Mast Cells Promote Nonalcoholic Fatty Liver Disease Phenotypes and Microvesicular Steatosis in Mice Fed a Western Diet

Abstract

Background and aims: Nonalcoholic fatty liver disease (NAFLD) is simple steatosis but can develop into nonalcoholic steatohepatitis (NASH), characterized by liver inflammation, fibrosis, and microvesicular steatosis. Mast cells (MCs) infiltrate the liver during cholestasis and promote ductular reaction (DR), biliary senescence, and liver fibrosis. We aimed to determine the effects of MC depletion during NAFLD/NASH.

Approach and results: Wild-type (WT) and Kit^W-sh (MC-deficient) mice were fed a control diet (CD) or a Western diet (WD) for 16 weeks; select WT and Kit^W-sh WD mice received tail vein injections of MCs 2 times per week for 2 weeks prior to sacrifice. Human samples were collected from normal, NAFLD, or NASH mice. Cholangiocytes from WT WD mice and human NASH have increased insulin-like growth factor 1 expression that promotes MC migration/activation. Enhanced MC presence was noted in WT WD mice and human NASH, along with increased DR. WT WD mice had significantly increased steatosis, DR/biliary senescence, inflammation, liver fibrosis, and angiogenesis compared to WT CD mice, which was significantly reduced in Kit^W-sh WD mice. Loss of MCs prominently reduced microvesicular steatosis in zone 1 hepatocytes. MC injection promoted WD-induced biliary and liver damage and specifically up-regulated microvesicular steatosis in zone 1 hepatocytes. Aldehyde dehydrogenase 1 family, member A3 (ALDH1A3) expression is reduced in WT WD mice and human NASH but increased in Kit^W-sh WD mice. MicroRNA 144-3 prime (miR-144-3p) expression was increased in WT WD mice and human NASH but reduced in Kit^W-sh WD mice and was found to target ALDH1A3.

Conclusions: MCs promote WD-induced biliary and liver damage and may promote microvesicular steatosis development during NAFLD progression to NASH through miR-144-3p/ALDH1A3 signaling. Inhibition of MC activation may be a therapeutic option for NAFLD/NASH treatment.

FIG. 1.

Biliary IGF-1 secretion and expression. (A) Cytokine multiELISA showed increased release of IGF-1, IL-6, FGF-β, IFN-γ, EGF, leptin, IL-17A, resistin, and IL-10 in WT WD mice compared to WT CD mice. (B) Fold change of the increased cytokines. (C) IgE levels in liver lysates increased in WT WD mice compared to WT CD mice. (D) IGF-1 secretion is increased in cholangiocyte supernatants and serum from WT WD mice compared to WT CD mice. (E) IGF-1 immunoreactivity is enhanced in WT WD mice compared to WT CD mice. (F) IGF-1 immunoreactivity is enhanced in human NASH compared to normal. n = 3 reactions from n = 10–15 mice for multiELISA, n = 4 reactions from n = 4 mice for IgE ELISA, n = 4 reactions from n = 8–10 mice for IGF-1 enzyme immunoassay (EIA). Data are mean ± SEM. *P < 0.05 versus WT CD mice. IGF-1 immunostaining shown at ×20 and ×60. Abbreviations: G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte-macrophage colony-stimulating factor; IFN-γ, interferon gamma; MIP-1α, macrophage inflammatory protein 1 alpha; NGF, nerve growth factor; OD, optical density; PDGF, platelet-derived growth factor; RANTES, regulated upon activation, normal T cell expressed, and secreted; SCF, stem cell factor.

FIG. 2.

MC presence and activation. WT WD mice had increased (A) mMCP-1 (MC marker, red arrows) expression; (B) chymase, tryptase, and c-Kit mRNA expression; and (C) serum histamine levels. (D) MCs (tryptase) and DR (CK-19) increased in patients with NAFLD or NASH. (E) Chymase, tryptase, and c-Kit mRNA expression increased in human NASH. n = 4 reactions from n = 10–15 mice for quantitative PCR; n = 2–4 reactions per sample from n = 9 normal, n = 9 NAFLD, and n = 17 NASH human samples for quantitative PCR; n = 4 reactions from n = 10–15 mice for EIA. Data are mean ± SEM. *P < 0.05 versus WT CD mice or human normal. mMCP-1 immunostaining shown at ×20 and ×40, tryptase/CK-19 coimmunostaining shown at ×20.

FIG. 3.

Liver steatosis and hepatocyte lipogenesis. (A) H&E showed increased steatosis, inflammation, and DR in WT WD mice, which was reduced in Kit^W-sh WD mice. (B) Serum ALT increased in WT WD mice compared to WT CD mice but decreased in Kit^W-sh WD mice compared to WT WD mice. (C) By oil red O staining, percentage lipid droplet area increased in WT WD mice, which was reduced in Kit^W-sh WD mice. Similar findings were noted for hepatic TG content. (D) Lipogenesis markers FASN, Acaca, and PPAR-γ were increased in hepatocytes from WT WD mice compared to WT CD mice but reduced in Kit^W-sh WD mice compared to WT WD mice. mRNA expression of β-oxidation markers carnitine palmitoyltransferase 1a (Cpt1a), Cpt2, and PPAR-α decreased in hepatocytes from WT WD mice compared to WT CD mice but increased in Kit^W-sh WD mice compared to WT WD mice. n = 4 reactions from n = 8–15 mice for ALT, n = 10 images from n = 8–15 mice for oil red O, n = 4 reactions from n = 8–15 mice for TG EIA, n = 4 reactions from n = 8–15 mice for quantitative PCR. Data are mean ± SEM. *P < 0.05 versus WT CD mice; ^#P < 0.05 versus WT WD mice. H&E shown at ×10, oil red O shown at ×20. Abbreviations: Acaca, acetyl-CoA carboxylase 1; FASN, fatty acid synthase.

