MicroRNA-214 Suppresses Gluconeogenesis by Targeting Activating Transcriptional Factor 4

Background: miR-214 targets ATF4 involved in glucose metabolism; however, the role of miR-214 and ATF4 in hepatic gluconeogenesis is unknown.

Results: Overexpression of miR-214 suppresses gluconeogenesis in hepatocytes in vitro and in vivo via the ATF4-dependent pathway.

Conclusion: miR-214 suppresses gluconeogenesis via targeting ATF4.

Significance: Our studies reveal that the miR-214-ATF4 axis is a novel pathway for the regulation of hepatic gluconeogenesis.

Abstract

Although the gluconeogenesis pathway is already a target for the treatment of type 2 diabetes, the potential role of microRNAs (miRNAs) in gluconeogenesis remains unclear. Here, we investigated the physiological functions of miR-214 in gluconeogenesis. The expression of miR-214 was suppressed by glucagon via protein kinase A signaling in primary hepatocytes, and miR-214 was down-regulated in the livers of fasted, high fat diet-induced diabetic and leptin receptor-mutated (db/db) mice. The overexpression of miR-214 in primary hepatocytes suppressed glucose production, and silencing miR-214 reversed this effect. Gluconeogenesis was suppressed in the livers of mice injected with an adenovirus expressing miR-214 (Ad-miR-214). Additionally, Ad-miR-214 alleviated high fat diet-induced elevation of gluconeogenesis and hyperglycemia. Furthermore, we found that activating transcription factor 4 (ATF4), a reported target of miR-214, can reverse the suppressive effect of miR-214 on gluconeogenesis in primary hepatocytes, and this suppressive effect was blocked in liver-specific ATF4 knock-out mice. ATF4 regulated gluconeogenesis via affecting forkhead box protein O1 (FOXO1) transcriptional activity. Finally, liver-specific miR-214 transgenic mice exhibited suppressed gluconeogenesis and reduced expression of ATF4, phosphoenolpyruvate carboxykinase, and glucose-6-phosphatase in liver. Taken together, our results suggest that the miR-214-ATF4 axis is a novel pathway for the regulation of hepatic gluconeogenesis.

Introduction

The liver has a unique role in maintaining blood glucose homeostasis. In the fed state, the liver increases glucose uptake, and this is converted into glycogen and triglycerides for energy storage. During fasting, hepatic gluconeogenesis is an important process for the regulation of blood glucose levels. Abnormally elevated hepatic gluconeogenesis is largely responsible for the overproduction of glucose in patients with type 2 diabetes (T2D)³ mellitus (1–3). Recently, the transcriptional regulators of hepatic gluconeogenesis have been highlighted as potential therapeutic targets for the treatment of T2D (4). Thus, inhibition of the gluconeogenesis pathway could form part of the strategy for combating T2D-induced hyperglycemia.

Hepatic gluconeogenesis is controlled through the transcriptional modulation of phosphoenolpyruvate carboxykinase (PCK) and glucose-6-phosphatase (Glc-6-P), the rate-limiting enzymes in the process, which have been shown to be regulated by glucagon and insulin (5). Glucagon promotes hepatic glucose production by activating cyclic AMP/protein kinase A (cAMP/PKA) signaling, which induces the formation of complex transcriptional machinery, consisting of transcription factors such as cAMP-response element-binding protein 1 (CREB) and coactivators such as CREB-regulated transcription coactivator 2 (CRTC2), on the promoters of PCK and Glc-6-P (6). In contrast, insulin suppresses hepatic gluconeogenesis by repressing the transcriptional activity of forkhead box protein O1 (FOXO1) via protein kinase B (Akt) phosphorylation-mediated protein degradation (7). The unphosphorylated form of FOXO1 localizes to the nucleus and interacts with the insulin-response element in the promoter of PCK (8).

