Skip to main content
Download PDF
Download ePub

Unveiling prognostic value of JAK/STAT signaling pathway related genes in colorectal cancer: a study of Mendelian randomization analysis

Infectious Agents and Cancer volume 20, Article number: 9 (2025) Cite this article

Abstract

Background

Colorectal cancer (CRC) ranks among the frequently occurring malignant neoplasms affecting the gastrointestinal tract. This study aimed to explore JAK-STAT signaling pathway related genes in CRC and establish a new prognostic model.

Methods

The data set used in this study is from a public database. JAK-STAT-differentially expressed genes (DEGs) were identified through differential expression analysis and weighted gene co-expression network analysis (WGCNA). Prognostic genes were selected from JAK-STAT-DEGs through Mendelian randomization (MR), univariate Cox regression, and least absolute shrinkage and selection operator (LASSO) analyses. The expressions of prognostic genes were verified by RT-qPCR. Then, a risk model was built and validated by the GSE39582. Independent prognostic factors were screened underlying risk scores and different clinical indicators, resulting in the construction of a nomogram. Additionally, immune infiltration, immune scores and immune checkpoint inhibitors analyses and gene set enrichment analysis (GSEA) were carried out.

Results

The 3,668 JAK-STAT-DEGs were obtained by intersection of 5826 CRC-DEGs and 9766 JAK-STAT key module genes. Five prognostic genes were selected (ANK3, F5, FAM50B, KLHL35, MPP2), and their expressions were significantly different between CRC and control groups. A risk model was constructed according to prognostic genes and verified by GSE39582. In addition, the nomogram exhibited superior predictive accuracy for CRC. Furthermore, immune analysis results indicated a notable positive correlation between risk score and the scores of immune (R = 0.486), stromal (R = 0.309), and ESTIMATE (R = 0.422). Immune checkpoint inhibitor ADORA2A (Cor = 0.483263) exhibited the strongest positive correlation with risk score. And MPP2 exhibited the most potent activating influence on the cell cycle pathway, whereas ANK3 demonstrated the most significant inhibitory effect within the apoptosis pathway.

Conclusions

A new JAK-STAT related CRC prognostic model was constructed and validated, which possessed an underlying predictive potential for CRC patients’ prognosis and could potentially enhance tailored guidance for immunotherapy.

Background

Colorectal cancer (CRC) is a common neoplasm of the gastrointestinal system. According to 2023 statistics, CRC has the third incidence and mortality rates globally [1]. The occurrence and development of CRC involve various genetic and environmental factors, and its molecular mechanism remain poorly understood [2]. At present, the treatment of CRC mainly counts on surgical resection, chemotherapy, radiotherapy and targeted treatment, but the efficacy of these treatment methods is limited due to the heterogeneity and drug resistance of CRC. Thus, the identification of new prognostic biomarkers and therapeutic targets, and establish efficacious prognostic models, are important significance on improving the quality of life and prolonging the survival of CRC patients.

The Janus kinase signal transducer and activator of transcription (JAK-STAT) is an essential cell signal transduction pathway, which is involved in regulating a plethora of physiological processes, containing regulating cellular proliferation, differentiation, immunoreaction, apoptosis and other physiological processes [3]. The JAK-STAT pathway consists of cell surface receptors, intracellular JAK family kinases and STAT family transcription factors. When the receptor binds to the corresponding ligand, it activates JAK kinase, which in turn phosphorylates STAT transcription factor, causing it to form dimers or multimers, transport to the nucleus, bind to specific DNA sequences, and regulate the expression of downstream genes [4, 5]. The JAK-STAT pathway interacts with various biological pathways in CRC, such as cell cycle, apoptosis, angiogenesis, inflammation, epigenetic regulation, etc. While abnormal activation or inactivation of the JAK-STAT pathway is related to the occurrence and development of various tumors, including CRC [6]. Therefore, the research on CRC and JAK-STAT pathway can help to enhance the understanding of the molecular mechanism of CRC, and also supply new targets for the prevention and therapy of CRC.

Mendelian randomization (MR) is a method which uses gene variants as natural experiments to investigate the causal relationship of exposure to factors and the occurrence of diseases [7]. The basic principle of MR is that genetic variants are randomly assigned to individuals, not subject to confounding factors and reverse causality. The three core principles of MR are as follows: in associational assumption, genetic variants used as instrumental variables (IVs) are strongly associated with the exposure factors. With independence assumption, the relationship between IVs and confounding factors are independent, meaning they do not share common causes. In addition, IVs only influence the outcome by virtue of exposures and have no direct influence any other biology processes related to the outcome in the exclusivity assumption. These principles allow researchers to assess causal relationships between exposure factors and outcomes. Meanwhile, the application of MR can avoid some constraints of traditional observational studies, for instance measurement error, confounding factors and reverse causality. Nowadays, MR analysis is widely used to investigate various diseases, containing malignant tumors such as CRC, hepatocellular carcinoma and breast cancer [8]. Furthermore, in Liu’s research, the Mendelian randomization method was further employed to select disease genes with causal relationships from candidate genes as potential biomarkers [9]. Therefore, in this study, we also drew on this approach and used MR analysis to screen out JAK-STAT signaling pathway genes that might be associated with the occurrence and development of colorectal cancer (CRC).

The aim of this study was to investigate the prognostic value of genes related to the JAK-STAT pathway in CRC. We used MR and COX analysis to screened and identified new JAK-STAT pathpathe-related prognostic models in CRC. To provide a new direction to improve the quality of life and prolong the survival of CRC patients.

Methods

Data source

Based on the cancer names colon adenocarcinoma (COAD) and rectum adenocarcinoma (READ), 383 CRC samples (371 CRC samples with clinical and survival information) and 51 normal samples were selected and treated as a training set (TCGA-CRC), which was obtained from the University of California Santa Cruz (UCSC) Xena (https://xenabrowser.net). The validation set (GSE39582) containing mRNA expression profile and survival information of 562 CRC samples (tissue of colorectal) was acquired from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/). The sequencing platform employed for GSE39582 was GPL570. The 155 JAK-STAT signaling pathway genes were derived from the Molecular Signatures Database (MsigDB) (https://www.gsea-msigdb.org/gsea/login.jsp) [10].

Identification of differentially expressed genes (DEGs) in CRC

The CRC-DEGs in the training set (CRC vs normal) were identified using the edgeR package (v 3.36.0) [11], with a standard of |Log2FC|> 1 and p.adj < 0.05. The ggVolcano (v 0.0.2) [12] and ComplexHeatmap (v 2.14.0) [13] packages were used to generate volcano map and heat map, respectively.

Weighted gene co-expression network analysis (WGCNA)

The JAK-STAT score for all samples in the training set was computed using the single sample gene set enrichment analysis (ssGSEA) algorithm based on 155 JAK-STAT signaling pathway genes, utilizing the GSVA package (v 1.46.0) [14]. A violin plot of JAK-STAT scores among CRC samples and normal samples was generated using the wilcox.test method with the ggplot2 package (v 3.3.5) [15]. Firstly, all samples within the training set were hierarchically clustered using the goodSamplesGenes function in WGCNA (v 1.70–3), and outlier samples were separately removed [16]. The optimal soft threshold (β) approaching a scale-free distribution was computed based on the expression matrix of all genes as input data. The adjacency and similarity were employed to construct a hierarchical clustering tree of genes. With a minimum number of 70 genes, modules were merged using the dynamic tree cutting algorithm. The association between the modules and JAK-STAT score was assessed using Pearson correlation analysis. Subsequently, the key modules were identified based on the correlation with the JAK-STAT score (p < 0.05, Cor ≥ 0.3), and the JAK-STAT key module genes were extracted. In determining these two thresholds, we considered the balance between reducing false—positive results and avoiding missing important associations (false—negatives). Modules that meet these thresholds are more likely to have meaningful interactions with the JAK—STAT pathway at the molecular level.

