Identification of Genes Related to 5-Fluorouracil Based Chemotherapy for Colorectal Cancer

Abstract

Background: Colorectal cancer (CRC) is one of the most common malignancies and its incidence and mortality are increasing yearly. 5-Fluorouracil (5-FU) has long been used as a standard first-line treatment for CRC patients. Although 5-FU-based chemotherapy is effective for advanced CRC, the consequent resistance remains a key problem and causes the poor prognosis of CRC patients. Thus, there is an urgent need to identify new biomarkers to predict the response to 5-FU-based chemotherapy.

Methods: CRC samples were retrieved from Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA). The immune-related genes were retrieved from the ImmPort database. Single-cell sequencing results from colorectal cancer were obtained by the ArrayExpress database. 5-FU resistance-related genes were filtered and validated by R packages. ESTIMATE algorithms were used to assess the tumor microenvironment (TME). KEGG and GO analysis were performed to explore the biological signaling pathway for resistant-response patients and sensitive-response patients in the tumor microenvironment. pRRophetic algorithms were used to predict 5-FU sensitivity. GSEA and GSVA analysis was performed to excavate the biological signaling pathway of the RBP7 gene.

Results: Nine immune-related genes were identified to be associated with 5-FU resistance and poor disease-free survival (DFS) of CRC patients and the signature of these genes was developed in a DFS-prognostic model. Four immune-related genes were determined to be associated with 5-FU resistance and overall survival (OS) of CRC patients. The signature of these genes was developed an OS-prognostic model. ESTIMATE scores showed a significant difference between 5-FU resistant and 5-FU sensitive CRC patients. Resistant-response patients and sensitive-response patients to 5-FU based chemotherapy showed different GO and KEGG enrichment on the tumor microenvironment. RBP7, as a tumor immune microenvironment (TIME) related gene, was found to have the potential of predicting chemotherapy resistance and poor prognosis of CRC patients. GSEA analysis showed multiple signaling differences between the high and low expression of RBP7 in CRC patients. Hypoxia and TNFα signaling via NFκB gene sets were significantly different between chemotherapy resistant (RBP7^High) and chemotherapy sensitive (RBP7^Low) patients. Single-cell RNA-seq suggested RBP7 was centrally distributed in endothelial stalk cells, endothelial tip cells, and myeloid cells.

Conclusions: Immune-related genes will hopefully be potential prognostic biomarkers to predict chemotherapy resistance for CRC. RBP7 may function as a tumor microenvironment regulator to induce 5-FU resistance, thereby affecting the prognosis of CRC patients.

Keywords: 5-FU resistance; colorectal cancer; immune-related genes; prognosis; tumor microenvironment.

Figure 1

The workflow for analyzing the tumor immune microenvironment related gene to 5-FU resistance in CRC.

Figure 2

The VENN diagram for the intersection of DEGs and IRGs among GEO database and ImmPort database. The blue part was DEGs in GEO datasets, the red part was immune genes in ImmPort database, and the intersection part of the two was immune-related genes of 5-FU chemotherapy sensitivity.

Figure 3

Relationships between the module and clinical traits for four GEO datasets. Each row represents a color module and column corresponds to 5-FU resistant or 5-FU sensitive. Each cell contains the corresponding correlation and p-value.

Figure 4

Multivariate Cox regression analysis in GSE106584 and TCGA cohort. (A) There were 9 genes related to DFS in GSE106584. (B) There were 4 genes related to OS in TCGA cohort *p-value < 0.05; **p-value < 0.001.

Figure 5

Kaplan-Meier survival based on the integrated classifier in the GSE106584 and TCGA cohort. (A) KM curve of nine-genes DFS-prognostic signature in GSE106584. (B) KM curve of four-genes OS-prognostic signature in TCGA cohort.

Figure 6

Tumor microenvironment score in CRC with different chemotherapy responses and tumor microenvironment related genes to 5-FU resistance in GSE69657. (A) 5-FU resistant patients showed statistically significant lower StromalScore. (B) 5-FU resistant patients showed lower ImmuneScore, but not statistically significant. (C) 5-FU resistant patients showed statistically significant lower ESTIMATEScore. (D) RBP7 was up-regulated in CRC patients with 5-FU resistance. (E) RBP7 was down-regulated in COAD compared with normal tissue. (F) RBP7 was down-regulated in READ compared with normal tissue. (G) Compared high StromalScore group with low StromalScore group, there were 355 down-expressed genes and 633 up-expressed genes in GSE69657. (H) The predicted 5 fluorouracil sensitivity in RBP7 subgroups. (I) Intersecting previously screened immune-related drug resistance genes with tumor microenvironment related genes and obtained two immune-related genes to 5-FU resistance genes in tumor microenvironment.

Figure 7

GO and KEGG enrichment analysis was performed in the tumor microenvironment to 5-FU resistance. (A) GO enrichment analysis, there were mostly enriched in organelle fission, extracellular matrix organization and extracellular structure organization on BP enrichment. There were mainly involved in collagen−containing extracellular matrix and spindle on CC enrichment. There were mainly enriched in extracellular matrix, tubulin binding and actin binding on MF enrichment. (B) KEGG enrichment, top10 pathways: Proteoglycans in cancer, Focal adhesion, Cell adhesion molecules, Staphylococcus aureus infection, Phagosome, Complement and coagulation cascades, Hematopoietic cell lineage, Viral myocarditis, Viral protein interaction with cytokine and cytokine receptor, and Cell cycle.

Figure 8

GSEA and GSVA analysis to explore RBP7 function enrichment based on GSE19860 and Single-Cell RNA-seq Analysis results from ArrayExpress databases. (A) The top 30 significant GO terms. (B) The top 30 significant KEGG pathways. (C) Hypoxia and TNFα signaling via NFκB gene sets were significantly different between chemotherapy resistant (RBP7^High) and chemotherapy sensitive (RBP7^Low) patients in GSE19860. (D) 30 clusters of the Single-Cell RNA-seq Analysis. (E) Distribution of RBP7 in colorectal cancer patients.

Figure 9

RBP7 high expressed in the 5-FU resistant Lovo cells. (A) The mRNA expression of RBP7 increased in the 5-FU resistant Lovo cells. (B) The protein expression of RBP7 increased in the 5-FU resistant Lovo cells. (C) The relative gray value of RBP7 for western blot analysis. **p-value < 0.001.