The role of microRNAs (miRNAs), the small RNA molecules (∼18–24 nucleotides) that regulate gene expression post-transcriptionally via mRNA destabilization or translational repression (9, 10), in gluconeogenesis is currently unclear. However, recent studies suggest that two miRNAs, miR-29a-c and miR-23a, may be critical for gluconeogenesis (11, 12). Furthermore, miR-375 and miR-223 have substantial effects on glucose metabolism (13–15). Currently, restoration of miRNA expression is a potential therapeutic approach for treating metabolic diseases (16). One particular miRNA, miR-214, has been extensively studied in the context of cell survival (17, 18) and tumor formation (19–21). miR-214 can directly target phosphatase and tensin homolog deleted on chromosome 10 and fibroblast growth factor-21 receptor 1 (FGFR1) (21, 22). Phosphatase and tensin homolog-mediated Akt/β-catenin-Foxo1 axis is a key regulator of innate inflammatory response in the mouse liver (24). Furthermore, FOXO1 is a key regulator of hepatic gluconeogenesis (25). FGFR1 is involved in the effect of fibroblast growth factor-21 (FGF21) on gluconeogenesis (26). All these implied that miR-214 could be involved in the regulation of gluconeogenesis. Otherwise, miR-214 targets activating transcription factor 4 (ATF4) and thereby inhibits bone formation (27). In our previous study, we found that ATF4 deletion mice exhibit lower levels of fasting blood glucose, and ATF4 plays a key role in high carbohydrate diet-induced insulin resistance (28). Furthermore, ATF4 is also involved in the regulation of CREB phosphorylation (29). Additionally, previous studies suggest that cooperative interaction between ATF4 and FOXO1 affects glucose metabolism (30). Given that CREB and FOXO1 are key regulators of hepatic gluconeogenesis (25), we postulated that miR-214 might be involved in the regulation of hepatic gluconeogenesis via targeting ATF4.

EXPERIMENTAL PROCEDURES

Antibodies and Chemicals

The primary antibodies against ATF4 (catalog number sc-200), PCK (catalog number sc-32879), and TRB3 (catalog number sc-34211) were purchased from Santa Cruz Biotechnology. The primary antibodies against p-PKA-substrate (catalog number 9621S) were purchased from Cell Signaling Technology. The primary antibody against Glc-6-P (ab83690) was purchased from Abcam (Cambridge, UK). Antibody against actin (A5316), glucagon, insulin, dexamethasone, and H89 were purchased from Sigma.

Animals and Treatment

The 8-to-10-week-old C57BL/6J male wild-type (WT) mice were purchased from the Model Animal Research Center of Nanjing University (Nanjing, China). The 4-week-old C57BL/6J male WT mice were fed a HFD (Research Diet Inc.) or control diet for 16 weeks. Genetically diabetic leptin receptor-mutated (db/db) mice were kindly provided by Xiang Gao from Nanjing University. Mice were maintained on a 12-h light/dark cycle at 25 °C and provided with free access to commercial rodent chow and tap water before the experiments. Mice were injected intravenously through the tail vein with adenovirus-expressing green fluorescent protein (Ad-GFP) or miR-214 (Ad-miR-214) at a dose of 1 × 10⁹ plaque-forming units in 0.2 ml of PBS. HFD-fed mice were injected with Ad-GFP or Ad-miR-214 after HFD feeding for 14 weeks, following by another 2 weeks of HFD feeding.

ATF4-floxed mice were generated using the transgenic method. Briefly, exons 2–3 of ATF4 (the entire coding region) were floxed with loxP sites and removed by crossing the floxed mice to CRE-expressing strains. The linearized construct was electroporated into embryonic stem (ES) cells. Chimeric mice were generated by injection of ES cells into mouse blastocysts at the Model Animal Research Center of Nanjing University (Nanjing, China). Mouse tail biopsies were analyzed by genomic PCR using the primers targeting the genome sequence on both sides of the first loxP site as primer ATF4-F, 5′-ggttgtcggccttgtttgcgttgc-3′, and primer ATF4-R, 5′-atcgccacgttcgcaggatgacac-3′; the WT allele results in an ∼150-bp PCR product, and floxed allele gives rise to an ∼180-bp PCR product. Liver-specific ATF4 knock-out mice (LV-ATF4 KO) were achieved by crossing ATF4 floxed mice with albumin-Cre mice.

