MicroRNA-137-3p Improves Nonalcoholic Fatty Liver Disease through Activating AMPKα

Nonalcoholic fatty liver disease (NAFLD) is one of the most common chronic liver diseases worldwide and can develop to nonalcoholic steatohepatitis and later hepatic cirrhosis with a high prevalence to hepatocellular carcinoma. Oxidative stress and chronic hepatic inflammation are implicated in the pathogenesis of NAFLD. MicroRNA-137-3p (miR-137-3p) are associated with oxidative stress and inflammation; however, its role and mechanism in NAFLD remain unclear. Mice were fed with a high-fat diet (HFD) for 24 weeks to establish the NAFLD model. To overexpress or suppress hepatic miR-137-3p expression, mice were intraperitoneally injected with the agomir, antagomir, or respective controls of miR-137-3p at a dose of 100 mg/kg weekly for 6 consecutive weeks before the mice were sacrificed. To validate the involvement of AMP-activated protein kinase alpha (AMPKα) or cAMP-specific phosphodiesterase 4D (PDE4D), HFD mice were intraperitoneally injected with 20 mg/kg compound C or 0.5 mg/kg rolipram every other day for 8 consecutive weeks before the mice were sacrificed. Hepatic miR-137-3p expression was significantly decreased in mice upon HFD stimulation. miR-137-3p agomir alleviated, while miR-137-3p antagomir facilitated HFD-induced oxidative stress, inflammation, and hepatic dysfunction in mice. Mechanistically, we revealed that miR-137-3p is directly bound to the 3′-untranslated region of PDE4D and subsequently increased hepatic cAMP level and protein kinase A activity, thereby activating the downstream AMPKα pathway. In summary, miR-137-3p improves NAFLD through activating AMPKα and it is a promising therapeutic candidate to treat NAFLD.


Introduction
Nonalcoholic fatty liver disease (NAFLD) is characterized as metabolic disorder and hepatic steatosis and emerges as one of the most common chronic liver diseases worldwide that can develop to nonalcoholic steatohepatitis (NASH) and later hepatic cirrhosis with a high prevalence to hepatocellular carcinoma [1][2][3][4]. Multiple factors are implicated in the pathogenesis of NAFLD, including oxidative stress and chronic hepatic inflammation [5,6]. During the simple hepatic steatosis stage, overwhelming nonestesterified fatty acids (NEFAs) in the hepatocytes destroy the electron transport chain and impair mitochondrial function, eventually leading to the accumulation of excessive reactive oxygen species (ROS) [7]. In addition, these NEFAs and the intracellular substances released from the injured hepatocytes also activate the intrahepatic Kupffer cells, recruit the infiltration of other leukocytes, and generate a proinflammatory microenvironment, thereby facilitating the progression of NAFLD. Therefore, it is reasonable to treat NAFLD via suppressing oxidative stress and inflammation.
AMP-activated protein kinase alpha (AMPKα) is a highly conserved energy sensor in eukaryotic cells, and AMPKα activation has become an attractive strategy to treat metabolic diseases [8,9]. Beyond the function in maintaining intracellular energy homeostasis, AMPKα also plays critical roles in modulating oxidative stress and inflammation [10]. Hu et al. showed that AMPKα activation upregulated the expression of uncoupling protein 2 and thus restrained myocardial ROS generation [11]. Nuclear factor erythroid-2-related factor-2 (NRF2) is identified as a core regulator of antioxidant defenses through binding to the antioxidant responsive elements of various antioxidant genes, including glutathione (GSH), superoxide dismutase (SOD), and catalase (CAT) [12]. Qu et al. showed that AMPKα activation by gastrodin increased the expression and nuclear translocation of intrahepatic NRF2 and subsequently ameliorated oxidative stress and inflammatory response in NAFLD rodents [13]. A complex molecular network is orchestrated to converge on nuclear factor-kappa B (NF-κB) to trigger hepatic inflammation upon different stresses [14]. Zhang et al. reported that NF-κB was robustly activated in hepatocytes and the liver upon NAFLD stimuli, which in turn aggravated hepatic damage via promoting the generation of multiple proinflammatory cytokines. Intriguingly, AMPKα activation could block NF-κB transcription activity and attenuate hepatic inflammation and fibrosis [15]. Particularly, genetic liver-specific AMPKα activation significantly reduced hepatic steatosis and inflammation, thereby alleviating high-fat diet-(HFD-) induced NAFLD [16]. These findings identify AMPKα as a promising therapeutic candidate for NAFLD.
MicroRNAs (miRNAs) are kinds of endogenously expressed, single-stranded, small noncoding RNAs and involve in multiple biological processes through binding to the 3 ′ -untranslated region (UTR) of messenger RNAs for degradation or translation repression [17][18][19]. Various miR-NAs are associated with intracellular ROS generation and inflammatory response and have been implicated in the development of NAFLD [20,21]. miR-378 expression was elevated in the liver of dietary obese mice and NASH patients, and this elevation facilitated hepatic inflammation and fibrosis in mice [15]. Hanin et al. reported that hepatic miR-132 level was dramatically increased in NAFLD patients and mice and that the transgenic mice with miR-132 overexpression exhibited a more severe fatty liver phenotype and metabolic syndrome at basal conditions [22]. miR-137-3p is widely expressed across species and organs and also enriched in the liver. Tian et al. recently determined that miR-137-3p is directly bound to the 3 ′ -UTR of Src to inactivate the downstream mitogen-activated protein kinase (MAPK) pathway, thereby preventing inflammatory response, oxidative stress, and neuronal injury in ischemic stroke [23]. Studies on the role of miR-137-3p in hepatic diseases are primarily confined to hepatocellular carcinoma [24,25]. A previous study based on miRNA transcriptome sequencing identified miR-137-3p as a promising antioxidant target in the liver [26]. In the current study, we tried to investigate the role and molecular basis of miR-137-3p in NAFLD.