Identify of JAK-STAT-DEGs and enrichment analysis

Utilizing the VennDiagram package (v 1.7.3) [17], we determined the overlap of CRC-DEGs and the JAK-STAT key module genes to identify JAK-STAT-DEGs. Enrichment analysis of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway of JAK-STAT-DEGs was performed utilizing DAVID (v 6.8.0) [18] (p < 0.05, count ≥ 2).

Mendelian randomization (MR) analysis

The following three assumptions must be satisfied in MR studies: (a) instrumental variables (IVs) are signally correlated with exposure, (b) IVs are unaffected by confounding factors, and (c) IVs solely impact the outcome by means of exposure [19]. The OpenGWAS database (https://gwas.mrcieu.ac.uk/) was utilized for retrieving the GWAS ID and GWAS data pertaining to JAK-STAT-DEGs and CRC, utilizing JAK-STAT-DEGs instrumental variables with significant correlations as the exposure factor and CRC as the outcome variable (ieu-b-4965). The ieu-b-4965 contained 5657 CRC samples and 372,016 controls, with 11,738,639 single nucleotide polymorphisms (SNPs). The TwoSampleMR package (v 0.5.6) [20] was employed the function extract instruments to select SNPs as IVs demonstrating significant associations with JAK-STAT-DEGs (p < 5 × 10–8). In addition, independent IVs were obtained genetically by removing the IVs for linkage disequilibrium (LD) (clump = TRUE, R2 = 0.001, kb = 10,000). Marker genes were identified from JAK-STAT-DEGs by leveraging the mr function in conjunction with five algorithms (MR Egger [21], Weighted median [22], Inverse variance weighted (IVW) [23], Simple mode [24], and Weighted mode [25]. The results of IVW were the main reference (p < 0.05). The scatter plots, forest plots, and funnel plots were utilized to present the analytical outcomes. The reliability of MR analysis was evaluated by sensitivity analysis, including heterogeneity test (p ˃ 0.05), horizontal pleiotropy test (p ˃ 0.05), and Leave-One-Out (LOO) analysis.

Univariate Cox analysis and LASSO analysis

Firstly, univariate Cox analysis was implemented based on survival data and the expression of marker genes in TCGA-CRC, with marker genes related to prognosis being identified (p < 0.05). Secondly, utilizing the glmnet package (v 4.0–2) [26], we conducted LASSO regression analysis on marker genes related to prognosis with the specified parameter settings of family = cox and nfold = 20, which entailed performing 20-fold cross-validation to identify prognostic genes.

Evaluation and verification of risk model

The prognostic risk model was constructed in accordance with the expression of prognostic genes (Xi) and regression coefficient (coefi), as determined by LASSO. A risk score was then calculated using the following formula.

$$risk\;score = \sum\nolimits_{n = 1}^{n} {\left( {coefi * Xi} \right)}$$

The risk scores for 371 CRC samples within the training set were calculated. Subsequently, all samples were classified into groups of high/low-risk based on the optimal threshold of risk score. A Kaplan–Meier (KM) survival analysis was completed for these two groups by survminer package (v 0.4.9) [27]. Based on the risk score, the survivalROC (v 1.0.3) [28] package was employed to calculate false and true positives. The results were then applied for receiver operating characteristic (ROC) curves at 1-, 3-, and 5-years and determining area under curves (AUC) values. AUC was used to indicate the accuracy of the prediction. A higher AUC value means a larger area under the curve (and a smoother curve), indicating a higher prediction accuracy. Moreover, validation of risk model was completed in GSE39582.

Independent prognostic analysis and nomogram creation

In univariate and multivariate Cox analyses, risk scores, tumor stage, age, pathologic T, N, M, and gender were incorporated as factors to confirm their status as independent prognostical predictors (p < 0.05). The rms package (v 6.2–0) [29] was utilized to develop a nomogram that predicted the survival rates for CRC patients at 1, 3, and 5 years. Afterwards, the reliability of the nomogram predictions was evaluated applying calibration curves and ROC curves.

Immune infiltration analysis

According to the gene expression, stromal, immune, and ESTIMATE scores between the risk groups were calculated by the ESTIMATE package (v 1.0.13) [30]. The ggplot2 package was utilized to visualize the results through the Wilcox.test method. The Spearman correlation between risk scores with above three scores was assessed. After that, the ssGSEA algorithm was utilized to assess the prevalence of 28 immune cells types across all samples in risk groups (p < 0.05). Subsequently, a box plot was generated by the ggplot2 package, employing the wilcox.test method for statistical analysis, based on the immune cell infiltration abundance of each sample. Furthermore, the Spearman correlation was computed of differential immune cells themselves, as well as between risk scores and differential immune cells. Finally, a box plot was created utilizing the ggplot2 package and the Wilcox.test method to visually represent the differences in immune checkpoint inhibitors between two risk groups (p < 0.05). Subsequently, the correlation between risk score and differential immune checkpoint inhibitors was determined applying Spearman method. All analyses were performed in the training set. In order to identifed the correlation between differential immune cells and differential immune checkpoints, Spearman’s correlation analysis of differential immune cells with differential immune checkpoints was performed using the R package “psych” (threshold minimum |Cor|> 0.5, p < 0.05) and visualized by heatmap. To investigated whether risk scores can be used for prognostic analysis in colorectal cancer patients treated with immune checkpoint inhibitors. The relationship between risk scores and patient response to immunotherapy (tumor shrinkage, overall survival, etc.) was analyzed using data from a public immunotherapy cohort, IMvigor210.

Gene set enrichment analysis (GSEA)

In the training set, we employed the edgeR package for intergroup analysis of differential gene expression, ranking genes by their log-fold change (High-Risk vs Low-Risk). Subsequently, using biological processes (BP), molecular functions (MF) and cellular components (CC), as well as KEGG gene sets as enrichment backgrounds, GSEA analysis was conducted with clusterProfiler package (v 4.7.1) [31] (p.adj < 0.05).

The gene–gene interaction network (GGI) and carcinogenic pathways analysis for the prognostic genes

Based on the GeneMANIA platform, a GGI network was conducted to identify genes and pathways related to prognostic genes. To explore the correlation between prognosis genes and cancer-related pathways, we evaluated the contribution utilizing the gene set cancer analysis Lite (GSCALite) database from the training set of CRC samples.

The expression validation of prognostic genes

To further validate the prognostic genes, we procured 5 pairs CRC and control tissue samples from Zhengzhou Central Hospital Affiliated to Zhengzhou University for reverse transcription quantitative polymerase chain reaction (RT-qPCR).

The studies involving human participants were reviewed and approved by Zhengzhou Central Hospital (NO.ZXYY2024186). The participants provided their written informed consent to participate in this study.

The TRIzol (Ambion, Austin, USA) was employed to isolate total RNA according to the manufacturer’s instructions. Subsequently, cDNA synthesis was performed utilizing the SureScript-First-strand-cDNA-synthesis-kit (Servicebio, Wuhan, China) as recommended by the producer. Then, RT-qPCR was performed utilizing the 2 × Universal Blue SYBR Green qPCR Master Mix (Servicebio) with the primer sequences (Table 1). The internal reference was GAPDH, and the expression was computed employing the 2−ΔΔCt method [32].