FIG. 4.

DR and biliary senescence. Immunostaining for (A) CK-19 and (B) Ki67 indicated increased DR and biliary proliferation (red arrows) in WT WD mice compared to WT CD mice, which decreased in Kit^W-sh WD mice compared to WT WD mice. Immunofluorescence for (C) p16 and (D) SA-β-Gal (both costained with CK-19) showed enhanced biliary senescence in WT WD mice compared to WT CD mice, which decreased in Kit^W-sh WD mice compared to WT WD mice. (E) Expression of senescence markers increased in isolated cholangiocytes from WT WD mice compared to WT CD mice but decreased in Kit^W-sh WD mice compared to WT WD mice. n = 10 images from n = 8 mice for CK-19 and Ki67, n = 4 reactions from n = 8–15 mice for quantitative PCR. Data are mean ± SEM. *P < 0.05 versus WT CD mice; ^#P < 0.05 versus WT WD mice. CK-19 and Ki67 staining shown at ×20, p16/CK-19 staining shown at ×160, SA-β-Gal/CK-19 staining shown at ×80.

FIG. 5.

KC presence, liver inflammation, and liver fibrosis. WT WD mice had increased (A) KC presence; (B) CCL3, CCL4, and CCL5 mRNA expression; and (C) IL-6 expression compared to WT CD mice, which decreased in Kit^W-sh WD mice. WT WD mice had increased (D) collagen deposition (fast green/sirius red); (E) Col1a1 mRNA expression and hydroxyproline levels; and (F) desmin expression (HSC marker) compared with WT CD mice; but these parameters were decreased in Kit^W-sh WD mice compared to WT WD mice. n = 4 reactions from n = 8–10 mice for quantitative PCR, n = 10 images from n = 8–10 mice for staining, n = 6 reactions from n = 8–10 mice for hydroxyproline assay. Data are mean ± SEM. *P < 0.05 versus WT CD mice; ^#P < 0.05 versus WT WD mice. Immunostaining shown at ×10 for F4/80 and ×20 for IL-6, sirius red/fast green shown at ×10, desmin staining shown at ×20.

FIG. 6.

miR-144-3p/ALDH1A3 signaling mechanism. (A) ALDH1A3 expression is reduced in WT WD mice, concomitant with increased MC presence (white arrows), compared to WT CD mice; but this is reversed in Kit^W-sh WD mice compared to WT WD mice, as indicated by immunostaining and western blotting. (B) ALDH1A3 expression is unchanged in human NAFLD compared to normal but reduced in human NASH, as determined by immunostaining and western blotting. (C) TargetScan software identified miR-144-3p as a highly conserved miRNA targeting ALDH1A3, and miR-144-3p targeting of ALDH1A3 was confirmed by IPA. (D) miR-144-3P expression is increased in total liver and isolated hepatocytes from WT WD mice compared to WT CD mice but decreased in Kit^W-sh WD mice compared to WT WD mice. (E) Similarly, miR-144-3P expression is increased in human NASH compared to normal; there is no change between NAFLD and normal. (F) By luciferase assay, miR-144-3P significantly decreases Aldh1a3-WT expression, but this down-regulation is ablated by the Aldh1a3 mutant form. n = 6 bands from n = 10–15 mice for western blotting, n = 1 band per sample from n = 4 normal, n = 9 NAFLD and n = 7 NASH total liver samples for western blotting, n = 9–12 reactions for total liver and n = 3 reactions for isolated hepatocytes from n = 10–15 mice for quantitative PCR, n = 3 reactions per sample from n = 4 normal, n = 12 NAFLD and n = 14 NASH for human total liver quantitative PCR, n = 5 reactions for luciferase assay. Data are mean ± SEM. *P < 0.05 versus WT CD mice, miRNA normal control, or normal human patient; ^#P < 0.05 versus WT WD mice. Mouse staining shown at ×80, and human staining shown at ×40. Abbreviations: NC, normal control; UTR, untranslated region.

FIG. 7.

Changes in liver damage and steatosis following MC injections. (A) Microvesicular steatosis, inflammation, and DR increased in WT WD+MC mice compared to WT WD mice. Kit^W-sh WD+MC mice had increased microvesicular steatosis, inflammation, and DR compared to Kit^W-sh WD mice. (B) Serum levels of ALT are significantly increased in WT WD and Kit^W-sh WD mice following MC injections compared to controls. (C) Oil red O staining showed a reduction in lipid droplet area (indicative of microvesicular steatosis development) in WT WD+MC mice compared to WT WD mice; however, Kit^W-sh WD+MC mice had an increase in lipid droplet area compared to Kit^W-sh WD mice. Hepatic TGs are unchanged in WT WD+MC mice compared to WT WD mice but increased in Kit^W-sh WD+MC mice compared to Kit^W-sh WD mice. n = 5 reactions for serum chemistry from n = 10–13 mice, n = 10 images from n = 10–13 mice for oil red O staining, n = 4 reactions from n = 10–13 mice for TG EIA. Data are mean ± SEM. *P < 0.05 versus WT WD mice or Kit^W-sh WD mice. H&E staining shown at ×10, oil red O staining shown at ×20.

FIG. 8.

Working model. During NAFLD progression to NASH, bile ducts are injured and increase secretion of IGF-1, which promotes MC migration to the portal areas. Enhanced MC migration and activation increase miR-144-3P expression in hepatocytes, which inhibits and down-regulates ALDH1A3 expression. Image made with Bio Render.