Animals and Study
Design. Two hundred eight-week-age male C57BL/6 mice were purchased from Huafukang Biosci-ence Co., Ltd. (Beijing, China) and adaptively fed in a SPF barrier facility with free access to water and normal diet (ND, 10% kcal fat, 70% kcal carbohydrates, and 20% kcal protein) for 1 week before the study commenced. The NAFLD model was established by feeding mice with a HFD (60% kcal fat, 20% kcal carbohydrates, and 20% kcal protein) for 24 weeks, whereas the matched mice were maintained on a ND [27]. To overexpress or suppress hepatic miR-137-3p expression, mice were intraperitoneally injected with the agomir, antagomir, or respective controls of miR-137-3p at a dose of 100 mg/kg weekly for 6 consecutive weeks before the mice were sacrificed. The miR-137-3p agomir (#miR40000149-4-5), antagomir (#miR30000149-4-5), and respective controls were purchased from Guangzhou RiboBio Co., Ltd. (Guangzhou, China). To validate the involvement of AMPKα or cAMP-specific phosphodiesterase 4D (PDE4D), HFD mice were intraperitoneally injected with 20 mg/kg compound C (CpC, #171260; Sigma, St. Louis, MO, USA), 0.5 mg/kg rolipram (#R6520; Sigma), or an equal volume of vehicle every other day for 8 consecutive weeks before the mice were sacrificed [28,29]. All mice were euthanatized by cervical dislocation. The animal experiments were approved by the Animal Care and Use Committee of Renmin Hospital of Wuhan University and were also in compliance with the Animal Research: Reporting of In Vivo Experiments guidelines.
2.6. Luciferase Reporter Assay. The wild-type (WT) or truncated (TRU) 3′-UTR of PDE4D were constructed and cloned into the pGL3-basic luciferase reporter plasmid (Promega, Madison, Wisconsin, USA), which were then cotransfected with miR-137-3p agomir (50 nmol/L) into HEK293T cells for 48 h. Subsequently, the cells were collected, and the firefly and Renilla luciferase activities were determined using the Dual-Luciferase Reporter Assay System (Promega), with the results expressed as the ratio of firefly to Renilla luciferase activity [43,44].

