A metabolic stress-inducible miR-34a-HNF4α pathway regulates lipid and lipoprotein metabolism

Abstract

Non-alcoholic fatty liver disease (NAFLD) is one of the most common liver diseases, but its underlying mechanism is poorly understood. Here we show that hepatocyte nuclear factor 4α (HNF4α), a liver-enriched nuclear hormone receptor, is markedly inhibited, whereas miR-34a is highly induced in patients with non-alcoholic steatohepatitis, diabetic mice and mice fed a high-fat diet. miR-34a is essential for HNF4α expression and regulates triglyceride accumulation in human and murine hepatocytes. miR-34a inhibits very low-density lipoprotein secretion and promotes liver steatosis and hypolipidemia in an HNF4α-dependent manner. As a result, increased miR-34a or reduced HNF4α expression in the liver attenuates the development of atherosclerosis in Apoe(-/-) or Ldlr(-/-) mice. These data indicate that the miR-34a-HNF4α pathway is activated under common conditions of metabolic stress and may have a role in the pathogenesis of NAFLD and in regulating plasma lipoprotein metabolism. Targeting this pathway may represent a novel approach for the treatment of NAFLD.

Figure 1. Hepatic HNF4α and miR-34a expression is inversely regulated in NASH patients and diabetic or HFD-fed mice

(a–e) Hepatic levels of TG (a), cholesterol (b), Hnf4α mRNA (c), HNF4α protein (d) and miR-34a (e) were determined in normal individuals or NASH patients (n=8). (f–h) Hepatic HNF4α protein levels were determined using Western blots in diabetic ob/ob or db/db mice (f; n=5) or STZ-treated mice (f; n=6), high fat diet (HFD)-fed mice (g; n=5) or high fat, high cholesterol (HFHC) diet-fed mice (g; n=8). Protein levels were quantified by ImageJ (h). (i–l) Hepatic miR-34a levels were quantified in ob/ob or db/db mice (i), STZ-treated mice (j), HFD-fed mice (k) or HFHC diet-fed mice (l). Results are shown as mean±SEM. Two-sided student t-test was performed. * P<0.05. ** P<0.01.

Figure 2. miR-34a regulates lipid metabolism and inhibits HNF4α expression in mice and HepG2 cells

(a–f) C57BL/6 mice were injected i.v. with either Ad-empty (control) or Ad-miR-34a (miR-34a) (n=6). After 7 days, plasma TG (a), cholesterol (chol; b) and hepatic TG (c) levels were determined. Hepatic Hnf4α mRNA level was quantified by qRT-PCR (d). Hepatic protein levels were determined by Western blots (e) and then quantified (f). (g–k) Wild-type and miR-34a^−/− mice were fed a Western diet for 12 weeks (n=5). Plasma TG (g), plasma cholesterol (h) and hepatic TG levels (I) were determined. Hepatic protein levels were determined by Western blotting (J) and HNF4α protein levels quantified (k). (l, m) ob/ob mice (l) or HFD-fed mice (m) were injected i.v. with anta-scr (scramble antagomir) or anta-miR-34a (miR-34a antagomir) once every 6 days (10 mg per kg) (n=4–5). After 3 injections, hepatic HNF4α protein levels were determined. (n, o) HepG2 cells were treated with Ad-empty (control), Ad-miR-34a or Ad-anti-miR-34a (anti-miR-34a). After 48 h, protein levels were determined by Western blotting (n) and then quantified (o) (n=3). (p, q) HepG2 cells were infected with Ad-empty or Ad-miR-34a for 48 h. Neutral lipids were stained by oil red O (p) and TG levels quantified (q) (n=6). The transfection assays were repeated once and similar results were obtained. Scale bar, 20 μm. Values are expressed as mean±SEM. Two-sided student t-test was performed. * P<0.05, ** P<0.01.