Table 1 RT-qPCR was performed utilizing the 2 × Universal Blue SYBR Green qPCR Master Mix (Servicebio) with the primer sequences
Full size table

Results

Acquisition of JAK-STAT-DEGs

A number of 5826 CRC-DEGs in the training set were observed, including 2201 up-regulated DEGs and 3625 down-regulated DEGs. According to the sorting by log2FC, the heat map illustrated the top 10 up/down-regulated DEGs (Fig. 1a, b).

Fig. 1
figure 1

Acquisition of JAK-STAT-DEGs. a, b The expression of differential genes between CRC samples and normal samples. c JAK-STAT scores were different between the two sample groups. d, e Data sample clustering and phenotypic information. f Scale-free soft threshold distribution. (Left) The square of the correlation coefficient is above 0.85, indicating that the network is approaching the scale-free distribution. (Right) The vertical axis represents the mean of all gene adjacency functions in the corresponding gene module. g–i Clustering of module eigengenes, modules with high correlations are grouped together. j Correlation between module genes and JAK-STAT score

Full size image

The JAK-STAT scores were found to be lower in CRC samples (Fig. 1c). The outcome of the sample clustering analysis revealed an outlier, which would be excluded to facilitate the construction of the sample clustering and clinical trait heat map (Fig. 1d, e). Employing a threshold of scale-free R2 close to 0.85, we selected β as 5 and employed the WGCNA approach to construct gene modules in the training set (Fig. 1f). Consequently, 20 modules were established. The MEDissThres was configured to 0.3 for the purpose of merging analogous modules, resulting in an aggregate of 14 modules (Fig. 1g, h, i). Subsequently, the brown, blue and black modules were identified based on the correlation with the JAK-STAT score (p < 0.05, Cor ≥ 0.3) as key modules, and the 9,766 JAK-STAT key module genes were extracted (Fig. 1j).

Finally, the intersection of CRC-DEGs and JAK-STAT key module genes was determined as 3668 JAK-STAT-DEGs using the ggvenn package (Fig. 2a). The functional enrichment analysis indicated that 3668 JAK-STAT-DEGs were linked to 621 pathways in BP entries, 168 pathways in CC entries, 140 pathways in MF entries (Fig. 2b) and 82 pathways in KEGG entries (Fig. 2c). These pathways were significantly enriched in cell-to-cell functional pathways and Immunoinflammatory pathways such as cell IL-17, PI3K-Akt, cAMP, Ras, and Rap1 signaling pathways, as well as adhesion molecules.

Fig. 2
figure 2

Acquisition of JAK-STAT-DEGs. a 3668 JAK-STAT-DEGs were determined. b Gene ontology (GO) top10 enrichment analysis results (Ranked according to p. value from small to large). c KEGG top10 enrichment analysis results

Full size image

MR analysis for JAK-STAT-DEGs and CRC

Utilizing 3668 JAK-STAT-DEGs as the exposure factor and CRC as the outcome variable, the IVM was employed to identify 66 JAK-STAT-DEGs exhibiting a significant causal relationship with CRC (p < 0.05). These 66 JAK-STAT-DEGs were served as marker genes for subsequent analysis. Of these 66 marker genes, 11 genes served as risk factors, while 55 genes function as safety factors (Fig. 3a, Additional file 1). ANK3 was taken as an example to show the subsequent analysis results (Fig. 3b). In the scatter plot, a negative correlation was observed between ANK3 and CRC slopes overall, suggesting that a decrease in ANK3 is associated with an increased risk of CRC (Fig. 3c). In the forest map, ANK3 was a safety factor for CRC (The effect size greater than 0) (Fig. 3d). The funnel plot illustrated that SNPs were roughly symmetrical and conformed to Mendel’s second law (Fig. 3e). The Cochran’s Q test revealed no evidence of heterogeneity (p = 0.727). Additionally, the horizontal pleiotropy test yielded a p-value of 0.461, suggesting the absence of confounding factors in MR analysis. Finally, the impact of SNPs on the dependent variable remained consistent, suggesting that the findings from the MR analysis were robust and dependable (Fig. 3f).

Fig. 3
figure 3

MR analysis for JAK-STAT-DEGs and CRC. a Results of MR analysis of 66 JAK-STAT-DEGs. b Using RT-qPCR assay, we found that ANK3 gene was significantly down-regulated in CRC samples compared to control samples. c The effect of SNPs of ANK3 on CRC correlated to overall. d Relationship between ANK3 and prognosis of CRC. e MR analysis of ANK3 and CRC. f Sensitivity analysis shows that the results are reliable

Full size image

Constructed, evaluated, and validated a valid risk model

Relevant to CRC, a set of 5 prognostic genes (ANK3, F5, FAM50B, KLHL35 and MPP2) were identified through the utilization of univariate Cox regression analysis and LASSO analysis from 66 marker genes (Fig. 4a, b, c). To verify the expression of these prognostic genes, RT-qPCR assay was used. The findings revealed significant down regulation of five prognostic genes in CRC samples compared to control samples (Fig. 4d). Subsequently, a risk model was structured around the expression and regression coefficient (− 0.194420331, 0.03154147, 0.017450427, 0.086191749, 0.197669923) of 5 prognostic genes. And 371 CRC patients were sorted into high/low-risk groups by optimum threshold value (− 0.6679441). There were 161 CRC patients in the high one and 210 CRC patients in the low one (Fig. 4e). The expression of 5 prognostic genes in these risk groups were visualized using a heatmap (Fig. 4f). The prognosis of CRC patients with high-risk was found to be unfavorable in both TCGA-CRC and GSE39582, as evidenced by the KM curves. (Fig. 4g, h). Further, the AUC values for 1, 3, and 5 years exceeded 0.6, suggesting a promising predictive performance of the model (Fig. 4i, j).

Fig. 4
figure 4

Constructed, evaluated, and validated a valid risk model. a Results of univariate Cox regression analysis. b Effective coefficient feature of LASSO model (For the accuracy of the model, lambda.min (0.008711847) was chosen as λ for model construction). c λ selection graph in LASSO model (The two dotted lines indicate two special lambda values, lambda.min on the left and lambda.1se on the right. lambda.min is more accurate and uses a larger number of genes). d The expression of five prognostic genes in tumor and normal tissues. e Risk curve for high and low risk groups (The risk graph is composed of two parts, with the horizontal coordinate being the same, which the patient’s risk value increasing from left to right). f Heat map of two risk groups (The expression of biomarkers in high and low risk groups was mapped using pheatmap package of R software). g, h K-M survival curve of the two risk groups. i, j ROC curves evaluate the effectiveness of risk models (The higher the AUC value, the higher the prediction accuracy)

Full size image

Nomogram exhibited strong predictive capability

Age, pathologic T and risk score were determined as independent prognostic factors by univariate Cox analysis, PH assumption test, and multivariate Cox analysis to develop a nomogram (Fig. 5a, b, c, d). The predictions for survival at 1, 3, and 5 years closely aligned with the calibration curve, while the AUC values for these timeframes exceeded 0.7, indicating a strong predictive capability of the nomogram (Fig. 5e, f).

