A Novel Function of MicroRNA 130a-3p in Hepatic Insulin Sensitivity and Liver Steatosis

In recent years, there has been an increase in the global prevalence of type 2 diabetes (T2D) and nonalcoholic fatty liver disease (NAFLD). Insulin resistance is one of the major pathological changes of T2D, and many studies have indicated a close relationship between insulin resistance and NAFLD. Various strategies have been proposed to treat both diseases, including lifestyle modifications and pharmacologic interventions. Because liver is a major organ in regulating glucose and lipid metabolism, many hepatic genes play important roles in the development of insulin resistance and fatty liver.

MicroRNAs (miRNAs) are endogenous, noncoding, short, single-stranded RNAs of ∼22 nucleotides (1). Mature miRNAs are produced by enzymatic processing of pre-miRNA sequences that involves the enzymes Drosha and Dicer. When two mature miRNAs originate from opposite arms of the same pre-miRNA, they are denoted with a -3p or -5p suffix. The mature miRNAs regulate gene expression at the posttranscriptional levels by stimulating the degradation or inhibiting translation of target mRNAs. Many miRNAs are evolutionarily conserved and believed to play a role in controlling a variety of biological process, including developmental patterning, cell differentiation, and cell proliferation (2). Recently, several studies have highlighted the significance of miRNAs in maintaining metabolic homeostasis (3). For example, miR-103, miR-107, miR-29, miR-320, and mmu-miR-183-96-182 (cluster) have been shown to regulate insulin sensitivity (4), and miR-122 and miR-370 are involved in the regulation of liver steatosis (5,6). These results suggest that regulation of these miRNAs may serve as potential therapeutic targets in metabolic disorders.

miR-130a is first identified in mouse (7) and later verified in human embryonic stem cells (8). miR-130a appears to be vertebrate-specific miRNA and has now been predicted or experimentally confirmed in a range of vertebrate species. It exerts important functions on cell cycle, angiogenesis, and so on (9). Interestingly, miR-130a has been implicated in glucose homeostasis. For example, miR-130a expression is downregulated by high glucose in pancreatic islets (10). Knockdown of miR-130a decreases the capability of Min6 cells to secrete insulin in response to glucose stimulation (11). It is unknown, however, whether miR-130a is involved in the regulation of insulin sensitivity, another critical aspect in the maintenance of glucose homeostasis. Furthermore, a study showed that miR-130a is aberrantly expressed in a rat model of NAFLD (12), suggesting the possible involvement of miR-130a in liver steatosis. The aim of our current study was to investigate these possibilities and elucidate underlying mechanisms. By overexpression or inhibiting expression of miR-130a in vitro and in vivo, our results identify a novel function for hepatic miR-130a in the regulation of insulin sensitivity and liver steatosis.

Male C57BL/6J mice were obtained from Shanghai Laboratory Animal Co. Ltd. (Shanghai, China). Leptin receptor-mutated (db/db) mice were kindly provided by Xiang Gao of Nanjing University. Eight- to 10-week-old mice were maintained on a 12-h light/dark cycle at 25°C and provided free access to commercial rodent chow and tap water prior to initiation of the experiments. For oral administration, 120 mg palmitic acid (Sigma-Aldrich) was dissolved in 1 mL control solution containing 80% ethanol, 10% Tween 80, and 10% polyethylene glycol. Male db/db mice (body weight ∼40 g) were infected with adenovirus-expressing miR-130a-3p (Ad-miR-130a-3p) via tail-vein injection for 2 days prior to being orally supplemented with palmitic acid (600 mg/kg body weight), where each mouse imbibed ∼200 μL solution (∼160 μL ethanol) every afternoon for 8 days before measurement. This dose is smaller and the period is shorter than those used in other studies investigating the effects of ethanol on glucose metabolism (13,14). Furthermore, minor effects can still be disregarded since both groups of mice were provided with the same amount of ethanol for the same period of time. These experiments were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee of the Institute for Nutritional Sciences, Shanghai Institute for Biological Sciences, Chinese Academy of Sciences.