Figure 3. miR-34a regulation of VLDL secretion and lipid metabolism depends on inhibition of hepatic HNF4α

(a–c) C57BL/6 mice were i.v. injected with Ad-empty or Ad-miR-34a (n=6). Hepatic mRNA (a) and protein levels (b, c) were quantified. B100, ApoB100. B48, ApoB48. (d) VLDL secretion was determined in C57BL/6 mice after i.p. injection of Tyloxapol (50 mg/kg) (n=6). (e–h) C57BL/6 mice were injected i.v. with Ad-empty, Ad-miR-34a, Ad-HNF4α or Ad-miR-34a+Ad-HNF4α (n=7). Hepatic protein levels (e), plasma TG (f), plasma cholesterol (g) and hepatic TG levels (h) were determined. (i) HepG2 cells were transfected with a control mimic or miR-34a mimic, together with a pMIR-Report construct containing wild-type or mutant 3′ UTR of HNF4α (n=4). After 36 h, luciferase activity was determined and normalized to β-gal activity. MutA and mutB stand for the first or second mutant binding site for miR-34a, respectively. (j) Liver-specific Hnf4α^−/− mice were injected i.v. with Ad-HNF4α-3′UTR (3′UTR_WT), Ad-HNF4α-3′UTR_mutA (3′UTR_mutA) or Ad-HNF4α-3′UTR_mutB (3′UTR_mutB) plus or minus Ad-miR-34a. After 7 days, hepatic protein levels were determined. WT, wild type. Values are expressed as mean±SEM. Two-sided student t-test was performed. * P<0.05, ** P<0.01.

Figure 4. Loss of hepatic HNF4α improves energy homeostasis and protects against atherosclerosis

(a–d) Apoe^−/− mice were fed a Western diet for 4 weeks, followed by injection of Ad-shLacZ or Ad-shHNF4α (n=6). Three weeks following adenoviral injection, plasma cholesterol (a) and TG (b) lipoprotein profiles were determined. En face aortic lesion sizes were analyzed (c) and hepatic TG levels quantified (d). (e–t) Hnf4α^fl/flLdlr^−/− (Ldlr^−/−) mice and Hnf4α^fl/flAlb-CreLdlr^−/− (DKO) mice were fed a Western diet for 16 weeks (n=6). Body weight gain (e) and body fat content from gonadal (gon) fat, retroperitoneal (ret) fat and subcutaneous (sub) fat (f) were determined. O₂ consumption (g, i), CO₂ production (h, i) and heat production (j) were assessed by CLAMS. Plasma cholesterol (k), TG (l), cholesterol lipoprotein profile (m) and TG lipoprotein profile (n) were analyzed. Aortic root was stained by oil red O (o) and lesion size quantified (p). Aorta was isolated (q), stained by oil red O (r) and the en face lesion size quantified (s). VLDL secretion was performed when mice were fed a Western diet for 12 weeks (t). In (Q), the arrow shows the location of the brachiocephalic artery. Scale bar, 200 μm. Values are expressed as mean±SEM. Two-sided student t-test was performed. * P<0.05, ** P<0.01.

Figure 5. miR-34a reduces the development of atherosclerosis in Ldlr^−/− mice

Ldlr^−/− mice were fed a Western diet for a total of 7 weeks (n=6). At the end of week 4, the mice were injected i.v. with Ad-empty or Ad-miR-34a. Plasma cholesterol (a), TG (b), cholesterol lipoprotein profile (c) and TG lipoprotein profile (d) were determined. Aortic root was stained by oil red O (e) and lesion size quantified (f). Aorta was also stained by oil red O (g) and the en face lesion size quantified (h). Hepatic protein levels were determined by Western blot assays (i). Scale bar, 200 μm. Values are expressed as mean±SEM. Two-sided student t-test was performed. * P<0.05, ** P<0.01.

Figure 6. p53, fatty acids and cholesterol regulate the miR-34a-HNF4α pathway

(a, b) Nuclear p53 protein levels in the livers of normal individuals and NASH patients were determined by Western blot assays (a) and then quantified (b). (c) HepG2 cells were infected with Ad-empty or Ad-p53. After 48 h, protein levels were determined by Western blot assays. (d–i) HepG2 cells were treated with either vehicle (Ve), palmitate (PA; 300 μM) (d), linoleic acid (LA; 300 μM) (e), oleic acid (OA; 300 μM) (f) or cholesterol (Ch;10 μg per ml) (g). Western blot assays were performed (d–g) and then quantified (h). qRT-PCR was used to quantify miR-34a levels (i). (j) Role of the miR-34a-HNF4α pathway in regulating lipid and lipoprotein metabolism. Under metabolic stress, miR-34a is highly induced via both p53-dependent and -independent pathways. The induction of miR-34a causes a reduction in hepatic HNF4α expression and VLDL secretion, leading to liver steatosis and inhibition of atherogenesis. Values are expressed as mean±SEM. Two-sided student t-test was performed. * P<0.05, ** P<0.01.