Multiomic Sequencing Reveals Distinctive Gene Expression and Epigenetic Alterations Associated With Primary Sclerosing Cholangitis Development in Treatment-Naïve Pediatric Ulcerative Colitis

Abstract

Background and aims: Primary sclerosing cholangitis (PSC) is a progressive cholestatic disease with up to 80% of patients also suffering from ulcerative colitis (PSC-UC). The difficulty in the diagnosis along with the increased risk for developing cancer represents a clinical challenge. Furthermore, the precise molecular factors regulating the phenotype of this disease subtype remain unknown.

Methods: We applied methyl-capture sequencing and mRNA sequencing to colonic mucosal biopsies from 3 groups of treatment-naïve children at diagnosis from the Determinants and Outcomes in CHildren and AdolescentS study: UC (n = 10), PSC-UC (n = 10), and healthy controls (n = 10).

Results: Differential gene expression between UC and PSC-UC showed significantly higher gene expression changes in PSC-UC patients when compared to UC. Specifically, expression of these genes was regulated by master transcriptional regulators (NLRP3, DLL1) and transcription factors (RELA, Myogenin, and FOXO1), which are shown to regulate expression of inflammatory response and immune-associated genes in PSC-UC patients exclusively. Differential methylation analysis between PSC-UC and UC demonstrated >2000 differentially methylated regions with a large proportion of them enriched in gene promoter and enhancer regions. We further show no difference in epigenetic age between PSC-UC and UC. Finally, we identify KLHL17 as hypomethylated and upregulated in PSC-UC patients.

Conclusion: Our study, for the first time, identifies distinct gene expression and DNA methylation alterations that differentiate UC from PSC-UC at diagnosis in treatment-naïve pediatric patients. We show the gene expression differences observed between PSC-UC and UC are modulated by intricate molecular mechanisms involving master transcriptional regulator-mediated signaling through transcription factors. These findings suggest the potential utility of these molecular markers as predictive biomarkers for PSC development in UC at an early stage of development. Further validation in larger patient cohorts is warranted.

Keywords: Methylation Profiling; Primary Sclerosing Cholangitis; Transcriptional Regulation; Ulcerative Colitis.

Graphical abstract

Figure 1

Volcano plots and Pathway enrichment analysis of differential expressed genes between each group. (A) PSC-UC versus Healthy Control (1180 upregulated and 892 downregulated), UC versus Control (249 upregulated 149 downregulated); PSC-UC versus UC (9 upregulated and 5 downregulated). Y-axis show -log10(padj) and x-axis shows FC. Cut off values of significant DEG (log2fold change ≥ 0.584 and P-value < .05). Significant genes are indicated in red, with upregulated represented on the right of the central axis (dotted line) and downregulated on the left of the central axis. (B) Each figure representing KEGG/GO pathways, FDR < 0.01. GO term was selected with the FDR value < 0.01 (blue to red: significant), gray (not significant). Y-axis shows pathway terms and X-axis, number of hits for each comparison: PSC-UC vs UC; PSC-UC vs control; and UC vs control. “Count”: number of genes belonging to a gene set. KEGG, Kyoto Encyclopaedia of Genes and Genomes.

Figure 2

Signal transduction and gene regulatory networks examined by GeneXplain genome browser: The network diagrams were constructed using the top 2 MTRs (lowest total rank)—highlighted in pink at the top of the diagram identified as master regulators of both upregulated and downregulated genes derived from each comparative analysis: (A) Upregulated genes in PSC-UC vs healthy control; (B) Downregulated genes in PSC-UC vs healthy control; (C) Upregulated genes UC vs healthy control; (D) Downregulated genes in UC vs healthy control. (E) Upregulated genes in PSC-UC vs UC. For each comparison, the networks further illustrate the intermediate molecules (highlighted in green), TF modules (highlighted in purple) enriched for the DEGs and corresponding genes (light blue).

Figure 3

Differential methylation analysis using DMRcate. (A) PSC-UC versus healthy control (937 hypermethylated and 1026 downregulated); (B) UC versus healthy control (1241 hypermethylated and 1039 hypomethylated); (C) PSC-UC versus UC (939 hypermethylated and 1327 hypomethylated). Y-axis shows -log10(Stouffer score) and X-axis shows mean difference in methylation. Cut-off values of significant DMRs at log10 (0.05). Central axis (dotted black line) separates upregulated (red) to the right and downregulated (blue) to the left. Bar graphs shown in (D–F) highlight number (Y-axis) of hyper (in green) and hypo (in orange) methylated promoter and enhancers for each comparison.

Figure 4

Integration of differential expression and differential methylation and correlation between epigenetic age and chronological age. Each plots (A–C) represent the genes that are both differentially methylated and expressed at the same time for each of the 3 comparisons. PSC-UC vs Control, UC vs Control, and PSC-UC vs UC respectively. The X-axis corresponds to the—log2FC(methylation) and the y-axis to log2FC(methylation). Genes identified as both significantly differentially expressed and methylated are displayed in the 4 quadrants: hypermethylated and upregulated are represented in yellow; hypermethylated downregulated in red; hypomethylated upregulated in green; and hypomethylated and downregulated in blue. Plots (D–F) represent the correlation between the epigenetic age and the chronological age for each group of patients; PSC-UC, UC, and controls. The X-axis corresponds to the chronological age and the Y-axis corresponds to pcgtAge which quantifies the rate of cellular divisions and level of biological stress.

Figure 5

Graphical summary of the study. The diagram illustrates increased gene expression and lower DNA methylation alterations in treatment naïve PSC-UC when compared to UC patients. Two distinct regulatory mechanisms identified to modulate expression differences (upregulation or downregulation, indicated by green or red arrows) between PSC-UC and UC patients: (A) Master transcriptional regulator—TF networks, and (B) Loss of DNA methylation in PSC-UC patients can drive expression of genes such as KLHL17.