The DNA fragment encoding miR-130a pre-miRNA was amplified using HEK293T genomic DNA as the PCR template and inserted into the expressing vector pSilencer4.1-CMV puro. DNA fragments of 3′-untranslated regions (UTRs) from growth factor receptor–bound protein 10 (GRB10), ubiquitin-conjugating enzyme E2 D3 (UBE2D3), and sterile alpha motif and SH3 domain–containing protein 1 (SASH1) containing potential miR-130a-3p binding sites were amplified and cloned into Xbal site immediately downstream of the stop codon in the pGL3-promoter vector (Promega). Mutated GRB10 3′UTR reporter plasmids (Mut1, Mut2, and Mut3) were constructed using the overlapping four-primer PCR to produce mutated 3′UTR pGL3 reporter plasmid, as previously described (15–17). miRNA double-stranded mimics for miR-130a-3p or miR-130a-5p and miR-130a-3p or miR-130a-5p inhibitors were purchased from GenePharma (Shanghai, China) with the following sequences: miR-130a-3p mimics (5′-cagugcaauguuaaaagggcau-3′), miR-130a-5p mimics (5′-uucacauugugcuacugucug c-3′), miR-130a-3p inhibitors (5′-augcccuuuuaacauugcacug-3′, 2′Ome modification), and miR-130a-5p inhibitors (5′-gcagacaguagcac aaugugaa-3′, 2′Ome modification). The cDNAs of mouse GRB10 was amplified from mouse liver genomic DNA and human fatty acid synthase (FAS) was provided by Massimo Loda (Dana-Farber Cancer Institute).

Primary hepatocytes were prepared by collagenase perfusion as described previously (18). Cells were transfected with plasmids expressing miR-130a pre-miRNA and FAS by Effectene transfection reagent (Qiagen, Hilden, Germany) or miR-130a-3p mimics, miR-130a-5p mimics, miR-130a-3p inhibitors, and miR- 130a-5p inhibitors using X-tremeGENE siRNA transfection reagent (Roche Diagnostics, Mannheim, Germany). Cells were incubated with 200 μmol/L sodium oleate (SO; Sigma-Aldrich) for 24 h before examination of triglyceride (TG) content.

Recombinant adenoviruses for expression of miR-130a-3p (Ad-miR-130a-3p) or control scrambled short hairpin RNA (Ad-scrambled) were generated using the BLOCK-iT adenoviral RNAi expression system (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Recombinant adenoviruses expressing GRB10 (Ad-GRB10) or green fluorescent protein (Ad-GFP) were generated using the AdEasy adenoviral vector system (Qbiogene, Irvine, CA) according to the manufacturer’s instructions. High-titer stocks of amplified recombinant adenoviruses were purified as previously described (19). Viruses were diluted in PBS and administered at a dose of 10⁷ plaque-forming units per well in 12-well plates, 10⁹ plaque-forming units per wild-type (WT) mice or 2 × 10⁹ plaque-forming units per db/db mouse via tail-vein injection.

HEK293T cells were cotransfected with internal control vector pRL-Renilla (Promega), different 3′UTR pGL3-promoter reporters, and miR-130a-3p mimics (or negative control) using Lipofectamine 2000. Forty-eight hours after transfection, the firefly and Renilla luciferase activities were assayed using Dual-Glo Luciferase assay system (Promega).

Levels of blood glucose and serum insulin were measured using a Glucometer Elite monitor or Mercodia Ultrasensitive Rat Insulin ELISA kit (ALPCO Diagnostics, Salem, NH), respectively. An L-alanine tolerance test was performed by intraperitoneal injection of 2 g/kg L-alanine after 24-h fasting as described previously (20). Glucose tolerance tests (GTTs) and insulin tolerance tests (ITTs) were performed by intraperitoneal injection of 2 g/kg glucose after overnight fasting and 0.75 units/kg insulin after 4-h fasting, respectively. Homeostasis model assessment of insulin resistance (HOMA-IR) index was calculated according to the formula: [fasting glucose levels (mmol/L)] × [fasting serum insulin (μU/mL)]/22.5.