Statistical
Analysis. Data were expressed as the means ± standard deviations and analyzed using SPSS software (Version 19.0). Comparisons between two groups with a normal distribution and homogeneity of variance were determined by an unpaired two-tailed Student's t-test. For multiple comparisons, one-way analysis of variance followed by the Tukey post hoc test was performed. A p value < 0.05 was considered significant.

miR-137-3p Agomir Improves HFD-Induced NAFLD in
Mice. We first examined hepatic miR-137-3p expression in mice upon HFD stimulation and observed a significant decrease of miR-137-3p in the liver from HFD-treated mice (Figure 1(a)). Intriguingly, miR-137-3p expression was transiently elevated in primary hepatocytes with PO stimulation at 6 h but significantly decreased at 18 h and 24 h (Figure 1(b)). To clarify the function of miR-137-3p in HFD-induced NAFLD, mice were treated with miR-137-3p agomir (Figure 1(c)). As shown in Figure 1(d), miR-137-3p agomir treatment reduced the body weight of HFD mice. The weight of adipose tissue was also decreased by miR-137-3p agomir, as determined by the decreased weights of epididymal and inguinal fat pads (Figure 1(e)). However, miR-137-3p agomir treatment did not affect food intake either under basal conditions or upon HFD stimulation ( Figure 1(f)). Systemic metabolic disorder (e.g., hyperglycemia, insulin resistance, and hyperlipemia) is a key feature of NAFLD and extremely facilitates the progression of hepatic injury [1]. Consistently, HFD-treated mice had higher FBG and serum insulin levels, and the HOMA-IR, an index of insulin resistance, was also increased in HFD mice, which were all reduced by miR-137-3p agomir treatment (Figures 1(g)-1(i)). In addition, HFD-induced elevations of serum TG and TC were also decreased in mice treated with miR-137-3p agomir (Figure 1(j)). In line with the systemic alterations, miR-137-3p agomir also reduced lipid accumulation in the liver upon HFD stimulation, as evidenced by the decreased hepatic TG, TC, and NEFA contents (Figures 1(k) and 1(l)). As expected, HFD-induced gain of liver weight was blocked in mice with miR-137-3p agomir treatment (Figure 1(m)). Hepatic fibrosis is an important pathological change during NAFLD progression, and our data showed that miR-137-3p agomir significantly prevented HFD-induced hepatic fibrosis in mice, as confirmed by the decreased hepatic hydroxyproline content and mRNA levels of fibrotic markers, including Col1α1, Col3α1, CTGF, and TGF-β1 (Figures 1(n) and 1(o)). Moreover, serum ALT and AST levels were also decreased by miR-137-3p agomir treatment, indicating an improved liver function (Figure 1(p)). Collectively, these findings indicate that miR-137-3p agomir improves HFD-induced NAFLD in mice.

miR-137-3p Agomir Reduces Hepatic Oxidative Stress
and Inflammation in HFD Mice. Oxidative stress is implicated in the pathogenesis of NAFLD [45]. As shown in Figures 3(a) and 3(b), hepatic ROS and H 2 O 2 levels were significantly elevated in HFD mice, which were reduced by miR-137-3p agomir. Accordingly, HFD-induced lipid peroxidation was suppressed in miR-137-3p agomir-treated mice, as evidenced by the decreased hepatic MDA content (Figure 3(c)). NRF2 is the core regulator of antioxidant defenses through upregulating the downstream antioxidant genes [45,46]. As shown in Figure 3(d), hepatic NRF2 expression, as well as the antioxidant GSH, total SOD activity, and CAT activity were all decreased upon HFD stimulation, which were significantly prevented in the presence of miR-137-3p agomir (Figures 3(d)-3(f)). Chronic inflammation is the other pathogenic factor of NAFLD [15]. As shown in Figure 3(g), HFD treatment promoted NF-κB activation in the liver that was significantly suppressed by miR-137-3p agomir, as confirmed by the decreased p65 phosphorylation. Consistently, the levels of proinflammatory cytokines, including IL-1β, IL-6, MCP-1, and TNF-α, were decreased, while the anti-inflammatory IL-10 was increased in the liver from mice with miR-137-3p agomir treatment (Figures 3(h) and 3(i)). Taken together, we conclude that miR-137-3p agomir reduces hepatic oxidative stress and inflammation in HFD mice.
3.6. miR-137-3p Activates AMPKα through Downregulating PDE4D. Finally, we investigated the potential mechanism through which miR-137-3p activated AMPKα. Using the online TargetScan software, PDE4D was selected for further examination because of its role in promoting cAMP  Data were expressed as the means ± standard deviations, and * p < 0:05 was considered significant. 6 Analytical Cellular Pathology hydrolysis and subsequently inactivating the cAMP/PKA pathway, the classic upstream mechanism of AMPKα [47,48]. As shown in Figure 6(a), a predicted binding between miR-137-3p and the 3′-UTR of PDE4D was found. And the results from luciferase reporter assay further validated this direct interaction (Figure 6(b)). Moreover, we found that miR-137-3p antagomir increased, while miR-137-3p agomir decreased hepatic PDE4D mRNA and protein levels in HFD mice (Figures 6(c)-6(e)). As expected, hepatic cAMP level and PKA activity were decreased by miR-137-3p antagomir, yet increased by miR-137-3p agomir in HFD mice (Figures 6(f) and 6(g)). AC catalyzes the generation of cAMP from ATP; however, our data suggested that neither the antagomir nor the agomir of miR-137-3p affected hepatic AC activity in HFD mice (Figure 6(h)). To confirm the necessity of PDE4D in miR-137-3p antagomir-mediated inactivation of AMPKα, primary hepatocytes were pretransfected with siPDE4D to knock down the endogenous PDE4D expression (Figure 6(i)). As shown in Figure 6(j), miR-137-3p antagomir significantly reduced AMPKα phosphorylation in POtreated hepatocytes, which was prevented in those with PDE4D silence. In addition, we also used a specific PDE4D inhibitor, rolipram, to suppress PDE4D activity in HFD mice. As expected, miR-137-3p antagomir-mediated deleterious effects on NAFLD were attenuated by rolipram, as evidenced by the decreased serum ALT and AST levels ( Figure 6(k)). In conclusion, we reveal that miR-137-3p activates AMPKα through downregulating PDE4D.

