 779 normal samples to obtain the expression profile of TRP channels in CRC. As shown in Fig. 1 and 19 TRGs exhibited substantial expression differences between tumor and normal groups (Fig. 1A). Specifically, the overexpression of TRPM2, TRPV4, TRPV5, TRPM4, TRPM5, TRPM6, TRPC5, and TRPC7 was observed in CRC tissues (p < 0.01).Through a comparative analysis of the transcriptional data, it was observed that the expression levels of the remaining 11 TRP genes (TRPV1, TRPC1, TRPC6, TRPC4, TRPV3, TRPC3, TRPC2, TRPM3, TRPA1, TRPV6, and TRPM6) showed that compared with to normal tissues, a significant decrease in colorectal cancer (CRC) (p < 0.001). These data revealed that TRGs have crucial biological functions in CRC formation and development (Fig. 1B).

Fig. 1.

The expressions and correlation of TRP family gens between CRC and Normal groups. A Expression levels of TRP genes in 620 CRC samples and adjacent normal pairs. B The correlations between multiple TRP genes in 620 CRC samples and adjacent normal pairs. *p < 0.05, **p < 0.01, and ***p < 0.001

The independent prognostic value of the tryptophan pathway-related risk score

We resorted to a survival nomogram to construct a clinical tool to predict patients’ survival probabilities. KM survival analysis proved that there are four genes significantly associated with survival outcomes (Fig. 2A). After data analysis, it was shown that TRPA1 acts as a protective factor (P = 0.016, HR = 0.815), while TRPM5 (HR = 1.349), TRPV4 (HR = 1.289), and TRPV3 (HR = 1.442) are key predictive factors with statistically significant differences (Fig. 2B, C). Considering the TNM staging in clinical practice, we incorporated it in the survival nomogram (Fig. 2D). An optional agreement between observation and prediction was observed in the calibration curve (Fig. 2E), suggesting that clinical outcomes for patients with CRC are regulated by TRP risk scores.

Fig. 2.

The consturction of the predictive nomogram of TRGs. A Kaplan-Meier survival analysis of TRPV4, TRPA1, TRPM5 and TRPV1. B, C Hazard ratio and P-value of constituents involved in univariate and multivariate Cox regression and some parameters of the TRGs. D Nomogram to predict the 1-y, 2-y and 3-y overall survival of CRC patients. E Calibration curve for the overall survival nomogram model in the discovery group. A dashed diagonal line represents the ideal nomogram, and the blue line, red line and orange line represent the 1-y, 2-y and 3-y observed nomograms

TRG-corresponding prognostic signature development and validation

LASSO regression identified TRG genes showed significant correlation with overall survival (OS) (Fig. 3A and C). Based on BSR analysis, three prognostic factors TRPV4, TRPA1, and TRPM5 were identified. The high expression of TRPV4 and TRPM5, as well as the low expression of TRPA1, were associated with poor prognosis in CRC patients (p < 0.05). According to KM survival curves, the high-risk group had worse clinical outcomes in terms of OS compared to the low-risk group (Fig. 3D). When comparing the area under the curve (AUC) values of the prognostic markers of the three genes with each individual gene mentioned above, the combined marker of the three genes showed a larger AUC value, indicating that the risk scoring model composed of TRG genes had better stability in predicting 1-year (AUC, 0.642), 2-year (AUC, 0.588), and 3-year survival (AUC, 0.576) outcomes (Fig. 3E). In GSE14333 and GSE38832, the AUC values of the TRG marker were 0.58 (Fig. 4A) and 0.59 (Fig. 4B), respectively. In survival analysis, significant interactions were observed in GSE14333 (p = 0.02) and GSE38832 (p = 0.01). This marker further confirmed the stability in predicting survival outcomes (Fig. 4C, D)

Fig. 3.

Construction and Validation of Prognostic Signatures of TRGs. A Coefficients of selected features are shown by lambda parameter; B partial likelihood deviance versus log (λ) was drawn using LASSO Cox regression model. C Prognostic analysis of gene signature in the TCGA set and the heatmap of the expression profiles of the prognostic genes in low-risk and high-risk group. D Kaplan-Meier survival analysis of the signature. E Time-dependent ROC analysis the of the gene signature. ROC receiver operating characteristic

Fig. 4.

ROC curves and KM survival analysis of Prognostic Signatures of TRGs in GSE14333 and GSE38832. A, B The figure shows the ROC curves of the prognostic signatures of TRGs in GSE14333 (A) and GSE38832 (B). C, D The figure show the result of KM analysis of GSE14333 (C) and GSE38832 (D).

Significant correlation of consensus clustering for TRGs with PFS and DFS of patients with CRC

According to the similarity analysis, k = 2 provides optimal clustering stability from k = 2–6. By clustering measures, expression levels of TRGs and the abundance of ambiguity were displayed (Fig. 5A, B). A total of 620 CRC samples were clustered into two subtypes, namely, cluster 1 (C1, n = 498) and cluster 2 (C2, n = 122) (Fig. 5C, D). A close association has been found between cluster 1/2 subtypes defined by TRGs and CRC patient heterogeneity. First, the expression of TRGs was higher in cluster 2 than in cluster 1. Second, the progression-free survival (PFS, p = 0.043, HR = 0.691) (Fig. 5E) and disease-free survival (DFS, p = 0.042, HR = 0.459) (Fig. 5F) of cluster 1 were longer than that of patients in cluster 2.

