Non-alcoholic fatty liver disease in mice with hepatocyte-specific deletion of mitochondrial fission factor

Aims/hypothesis Mitochondria are highly dynamic organelles continuously undergoing fission and fusion, referred to as mitochondrial dynamics, to adapt to nutritional demands. Evidence suggests that impaired mitochondrial dynamics leads to metabolic abnormalities such as non-alcoholic steatohepatitis (NASH) phenotypes. However, how mitochondrial dynamics are involved in the development of NASH is poorly understood. This study aimed to elucidate the role of mitochondrial fission factor (MFF) in the development of NASH. Methods We created mice with hepatocyte-specific deletion of MFF (MffLiKO). MffLiKO mice fed normal chow diet (NCD) or high-fat diet (HFD) were evaluated for metabolic variables and their livers were examined by histological analysis. To elucidate the mechanism of development of NASH, we examined the expression of genes related to endoplasmic reticulum (ER) stress and lipid metabolism, and the secretion of triacylglycerol (TG) using the liver and primary hepatocytes isolated from MffLiKO and control mice. Results MffLiKO mice showed aberrant mitochondrial morphologies with no obvious NASH phenotypes during NCD, while they developed full-blown NASH phenotypes in response to HFD. Expression of genes related to ER stress was markedly upregulated in the liver from MffLiKO mice. In addition, expression of genes related to hepatic TG secretion was downregulated, with reduced hepatic TG secretion in MffLiKO mice in vivo and in primary cultures of MFF-deficient hepatocytes in vitro. Furthermore, thapsigargin-induced ER stress suppressed TG secretion in primary hepatocytes isolated from control mice. Conclusions/interpretation We demonstrated that ablation of MFF in liver provoked ER stress and reduced hepatic TG secretion in vivo and in vitro. Moreover, MffLiKO mice were more susceptible to HFD-induced NASH phenotype than control mice, partly because of ER stress-induced apoptosis of hepatocytes and suppression of TG secretion from hepatocytes. This study provides evidence for the role of mitochondrial fission in the development of NASH. Graphical abstract Supplementary Information The online version contains peer-reviewed but unedited supplementary material available at 10.1007/s00125-021-05488-2.

High-Technologies Co., Ibaraki, Japan). Data analysis was performed with 35-52 sections for mice fed NCD (MffLiKO and control mice; n = 3 each) and 36 sections for those fed HFD (MffLiKO and control mice; n = 3 each). The mitochondrial size, the width of mitochondrial cristae, and the length of mitochondria-associated ER membranes (MAM) with the distance between endoplasmic reticulum (ER) and mitochondria being within 30 nm [6] were all measured using the free image analysis software (ImageJ 1.52a, NIH, USA).

Microarray analysis
Total RNA was extracted from the liver and purified using SV Total RNA Isolation System (Promega Corporation, Madison, WI). Microarray analysis was performed by SurePrint G3 Mouse Gene Expression Microarray 8×60K v2 (Agilent Technologies, Santa Clara, CA). The result was analysed using Feature Extraction Software version 9.5.1.1 (Agilent Technologies, Inc.). To identify up-regulated or down-regulated genes, we calculated the Z-scores and ratios (non-log-scaled fold-change) from the normalised signal intensities of each probe and compared to the controls. In this study, Z-score ≥ 2.0 and ratio ≥1.5-fold and Z-score ≤ -2.0 and ratio ≤ 0.66 were considered as up-regulated and down-regulated genes, respectively. To determine significantly over-represented gene ontology (GO) biological process (BP) terms and significant enrichment of pathways, we used Database for Annotation, Visualization and Integrated Discovery (DAVID) version 6.8 (http://david.abcc.ncifcrf.gov/home.jsp).

Quantitative real-time PCR
Total RNA was extracted from the liver tissues and primary hepatocytes of mice.
Quantitative real-time PCR analysis was performed as described [1]. The mRNA levels were normalised to that of Gapdh mRNA. Primers used in this study are listed in ESM Table 1.