Fig. 5
figure 5

Independent prognostic analysis of risk models. a Independent prognosis -univariate cox analysis results. b The relationship between Schoenfeld residuals and time (Each covariate satisfies the PH risk hypothesis). c Independent prognosis -multivariate cox analysis results. d The 1-, 3-, and 5-year survival rates were predicted by the nomogram (The higher the score, the lower the survival rate). e Calibration curve of nomogram (The closer the slope is to 1, the more accurate the prediction). f ROC curve of the nomogram model

Full size image

Differences in the immune microenvironment of different risk groups

Immune analysis results suggested that the scores of immune (R = 0.486), stromal (R = 0.309), and ESTIMATE (R = 0.422) were significantly different (p < 0.05) in risk groups (Fig. 6a). Significant differences in these scores were suggestive of marked alterations in the tumor microenvironment, which has a critical impact on tumor progression and prognosis, in different risk states. A noteworthy and favorable correlation was found between risk score and the scores of immune, stromal, and ESTIMATE (Fig. 6b, c, d). This suggests that as risk increases, the immune and mesenchymal status of the tumor microenvironment changes accordingly, the more unfavorable the prognosis of the tumor patients. In addition, there was a notable variance observed in the infiltration levels of 22 immune cell types between the two groups (Fig. 6e). The risk score existed a noteworthy positive association and Natural killer T cells, while a strong negative correlation with Type 17 T helper cells (Fig. 6f, g). In addition, in the training set, 21 immune checkpoint inhibitors exhibited notable variances (p < 0.05) among distinct risk categories (Fig. 6h). In the low-risk group, if the expression level of a particular immune checkpoint inhibitor was high, this may be associated with a better prognosis. Conversely, a higher expression level in a high-risk group is associated with a poorer prognosis. The analysis of correlation between risk score and 21 immune checkpoint inhibitors indicated that ADORA2A (Cor = 0.483263) displayed the highest positive correlation (Fig. 6i). The correlation analysis between the differential immune cells and the differential immune checkpoints showed that the immune cells with strong correlation with the differential immune checkpoints were 2-Methylbutyrate, Isovaleric acid, and the immune checkpoints with strong correlation between the differential immune cells and the differential immune checkpoints were 2-Methylbutyrate, Acetic acid, Valeric acid (Additional file 2a). We used the public immunotherapy cohort data, IMvigor210, to analyze the relationship between risk scores and patient response to immunotherapy (tumor shrinkage, overall survival). The results showed a p-value of 0.25 for risk score versus tumor shrinkage (the difference in the ratio of CR/PR versus SD/PD), suggesting that a significant correlation between risk score and tumor shrinkage was not demonstrated in this dataset (Additional file 2b). Specifically, the difference in the proportion of tumor shrinkage was not significant between the high and low risk score groups (p = 0.581), which may suggest that there was no direct correlation between risk score and the effect of immunotherapy on tumor shrinkage (Additional file 2c). However, there was some correlation between the risk score and overall survival, especially at the optimal stage value of − 0.2890324, which was shown to have a p-value of 0.046 by the Kaplan–Meier survival curve analysis, suggesting that the risk score has some clinical value in assessing the survival prognosis of colorectal cancer patients (Additional file 2d).

Fig. 6
figure 6

Immune microenvironment analysis. a The scores of immune, stromal, and ESTIMATE between two risk group. b–d Correlation between risk scores and 3 immune scores. e Infiltration abundance of 28 immune cells in the 2 risk groups (* indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001, **** indicates p < 0.0001). f Correlation between different immune cells. g Associated with RiskScore and different immune cells. h Expression of immune checkpoint inhibitors between 2 risk groups. i Correlation between risk score and different immune checkpoint inhibitors

Full size image

GSEA enrichment analysis of DEGs in different risk groups

We want to identify sets of genes or pathways that are associated with specific prognoses through GSEA analysis of differentially expressed genes in different risk groups. GSEA enrichment analysis demonstrated that pathways related to cancer exhibited a profound association with DEGs in two risk groups. The pathways of GO included uronic acid metabolic process, collagen trimer, synaptic membrane, cation channel complex and cation channel activity. According to the sorting by p.adjust, the top 5 pathways of KEGG were cell adhesion molecules cams, ascorbate and aldarate metabolism, neuroactive ligand receptor interaction, pentose and glucuronate interconvs and cytokine-cytokine receptor interaction (Fig. 7a, b, c, d).

Fig. 7
figure 7

GSEA enrichment analysis of DEGs in different risk groups. a–d Enrichment results of Top10 functional pathways between 2 risk groups. From a–d are BP, CC, MF, KEGG. A positive Enrichment Score indicates that the functional pathway is enriched in front of the gene sequencing sequence, while a negative Enrichment Score indicates that the functional pathway is enriched in the rear of the gene sequencing sequence. The horizontal coordinate represents the sequence of gene differences, and each small vertical line represents a gene

Full size image

Constructed a GGI network and carcinogenic pathways analysis for prognostic genes

GGI network could help researchers reveal interactions between prognostic genes. Conducting an analysis of the regulatory network and performing functional enrichment analysis on a set of 5 prognostic genes along with their 20 interacting genes in order to construct a GGI network (Additional file 3a). F5 participated in hemostasis, coagulation and blood coagulation pathways. The investigation into the influence of 5 prognostic genes on ten prevalent cancer pathways (Hormone AR, TSC/mTOR, EMT, PI3K/AKT, RTK, Hormone ER, DNA Damage Response, RAS/MAPK, cell cycle, and apoptosis pathways) indicated that MPP2 exhibited the most potent activating influence on the cell cycle pathway, whereas ANK3 demonstrated the most significant inhibitory effect within the apoptosis pathway (Additional file 3b).

Discussion

Colorectal cancer (CRC) is one of the most common malignant tumors, it is associated with high morbidity and mortality rates worldwide [6]. Currently, the main treatment for CRC are surgical resection, chemotherapy, and radiotherapy. Unfortunately, due to delayed diagnosis, metastasis, recurrence, and drug resistance, the CRC prognosis remains poor [33]. Consequently, the exploration of CRC pathogenesis and search for effective prognostic markers and treatment targets are significant importance for improving the quality of life and prolonging the survival period of CRC patients.

The JAK-STAT signaling pathway is an essential cellular cytokine signaling pathway which is tightly linked to the progression of CRC [34]. Aberrant activation of the JAK-STAT pathway is closely related to the occurrence and development of various tumors, including CRC [6]. The upstream receptors, downstream transcription factors and negative feedback regulators of the JAK-STAT signaling pathway may be undergo mutations or exhibit abnormal expression, leading to sustained activation or imbalance of the pathway, thereby promoting malignant transformation and immune evasion in CRC cells [35]. Therefore, this study has explored the genes associated with the JAK-STAT signaling pathway in CRC and established a new prognostic model.

We established a risk model that was based on the expression of five prognostic genes (ANK3, F5, FAM50B, KLHL35, and MPP2) to evaluate the prognostic risk of CRC patients. We divided the patients in the TCGA-CRC data set into groups of high- and low-risk according to the risk score, and found that the survival time of the high-risk group was significantly lower. Validation of the GSE39582 dataset also supported this finding. In addition, we developed a prognostic model based on risk scores and different clinical indicators, to predict 1-, 3-, and 5-year survival probabilities for CRC patients. These results indicate that our risk model and nominal graph have potential diagnostic value and can serve as auxiliary tools for prognostic assessment of CRC patients.

