Gene coexpression networks reveal a broad role for lncRNAs in inflammatory bowel disease
John L. Johnson, Davit Sargsyan, Eric M. Neiman, Amy Hart, Aleksandar Stojmirovic, Roman Kosoy, Haritz Irizar, Mayte Suárez-Fariñas, Won-Min Song, Carmen Argmann, Stefan Avey, Liraz Shmuel-Galia, Tim Vierbuchen, Gerold Bongers, Yu Sun, Leonard Edelstein, Jacqueline Perrigoue, Jennifer E. Towne, Aisling O’Hara Hall, Katherine A. Fitzgerald, Kasper Hoebe
John L. Johnson, Davit Sargsyan, Eric M. Neiman, Amy Hart, Aleksandar Stojmirovic, Roman Kosoy, Haritz Irizar, Mayte Suárez-Fariñas, Won-Min Song, Carmen Argmann, Stefan Avey, Liraz Shmuel-Galia, Tim Vierbuchen, Gerold Bongers, Yu Sun, Leonard Edelstein, Jacqueline Perrigoue, Jennifer E. Towne, Aisling O’Hara Hall, Katherine A. Fitzgerald, Kasper Hoebe
View: Text | PDF
Resource and Technical Advance Gastroenterology Immunology

Gene coexpression networks reveal a broad role for lncRNAs in inflammatory bowel disease

Abstract

The role of long noncoding RNAs (lncRNAs) in disease is incompletely understood, but their regulation of inflammation is increasingly appreciated. We addressed the extent of lncRNA involvement in inflammatory bowel disease (IBD) using biopsy-derived RNA-sequencing data from a large cohort of deeply phenotyped patients with IBD. Weighted gene correlation network analysis revealed gene modules of lncRNAs coexpressed with protein-coding genes enriched for biological pathways, correlated with epithelial and immune cell signatures, or correlated with distal colon expression. Correlation of modules with clinical features uncovered a module correlated with disease severity, with an enriched interferon response signature containing the hub lncRNA IRF1-AS1. Connecting genes to IBD-associated single nucleotide polymorphisms (SNPs) revealed an enrichment of SNP-adjacent lncRNAs in biologically relevant modules. Ulcerative colitis–specific SNPs were enriched in distal colon–related modules, suggesting that disease-specific mechanisms may result from altered lncRNA expression. The function of the IBD-associated SNP-adjacent lncRNA IRF1-AS1 was explored in human myeloid cells, and our results suggested IRF1-AS1 promoted optimal production of TNF-α, IL-6, and IL-23. A CRISPR/Cas9-mediated activation screen in THP-1 cells revealed several lncRNAs that modulated LPS-induced TNF-α responses. Overall, this study uncovered the expression patterns of lncRNAs in IBD that identify functional, disease-relevant lncRNAs.

Authors

John L. Johnson, Davit Sargsyan, Eric M. Neiman, Amy Hart, Aleksandar Stojmirovic, Roman Kosoy, Haritz Irizar, Mayte Suárez-Fariñas, Won-Min Song, Carmen Argmann, Stefan Avey, Liraz Shmuel-Galia, Tim Vierbuchen, Gerold Bongers, Yu Sun, Leonard Edelstein, Jacqueline Perrigoue, Jennifer E. Towne, Aisling O’Hara Hall, Katherine A. Fitzgerald, Kasper Hoebe

×

Figure 8

Relationship of modules to IBD genetics.

Options: View larger image (or click on image) Download as PowerPoint
Relationship of modules to IBD genetics. (A) The percentage of SNP-adjac...
(A) The percentage of SNP-adjacent lncRNA genes and protein-coding genes (within 50 kb) was calculated for each module (filled circles). Significance was determined using a Fisher’s exact test. The average percentage of SNP-adjacent lncRNA genes and protein-coding genes for all modules was also calculated (dashed lines). *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001. (B) The proportion of disease-specific (i.e., significant in one but not the other) SNP-adjacent lncRNAs was calculated for modules having at least 3 UC- or CD-specific SNP-adjacent lncRNAs. (C) Differential gene expression analysis was performed on UC versus CD samples in the rectum and nonrectum colon. All significant genes with adjusted P value below 0.01 and absolute log2 fold-change of 1 were intersected with the genes in the yellow and lightyellow modules. Significance testing of the overlap between the 2 gene sets was performed using a hypergeometric distribution with all genes included in the WGCNA as the total gene set. (D) For yellow module genes, the gene significance of each yellow module gene to the rectum versus nonrectum samples was calculated and plotted against the module membership of each yellow module gene. (E) For yellow module genes, the gene significance of each gene to active versus inactive IBD status was calculated and plotted against the module membership of each yellow module gene. For D and E, significance was calculated using Pearson correlation.