Western blot analysis
The liver samples were obtained from 14 weeks old mice. Western blot analysis was performed as described [1]. The primary and secondary antibodies used are listed in ESM and Hoechst 33342 (Dojindo Laboratories, Kumamoto, Japan) were used to stain Takeichi and Associates 5 mitochondria and nuclei, respectively, and analysed on Zeiss LSM700 laser scanning confocal microscope (Carl Zeiss, Oberkochen, Germany).

TG secretion assays:
For TG secretion assay, primary hepatocytes (25 × 10 4 cells/well) were plated on collagen-coated 6-well plates with growth medium (25 mmol/l glucose, oleic acid-free), washed with the pre-incubation solution (glucose-free Ringer's solution added 300 µmol/l oleic acid) 3 times, and incubated with the pre-incubation solution for 20 minutes.
The pre-incubation solution was replaced with Ringer's solution added 300 µmol/l oleic acid with or without 25 mmol/l glucose, after which the supernatant was collected at 0, 20, 40, 60, 80, and 120 minutes. The TG content was measured using TG E-Test Wako (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan).
Thapsigargin is a non-competitive inhibitor of sarco/endoplasmic reticulum Ca 2+ ATPase (SERCA), which has often been used as a potent inducer of ER stress. For thapsigargin experiment, primary hepatocytes from control mice (35 × 10 4 cells/well) were plated on collagen-coated 12-well plates, incubated with serum-depleted DMEM for 3 h, and treated with growth medium (25 mmol/L glucose, oleic acid-free) plus DMSO (Sigma-Aldrich Co.) with or without thapsigargin (450 nmol/L, Merck KGaA) for 12 h.
The TG secretion was examined as described above.
For Oil Red O staining, primary hepatocytes (25 × 10 4 cells/well) before and 120 min after glucose stimulation were fixed in 10% (wt/vol.) paraformaldehyde for 10 minutes and treated with 60% (vol./vol.) isopropanol for 1 minute at room temperature.

Isolated mitochondria experiments
The mitochondria from hepatocytes were isolated as described [7]. To measure the respiration rate as oxygen consumption, we used Oxytherm electrode unit (Hansatech Instruments Ltd., Norfolk, UK) as described [8]. In this study, oxygen consumption of  Table 1  Values are expressed as means ± SEM. *** p <0.001, calculated by Student' s t test (a-e). Fis1 encoding fission, mitochondrial 1, MiD49 encoding mitochondrial dynamic protein of 49 kDa (also known as Mief2) and MiD51 encoding mitochondrial dynamic protein of 51 kDa (also known as Mief1).

MffLiKO-hepatocytes
Immune system process  mRNA levels were normalised to Gapdh and presented relative to control set at 1 by quantitative PCR. The genes related to inflammatory markers (c) and ER stress (d).
Values are expressed as means ± SEM. * p <0.05, calculated by Student' s t test (c, d).

Inflammatory response
The most significant GO terms in upregulated gene-BP (HFD30w) -Log 10 (p-value) Innate immune response Immune system process Positive regulation of inflammatory response Antigen processing and presentation Neutrophil chemotaxis Response to virus Response to unfolded protein Cell-matrix adhension Mast cell activation a 0 2 4 6 8 b The most significant GO terms in downregulated gene-BP (HFD30w) Oxidation-reduction process Response to mechanical stimulus Steroid metabolic process Lipid metabolic process Epoxygenase P450 pathway Cellular response to hormone stimulus Fatty acid metabolic process Steroid hormone mediated signaling pathway Histidine metabolic process Long-chain fatty acid metabolic process mRNA levels were normalised to Gapdh and presented relative to control set at 1 by quantitative PCR. The genes related to inflammatory markers (c) and fibrosis markers (d).
Values are expressed as means ± SEM. * p <0.05 and ** p <0.01, calculated by Student' s t test (c, d).