Abstract
Background/Aim: Glycolysis has a role in regulating the tumor immune microenvironment. However, the functions and clinical role for facilitating the prognosis prediction of colorectal cancer (CRC) based on glycolysis and immune-related genes remain to be identified. Materials and Methods: Genes associated with glycolysis and immunity (GI) were identified from established databases (MSigDB and ImmPort). The TCGA (training cohort) and GSE39582 (validation cohort) datasets were used. Cox regression and least absolute shrinkage and selection operator (LASSO) Cox regression analyses were applied for model construction. The prognostic power of the GI signature was examined by multivariate Cox regression analysis. The correlations between the GI signature, immune cell infiltration and immune checkpoint blockade (ICB) genes were analyzed. To further validate the identified gene signature, quantitative RT-PCR was performed. Cell proliferation assays were conducted for CCK8 detection. Results: A new GI model was constructed, and this signature may serve as an independent prognostic biomarker in CRC. The GI signature remained an effective tool for predicting prognosis among each clinical subgroup. This signature was related to immune cell infiltration of myeloid dendritic cells, cancer-associated fibroblasts (CAFs), CD4+ and CD8+ T cells and response to the ICB immunotherapy-related genes IDO1, BTLA, PD-L1 and PD-L2. In addition, our findings showed that PMM2, IL20RB, and NTF4 exhibited high expression levels in CRC. The upregulation of these genes resulted in the promotion of the proliferation of CRC cells. Conclusion: This novel prognostic signature contributed to CRC risk stratification and survival prediction based on glycolysis and immune status.
Colorectal cancer (CRC) is the third common malignant tumor and ranks second among death-related causes in all cancers (1). Based on global cancer statistics in 2020, more than 1.9 million new cancer cases are diagnosed with CRC, with approximately 935,000 deaths due to this disease (1). Recent advances in surgical resection, chemotherapy, radiotherapy, molecular targeted therapy, and immunotherapy have significantly improved outcomes for patients with CRC. Although these treatment approaches have improved patients’ outcomes, the 5-years survival rates of many CRC patients remain poor (2-4). Thus, it is needed to determine molecular biomarkers to predict the prognosis of CRC and develop personalized treatment regimens.
Metabolic reprogramming is an important hallmark of cancer, which promotes malignant transformation and tumor development, including tumor cell invasion and metastatic abilities (5). Tumor metabolic reprogramming has an important role in regulating the antitumor immune response in the tumor microenvironment (TME) (6). The “Warburg effect” represents a prevalent form of metabolic reprogramming characterized by aerobic glycolysis. It is featured by the process of transformation of the excess glucose into the production of lactate (7, 8). Tumor glycolysis can facilitate the aggressiveness, progression, and metastasis of cancer cells and has been implicated in resistance to therapy, such as conventional chemotherapy and radiotherapy (7, 9). Tumor glycolysis can impact the capability of T cells to inhibit tumors and plays a crucial role in regulating myeloid-derived suppressor cells (MDSCs), tumor-associated macrophages (TAMs), and tumor-associated neutrophils (TANs) on tumor immunity (10, 11). Thus, tumor glycolysis could be a potential biochemical target for the treatment of cancer. Recently, some studies have reported on the role of a combined signature of glycolysis and immune prognostic model in several cancers, such as melanoma, hepatocellular carcinoma, and lung squamous cell carcinoma (12-14). However, the relationship and clinical significance of glycolysis with immunity in CRC remains largely unclear.
In the current study, we conducted a comprehensive analysis to construct a new glycolysis and immune (GI) signature based on glycolysis and immune-related genes. We analyzed the association of the GI signature with immune checkpoint molecules and immune infiltration in the TME in CRC. We also investigated the prognostic and predictive values of the GI signature for CRC patients.
Patients and Methods
Ethics approval and consent to participate. All procedures performed in studies involving human participants were in accordance with the ethical standards of the Gene Expression Omnibus and The Cancer Genome Atlas Human Subjects Protection and Data Access Policies, adopted by the National Cancer Institute (NCI) and the National Human Genome Research Institute (NHGRI). Informed consent was obtained from all individual participants included in the study.
Transcriptome and clinical data. CRC expression patterns of mRNA and available clinical data were downloaded from The Cancer Genome Atlas (TCGA) database (https://cancergenome.nih.gov). Patients who had no prognostic data were excluded. Expression data were normalized through a log2(x+1) transformation. A total of 618 tumors with CRC were identified in the TCGA data. In addition, GSE39582 of CRC for the Cartes d’Identité des Tumeurs (CIT) program was obtained from the Gene-Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/). This dataset was based on the GPL570 platform [(HG-U133_Plus_2) Affymetrix Human Genome U133 Plus 2.0 Array]. The multichip Average (RMA) algorithm was applied to normalize the expression data and the GSE39582 dataset, performed using a log2 transformation. Patients with no survival data were excluded and 562 CRC samples were eventually included. The TCGA data were used as a training cohort and the CIT data were used as a validation cohort. The detailed study characteristics are listed in Table I.
Detailed study characteristics.
Identification of the GI signature. Glycolysis-related genes (n=200) were collected from the Molecular Signatures Database (MSigDB) (15). Immune-related genes (n=1,793) were collected from the immunology database and analysis portal (ImmPort) system (https://www.immport.org/home).
We performed the following three steps to construct a prognostic model. Firstly, univariate Cox regression analysis was used to screen for significant prognostic results. p-Values <0.05 indicated statistical significance. Secondly, for significant genes, we then used the least absolute shrinkage and selection operator (LASSO) regression method to prevent model overfitting within them. Finally, we determined the risk signature using the multivariate Cox proportional hazards regression model (two-step method). Seven important genes were applied to establish the GI signature, as shown below: GI score=∑Coefficienti × Expressioni.
Functional enrichment analysis. To investigate the possible biological functions of the GI signature, Gene ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were performed using the “clusterProfiler” R package. An adjusted p<0.05 was used as the cut-off criterion. Associations between the GI signature and genes were calculated using Spearman’s correlation coefficient. Significant genes (correlation coefficient > |0.3|, p<0.001) were selected, and these genes were further used for the GO and KEGG enrichment analyses.
Immune cell analysis. To evaluate the composition and density of immune cells in the TME, we used the CIBERSORT, TIMER and MCP counter methods. CIBERSORT, the ν-support vector regression algorithm, was used to quantify the fractions of cell types (16), including 22 types of immune cells (i.e., B cells, T cells, and NK cells subsets, monocytes, eosinophils and neutrophils etc.). TIMER, the linear least square regression algorithm, was applied to figure out the abundance of six immune cell types (17) (http://timer.cistrome.org/), including CD4+ T cells, CD8+ T cells, B cells, neutrophils, macrophages, and myeloid dendritic cells (DC). The MCPcounter algorithm is based on the mean of marker gene expression and was performed to evaluate the population abundance of 8 immune cell types, cytotoxicity score, cancer associated fibroblasts (CAFs), and endothelial cells (18).
ICB analysis. Treatment with the use of immune checkpoint blockade (ICB) has achieved significant progress in the field of cancer treatment (19). Several crucial targets of immunotherapy have been reported, including cytotoxic T-lymphocyte antigen 4 (CTLA-4), programmed death 1 (PD-1), programmed death ligand 1 (PD-L1), programmed death ligand 2 (PD-L2), B and T lymphocyte associated (BTLA), and indoleamine 2,3-dioxygenase 1 (IDO1) (20, 21). The relationship between the GI signature and ICB immunotherapy-related genes was analyzed to investigate the role of the GI signature for response to immunotherapy.
RNA isolation and RT-PCR analysis. Ruyao Biotechnology (Zhejiang, PR China) provided the human normal colonic epithelial cell line NCM460. The human colon adenocarcinoma cell line SW480 was acquired from the Chinese Academy of Sciences (Shanghai, PR China).
Total RNA from the cells was extracted using TRIzol (Aidlab, PR China), following the instructions provided by the manufacturer. cDNA amplification and synthesis were performed using a TaqMan miRNA reverse transcription kit (Thermo Fisher Scientific). Gene expression levels were standardized to GAPDH levels. Quantitative real-time PCR (qRT-PCR) was used to measure the expression levels of the identified gene signature. The primers used in the upstream and downstream experiments had the following sequences: CD1B-forward (F): 5′-GTTCTCTTTCCTGGTGGTAACAG-3′; CD1B-reverse (R): 5′-GGACGAGGTCTGGATAACATGAA-3′; INSL3-F: 5′-ACCCCAGA GATGCGTGAGAA-3′; INSL3-R: 5′-CTCCAGCCACTGTAGCAA CTC-3′; IL20RB-F: 5′-GGCCACTGTGCCATACAAC-3′; IL20RB-R: 5′-TCTTTGGTGATCTCCATCCCA-3; FABP4-F: 5′-ACTGGGCC AGGAATTTGACG-3′; FABP4-R: 5′-CTCGTGGAAGTGACGCCTT-3′; HSPA1A-F: 5′-TGGTGCAGTCCGACATGAAG-3′; HSPA1A-R: 5′-GCTGAGAGTCGTTGAAGTAGGC-3′; PMM2-F: 5′-CTTCGA CGTGGATGG GACC-3′; PMM2-R: 5′-CGCCTACCACTCCG ATTTTG-3′; NTF4-F: 5′-GTACTTCTTTGAAACCCGCTG-3′; NTF4-R: 5′-GCAGT GTCAATTCGAATCCATC-3′; GAPDH-F: 5′-CAGTGCCAGCCTCGTCTAT-3′; GAPDH-R: 5′-AGGGGCCATCCA CAGTCTTC-3′. The 2−ΔΔCT approach was utilized for quantifying relative fold changes in expression.
Cell proliferation assay. Cells were incubated in normal culture medium (DMEM) diluted with 10% CCK-8 reagent (Yihsheng Biotechnology, Shanghai, PR China) at a temperature of 37°C. The proliferation rates were measured 0, 1, 2, and 3 days post-transfection. The absorbance of each well was recorded using a microplate reader. The optical density (OD) value was measured using a microplate reader at a wavelength of 450 nm.
Statistical analysis. Statistical analyses were conducted using the R software version 4.1.2 (R Foundation for Statistical Computing, Vienna, Austria) and the IBM SPSS22.0 software. Comparisons between two groups were performed with Student t-test. Spearman correlation analysis was used to evaluate the potential relevance of the GI signature with immune cell infiltration and ICB therapy-related targets. The high- and low-risk groups were divided based on the median score of the training cohort (cut-off value=0.945), and Kaplan–Meier (KM) analysis was used to compare differences in overall survival (OS) between the high-risk and low-risk group. Univariate and multivariate COX regression analyses were used to evaluate the association between the GI signature and other variables and the prognosis of CRC. We calculated hazard ratios (HRs) and the corresponding 95% confidence intervals (CIs) for assessing the prognostic significance. The time-dependent receiver operating characteristics (ROC) curve and the area under the curve (AUC) values were calculated to evaluate the predictive power.
Results
Development of the GI signature. In the training set, we performed univariate Cox regression analysis to identify significant prognostic genes. This analysis was aimed at selecting candidate genes that were associated with survival. A total of 180 prognostic genes underwent further LASSO regression analysis, resulting in the identification of ten final genes. To optimize our model and identify only the most significant genes, a two-step multivariate Cox regression analysis was conducted for these ten genes. Eventually, we found the seven most important genes used in our GI signature, including CD1B, INSL3, IL20RB, FABP4, HSPA1A, PMM2 and NTF4. The GI signature was developed as following: (−0.293 × CD1B expression) + (0.297 × INSL3 expression) + (0.179 × IL20RB expression) + (0.114 × FABP4 expression) + (0.165 × HSPA1A expression) + (−0.367 × PMM2 expression) + (0.115 × NTF4 expression).
Biological function analysis of the GI signature. We conducted GO and KEGG enrichment analyses to uncover the potential function and biological pathways of the GI signature. The results showed that the GI signature was significantly enriched in some biological functions, such as wound healing, extracellular matrix organization, regulation of cellular response to growth factor stimulus, regulation of leukocyte migration, myeloid leukocyte migration, cell adhesion (focal adhesion, cell–cell junction, and cell–substrate adhesion), proteoglycans in cancer, extracellular matrix (ECM)−receptor interaction, and leukocyte transendothelial migration (Figure 1). This suggested that the GI signature may play a key role in CRC progression.
Biological functions of the glycolysis and immune-related gene (GI) signature. (A) Results of gene ontology (GO) enrichment analysis, including biological process (BP), cellular component (CC), and molecular function (MF). (B) Results of Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis.
Prognostic analysis of the GI signature. CRC tumors were divided into high-risk and low-risk groups based on the GI signature. The distribution of the GI signature and patients’ survival status are summarized in Figure 2A; these indicated that a high-risk score may have a markedly increased death rate compared with a low-risk score. In the KM survival analysis, patients in the high GI score group had a significantly worse prognosis than those in the low GI score group in the training and validation cohorts (all p-values <0.001) (Figure 2B). In the ROC analyses, the AUC values of the 1-, 3-, and 5-year survival rates were calculated to be 0.77, 0.74, and 0.65 for the TCGA training set, whereas they were 0.69, 0.65, and 0.65 for the validation cohort, respectively (Figure 2C). This suggests that the GI signature may act as a potential marker with a predictive value for the prognosis of CRC.
Prognostic and predictive significance of the glycolysis and immune-related gene (GI) signature in colorectal cancer (CRC). (A) The distribution of the GI signature risk score (upper panels) and patients’ risk survival status (lower panels). (B) Kaplan–Meier (KM) survival analysis showed the prognosis of the GI signature in the TCGA training set and the independent validation cohort (p<0.001). (C) Time-dependent receiver operating characteristics (ROC) curve analyses in the TCGA training set and validation cohort.
The GI signature as an independent factor for prognosis. To evaluate whether the GI signature was significant in independently predicting survival, univariate and multivariate Cox regression analyses were conducted for this signature along with clinical features (Table II). Results from the univariate COX analysis showed that the GI signature was significantly correlated with poor prognosis in both the training (HR=3.516, 95% CI=2.336-5.292, p<0.001) and validation (HR=1.824, 95% CI=1.355-2.454, p<0.001) sets. After correcting for other clinical characteristics, multivariate Cox regression analysis demonstrated that the GI signature was still a significant biomarker, independently predicting the prognosis of patients with CRC in the training (HR=2.707, 95% CI=1.675-4.372, p<0.001) and validation cohort (HR=1.796, 95% CI=1.318-2.448, p<0.001).
Univariate and multivariate Cox regression analyses of the GI signature along with clinical features.
Prognostic and predictive value of the GI signature in different clinical subgroups. To better investigate the clinical application value of the GI signature, the prognostic results of the GI signature with different clinical features were evaluated through stratification analyses. As a result, the high-GI score group showed significantly poorer survival than the low-GI score group among CRC patients with different clinical characteristics, including age (≥60 vs. <60 years), sex (male vs. female), tumor stage (stage 1-2 vs. stage 3-4) (Figure 3, all p-values <0.01), depth of invasion (T 1-2 vs. T3-4), lymph node metastasis (N: positive vs. negative), and distant metastasis (M: positive vs. negative) (Figure 4, all p-values <0.01). These results suggested that the GI signature was still effective for predicting the prognosis of CRC tumors within different subgroups.
The glycolysis and immune-related gene (GI) signature for predicting survival in the clinical subgroups of age, sex, and tumor stage (p<0.01 for all). (A) ≥60 years. (B) <60 years. (C) Male patients. (D) Female patients. (E) Stage 3-4. (F) Stage 1-2.
The glycolysis and immune-related gene (GI) signature for predicting survival in clinical subgroups of depth of invasion (T stage), lymph node metastasis, and distant metastasis (M) (p<0.01 for all). (A) T 3-4. (B) T 1-2. (C) Lymph node-positive tumors. (D) Lymph node-negative tumors. (E) Patients with distant metastasis. (F) Patients without distant metastasis.
Relationship between the GI signature, immune micro-environment and ICB therapy. To investigate the correlation between the GI signature and the immune microenvironment, we performed analyses between the GI signature and immune infiltrate cells. The infiltration levels of various immune cell types, endothelial cells and CAFs were obtained. The GI signature was negatively correlated with immune infiltration of ‘Myeloid DC resting’ (r=−0.231, p<0.001), ‘Myeloid DC’ (r=−0.171, p<0.001), ‘CD4 T cell memory activated’ (r=−0.159, p<0.001), ‘CD8 T cell’ (r=−0.154, p<0.001), ‘CD4 T cell memory resting’ (r=−0.145, p<0.001), and ‘B cell’ (r=−0.106, p=0.008) (Figure 5A-F), whereas this signature was positively correlated with immune infiltration of ‘CAF’ (r=0.163, p<0.001), ‘T cell regulatory (Tregs)’ (r=0.112, p=0.005), and ‘Macrophage M0’ (r=0.091, p=0.024) (Figure 5G-I). These analyses indicated that the GI signature may have a role in the regulation of the immune microenvironment.
Association between the glycolysis and immune-related gene (GI) signature and immune cell microenvironment. (A) Myeloid dendritic cell (DC) resting (r=−0.231, p<0.001). (B) Myeloid DC (r=−0.171, p<0.001). (C) CD4 T cell memory activated (r=−0.159, p<0.001). (D) CD8 T cell (r=−0.154, p<0.001). (E) CD4 T cell memory resting (r=−0.145, p<0.001). (F) B cell (r=−0.106, p=0.008). (G) Cancer-associated fibroblast (CAF) (r=0.163, p<0.001). (H) T cell regulatory (Tregs) (r=0.112, p=0.005). (I) Macrophage M0 (r=0.091, p=0.024).
Immunotherapy with ICB therapy is becoming a well-recognized cancer treatment, which has improved survival in advanced tumors. However, some cancer patients become resistant to immunotherapy (22). Therefore, this study investigated the expression of common ICB-related genes (CTLA-4, PD-1, PD-L1, PD-L2, BTLA, and IDO1) in CRC. The results demonstrated that the GI signature was negatively associated with the expression levels of IDO1 (r=−0.168, p<0.001), PD-L1 (r=−0.113, p=0.005), BTLA (r=−0.095, p=0.019), and PD-L2 (r=−0.093, p=0.02) (Figure 6). Despite the weak correlation coefficients observed between the GI signature, immune microenvironment, and ICB-related genes, the large sample size of the current study ensures that even these small correlations can achieve statistical significance.
Association between the glycolysis and immune-related gene (GI) signature and immune checkpoint blockade (ICB) therapy-related genes. (A) Indoleamine 2,3-dioxygenase 1 (IDO1) (r=−0.168, p<0.001). (B) Programmed death ligand 1 (PD-L1) (r=−0.113, p=0.005). (C) T lymphocyte associated (BTLA) (r=−0.095, p=0.019). (D) Programmed death ligand 2 (PD-L2) (r=−0.093, p=0.02).
Verification of the expression of the 7 genes. To investigate the gene expression patterns in both CRC and normal cell lines, RT-PCR was employed to detect mRNA expression levels. The results indicated that the expression of INSL3, FABP4, HSPA1A and CD1B did not differ significantly between the cancer and normal groups. The expression levels of NTF4, IL20RB, and PMM2 in the cancer group were significantly higher than those in the normal group (Figure 7).
Gene expression in colorectal cancer (CRC) cell lines was conducted using quantitative real-time PCR (qRT-PCR), for INSL3 (A), FABP4 (B), HSPA1A (C), CD1B (D), NTF4 (E), IL20RB (F) and PMM2 expression (G). NS: not significant, *p<0.05, **p<0.01, ***p<0.001.
Knockdown of NTF4, IL20RB and PMM2 suppressed CRC cell proliferation. By using si-NTF4, si-IL20RB and si-PMM2, we built NTF4, IL20RB, and PMM2-knockdown cell lines and validated the knockdown efficiency through RT-PCR analysis. We observed that the expression of the corresponding genes in CRC was significantly inhibited after interfering with PMM2, IL20RB, and NTF4, respectively (Figure 8A-C). In addition, the results obtained from CCK-8 assays demonstrated that the knockdown of PMM2, IL20RB, and NTF4 significantly inhibited the proliferation of CRC cells (Figure 8D-F). Our findings reveal that PMM2, IL20RB, and NTF4 may impact the prognosis of CRC patients by promoting CRC proliferation.
Knockdown of genes suppressed the proliferation of colorectal cancer (CRC) cells. (A-C) PMM2, IL20RB and NTF4 expression after transfection with siRNA. (D-F) CCK8 assays evaluated the proliferation of SW480 cells after PMM2, IL20RB, and NTF4 knockdown. **p<0.01, ***p<0.001.
Discussion
CRC is largely an asymptomatic tumor, and many patients are usually diagnosed at an advanced stage (23). Recently, Pawelka et al. investigated the diagnostic and predictive potential of miRNA expression profiles in CRC (24), highlighting the necessity to find useful molecular biomarkers to accurately and timely improve the ability of early detection and predict clinical outcomes of CRC patients. Tumor glycolysis is not only closely correlated with tumor invasion and metastasis, and response to drug resistance but also may have an inhibitory role in the immune environment (10, 25). For example, the overexpression of PDK1, a glycolysis enzyme, is associated with poor prognosis of CRC patients, and knockdown of this gene decreases the liver metastasis of CRC in mice (26). Tumor glycolysis under the regulation of c-Myc has an effect on NK cell immune activity in CRC (27). The complex crosstalk between glycolysis and the immune microenvironment reflects a highly promising approach for cancer immunotherapy and molecular targeted therapy (10, 28). In this study, we constructed the GI signature using glycolysis- and immune-related genes to explore its predictive efficacy and prognostic significance for CRC, which provided a reference for clinical decision-making for the treatment of patients with CRC.
In the present study, the GI signature identified was correlated with a worse prognosis of CRC. This signature can act as an independent prognostic factor for predicting the survival of patients with CRC after adjusting for other clinical features, such as tumor stage, depth of invasion, lymph node metastasis, and distant metastasis. The GI signature may be involved in the regulation of immune infiltration and predict response to ICB immunotherapy in CRC. Therefore, the signature could be leveraged to develop novel therapeutic strategies and demonstrates potential value for clinical applications.
Yang et al. reported that a GI signature was correlated with poor prognosis and can be an independent prognostic factor for predicting melanoma patient outcomes (14). Hu et al. found that the glycolysis- and immune-related prognosis signature in the high-risk group was significantly associated with worse survival and could be used to predict the prognosis of hepatocellular carcinoma. It also can help in risk stratification of hepatocellular carcinoma (12). Huang et al. reported that an integrated glycolysis-immune signature was associated with poor survival and can accurately predict patient outcomes in lung squamous cell carcinoma (13). Our results were consistent with what had been previously reported in the literature, and we found that the GI signature was related to unfavorable prognosis. This signature can stratify tumors into high and low-risk groups and has potential value for survival prediction in CRC. Further subgroup analyses demonstrated that our signature was still capable of accurately predicting patient outcomes within each subgroup of age, sex, tumor stage, depth of invasion, lymph node metastasis, and distant metastasis status. In addition, biological function analysis showed that the GI signature was associated with malignant behaviors, such as wound healing, cell adhesion, and ECM–receptor interaction, which verified that the signature may play a key role in the progression of CRC. Recent studies have highlighted the crucial role of glycolysis in regulating the immune environment in cancer (10, 29). Glycolysis can suppress T cell-mediated antitumor activity by the blockade of effective T cell trafficking to tumors, contributing to tumor immune evasion (30). Elevated glycolysis can increase the accumulation of Tregs (immunosuppressive cells), which promotes tumor immune evasion, tumor progression and metastasis (29). MDSCs and Tregs, as immunosuppressive cells, promote tumor growth and immune suppression in myeloma (31). MDSCs play a key role in tumor progression, angiogenesis, and formation of metastases of CRC (32). CAFs have important functions in regulating cancer cell growth and ECM remodeling, promoting angiogenesis, and suppressing anti-tumor immune responses (33). CAFs are associated with tumor progression and poor prognosis in CRC (34). Increased M0 macrophages are reported to be related to poor prognosis in lung adenocarcinoma (35). In our study, the GI signature was observed to be positively correlated with immune infiltration of ‘CAF’, ‘Tregs’, and ‘Macrophage M0’ and negatively correlated with immune infiltration of ‘Myeloid DC resting’, ‘Myeloid DC’, ‘CD4 T cell memory activated/resting’, ‘CD8 T cell’, and ‘B cell’ types, which suggested that the GI signature may inhibit anti-tumor immune immunity. Even though the correlation coefficients were small in the present study, this can be attributed to the fact that statistical significance is influenced by sample size: larger samples yield more precise estimates of the population correlation, thereby facilitating the detection of small effects.
ICB immunotherapy has become a promising treatment option in advanced tumors. However, only a subset of the population benefits from ICB immunotherapy, which may be due to resistance to ICB and immune-related adverse events (22, 36, 37). Some molecular predictive biomarkers for ICB efficacy have been reported. Berle et al. highlighted a distinct immune system signature in recurrent, microsatellite stable (MSS) colon cancer, revealing specific patient subgroups within the MSS category that might benefit from immunotherapy (38). In addition, an exploratory study by Zor et al. identified potential biomarkers, such as kynurenine and soluble OX40 within tumor-infiltrating lymphocytes, suggesting their potential as immunotherapeutic targets for CRC (39).
Tumor mutational burden (TMB) does not substantiate its application as a biomarker for ICB efficacy (40). Identification of reliable predictive biomarkers is needed for improving the prediction accuracy of ICB therapy efficacy and optimizing treatment strategies (41). In the present study, we observed that the GI signature was related to immune checkpoint molecules IDO1, PD-L1, BTLA, and PD-L2, suggesting that this signature may have the potential to predict response to ICB therapy. Further research is necessary to validate these results.
The results of RT-PCR confirmed the high expression of NTF4, IL20RB and PMM2 genes in CRC cells. Furthermore, we found that knockdown of NTF4, IL20RB and PMM2 suppressed the proliferation of CRC cells. Yang et al. reported that NTF4 was significantly overexpressed in CRC and upregulated NTF4 promoted CRC development (42). IL20R has two subunits, IL20RA and IL20RB. IL20RA is overexpressed in CRC tissues and is associated with CRC development and progression (43, 44). Similarly, the overexpression of IL20RB is observed in clear renal cell carcinoma and knockdown of IL20RB inhibits the proliferation of tumor cells (45). In addition, IL20RB contributes to bone metastasis of lung cancer (46). Kiparissi et al. elucidated the involvement of PMM2 variants in the development of inflammatory bowel disease (IBD), suggesting a potential link to CRC pathogenesis (47). The study by Ye et al. on the glycolytic crosstalk genes P4HA1 and PMM2 between IBD and CRC offers foundational insights into the involved metabolic pathways (48), supporting our hypothesis about PMM2’s role in CRC. The expression of PMM2 is significantly upregulated in tissues of renal cell carcinoma and colon cancer (49, 50). Knockdown of PMM2 in renal cell carcinoma cells results in the inhibition of tumor cell migration and invasion (50). Our findings are in line with these previous publications.
Conclusion
This work established a new GI signature in CRC. The GI signature was associated with a worse prognosis of CRC and can serve as an effective prognostic tool for predicting patient outcomes. This signature can aid clinicians in the selection of individualized treatment strategies and thus improve the efficacy of treatment. Future studies are necessary to substantiate these findings, which could involve comprehensive experiments and larger cohorts for validation.
Acknowledgements
We gratefully acknowledge the Gene Expression Omnibus and The Cancer Genome Atlas.
Footnotes
Authors’ Contributions
Min Jiang and Chunying Jiang contributed to the conception and design of this research. Yong Liu, Jianzhong Xu, Zhen Xu and Min Jiang performed data analyses. Tao Ye and Shiyang Li contributed to the interpretation and completion of the figures and tables. All Authors contributed to the drafting of the article and final approval of the submitted version.
Conflicts of Interest
The Authors declare that they have no conflicts of interest.
- Received September 20, 2023.
- Revision received November 22, 2023.
- Accepted November 23, 2023.
- Copyright © 2024 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