Discussion
The liver is one of the most important organs to modulate glycolipid metabolism and the whole-body energy homeostasis, whose function would be extremely impaired due to the excessive hepatic lipid accumulation. NAFLD consists of a spectrum of pathophysiological conditions that ranges from simple hepatic steatosis to advanced NASH associated with enhanced hepatic inflammation and fibrogenesis, then ). (f) Total SOD and CAT activity in the liver (n = 6). (g) p65 phosphorylation determined by western blot (n = 6). (h, i) Hepatic IL-1β, IL-6, MCP-1, TNF-α, and IL-10 levels determined by the commercial ELISA kits (n = 6). Data were expressed as the means ± standard deviations, and * p < 0:05 was considered significant. 7 Analytical Cellular Pathology to the end-stage liver disease such as cirrhosis and hepatocellular carcinoma [1,49]. Due to the high prevalence of metabolic syndrome nowadays, NAFLD has become one of the most common chronic liver diseases and affects over 25% of the global population worldwide [50]. Currently, no specific and effective therapies are available to prevent NAFLD progression, and transplantation is the only choice for these patients with end-stage NAFLD. In the present study, we found that hepatic miR-137-3p expression was significantly decreased in the development of NAFLD and that the treatment with miR-137-3p agomir could reduce hepatic oxidative stress and inflammation, thereby improving NAFLD in mice. Taken together, our findings identify miR-137-3p as a promising therapeutic target to treat NAFLD.
The exact mechanisms for the initiation and progression of NAFLD remain unclear. Emerging studies report that excessive hepatic ROS and chronic inflammation are closely associated with hepatocyte injury and hepatic fibrosis. Hepatic IL-1β, IL-6, MCP-1, TNF-α, and IL-10 levels determined by the commercial ELISA kits (n = 6). Data were expressed as the means ± standard deviations, and * p < 0:05 was considered significant.