Mice were fasted for 6 h prior to insulin injection as previously described (18). Sections of liver, white adipose tissue (WAT), or soleus muscles were excised from anesthetized mice and snap frozen and kept as untreated controls. Three to five minutes after injection with insulin at a dose of 2 units/kg/WT mouse or 5 units/kg/db/db mouse via the portal vein, pieces of liver, WAT, or soleus muscles section were excised and snap frozen for Western blot analysis.

mRNA (21) and miRNA (22) levels were examined by RT-PCR as previously described. Primer sequences used for RT-PCR are listed in the Supplementary Data. Western blot analysis was performed as previously described (19). Primary antibodies (anti-p-insulin receptor [IR; tyr1150/1151], anti-IR, anti-p-protein kinase B [AKT; ser473], anti-AKT, anti-p-glycogen synthase kinase [GSK] 3β [ser9], anti-GSK3β, anti-FAS, and anti–proliferator-activated receptor [PPAR] γ [all from Cell Signaling Technology, Beverly, MA]) were incubated overnight at 4°C, and specific proteins were visualized by ECL Plus (Amersham BioSciences Inc., Buckinghamshire, U.K.). Band intensities were measured using Quantity One (Bio-Rad Laboratories, Hercules, CA) and normalized to total protein or actin.

Hepatic and cellular lipids were extracted with chloroform-methanol (2:1) as described previously (23). TG was measured with a TG kit (BHKT Clinical Reagents Co., Beijing, China) according to the manufacturer’s instructions.

Frozen sections of liver were stained with Oil Red O. Paraformaldehyde-fixed, paraffin-embedded sections of liver were stained with hematoxylin and eosin for histology.

All data are expressed as means ± SEMs. Significant differences were assessed either by two-tailed Student t test or one-way ANOVA followed by the Student–Newman–Keuls (SNK) test. P < 0.05 was considered statistically significant.

To explore a role for miR-130a in the regulation of insulin sensitivity, we first examined the effects of pre-miR-130a overexpression on phosphorylation of IR (tyr1150/1151), AKT (ser473), and GSK3β (ser9), the three major components of insulin signaling (24), in HepG2 and primary cultured mouse hepatocytes transfected with plasmids expressing the pre-miR-130a or control plasmid. We found that insulin-stimulated phosphorylation of IR, AKT, and GSK3β was significantly elevated in both HepG2 and primary cultured hepatocytes cells overexpressing pre-miR-130a compared with control cells (Supplementary Fig. 1A and B). Although both miR-130a-3p and miR-130a-5p levels were significantly increased by pre-miR-130a compared with control cells, basal levels of miR-130a-5p were much lower compared with miR-130a-3p in HepG2 cells (Supplementary Fig. 2A and B).

To distinguish the effects of miR-130a-3p and miR-130a-5p on insulin signaling, we examined the effects of miR-130a-3p and miR-130a-5p overexpression on phosphorylation of IR, AKT, and GSKβ in cells transfected with miR-130a-3p or miR-130a-5p mimics. Consistent with the experiments using pre-miR-130a plasmids, insulin-stimulated phosphorylation of IR, AKT, and GSK3β was also significantly elevated in cells transfected with miR-130a-3p mimics (Fig. 1A and B). Similar results were obtained using adenoviruses expressing miR-130a-3p (Ad-miR-130a-3p) (Supplementary Fig. 1C and D). Furthermore, insulin-stimulated phosphorylation of IR, AKT, and GSK3β was impaired when endogenous miR-130a-3p levels were reduced by miR-130a-3p inhibitors (Fig. 1C and D). However, insulin-stimulated phosphorylation of IR, AKT, and GSK3β was not affected by miR-130a-5p mimics (Supplementary Fig. 3A and B). Similar results were observed when endogenous miR-130a-5p levels were reduced by miR-130a-5p inhibitors (Supplementary Fig. 3C and D). These results suggest that miR-130a-5p has no effect in the regulation of insulin signaling.