Floxed miR-214 transgenic mice were constructed from Shanghai Biomodel Organism Science & Technology Development (Shanghai, China). Briefly, the transgenic vector piZEG-pre-miR-214 mouse was constructed based on the piZEG plasmid. piZEG contains a CAG promoter driving the enhanced green fluorescent protein coding regions linked by an internal ribosomal entry site. The CAG promoter is followed by a loxP-flanked sequence containing LacZ, a neomycin-bpA selection cassette, and a transcriptional STOP sequence (31, 32). To produce the piZEG-pre-miR-214 expression vector, the DNA fragment encoding miR-214 pre-miRNA (flanking upstream and downstream 70 nucleotides) was amplified and flanked by BglII and XhoI restriction enzyme sites, then cloned and inserted into the BglII-XhoI sites of piZEG. The piZEG-pre-miR-214 construct was purified and used for microinjection. The founder mice were genotyped by PCR using primers to detect LacZ and confirmed by LacZ detection in tail biopsies as described on line. Liver-specific miR-214 transgenic (LV-miR-214 TG) mice were produced by the founder mice intercrossing with albumin-cre transgenic mice.

All transgenic mice were backcrossed to C57BL/6J background before investigation. The Institutional Animal Care and Use Committee of the Institute for Nutritional Sciences (Shanghai, China) approved all the animal-related experimental procedures in this study (permit number INS09-1001).

Glucose Tolerance Test (GTT), Pyruvate Tolerance Test (PTT), and Insulin Tolerance Test (ITT) Assays

GTTs and PTTs were performed by intraperitoneal (i.p.) injection of d-glucose (2 g/kg) or pyruvate (2 g/kg), respectively (12), following overnight fasting. ITTs were performed by i.p. injection of 0.75 units/kg insulin after 4 of fasting. Blood glucose levels were measured using a glucometer Elite monitor.

Generation and Administration of Recombinant Adenoviruses

The recombinant adenoviruses used for miR-214 and ATF4 expression were generated using AdEasy^TM Vector System (Qbiogene), according to the manufacturer's instruction. Adenovirus-mediated overexpression of FOXO1 was a gift from Dr. Youfei Guan from Shenzhen University (Shenzhen, China). Viruses were diluted in PBS and administered at a dose of 1 × 10⁷ pfu/well in 12-well plate or through tail vein injection using 1 × 10⁹ pfu/mice.

Cell Culture and Treatments

Mouse primary hepatocytes were prepared from livers of male C57BL/6J mice as described previously (33). Mouse primary hepatocytes were maintained in Dulbecco's modified Eagle's medium (DMEM) with 25 mm glucose from Gibco, 10% FBS, 50 mg/ml penicillin and streptomycin at 37 °C, and 5% CO₂, 95% air. To investigate whether glucagon modulates miR-214 expression in a PKA-dependent manner, primary hepatocytes were treated with PKA inhibitor H89 from Sigma at a concentration of 10 μm (34). miRNA double-stranded mimics for miR-214 or miR-214 inhibitors were purchased from GenePharma (Shanghai, China) with the following sequences: miR-214 mimics (5′-acagcaggcacagacaggcagu-3′) and miR-214 inhibitors (5′-acugccugugccugcugu-3′, 2′Ome modification). miR-214 mimics were transfected into primary mouse hepatocytes at a dose of 0.5 ng/well in 12-well plate, which is more close to a physiological range. For small interfering RNA (si-RNA) transfection, double-stranded siRNA targeting mouse ATF4 (si-ATF4) (sense 5′-gaguuaguuugacagcuaatt-3′ and antisense 5′-uuagcugucaaacuaacucca-3′) were purchased from GenePharma (Shanghai, China). For ATF4 overexpression, primary hepatocytes were transfected with pCMV-HA-ATF4 for 48 h with plasmid pCMV-HA used as a control.