Analytical Cellular Pathology
Under physiological conditions, the liver contains multiple antioxidant defenses, including the antioxidant enzymes (e.g., SOD and CAT), which catalyze the transition from nocuous free radicals to innocuous water. Yet, the expressions of these enzymes in the liver were significantly reduced in the context of NAFLD [46,51]. NRF2 is a redox-sensitive transcription factor to induce the expression of these antioxidant enzymes, and our data revealed that hepatic NRF2 expression was significantly downregulated in HFD mice, accompanied with decreased GSH content, and total SOD and CAT activities. NF-κB functions as the major modulator of inflammatory response and is sequestered in the cytoplasm as an inactive form through binding to the inhibitor of NF-κB in quiescent cells, which then translocates to the nucleus to promote the transcription of various inflammatory cytokines under stressed conditions [52]. Herein, we found that NF-κB p65 phosphorylation was enhanced in the liver from HFD mice, and the levels of downstream IL-     (c-e) Mice were fed with a HFD for 24 weeks to establish NAFLD and were also intraperitoneally injected with the miR-137-3p agomir, antagomir, or respective controls (100 mg/kg weekly) at the last 6 consecutive weeks. Relative PDE4D mRNA and protein levels in HFD mice treated with miR-137-3p agomir, antagomir, or their respective controls (n = 6). (f, g) The cAMP level and PKA activity in HFD mice treated with miR-137-3p agomir, antagomir, or their respective controls (n = 6). (h) Relative AC activity (n = 6). (i) Primary hepatocytes were transfected with siPDE4D or siRNA for 24 h, and relative PDE4D mRNA level was detected (n = 6).
(j) Primary hepatocytes were preincubated with miR-137-3p antagomir (50 nmol/L) for 24 h and then stimulated with PO for 24 h. To knock down PDE4D, cells were pretransfected with siPDE4D or siRNA for 24 h before miR-137-3p antagomir treatment. AMPKα phosphorylation in miR-137-3p antagomir-treated primary hepatocytes with or without PDE4D silence (n = 6). (k) Mice were fed with a HFD for 24 weeks to establish NAFLD and were also intraperitoneally injected with the miR-137-3p agomir, antagomir, or respective controls (100 mg/kg weekly) at the last 6 consecutive weeks. To inhibit PDE4D, mice were intraperitoneally injected with 0.5 mg/kg rolipram or an equal volume of vehicle every other day for 8 consecutive weeks before the mice were sacrificed. Serum ALT and AST levels in miR-137-3p antagomir-treated HFD mice with or without PDE4D inhibition (n = 8). Data were expressed as the means ± standard deviations, and * p < 0:05 was considered significant. n.s. indicated no significance. 10 Analytical Cellular Pathology 1β, IL-6, MCP-1, and TNF-α were also increased. Intriguingly, miR-137-3p agomir significantly reduced, whereas miR-137-3p antagomir further promoted hepatic oxidative stress and inflammation upon HFD stimulation. miRNAs are a class of endogenous genetic regulators with pleiotropic functions, such as controlling ROS production and inflammatory response. Cheng et al. demonstrated that miR-421 overexpression significantly decreased hepatic MnSOD and CAT expression and subsequently aggravated oxidative damage upon NAFLD stimulation via directly targeting sirtuin 3 [53]. In addition, Zhang et al. recently showed that miR-96-5p was capable of suppressing p66shc expression, thereby reducing oxidative stress and hepatic steatosis [54]. Based on these contexts, we thus investigated the function of miR-137-3p, a potential antioxidant molecule in the liver, during NAFLD progression [26]. Intriguingly, our data revealed that miR-137-3p overexpression could reduce hepatic oxidative stress and inflammation, thereby improving NAFLD in mice. Mechanistically, miR-137-3p is directly bound to the 3′-UTR of PDE4D, a cAMP-specific phosphodiesterase, and subsequently increased hepatic cAMP level and PKA activity, thereby activating the downstream AMPKα pathway. Consistently, other findings also proved that the treatment with PDE inhibitors could protect against HFD-induced NAFLD [55,56]. Beyond AMPKα, many other downstream targets of miR-137-3p have been proposed. Tian et al. revealed that miR-137-3p directly bound to the 3 ′ -UTR of Src and subsequently inhibited the MAPK pathway [23]. Findings from Wei et al. identified AFM as a potential target of miR-137-3p to mediate its regulation on the progression of hepatocellular carcinoma [24]. More recently, Yang et al. reported that miR-137-3p negatively regulated peptidylprolyl isomerase C (Ppic) expression via direct targeting and that Ppic knockdown partially reversed the effects of miR-137-3p inhibition [57]. Whether other possible targets of miR-137-3p exist in the regulation of NAFLD remains unclear and needs further investigation.
In summary, we prove that miR-137-3p is a key player in regulating NAFLD via activating AMPKα.

Data Availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest
The authors declare that there are no conflicts of interest.

Authors' Contributions
Yuanjie Yu, Ming Li, and Qian Zhang conceived the hypothesis and designed the study. Yuanjie Yu, Chunping He, Shiyun Tan, and Ming Li carried out the experiments and acquired the data. Mengjun Huang and Yitian Guo conducted the data analysis. Yuanjie Yu and Ming Li drafted and revised the manuscript. Yuanjie Yu and Chunping He contributed equally to this work.