To investigate the effects of miR-130a-3p on insulin sensitivity in vivo, we infected male C57BL/6J WT mice with Ad-miR-130a-3p or control scrambled adenovirus (Ad-scrambled) via tail-vein injection and found that miR-130a-3p levels were significantly increased in the livers of these mice compared with Ad-scrambled mice (Fig. 2A). Although fed blood glucose levels remained unchanged, fasting blood glucose levels were significantly decreased in Ad-miR-130a-3p mice compared with the Ad-scrambled group (Fig. 2B). A reduction in fasting blood glucose suggests an effect on hepatic glucose production. Indeed, L-alanine tolerance tests suggest that Ad-miR-130a-3p downregulates gluconeogenesis in liver (Supplementary Fig. 4). Serum insulin levels were not affected by Ad-miR-130–3p under either fed or fasting conditions (Fig. 2C). That fasting blood glucose levels decreased while insulin remained normal in Ad-miR-130a-3p mice compared with control mice is the result of improved insulin sensitivity, which is another critical aspect regulating glucose homeostasis, in addition to an alteration in serum insulin levels. The decreased fasting blood glucose levels, however, resulted in a lower HOMA-IR index in Ad-miR-130a-3p mice (Fig. 2D). Glucose tolerance and clearance were further examined by GTTs and ITTs, respectively. As predicted, Ad-miR-130a-3p mice exhibited increased glucose tolerance and clearance (Fig. 2E). In addition, insulin-stimulated phosphorylation of IR, AKT, and GSK3β was also increased in the livers of these mice (Fig. 2F).

Although adenovirus-mediated gene transfer occurs preferentially in liver cells, it may also mediate weak overexpression in other tissues, including WAT and muscle, which could also contribute to whole-body insulin sensitivity. To test this possibility, we examined expression levels of miR-130a-3p and insulin signaling in WAT and muscle in mice injected with Ad-miR-130a-3p or control adenovirus. We found that miR-130a-3p expression or insulin signaling was not altered in these tissues in Ad-miR-130a-3p mice (Supplementary Fig. 5A and B), suggesting that the observed changes in insulin sensitivity in Ad-miR-130a-3p mice during GTT and ITT are mainly caused by the altered expression of miR-130a-3p in liver.

We identified mRNA targets for miR-130a-3p according to the following criteria: 1) the gene is predicted using miRGen targets, which provides access to unions and intersections of four widely used target prediction programs including TargetScan, PicTar, miRanda, and DIANA-microT (Table 1); and 2) the gene is reported to be associated with insulin sensitivity. In this way, three genes got our attention: GRB10, UBE2D3, and SASH1. The potential binding site for miR-130a-3p within GRB10, UBE2D3, and SASH1 3′UTRs is highly conserved in human, mouse, and rat (Fig. 3A and Supplementary Fig. 6A and B). To investigate whether these three genes can be regulated by miR-130a-3p, pGL3-promoter-based GRB10, UBE2D3, or SASH1 3′UTR reporter was cotransfected with miR-130a-3p mimics or negative control to HEK293T cells. Among the three genes assayed, only luciferase activity with GRB10 3′UTR was significantly inhibited by miR-130a-3p mimics (Fig. 3B). Furthermore, the mutated GRB10 3′UTR reporter was not inhibited (Fig. 3B). Correspondingly, mRNA and protein levels of GRB10 were decreased or increased by Ad-miR-130a-3p in vitro and in vivo or miR-130a-3p inhibitors in vitro, respectively (Fig. 3C–E). mRNA levels of Ube2d3 and Sash1 were not affected by miR-130a-3p in vitro and in vivo (Fig. 3C and D). These results exclude the possibility for UBE2D3 and SASH1 as target genes of miR-130a-3p.

The possible involvement of GRB10 in miR-130a-3p-stimulated insulin signaling was then investigated in HepG2 cells infected with Ad-miR-130-3p with or without adenoviruses expressing GRB10 (Ad-GRB10). As predicted, overexpression of GRB10 significantly blocked miR-130a-3p-enhanced insulin signaling (Fig. 3F). To gain further insights into the importance of GRB10 in regulating insulin sensitivity by miR-130a-3p in vivo, we first examined the effects of Ad-miR-130a-3p on GRB10 expression in mice and found protein levels of GRB10 were decreased by Ad-miR-130a-3p (Fig. 4A). Then we injected mice with Ad-miR-130a-3p with or without Ad-GRB10. Although overexpression of Ad-GRB10 (Fig. 4B) had no effects on serum insulin levels, injection of Ad-GRB10 increased fasting blood glucose levels and HOMA-IR index of Ad-miR-130a-3p mice (Fig. 4C–E). Correspondingly, mice injected with Ad-GRB10 largely attenuated miR-130a-3p-potentiated glucose and insulin tolerance (Fig. 4F). Consistent with previous reports (25), overexpression of GRB10 alone decreased insulin sensitivity and signaling (Fig. 4G–L).

Based on the above results, we speculated that miR-130a-3p might also be involved in the regulation of insulin sensitivity under insulin-resistant conditions. To test this possibility, we first examined miR-130a-3p and GRB10 expression in db/db mice, a genetic model for insulin resistance (26), and found that miR-130a-3p levels were decreased, but GRB10 protein levels were increased, in the livers of db/db mice compared with WT mice (Fig. 5A and B). To investigate whether decreased miR-130a-3p was responsible for the impaired insulin signaling in db/db mice, we injected these mice with Ad-miR-130a-3p or Ad-scrambled and examined the effects on insulin sensitivity. As expected, miR-130a-3p mRNA levels were significantly increased in the livers of these mice compared with control mice (Fig. 5C). Glucose tolerance and clearance were further examined by GTTs and ITTs, respectively. As observed in WT mice, Ad-miR-130a-3p db/db mice also exhibited improved glucose tolerance and clearance (Fig. 5D). Furthermore, insulin-stimulated phosphorylation of IR, AKT, and GSK3β was also increased in the livers of these mice (Fig. 5E). The improved insulin sensitivity is associated with decreased GRB10 expression in the livers of miR-130a-3p-overexpressed db/db mice (Fig. 5F and G).

Because insulin resistance is an almost universal finding in liver steatosis and our data indicated a significant decrease in miR-130a-3p expression in liver of db/db mice, an animal model with hepatic steatosis (Fig. 5A), we speculated that miR-130a-3p might be involved in the regulation of liver steatosis. To test this possibility, we injected these mice with Ad-miR-130a-3p or Ad-scrambled. Overexpression of miR-130a-3p (Fig. 6A) significantly decreased liver steatosis in db/db mice, as demonstrated by the decreased contents of hepatic and serum TG (Fig. 6B and C) and histological changes (Fig. 6D).

We then examined the impact of Ad-miR-130a-3p on levels of genes and proteins related to lipid metabolism in liver. We found that mRNA and protein levels of FAS were significantly decreased by Ad-miR-130a-3p in the livers of both WT and db/db mice compared with control mice (Fig. 6E and F). Except for ApoE gene, which was decreased by Ad-miR-130a-3p in the livers of db/db mice, the rest of the genes examined were not affected in either WT or db/db mice (Fig. 6E). We did not find FAS has complementary sites of miR-130a-3p in its 3′UTR. A previous study (27) and our current work (Fig. 6F), however, showed that miR-130a-3p inhibits expression of peroxisome PPARγ, which is upstream regulator of FAS (28), suggesting that miR-130a-3p may regulate FAS expression via PPARγ.

To assess a role of FAS in regulating liver steatosis by miR-130a-3p, we investigated the effects of miR-130a-3p on SO-induced steatosis (29) in HepG2 cells. Overexpression of miR-130a-3p significantly prevented SO-increased TG content in these cells compared with control cells (Fig. 6G), and this blocking effect was reversed by overexpression of FAS (Fig. 6H). A key role for FAS in mediating miR-130a-3p regulation of liver steatosis in vivo was further investigated in Ad-miR-130a-3p db/db mice orally provided with palmitic acid, the product of FAS action, or control vehicle. Palmitic acid supplementation blocked miR-130a-3p–inhibited hepatic lipid accumulation in db/db mice (Fig. 6I and J).

We have been interested in the role of miRNAs in regulating insulin sensitivity. Specifically, we concentrated on investigating the role of miR-130a, which has been implicated in glucose homeostasis by affecting insulin secretion (10,11). Besides, it has been shown that miR-130a is downregulated in endothelial progenitor cells from T2D patients (30). A direct effect of miR-130a on insulin sensitivity, however, has not been described previously. In this study, we observed that 1) overexpression of miR-130a-3p in vitro increases insulin-stimulated phosphorylation of IR, AKT, and GSK3β and knocking down miR-130a-3p expression has the opposite effect, whereas miR-130a-5p has no effect in the regulation of insulin sensitivity; and 2) Ad-miR-130a-3p mice exhibit increased glucose clearance and tolerance. Our study is the first to demonstrate a novel function of miR-130a-3p in the regulation of insulin sensitivity.

Mature miR-130a is of 22 nucleotides, and several proteins have been identified as its downstream targets (31–33). In order to fully understand the molecular mechanisms by which miR-130a-3p regulates insulin sensitivity, we explored other target genes of miR-130a-3p possibly associated with insulin sensitivity. Among the predicted target genes from open access software, we first focused on three proteins: GRB10, UBE2D3, and SASH1. GRB10 belongs to a superfamily of adapter proteins that are known to interact with a number of receptor tyrosine kinases and signaling molecules. It plays a role in diverse biological processes (34). Numerous studies have shown that GRB10 is implicated in the modulation of insulin responsiveness. For example, GRB10 has been shown to inhibit insulin signaling by disrupting the association of IRS1/IRS2 with IR (35). Overexpression of GRB10 was found to decrease insulin-stimulated glycogen synthase activity and glycogen synthesis in primary hepatocytes. In contrast, peripheral disruption of the GRB10 gene enhances insulin signaling and sensitivity in vivo (25). UBE2D3, also termed UBCH5C, has ubiquitin-specific protease activity and participates in protein metabolism. It is downregulated in human dermal fibroblasts by resveratrol (36), a type of natural phenol that could improve insulin sensitivity and decrease blood glucose. In addition, a recent work showed that SASH1 can significantly inhibit insulin’s ability to repress FOXO1A-mediated transcription by inhibiting insulin-reduced AKT phosphorylation (37). These studies suggest the possible involvement of UBE2D3 and SASH1 in glucose homeostasis regulation.

Our data confirmed that GRB10, but not UBE2D3 and SASH1, is a target gene of miR-130a-3p. Consistent with previous results (25), we also observed an inhibitory effect of Ad-GRB10 on insulin signaling in vitro and insulin sensitivity in vivo. Moreover, Ad-GRB10 also blocks miR-130a-3p–enhanced insulin sensitivity. These results suggest that GRB10 is a novel direct target gene of miR-130a-3p and is required for regulating miR-130a-3p–potentiated insulin sensitivity. Because GRB10 is involved in many different physiological processes, our results suggest that miR-130–3p may also have other novel functions, which will be studied in the future.

A recent study showed that miR-130a was aberrantly expressed in rats with NAFLD (12), and furthermore, studies show that insulin resistance is an almost universal finding in NAFLD, suggesting that miR-130a-3p might be involved in the regulation of NAFLD. A critical role of miR-130a-3p in regulating lipid metabolism in liver is confirmed by our observation that overexpression of miR-130a-3p effectively reverses liver steatosis in db/db mice. Consistent with a role of miR-130a-3p in the regulation of liver steatosis, other studies have shown that miR-122 and miR-370 are involved in the development of liver steatosis (5,6), suggesting that the importance of miRNAs in liver steatosis should not be ignored.

The reduced accumulation of hepatic TGs in db/db mice injected with Ad-miR-130a-3p is likely to reflect an imbalance in hepatic TG synthesis, β-oxidation, and uptake and/or secretion of fatty acids (21). For this reason, we examined changes in levels of mRNA expression or proteins related to each of these processes. Among these genes, levels of FAS mRNA and protein were greatly reduced in the livers of both WT and db/db mice injected with Ad-miR-130a-3p. Thus the suppressing effect of miR-130a-3p is most likely mediated by its inhibiting effects on FAS expression. The importance of miR-130a-3p–dependent repression of FAS was further demonstrated in vitro and in vivo by FAS overexpression or supplementation with palmitic acid, the product of FAS action, respectively.