ANK3 is a protein involved in cytoskeleton and signal transduction, which undergoes mutations or abnormal expression in various tumors [36]. In earlier studies, ANK3 has an anticancer effect in the development of papillary thyroid carcinoma [37], and in breast cancers with high AR protein expression, ANK3 is independently associated with good prognosis [38]. However, in the study by KS the prognostic gene ANK3 showed a significant low expression state in renal clear cell carcinoma compared to normal tissues [39]. This phenomenon provided important clues into the role of ANK3 in cancer. From the perspective of model construction, ANK3 gene had a negative contribution to model score. When we combine the low expression of ANK3 with this negative contribution, we can infer that ANK3 may play a specific role in the prognosis of cancer. Further research has found that ANK3 regulate the specific peroxisome proliferator -activated receptor (PPAR) signaling pathway in CRC [39]. Moreover, abnormal ANK3 ‘s function may lead to disrupted regulation of cell membrane ion channels, thereby affecting intracellular ion concentration and altering signaling pathways such as PPAR and JAK-STAT to affect the prognosis of CRC patients.

Coagulation factor V (F5) is an important factor involved in blood coagulation and fibrinolysis, is mutated or abnormally expressed in a variety of tumors [40, 41]. F5 may increase the growth and metastasis of CRC cells by promoting angiogenesis and thrombosis [42]. Generally speaking, F5 participates in the coagulation cascade reaction, which is crucial for the hemostasis process [43]. However, the association between F5 and the JAK-STAT signaling pathway in CRC may be indirect. Because coagulation abnormalities in the tumor microenvironment may affect the function of immune cells, and the signal transduction of immune cells was closely related to the JAK-STAT pathway [44]. Therefore, abnormal expression of the F5 gene may cause changes in the coagulation status of the tumor microenvironment, thereby affecting the activation of the JAK-STAT signaling pathway by immune cells.

FAM50B, a protein involved in cell cycle and DNA repair, is mutated or abnormally expressed in a variety of tumors [45]. FAM50B may affect the expression of key genes in the JAK-STAT signaling pathway by regulating gene transcription in CRC [46]. Its mutation or abnormal expression may be a poor prognostic factor, resulting in dysregulation of the expression of relevant signaling molecules in the JAK-STAT pathway. Enhance the proliferation, survival, and metastasis ability of cancer cells.

KLHL35 is a protein involved in cytoskeleton and signal transduction, and is mutated or abnormally expressed in a variety of tumors [47]. KLHL35 helped maintain the normal turnover and homeostasis of intracellular proteins in normal cells. KLHL35 may promote the ubiquitination degradation of negative regulatory factors in the JAK-STAT pathway in CRC, thereby enhancing the activity of the JAK-STAT pathway and giving cancer cells stronger proliferation and invasion abilities, leading to poor prognosis [48]. In other ways, KLHL35 may affect the autophagy process of cells. Abnormal autophagy is closely related to the occurrence, development, and prognosis of CRC [47].

MPP2 is a protein involved in cell adhesion and signal transduction, can maintain the epithelial phenotype of CRC cells by stabilizing e-cadherin, and inhibit epithelial-mesenchymal transformation and transfer [49]. It may regulate the localization or stability of signaling molecules in the JAK-STAT pathway, affect the normal signal transduction of cells, and lead to changes in the proliferation, invasion, and metastasis ability of colorectal cancer cells [50]. In our analysis, the expressions of these prognostic genes were significantly different between CRC and control groups, and a causal relationship was found in combination with MR. Then the confirmation of these prognostic genes were also verified by RT-qPCR. The results showed that five prognostic genes were downregulated significantly in CRC patients versus control samples. These genes could be serve as potential targets for the prognosis of CRC.

We found significant differences in multiple biological pathways between two risk groups through gene set enrichment analysis. According to the sorting by KEGG, top 5 pathways were cell adhesion molecules (CAMs), ascorbate and aldarate metabolism, neuroactive ligand receptor interaction, pentose and glucuronate interconvs and cytokine-cytokine receptor interaction. CAMs play a crucial role in mediating cell–cell interactions, including immune cell trafficking, activation, and progression [51]. Ascorbate (Vitamin C) is an antioxidant that detoxifies reactive oxygen species. Alterations in this pathway may influence CRC growth and progression [52]. Neuroactive ligands and their receptors modulate inflammation and tumor microenvironment. Understanding these interactions can aid in detecting CRC and predicting outcomes [53]. The pentose phosphate pathway converts glucose to 5-phosphoribosyl-1-pyrophosphate and generates NADPH. Dysregulation in these pathways may contribute to the disease progression [54]. Cytokines are critical for immune responses and inflammation. Their dysregulation in CRC affects prognosis and pathogenesis [55]. In conclusion, these pathways provide valuable insights into development and potential therapeutic targets of CRC.

By immunoassay, we found significant differences in immune infiltration, immune score, and immune checkpoint inhibitor expression between two risk groups. We found that the high-risk group had signally higher levels of immunocyte infiltration than another group, including activated B cells, gamma delta T cells (γδ T cells), myeloid-derived suppressor cells (MDSC), type 1 T helper cells, mast cells, macrophages, and natural killer cells. These immune cells have different roles in the tumor immune microenvironment, both inhibiting tumor growth and metastasis, and promoting tumor escape and drug resistance [56,57,58]. γδ T cells can recognize and kill tumor cells directly. They are also involved in the production of cytokines like IL-17, which can have both tumor-promoting and tumor-suppressing roles [59]. MDSC often accumulate in the tumor microenvironment and can suppress T cell responses, promoting tumor growth [60]. Therefore, the balance between these pro-tumor and anti-tumor immune cells can influence the progression and outcome of CRC.

We also found that the high-risk group had signally higher immune scores than the low-risk group. Immune checkpoint inhibitors were significantly higher expressed in the high-risk group than another group, including ADORA2A, BTLA, CD274, CTLA4, HAVCR2, IDO1, LAG3, PDCD1, PDCD1LG2 and TIGIT. Among these immune checkpoints, ADORA2A is a part of the G protein-coupled receptor superfamily that promotes CRC cell proliferation and inhibits apoptosis through activation of the PI3K/AKT signaling pathway [56]. High levels of ADORA2A expression in CRC cells correlate with increased cell proliferation, migration, and invasion, whereas its knockdown leads to the opposite effects [61]. This suggests that ADORA2A may facilitate CRC progression and could be a target for management and treatment.

Immune checkpoint inhibitors play an important role in tumor immunotherapy. By blocking the interaction between immune checkpoint inhibitors and their ligands, the activity of immune cells can be restored, thus enhancing the immune clearance of tumors [62]. Our results show that the expression levels of immune checkpoint inhibitors are higher in the high-risk group than in the low-risk group, which may mean that tumor cells in the high-risk group are more likely to evade the surveillance and kill of immune system, leading to a poorer prognosis. In summary, the results of our immunoassay showed the different immune microenvironment between the two risk groups, providing new guidance and basis for immunotherapy in CRC patients.

In cancer research, it is critical to understand the impact of prognostic genes on cancer pathways. This study explored the effects of 5 prognostic genes on ten prevalent cancer pathways, which indicated that MPP2 exhibits the most potent activating influence on the cell cycle pathway, whereas ANK3 demonstrates the most significant inhibitory effect within the apoptosis pathway. The role of these genes in these pathways may provide new targets for cancer therapy. For example, the PI3K/AKT/mTOR and RAF/MEK/ERK pathways are common targets in cancer therapy because they are often activated at key targets through mutations and chromosomal translocations [63, 64]. The interactions of these pathways are involved in tumor formation, and inhibitors targeting these pathways have shown antitumor activity in clinical trials [65]. In addition, this study highlights the influence of prognostic genes in cancer pathways, in particular the activation role of MPP2 in the cell cycle pathway and the inhibition role of ANK3 in the apoptotic pathway, providing new targets for cancer therapy.

In this study, a new prognostic model was constructed by screening and validating the prognostic genes associated with the JAK-STAT signaling pathway in CRC, and immunoassay and carcinogenic pathway analysis were performed. This prognostic model has high predictive accuracy and potential clinical application value, providing new ideas for guiding the prognosis and therapy of CRC patients.

The limitation of this study is that we only used publicly available transcriptome data, not genome, epigenome, proteome, etc. Secondly, due to the heterogeneity of tumors, instrumental variables in MR methods may not accurately capture all biological pathways associated with CRC. Besides, we also did not carry out more researches about in-depth experimental mechanism and clinical application. Therefore, in future research, we hope to establish a comprehensive method that combines the detection of blood samples and tissue samples. Simultaneously consider conducting case–control studies to enhance the robustness and credibility of the research. On the one hand, we will delve into the expression patterns of identified genes in the blood to determine their ability to reflect the situation in tumors. On the other hand, we will expand the research scope of tissue samples, including collecting more tissue samples from colon cancer patients from different regions, races, and treatment backgrounds, to gain a more comprehensive understanding of the expression stability and specificity of these genes in different patient populations. By comparing the expression of these genes in blood and tissue samples, a more comprehensive and accurate clinical risk stratification system can be constructed. Our study provides a new perspective and method for CRC prognostication and treatment, and new clues and targets for CRC molecular mechanism and immunotherapy.

Conclusion

Our study shows that five prognostic genes (ANK3, F5, FAM50B, KLHL35, MPP2) have been screened, which are promising targets for tumor immunotherapy. In addition, a prognostic model was developed, which can be used as an effective prognostic model for CRC patients.

Data availability

The datasets analysed during the current study are available in the [University of California Santa Cruz (UCSC) Xena] repository, [https://xenabrowser.net], [Molecular Signatures Database (MsigDB)], [https://www.gsea-msigdb.org/gsea/login.jsp] and [Gene Expression Omnibus (GEO) database], [https://www.ncbi.nlm.nih.gov/] with login numbers (GSE39582).

Abbreviations

CRC:

Colorectal cancer

JAK-STAT-DEGs:

JAK-STAT-differentially expressed genes

WGCNA:

Weighted gene co-expression network analysis

MR:

Mendelian randomization

LASSO:

Least absolute shrinkage and selection operator

GSEA:

Gene set enrichment analysis

JAK-STAT:

Janus kinase signal transducer and activator of transcription

IVs:

Instrumental variables

COAD:

Colon adenocarcinoma

READ:

Rectum adenocarcinoma

UCSC:

University of California Santa Cruz

GEO:

Gene expression omnibus

MsigDB:

Molecular signatures database

ssGSEA:

Single sample gene set enrichment analysis

GO:

Gene Ontology

KEGG:

Kyoto Encyclopedia of Genes and Genomes

SNPs:

Single nucleotide polymorphisms

LD:

Linkage disequilibrium

IVW:

Inverse variance weighted

LOO:

Leave-One-Out

KM:

Kaplan-Meier

ROC:

Receiver operating characteristic

AUC:

Area under curves

BP:

Biological processes

MF:

Molecular functions

CC:

Cellular components

GSCALite:

Gene set cancer analysis Lite

RT-qPCR:

Reverse transcription quantitative polymerase chain reaction

CAMs:

Cell adhesion molecules

γδ T cells:

Gamma delta T cells

MDSC:

Myeloid-derived suppressor cells

References

  1. Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin. 2023;73:17–48.

    Article  PubMed  Google Scholar 

  2. Fearon ER. Molecular genetics of colorectal cancer. Annu Rev Pathol. 2011;6:479–507.

    Article  CAS  PubMed  Google Scholar 

  3. Xin P, Xu X, Deng C, Liu S, Wang Y, et al. The role of JAK/STAT signaling pathway and its inhibitors in diseases. Int Immunopharmacol. 2020;80: 106210. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.intimp.2020.106210.

    Article  CAS  PubMed  Google Scholar 

  4. Huang IH, Chung WH, Wu PC, Chen CB. JAK-STAT signaling pathway in the pathogenesis of atopic dermatitis: an updated review. Front Immunol. 2022;13:1068260. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fimmu.2022.1068260.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Shao F, Pang X, Baeg GH. Targeting the JAK/STAT signaling pathway for breast cancer. Curr Med Chem. 2021;28(25):5137–51.

    Article  CAS  PubMed  Google Scholar 

  6. Ghasemian A, Omear HA, Mansoori Y, Mansouri P, Deng X, Darbeheshti F, et al. Long non-coding RNAs and JAK/STAT signaling pathway regulation in colorectal cancer development. Front Genet. 2023;14:1297093. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fgene.2023.1297093.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Bowden J, Holmes MV. Meta-analysis and Mendelian randomization: a review. Res Synth Methods. 2019;10(4):486–96. https://doiorg.publicaciones.saludcastillayleon.es/10.1002/jrsm.1346.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Murphy N, Song M, Papadimitriou N, Carreras-Torres R, Langenberg C, Martin RM, et al. Associations between glycemic traits and colorectal cancer: a Mendelian randomization analysis. J Natl Cancer Inst. 2022;114(5):740–52.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Liu J, Li J, Tang Y, Zhou K, Zhao X, Zhang J, et al. Transcriptome analysis combined with Mendelian randomization screening for biomarkers causally associated with diabetic retinopathy. Front Endocrinol (Lausanne). 2024;15:1410066. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fendo.2024.1410066.

    Article  PubMed  Google Scholar 

  10. Jiang F, Chen X, Shen Y, Shen X. Identification and validation of an m6A modification of JAK-STAT signaling pathway-related prognostic prediction model in gastric cancer. Front Genet. 2022;13: 891744. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fgene.2022.891744.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Li R, Yang YE, Yin YH, Zhang MY, Li H, Qu YQ. Methylation and transcriptome analysis reveal lung adenocarcinoma-specific diagnostic biomarkers. J Transl Med. 2019;17(1):324. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12967-019-2068-z.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Mullan KA, Bramberger LM, Munday PR, Goncalves G, Revote J, Mifsud NA. ggVolcanoR: a Shiny app for customizable visualization of differential expression datasets. Comput Struct Biotechnol J. 2021;19:5735–40.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics. 2016;32(18):2847–9.

    Article  CAS  PubMed  Google Scholar 

  14. Wang H, Zhao Y, Zhang C, Li M, Jiang C, Li Y. Rab27a was identified as a prognostic biomaker by mRNA profiling, correlated with malignant progression and subtype preference in gliomas. PLoS ONE. 2014;9(2): e89782. https://doiorg.publicaciones.saludcastillayleon.es/10.1371/journal.pone.0089782.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Mao W, Ding J, Li Y, Huang R, Wang B. Inhibition of cell survival and invasion by Tanshinone IIA via FTH1: A key therapeutic target and biomarker in head and neck squamous cell carcinoma. Exp Ther Med. 2022;24(2):521. https://doiorg.publicaciones.saludcastillayleon.es/10.3892/etm.2022.11449.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinform. 2008;9:559. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/1471-2105-9-559.

    Article  CAS  Google Scholar 

  17. Chen H, Boutros PC. VennDiagram: a package for the generation of highly-customizable Venn and Euler diagrams in R. BMC Bioinform. 2011;12:35. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/1471-2105-12-35.

    Article  Google Scholar 

  18. Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC. DAVID: database for annotation, visualization, and integrated discovery. Genome Biol. 2003;4(5):P3.

    Article  PubMed  Google Scholar 

  19. Burgess S, Cronjé HT. Incorporating biological and clinical insights into variant choice for Mendelian randomisation: examples and principles. eGastroenterology. 2024;2(1):e100042. https://doiorg.publicaciones.saludcastillayleon.es/10.1136/egastro-2023-100042.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7: e34408. https://doiorg.publicaciones.saludcastillayleon.es/10.7554/eLife.34408.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017;32(5):377–89.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304–14.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Burgess S, Scott RA, Timpson NJ, Davey Smith G, Thompson SG. Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors. Eur J Epidemiol. 2015;30(7):543–52.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Chen X, Kong J, Diao X, Cai J, Zheng J, Xie W. Depression and prostate cancer risk: a Mendelian randomization study. Cancer Med. 2020;9(23):9160–7. https://doiorg.publicaciones.saludcastillayleon.es/10.1002/cam4.3493.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Hu J, Song J, Chen Z, Yang J, Shi Q, Jin F. Reverse causal relationship between periodontitis and shortened telomere length: bidirectional two-sample Mendelian random analysis. Front Immunol. 2022;13:1057602. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fimmu.2022.1057602.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Engebretsen S, Bohlin J. Statistical predictions with glmnet. Clin Epigenetics. 2019;11(1):123. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13148-019-0730-1.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Liu TT, Li R, Huo C, Li JP, Yao J, Ji XL, Qu YQ. Identification of CDK2-related immune forecast model and ceRNA in lung adenocarcinoma, a pan-cancer analysis. Front Cell Dev Biol. 2021;9: 682002. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fcell.2021.682002.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Heagerty PJ, Lumley T, Pepe MS. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics. 2000;56(2):337–44.

    Article  CAS  PubMed  Google Scholar 

  29. Xu J, Yang T, Wu F, Chen T, Wang A, Hou S. A nomogram for predicting prognosis of patients with cervical cerclage. Heliyon. 2023;9(11): e21147. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.heliyon.2023.e21147.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Syst Biol. 2010;59(3):307–21.

    Article  CAS  PubMed  Google Scholar 

  31. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25(4):402–8. https://doiorg.publicaciones.saludcastillayleon.es/10.1006/meth.2001.1262.

    Article  CAS  PubMed  Google Scholar 

  33. Wang J, Zhang Y, Song H, Yin H, Jiang T, Xu Y, et al. The circular RNA circSPARC enhances the migration and proliferation of colorectal cancer by regulating the JAK/STAT pathway. Mol Cancer. 2021;20(1):81. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12943-021-01375-x.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Wang S, Zhang W. Genetic variants in IL-6/JAK/STAT3 pathway and the risk of CRC. Tumour Biol. 2016;37(5):6561–9.

    Article  CAS  PubMed  Google Scholar 

  35. Philips RL, Wang Y, Cheon H, Kanno Y, Gadina M, Sartorelli V, et al. The JAK-STAT pathway at 30: Much learned, much more to do. Cell. 2022;185(21):3857–76.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Tan Y, Meng W, Jiang Z, Li N, Zhang T, Zhang J, et al. A comprehensive analysis of the prognostic and immunological role of ANK3 in pan-cancer. Transl Cancer Res. 2024;13(2):1091–113.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Zeng C, Long J, Deng C, Xie L, Ma H, Guo Y, et al. Genetic alterations in papillary thyroid carcinoma with Hashimoto’s thyroiditis: ANK3, an indolent maintainer of papillary thyroid carcinoma. Front Oncol. 2022;12: 894786. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fonc.2022.894786.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Kurozumi S, Joseph C, Raafat S, Sonbul S, Kariri Y, Alsaeed S, et al. Utility of ankyrin 3 as a prognostic marker in androgen-receptor-positive breast cancer. Breast Cancer Res Treat. 2019;176(1):63–73.

    Article  CAS  PubMed  Google Scholar 

  39. Somsuan K, Aluksanasuwan S. Bioinformatic analyses reveal the prognostic significance and potential role of ankyrin 3 (ANK3) in kidney renal clear cell carcinoma. Genomics Inform. 2023;21(2): e22. https://doiorg.publicaciones.saludcastillayleon.es/10.5808/gi.23013.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Andresen MS, Sletten M, Sandset PM, Iversen N, Stavik B, Tinholt M. Coagulation factor V (F5) is an estrogen-responsive gene in breast cancer cells. Thromb Haemost. 2022;122(8):1288–95.

    Article  PubMed  Google Scholar 

  41. Leppert U, Eisenreich A. The role of tissue factor isoforms in cancer biology. Int J Cancer. 2015;137(3):497–503.

    Article  CAS  PubMed  Google Scholar 

  42. Vossen CY, Hoffmeister M, Chang-Claude JC, Rosendaal FR, Brenner H. Clotting factor gene polymorphisms and colorectal cancer risk. J Clin Oncol. 2011;29(13):1722–7. https://doiorg.publicaciones.saludcastillayleon.es/10.1200/JCO.2010.31.8873.

    Article  CAS  PubMed  Google Scholar 

  43. Tinholt M, Stavik B, Tekpli X, Garred Ø, Borgen E, Kristensen V, et al. Coagulation factor V is a marker of tumor-infiltrating immune cells in breast cancer. Oncoimmunology. 2020;9(1):1824644. https://doiorg.publicaciones.saludcastillayleon.es/10.1080/2162402X.2020.1824644.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Lind SM, Sletten M, Hellenes M, Mathelier A, Tekpli X, Tinholt M, et al. Coagulation factor V in breast cancer: a p53-regulated tumor suppressor and predictive marker for treatment response to chemotherapy. J Thromb Haemost. 2024;22(6):1569–82.

    Article  PubMed  Google Scholar 

  45. Thompson NA, Ranzani M, van der Weyden L, Iyer V, Offord V, Droop A, et al. Combinatorial CRISPR screen identifies fitness effects of gene paralogues. Nat Commun. 2021;12(1):1302. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41467-021-21478-9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Ding Q, Hou Z, Zhao Z, Chen Y, Zhao L, Xiang Y. Identification of the prognostic signature based on genomic instability-related alternative splicing in colorectal cancer and its regulatory network. Front Bioeng Biotechnol. 2022;10: 841034. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fbioe.2022.841034.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Zou J, Zhang H, Huang Y, Xu W, Huang Y, Zuo S, et al. Multi-omics analysis of the tumor microenvironment in liver metastasis of colorectal cancer identified FJX1 as a novel biomarker. Front Genet. 2022;13: 960954. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fgene.2022.960954.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Hu B, Chen Z, Yao F, Li B. Construction of a risk model for colon cancer prognosis based on ubiquitin-related genes. Turk J Gastroenterol. 2023;34(5):449–56.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Di Z, Zhou S, Xu G, Ren L, Li C, Ding Z, et al. Single-cell and WGCNA uncover a prognostic model and potential oncogenes in colorectal cancer. Biol Proced Online. 2022;24(1):13. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12575-022-00175-x.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Yamada T, Saitoh Y, Kametani K, Kamijo A, Sakamoto T, Terada N. Involvement of membrane palmitoylated protein 2 (MPP2) in the synaptic molecular complex at the mouse cerebellar glomerulus. Histochem Cell Biol. 2022;158(5):497–511.

    Article  CAS  PubMed  Google Scholar 

  51. Harjunpää H, Llort Asens M, Guenther C, Fagerholm SC. Cell adhesion molecules and their roles and regulation in the immune and tumor microenvironment. Front Immunol. 2019;10:1078. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fimmu.2019.01078.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Joly JH, Chew BTL, Graham NA. The landscape of metabolic pathway dependencies in cancer cell lines. PLoS Comput Biol. 2021;17(4): e1008942. https://doiorg.publicaciones.saludcastillayleon.es/10.1371/journal.pcbi.1008942.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Lin H, Xia L, Lian J, Chen Y, Zhang Y, Zhuang Z, et al. Delineation of colorectal cancer ligand-receptor interactions and their roles in the tumor microenvironment and prognosis. J Transl Med. 2021;19(1):497. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12967-021-03162-0.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Sun H, Zhang AH, Song Q, Fang H, Liu XY, Su J, et al. Functional metabolomics discover pentose and glucuronate interconversion pathways as promising targets for Yang Huang syndrome treatment with Yinchenhao Tang. RSC Adv. 2018;8(64):36831–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. UConn today. UConn immunology researcher unraveling relationship between cytokines and colorectal cancer. 2021 https://today.uconn.edu/2021/10/uconn-immunology-researcher-unraveling-relationship-between-cytokines-and-colorectal-cancer/ [Accessed October 18, 2021]

  56. Wu SY, Fu T, Jiang YZ, Shao ZM. Natural killer cells in cancer biology and therapy. Mol Cancer. 2020;19(1):120. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12943-020-01238-x.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Gao J, Liang Y, Wang L. Shaping polarization of tumor-associated macrophages in cancer immunotherapy. Front Immunol. 2022;13: 888713. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fimmu.2022.888713.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Liu X, Li X, Wei H, Liu Y, Li N. Mast cells in colorectal cancer tumour progression, angiogenesis, and lymphangiogenesis. Front Immunol. 2023;14:1209056. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fimmu.2023.1209056.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Ma R, Yuan D, Guo Y, Yan R, Li K. Immune effects of γδ T cells in colorectal cancer: a review. Front Immunol. 2020;11:1600. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fimmu.2020.01600.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  60. Galati D, Zanotta S, Bocchino M, De Filippi R, Pinto A. The subtle interplay between gamma delta T lymphocytes and dendritic cells: is there a role for a therapeutic cancer vaccine in the era of combinatorial strategies? Cancer Immunol Immunother. 2021;70(7):1797–809.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  61. Ran L, Mou X, Peng Z, Li X, Li M, Xu D, et al. ADORA2A promotes proliferation and inhibits apoptosis through PI3K/AKT pathway activation in colorectal carcinoma. Sci Rep. 2023;13(1):19477. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41598-023-46521-1.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  62. Bagchi S, Yuan R, Engleman EG. Immune checkpoint inhibitors for the treatment of cancer: clinical impact and mechanisms of response and resistance. Annu Rev Pathol. 2021;16:223–49.

    Article  CAS  PubMed  Google Scholar 

  63. Corrales E, Levit-Zerdoun E, Metzger P, Mertes R, Lehmann A, Münch J, et al. PI3K/AKT signaling allows for MAPK/ERK pathway independency mediating dedifferentiation-driven treatment resistance in melanoma. Cell Commun Signal. 2022;20(1):187. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12964-022-00989-y.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  64. Maharati A, Moghbeli M. PI3K/AKT signaling pathway as a critical regulator of epithelial-mesenchymal transition in colorectal tumor cells. Cell Commun Signal. 2023;21(1):201. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12964-023-01225-x.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Li Q, Li Z, Luo T, Shi H. Targeting the PI3K/AKT/mTOR and RAF/MEK/ERK pathways for cancer therapy. Mol Biomed. 2022;3(1):47. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s43556-022-00110-2.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We would like to acknowledge all the people who were involved in this project and supported it.

Funding

This study was supported by the Key Scientific Research Project of the University in Henan Province (23A310027).

Author information

Authors and Affiliations

  1. Department of Oncology and Rehabilitation, Zhengzhou Central Hospital Affiliated to Zhengzhou University, No. 16 Tongbai North Road, Zhengzhou, Henan, China

    Nan Zhang, Wenli Yue, Bihang Jiao, Duo Cheng, Jingjing Wang, Fang Liang, Yingnan Wang & Xiyue Liang

  2. Department of Gastroenterology, Zhengzhou Central Hospital Affiliated to Zhengzhou University, Zhengzhou, Henan, China

    Kunkun Li

  3. Medical Key Laboratory for Diagnosis and Treatment of Colorectal Cancer in Henan Province, Zhengzhou, Henan, China

    Kunkun Li

  4. Department of Anorectal Surgery, Zhengzhou Central Hospital Affiliated to Zhengzhou University, Zhengzhou, Henan, China

    Junwei Liu

  5. Department of Gastrointestinal Surgery, Zhengzhou Central Hospital Affiliated to Zhengzhou University, Zhengzhou, Henan, China

    Yadong Li

Authors
  1. Nan Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Wenli Yue
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Bihang Jiao
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Duo Cheng
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Jingjing Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Fang Liang
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Yingnan Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Xiyue Liang
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Kunkun Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Junwei Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Yadong Li
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

NZ: Data curation, Validation, Writing-original draft, Writing-review & editing. DC, JL and YL: Formal analysis, Validation, Data curation, Writing-original draft. WY, JW, BJ, YW, XL: Investigation, Methodology, Writing-review & editing. FL, KL: Funding acquisition, Project administration, Resources, Supervision, Writing—review & editing. All authors have read and agreed to the published version of the manuscript.

Corresponding author

Correspondence to Nan Zhang.

Ethics declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Zhengzhou Central Hospital (NO. ZXYY2024186). The participants provided their written informed consent to participate in this study.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

13027_2025_640_MOESM1_ESM.xlsx

Additional file 1:Results of MR analysis of 66 JAK STAT DEGs: 11 genes served as risk factors, while 55 genes function as safety factors

13027_2025_640_MOESM2_ESM.tif

Additional file 2:Immunocorrelation analysis results.Heatmap of the correlation between differential immune cells and differential immune checkpoints. Differential immune cells on the horizontal axis and differential immune checkpoints on the vertical axis, with the color indicating the size of the correlation, where the larger the correlation, the darker the color, with red being a positive correlation and blue being a negative correlation.Results of risk score.Tumor shrinkage difference analysis.KM curve between risk score and overall survival.

13027_2025_640_MOESM3_ESM.tif

Additional file 3:Carcinogenic pathways analysis.The GeneMANIA database constructs PPI regulatory networks for biomarkers.Intensity of biomarker action in the cancer-associated pathway.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhang, N., Yue, W., Jiao, B. et al. Unveiling prognostic value of JAK/STAT signaling pathway related genes in colorectal cancer: a study of Mendelian randomization analysis. Infect Agents Cancer 20, 9 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13027-025-00640-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13027-025-00640-8

Keywords

Infectious Agents and Cancer

ISSN: 1750-9378

Contact us