Glucose Output Assay

Mouse primary hepatocytes were infected with adenovirus or transfected with miR-214 mimics and inhibitor. Cells were stimulated with or without 100 nm glucagon and 1 μm dexamethasone for 10 h, then washed five times with PBS, and then the medium was replaced with Krebs-Ringer buffer (115 mm NaCl, 5.9 mm KCl, 1.2 mm MgCl₂, 1.2 mm NAH₂PO₄, 2.5 mm CaCl₂, 25 mm NaHCO₃, pH 7.4) supplemented with 20 mm lactate and 2 mm pyruvate. After 6 h of incubation, medium was collected, and the glucose concentration was measured using a glucose assay kit from Sigma. The readings were then normalized to the total protein content (35).

Glycogen Content Assay

Mouse primary hepatocytes were infected with adenovirus or transfected with si-RNA as indicated. Glycogen content in primary hepatocytes and livers can be determined by an acid hydrolysis method as described in a previous study (36).

RNA Isolation, Relative Quantitative-PCR (RT-PCR), and Western Blotting

RNA isolation and RT-PCR were performed as described previously (37). GAPDH was used as an internal control for each gene of interest. For miRNA detection, poly(A) tail was added to RNase-free DNase-digested total RNA as described previously (38). SYBR Green quantitative RT-PCR was used to assay miRNA expression with the specific forward primers and the universal reverse primer complementary (Hi-reverse 5′-ccagtctcagggtccgaggtattc-3′) to the anchor primer. U6 and 18 S were used as internal control. The sequences of primers mainly used in this study showed in Table 1. Western blotting was performed as described previously (29).

TABLE 1.

Primers for RT-PCR

Gene	Sequence (5′ to 3′)
U6-F	CTCGCTTCGGCAGCACA
miR-214-F	ACAGCAGGCACAGACAGGCAG
Pck
Forward	CTTCTCTGCCAAGGTCATCC
Reverse	TTTTGGGGATGGGCAC
G6p
Forward	ATGACTTTGGGATCCAGTCG
Reverse	TGGAACCAGATGGGAAAGAG
Atf4
Forward	CCTGAACAGCGAAGTGTTGG
Reverse	TGGAGAACCCATGAGGTTTCAA
Gsk3β
Forward	TTGGACAAAGGTCTTCCGGC
Reverse	AAGAGTGCAGGTGTGTCTCG
Gys
Forward	GCACGGAGAGGCTCTCAGAT
Reverse	AGGTGTCTGGCATGCTGGTAA
Gyp
Forward	GGTAGCCATCCAGCTGAATGAC
Reverse	TCAATGTCCACAAAAATCCTCATC

Data Analysis

Means ± S.E. shown are representative of at least two independent in vitro or in vivo experiments, with the number of dishes/per condition or mice included in each group in each experiment indicated. Significant differences were assessed by using a two-tailed Student's t test or one-way ANOVA followed by the Student-Newman-Keuls (SNK) test. p < 0.05 was considered statistically significant.

RESULTS

Expression of miR-214 Is Suppressed in the Livers of Fasted and Diabetic Mice

In an effort to explore the function of miR-214 in gluconeogenesis, we first examined miR-214 expression in primary hepatocytes treated with glucagon. Exposure of primary mouse hepatocytes in 100 nm glucagon resulted in significantly increased Pck and G6p mRNA (Fig. 1A) and a 50% decrease in miR-214 expression level at 1 h normalized by 18 S or U6 (Fig. 1B). In contrast, levels of miR-214 precursor (pre-miR-214) were not affected by glucagon (Fig. 1C). Furthermore, glucagon in the range from 0 to 100 nm suppressed the expression of miR-214 in primary hepatocytes in a concentration-dependent manner up to 2 h normalized by 18 S or U6 (Fig. 1D). Regulation of PCK and Glc-6-P expression by glucagon is mediated by PKA signaling (5). Therefore, we tested whether glucagon modulates miR-214 expression in a PKA-dependent manner. Exposure of primary mouse hepatocytes to glucagon decreased the expression of miR-214 and increased phospho-PKA substrate antiserum. These effects were blocked by the PKA inhibitor, N-[2-(P-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide (H89) (34) at a concentration of 10 μm (Fig. 1, E and F). Expression levels of miR-214 decreased in a gradient in the livers of mice after fasting for 12, 24, and 48 h and normalized by 18 S or U6 (Fig. 1G). In addition, miR-214 expression decreased in the livers of HFD-fed and db/db mice normalized by 18 S or U6 (Fig. 1, H and I).