We did not find a direct regulatory effect of miR-130–3p on FAS expression. It has been shown that miR-130a-3p potently repressed PPARγ expression by targeting both the mRNA coding and 3′UTRs, thereby blocking the expression of PPARγ-regulated gene (27). The PPARs are a group of nuclear receptor proteins that function as transcription factors regulating the expression of genes, including FAS (28). It has been shown that targeted deletion of PPARγ in hepatocytes protects mice against high-fat diet–induced hepatic steatosis (38). Consistent with these studies, we found decreased PPARγ expression in both WT and db/db mice injected with Ad-miR-130a-3p. As PPARγ could regulate FAS expression, we speculate that the effects of miR-130-3p on FAS expression may be mediated indirectly via inhibiting PPARγ expression.

Our data support a role of miR-130a-3p/GRB10 in the regulation of insulin sensitivity under insulin-resistant conditions in mice, as demonstrated by the reversal effects of miR-130a-3p on impaired glucose tolerance and clearance in db/db mice, which is associated with decreased GRB10 expression. Though it has not been tested, genome-wide association studies have provided some hints for the association of GRB10 single nucleotide polymorphisms and T2D patients. For example, it is reported that the GRB10 single nucleotide polymorphism rs2237457 is associated with T2D in an Amish population (39). Based on these results, we speculate that signal pathways mediated by miR-130a-3p/GRB10 are likely to be involved in insulin resistance in diabetic patients. This possibility, however, will require extensive additional work and collaboration with clinical researchers in the future.

In this study, we also investigated whether there is any possible link between miR-130a-3p regulation of insulin sensitivity and liver steatosis. We found that GRB10 alone decreased FAS expression and TG content (Supplementary Fig. 7A and B), suggesting that GRB10 is unlikely to be involved in miR-130a-3p regulation of liver steatosis. Surprisingly, GRB10 expression was increased by FAS overexpression compared with control cells (Supplementary Fig. 7C). Furthermore, overexpression of FAS significantly blocked miR-130a-3p–enhanced insulin signaling in these cells, and FAS overexpression alone also decreased insulin signaling (Supplementary Fig. 7D). These results suggest the possible involvement of FAS in miR-130a-3p regulation of insulin sensitivity.

Taken together, we show that overexpression of miR-130-3p improves insulin sensitivity in vitro and in vivo, via directly targeting GRB10, and it reverses hepatic steatosis in db/db mice via repressing FAS expression (Fig. 6K). These results demonstrate a novel function for miR-130a-3p in the regulation of insulin sensitivity and liver steatosis. Our results are also important in providing new perspective for the understanding of the molecular mechanisms underlying insulin resistance and liver steatosis and potential treatment strategy for metabolic disorders.

Funding. This work was supported by grants from the Ministry of Science and Technology of China (973 Program 2010CB912502), the China National Science Fund for Distinguished Young Scientists (81325005), the National Natural Science Foundation (81130076, 31271269, 81100615, 30890043, 81390350, and 81300659), the Basic Research Project of Shanghai Science and Technology Commission (13JC1409000), the Key Program of Shanghai Scientific and Technological Innovation Action Plan (10JC1416900), the Knowledge Innovation Program of the Chinese Academy of Sciences (KSCX2-EW-R-09), and a Chinese Academy of Sciences–funded project (2011KIP307). F.X. was supported by a China Postdoctoral Science Foundation–funded project (2012M520950 and 2013T60473) and a Chinese Academy of Sciences–funded project (2013KIP310). F.G. was supported by the One Hundred Talents Program of the Chinese Academy of Sciences.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. F.X. and J.Y. researched data and wrote, reviewed, and edited the manuscript. B.L., J.Z., C.W., S.C., Y.D., and Y.L. provided research material. Y.G., K.L., J.D., Y.X., and Z.Z. researched data. F.G. directed the project; contributed to discussion; and wrote, reviewed, and edited the manuscript. F.G. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.