Fig. 5.

Differential Clinicopathological Features and Survival of CRC in Cluster1/2 Subtypes. A Consensus CDF curve shows the cumulative distribution functions of the consensus matrix for each k. B Delta area curve of consensus clustering, indicating the relative change in area under the cumulative distribution function (CDF) curve for each category number k compared with k–1. The horizontal axis represents the category number k and the vertical axis represents the relative change in area under CDF curve. C, D Heatmap depicting consensus clustering solution (k = 2) for TRGs in CRC samples (n = 620) and heatmap of TRP-related gene expression in different subgroups, red represents high expression and blue represents low expression. Kaplan-Meier survival analysis of the different subtypes. E A PCA plot of the data showing no batch effect in the TCGA CRC dataset. Red nodes represent the C1 cluster while blue nodes represent the C2 cluster. F, G Kaplan-Meier curves of progression free survival (PFS, F) and disease-free survival (DFS, G) for patients with CRC in two clusters (cluster1/2). H The distribution of clinical characteristics in the samples from different subtypes. *p < 0.05, **p < 0.01, and ***p < 0.001

Using principal component analysis (PCA), we examined gene-expression patterns across cluster 1/2 and found that the two subtypes had distinct genomic profiles (Fig. 5G). Afterward, clinicopathological features including different classifications of the two clusters were compared (Fig. 5H).

Effect of genetic alterations of TRG signatures on immune cell infiltration

We focused on the relationship between the risk score and the invasion of certain immune cell types to assess the effect of these seven TRGs signatures on the CRC immune milieu (Fig. 6A, B). In the study, a notable inverse relationship was observed between the risk score and the quantity of B cells (p = 0.009), myeloid dendritic cells (p = 0.022), neutrophils (p = 0.001), and T cells (p = 7.15e⁻¹⁰). This result revealed that the TRGs signature was involved in the CRC immune microenvironment.

Fig. 6.

Immune landscape of colorectal cancer (CRC) patients in cluster1/2 group. A, B The infiltrating levels of 22 immune cell types in cluster1/2 subtypes. C, D Heatmap and abundance of immune cell infiltration in cluster1/2 subtypes. E A heatmap of the correlation between TRGs and immune score based on TIMER. F The expression distribution of SIGLEC15, TIGIT, CD274, HAVCR2, PDCD1, CTLA4, LAG3 and PDCD1LG2 in different subtypes. G The statistical table and distribution of immune response of samples in different subtypes based on the prediction results

Association of immune cell infiltration and immune checkpoint with subtypes derived by TRGs

Considering the diverse population of immune cells, we used CIBERSORT for further exploration of immune cells and validation of the results. Comparing the abundance of multiple immune cells between two subtypes, we found that plasma cells, CD8⁺ T cells, were significantly downregulated in subtype 2 (Fig. 6E). Meanwhile, M0 and M2 macrophages were upregulated in subtype 2 (Fig. 6C, D).

Based on the differential immunity infiltration characteristics of the two subtypes, we investigated whether immune checkpoint ligands were differentially expressed in each subtype. It was statistically significant that LAG3, TIGIT, SIGLEC15, and PD-L2 were highly expressed in cluster 2 (Fig. 6F). An algorithm dubbed Tumor Immunological Dysfunction and Exclusion (TIDE) was utilized to describe the presence of T cells in immune landscapes to predict the treatment response. With considerably higher TIDE scores in the G2 group compared to the G1 group, we could measure TRG expression prior to checkpoint blockade treatment and predict the likelihood of an ICB response (Fig. 6G).

mRNA expression of TRGs in CRC cell lines

Further validation of the three hub genes was conducted using RT-qPCR in colorectal cancer (CRC) and normal cell lines. The results showed a significant downregulation of TRPA1 (Fig. 7A) in SW480 and HCT116 compared to FHC. Additionally, the expression levels of TRPV4 (Fig. 7B) and TRPM5 (Fig. 7C) were notably higher in CRC cell lines compared to normal cell lines. The RT-qPCR results for the three hub genes were in agreement with those from the TCGA datasets, confirming their validity. Moreover, the expression of the other 16 TRGs was also validated in the mentioned cell lines (Supplementary Fig. 1).

Fig. 7.

The mRNA expression of TRPA1 (A), TRPV4 (B) and TRPM5 (C) in FHC, SW480 and FHC cell lines. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001

Discussion

Our study examined the transcriptional patterns, prognostic significance, and function of TRP channels in patients with CRC from the TCGA datasets. A total of 11 TRG decreases and 8 TRG increases in CRC tissue were observed. TRPA1 acted as a protective predictor, while TRPM5, TRPV4, and TRPV3 were identified as risk predictors. Furthermore, our prognostic model suggested that TRPA1, TRPV4, and TRPM5 are considered the key genes in CRC, which was supported not only by the ROC analysis and KM survival analysis of GSE14333 and GSE38832 but also, to some extent, by the results of RT-qPCR.

Among the three signature-associated genes, TRPA1 was downregulated in CRC and acted as a protective predictor involved in cancer cell migration and proliferation [18]. According to the reviewed literature, A549 lung cancer epithelial cells had extremely low TRPA1 expression, which, in turn, decreased the production of the pro-inflammatory cytokine interleukin (IL)-8 [19], which is consistent with our result. TRPA1 exhibited a similar tendency in both NCM640 and HT29 CRC cell lines [20].As a redox-sensing Ca²⁺-influx channel in neurons, TRPA1 is intriguing because it might protect inner-cell survival against reactive oxygen species (ROS) [21]. Also, TRPA1 signaling may facilitate hypoxia-induced COX-2 transcription and cell invasion [22]. The above-described evidence proves that TRPA1 plays a critical role by protecting cells against ROS and reducing inflammatory cytokines.

Another important finding was that TRPV4 increased the risk score of patients with CRC. Consistently, the overexpression of TRPV4 enhances the migration and inflammation in lung cancer (20), as well as liver cancer, bladder cancer, and skin cancer [23–25]. Emerging discoveries have advanced our understanding that TRPV4 mediates NLRP3 activation, triggering IL-1β production and then eliciting inflammatory responses in vivo [26]. TRPV4 induces macrophage phagocytosis and secretion of pro-resolution cytokines [27], corroborating numerous previous findings on TRPV4 in inflammation and immune infiltration of CRC. Experiments conducted in vitro and in vivo have shown that inhibiting the expression of TRPV4 can activate the PTEN pathway, leading to the inhibition of colorectal cancer development [28].

This study supports evidence from previous observations showing that TRPM5 is widely expressed in the digestive system [29], which is associated with a range of malignancies [30, 31]. Lung metastasis increase is observed when TRPM5 expression is enforced [32]. The use of an inhibitor of TRPM5 significantly reduces melanoma metastasis [33]. As with earlier studies, TRPM5 indicates a potential therapeutic application in patients with CRC.

According to the molecular categorization of TRP channels, the PFS and DFS of cluster 1 were superior, with upregulated M1 macrophages and activated memory CD4⁺ T cells contributing to this improvement. According to the already conducted research, TRP channels have the potential to prevent the polarization of macrophages toward an M1 phenotype [34]. M1 macrophages drive antitumor immune response by IL-12 production and IL-10 inhibition [35]. In addition, activated memory CD4⁺ T cells were observed in cluster 1, which expressed a lower level of TRP channels. Activated memory CD4⁺ T cell expansion elicits plasmacytic infiltrations to produce massive cytokines and enhances antigen presentation functions of the dendritic cell, promoting the activation of antitumor immunity.

With the highly heterogeneous performance of CRC, the tissue microenvironment (TME) plays a critical role in advancing immunotherapies. Positive feedback from eight immune checkpoints might open up new possibilities for the exploration of combination drugs in CRC patients under the G2 subtype. Although there is a conflict between an immune checkpoint and ICB, the main reason can be summarized as follows. First, based on in-depth research on the significance of T cells in the TME, the TIDE algorithm has been created [36]. However, plasma cells, macrophages, and dendritic cells are also involved in the immune landscape with the participation of T cells, indicating that a comprehensive algorithm is required to be developed. Moreover, the possible interference of other microbiology processes and molecules cannot be ruled out, requiring specific research on immune checkpoints with TRP channels in CRC.

Certainly, the study had potential limitations. First, this study did not assess clinical data for further excavation. Second, it is also important to use a larger sample size in further in vivo clinical samples and in vitro experiments to confirm the current findings. As a consequence of our study, we could discover a novel gene signature predicting CRC patients’ prognosis, giving a platform for future research on TRP channels in CRC.

Conclusions

The current study comprehensively revealed the prognostic value of TRGs, indicating that CRC patients with low expression of TRPA1 and overexpressed TRPV4 and TRPM5 have a worse prognosis. These three genes were considered hub genes in CRC and validated by the analysis of RT-qPCR.

Author contributions

YCL, XBY and WJZ contributed equally. YCL, XBY and WJZ conceived and designed the study. YCL and XBY performed data curation. JX, HYZ and TH performed experimental work. YCL, JJY, CT and CW prepared the original draft. QMW, QQY and WYH reviewed and edited the manuscript. All authors have read and approved the final manuscript.

Funding

This study is sponsored by 2023 Shanghai Science and Technology Innovation Action Plan-Medical Innovation Research Project (23Y11921800) and “14th Five-Year Plan” Traditional Chinese Medicine Specialty and Traditional Chinese Medicine Emergency Capacity Improvement Project (ZYTSZK1-8).

Data availability

RNA-sequencing data and clinical data were downloaded from The Cancer Genome Atlas (TCGA, https://cancergenome.nih.gov/) and The Genotype-Tissue Expression (GTEx, https://gtexportal.org/home/datasets) database. All data supporting the findings of this study were available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.