Identification of core gene networks and hub genes associated with progression of non-alcoholic fatty liver disease by RNA sequencing

Kikuko Hotta; Masataka Kikuchi; Takuya Kitamoto; Aya Kitamoto; Yuji Ogawa; Yasushi Honda; Takaomi Kessoku; Kaori Kobayashi; Masato Yoneda; Kento Imajo; Wataru Tomeno; Akihiro Nakaya; Yutaka Suzuki; Satoru Saito; Atsushi Nakajima

doi:10.1111/hepr.12877

Identification of core gene networks and hub genes associated with progression of non-alcoholic fatty liver disease by RNA sequencing

Hepatol Res. 2017 Dec;47(13):1445-1458. doi: 10.1111/hepr.12877. Epub 2017 Mar 29.

Authors

Kikuko Hotta¹, Masataka Kikuchi², Takuya Kitamoto^{3

4}, Aya Kitamoto^{3

5}, Yuji Ogawa⁶, Yasushi Honda⁶, Takaomi Kessoku⁶, Kaori Kobayashi^{2

7}, Masato Yoneda⁶, Kento Imajo⁶, Wataru Tomeno⁶, Akihiro Nakaya², Yutaka Suzuki⁸, Satoru Saito⁶, Atsushi Nakajima⁶

Affiliations

¹ Department of Medical Innovation, Osaka University Hospital, Suita, Japan.
² Department of Genome Informatics, Graduate School of Medicine, Osaka University, Suita, Japan.
³ Pharmacogenomics, Kyoto University Graduate School of Medicine, Kyoto, Japan.
⁴ Laboratory of Radiation Safety, National Center for Geriatrics and Gerontology, Obu, Japan.
⁵ Advanced Research Facilities and Services, Hamamatsu University School of Medicine, Hamamatsu, Japan.
⁶ Department of Gastroenterology and Hepatology, Yokohama City University Graduate School of Medicine, Yokohama, Japan.
⁷ Medical Solutions Division, NEC Corporation, Tokyo, Japan.
⁸ Department of Computational Biology and Medical Sciences, The University of Tokyo, Kashiwa, Japan.

PMID: 28219123
DOI: 10.1111/hepr.12877

Abstract

Aim: Non-alcoholic fatty liver disease (NAFLD) progresses because of the interaction between numerous genes. Thus, we carried out a weighted gene coexpression network analysis to identify core gene networks and key genes associated with NAFLD progression.

Methods: We enrolled 39 patients with mild NAFLD (fibrosis stages 0-2) and 21 with advanced NAFLD (fibrosis stages 3-4). Total RNA was extracted from frozen liver biopsies, and sequenced to capture a large dynamic range of expression levels.

Results: A total of 1777 genes differentially expressed between mild and advanced NAFLD (q-value <0.05) clustered into four modules. One module was enriched for genes that encode cell surface or extracellular matrix proteins, and are involved in cell adhesion, proliferation, and signaling. This module formed a scale-free network containing four hub genes (PAPLN, LBH, DPYSL3, and JAG1) overexpressed in advanced NAFLD. PAPLN is a component of the extracellular matrix, LBH and DPYSL3 are reported to be tumor suppressors, and JAG1 is tumorigenic. Another module formed a random network, and was enriched for genes that accumulate in the mitochondria. These genes were downregulated in advanced NAFLD, reflecting impaired mitochondrial function. However, the other two modules did not form unambiguous networks. KEGG analysis indicated that 71 differentially expressed genes were involved in "pathways in cancer". Strikingly, expression of half of all differentially expressed genes was inversely correlated with methylation of CpG sites (q-value <0.05). Among clinical parameters, serum type IV collagen 7 s was most strongly associated with the epigenetic status in NAFLD.

Conclusions: Newly identified core gene networks suggest that the NAFLD liver undergoes mitochondrial dysfunction and fibrosis, and acquires tumorigenic potential epigenetically. Our data provide novel insights into the pathology and etiology of NAFLD progression, and identify potential targets for diagnosis and treatment.

Keywords: DNA methylation; fibrosis; next-generation RNA sequencing; non-alcoholic fatty liver disease; weighted gene coexpression network analysis.