FIGURE 1.

miR-214 expression is down-regulated in glucagon-treated primary mouse hepatocytes and the livers of fasting and diabetic mice. A–F, primary mouse hepatocytes (n = 6 dishes/condition) were treated without (−Glu) or with 100 nm (or as indicated) glucagon (+ Glu) in the absence (− H89) or presence of 10 μm H89 (+ H89) for 10 h, or as indicated. G–I, miR-214 expression was analyzed in the livers of male C57/B6J wild-type (WT) mice fed or fasted for 12, 24, or 48 h, fed a HFD (+ HFD) or control diet (− HFD) for 16 weeks, or control WT (wt) or db/db mice (db) (n = 6 for each group). Means ± S.E. shown are representative of at least two independent in vitro or in vivo experiments, with the number of dishes/condition or mice included in each group in each experiment indicated. Statistical significance was determined by two-tailed Student's t test or one-way ANOVA followed by the SNK test for the effect of glucagon versus control vehicle, fasting, HFD, or db/db mice versus corresponding control mice (*, p < 0.05) or H89 versus control vehicle in glucagon-treated cells (†, p < 0.05). A, Pck and G6p mRNAs; B, D, F, and G–I, miR-214 level; C, pre-miR-214 level; E, Western blot of phosphorylated PKA substrate (p-PKA substrate) and nonspecific band (NS). G6P, Glc-6-P.

miR-214 Suppresses Glucose Production in Mouse Primary Hepatocytes

Given that miR-214 expression is regulated by glucagon, we investigated the effect of miR-214 on gluconeogenesis. Addition of miR-214 mimics resulted in about 4-fold increase in miR-214 expression in primary hepatocytes (Fig. 2A). As expected, miR-214 mimics significantly suppressed Pck and G6p mRNA and protein levels, as well as glucose production, under either basic or dexamethasone and glucagon-stimulated conditions (Fig. 2, B–D). Conversely, an miR-214 inhibitor significantly suppressed the expression of miR-214, which led to increased Pck and G6p mRNA and protein levels, as well as glucose production, under either basic or stimulated conditions (Fig. 2, E–H).

FIGURE 2.

miR-214 suppresses glucose production in primary mouse hepatocytes. A–H, primary mouse hepatocytes (n = 6 dishes/condition) were infected or transfected with negative control (− miR-214) or miR-214 mimics (+ miR-214) and negative control (− anti-miR-214) or miR-214 inhibitor (+ anti-miR-214), for 48 h in the absence (− Dex + Glu) or presence of 1 μm dexamethasone and 100 nm glucagon (+ Dex + Glu) for 10 h. Means ± S.E. shown are representative of at least two independent in vitro experiments, with the number of dishes/condition indicated. Statistical significance was determined by the Student's t test or one-way ANOVA followed by the SNK test for the effect of miR-214 mimics or anti-miR-214 versus corresponding control (*, p < 0.05) or with versus without Dex and Glu in miR-214 mimics or anti-miR-214 infected cells (†, p < 0.05). A and E, miR-214 level; B and F, Pck and G6p mRNAs; C and G, PCK and Glc-6-P (G6P) proteins (top, Western blot; bottom, quantitative measurements of PCK and Glc-6-P protein relative to actin); D and H, glucose output assay.

miR-214 Effectively Reverses HFD-induced Elevated Gluconeogenesis and Hyperglycemia

We then investigated the physiological function of miR-214 in the regulation of hepatic gluconeogenesis in vivo. miR-214 was 1.6-fold overexpressed in liver after adenovirus injection via the tail vein at day 9 (Fig. 3A). Both fed and fasting blood glucose levels were significantly lower in adenoviruses expressing miR-214 (Ad-miR-214)-treated mice compared with mice injected with control adenovirus (Ad-GFP) (Fig. 3B). Serum insulin levels were no different in the Ad-miR-214 group of mice compared with the Ad-GFP group (Fig. 3C). Glucose tolerance, insulin tolerance, and glucose generation were examined by using GTTs, ITTs, and PTTs, respectively. For ITTs, the difference between Ad-miR-214 and control mice was small; however, GTTs showed that exogenous glucose was cleared faster in Ad-miR-214 mice (Fig. 3D). Results of PTTs showed that gluconeogenic capacity in the livers of mice administered Ad-miR-214 was lower than in control mice (Fig. 3D). Consistent with the suppressed gluconeogenic capacity, PCK and Glc-6-P protein levels were significantly suppressed in the livers of mice injected with Ad-miR-214 (Fig. 3E).

FIGURE 3.

Overexpression of miR-214 reverses HFD-induced elevated gluconeogenesis and hyperglycemia. A–G, male C57/B6J wild-type (WT) mice (n = 6 for each group) were injected with Ad-GFP (− Ad-miR-214) or Ad-miR-214 (+ Ad-miR-214) for 9 days. H–K, male C57/B6J WT mice (n = 6 for each group) were fed on a HFD (+ HFD) or control diet (− HFD) for 14 weeks and then injected with Ad-GFP (− Ad-miR-214) or Ad-miR-214 (+ Ad-miR-214), followed by HFD for another 2 weeks. Means ± S.E. shown are representative of at least two independent in vivo experiments, with the number of mice included in each group in each experiment indicated. Statistical significance was determined by two-tailed Student's t test or one-way ANOVA followed by the SNK test for the effect of miR-214 overexpression or HFD versus the corresponding control (*, p < 0.05) or miR-214 overexpression versus control in HFD mice (†, p < 0.05). A and H, miR-214 level; B, fed and fasting blood glucose levels; C, serum insulin levels; D, ITTs, GTTs, and PTTs; E and K, PCK and Glc-6-P (G6P) protein (top or left, Western blot; bottom or right, quantitative measurements of PCK and Glc-6-P protein relative to actin); F, glycogen content; G, Gsk3β, Gys, and Gyp mRNAs; I, fasting blood glucose levels; J, GTTs and PTTs.

Besides gluconeogenesis, glucose homeostasis is also affected by glycogenolysis, a process that involves glycogen synthase kinase-3β (GSK3β), glycogen phosphorylase, and glycogen synthase (30). Therefore, we determined glycogen content to evaluate the contribution of glycogenolysis to the miR-214-mediated suppression of fed and fasting blood glucose levels. Hepatic glycogen content increased in the livers of mice injected with Ad-miR-214 (Fig. 3F). Additionally, miR-214 promoted the expression of Gsk3β and suppressed the expression of Gyp and Gys (Fig. 3G).

Using an HFD-induced diabetic mouse model (39), we investigated whether forced expression of miR-214 in the liver could ameliorate HFD-induced hyperglycemia. Expression levels of miR-214 were elevated ∼2.2-fold in the livers of HFD-fed mice injected with Ad-miR-214 compared with those injected with Ad-GFP (Fig. 3H). The higher fasting blood glucose level induced by HFD was reduced in mice injected with Ad-miR-214 (Fig. 3I). As observed in C57BL/6J mice, miR-214 overexpression also improved glucose tolerance and pyruvate tolerance in the HFD-induced diabetic mouse model (Fig. 3J). Additionally, PCK and Glc-6-P protein levels increased in the livers of HFD-induced diabetic mice compared with mice fed a control diet, and this increase was reversed by Ad-miR-214 (Fig. 3K).

ATF4 Mediates the Suppressive Effects of miR-214 on Gluconeogenesis in Vitro

To understand the mechanisms underlying the suppression of gluconeogenesis by miR-214, we investigated the direct target genes of miR-214. