Holistic exploration of CHGA and hsa-miR-137 in colorectal cancer via multi-omic data Integration

Abstract

Colorectal cancer (CRC) ranks among the most widespread malignancies globally, with early detection significantly influencing prognosis. Employing a systems biology approach, we aimed to unravel the intricate mRNA-miRNA network linked to CRC pathogenesis, potentially yielding diagnostic biomarkers. Through an integrative analysis of microarray, Bulk RNA-seq, and single-cell RNA-seq data, we explored CRC-related transcriptomes comprehensively. Differential gene expression analysis uncovered crucial genes, while Weighted Gene Co-expression Network Analysis (WGCNA) identified key modules closely linked to CRC. Remarkably, CRC manifested its strongest correlation with the turquoise module, signifying its pivotal role. From the cohort of genes showing high Gene Significance (GS) and Module Membership (MM), and Differential Expression Genes (DEGs), we highlighted the downregulated Chromogranin A (CHGA) as a notable hub gene in CRC. This finding was corroborated by the Human Protein Atlas database, which illustrated decreased CHGA expression in CRC tissues. Additionally, CHGA displayed elevated expression in primary versus metastatic cell lines, as evidenced by the CCLE database. Subsequent RT-qPCR validation substantiated the marked downregulation of CHGA in CRC tissues, reinforcing the significance of our differential expression analysis. Analyzing the Space-Time Gut Cell Atlas dataset underscored specific CHGA expression in epithelial cell subclusters, a trend persisting across developmental stages. Furthermore, our scrutiny of colon and small intestine Enteroendocrine cells uncovered distinct CHGA expression patterns, accentuating its role in CRC pathogenesis. Utilizing the WGCNA algorithm and TargetScan database, we validated the downregulation of hsa-miR-137 in CRC, and integrated assessment highlighted its interplay with CHGA. Our findings advocate hsa-miR-137 and CHGA as promising CRC biomarkers, offering valuable insights into diagnosis and prognosis. Despite proteomic analysis yielding no direct correlation, our multifaceted approach contributes comprehensive understanding of CRC's intricate regulatory mechanisms. In conclusion, this study advances hsa-miR-137 and CHGA as promising CRC biomarkers through an integrated analysis of diverse datasets and network interactions.

Keywords: Biomarker; CHGA; Colorectal cancer; Hsa-miR-137; Single-cell RNA sequencing; WGCNA.

Fig. 1

WGCNA analysis and hub-gene selection for mRNA microarray dataset. A) Module-trait relationship. Each row corresponds to a module eigengene, and the column corresponds to CRC status. The numbers in each cell represent the corresponding correlation and p-value; Module features of GS and MM. Each point represents an individual within each module, which are plotted by GS on the y-axis and gene MM on the x-axis; B) Molecular signature hallmarks identified within the turquoise module through EnrichR database; C) Visualization of GSE81558 and GSE110225 DEG analysis. Volcanic diagram shows the significance of –log₁₀(p.val) on the y-axis and the increase and decrease threshold of gene expression based on LogFC on the x-axis. Each gene is also marked as a dot (blue dots for upregulation and red dots for downregulation) on the graph; D) PPI network of CHGA gene using STRING database.

Fig. 2

CHGA expression characterization using external databases. A) The effect of changes in CHGA gene expression on survival and prognosis of the disease by GEPIA; B) The rate of change in gene expression. Red: tumor sample and gray: normal sample; C) Expression of hub genes in the Human Protein Atlas database, CHGA expression was downregulated in CRC tissues. (Left: normal tissue, right: cancerous tissue); D)CHGA expression profiles and clinicopathological data of patients with CRC from UCSC database; E) The analysis of the CCLE database revealed a high level of CHGA expression in primary (ECC4, SW1463, KM12, and TGBC18TKB) compared to metastatic (NCIH716, SNUC1, and SW626) cell lines; F) The expression pattern of CHGA in control and CRC tissues by qRT-PCR.

Fig. 3

The expression pattern of CHGA on single-cell RNA-seq data. A) specific expression of CHGA within epithelial cell sub-clusters in Space-Time Gut Cell Atlas dataset; B)CHGA showed a consistent trend across three developmental stages (Fetal, Pediatric, and Adult); C)CHGA expression was notably reduced in adenocarcinoma samples compared to normal cells.

Fig. 4

WGCNA analysis and hub-miRNA selection for non-coding dataset. A) Module-trait relationship. Each row corresponds to a module eigengene, and the column corresponds to CRC status. The numbers in each cell represent the corresponding correlation and p-value; Module features of GS and MM. Each point represents an individual within each module, which are plotted by GS on the y-axis and gene MM on the x-axis; B) Visualization of GSE108153 DEG analysis. Volcanic diagram shows the significance of –log₁₀(p.val) on the y-axis and the increase and decrease threshold of gene expression based on LogFC on the x-axis. Each gene is also marked as a dot (blue dots for upregulation and red dots for downregulation) on the graph; C) Distribution of hsa-miR-137 copy numbers across a diverse array of 57 bowel-related cell lines; D) uncovering potential transcriptomic correlations between CHGA and hsa-miR-137 within distinct CRC subtypes; E) Evaluation of significant correlation between CHGA and hsa-miR-137 at protein level.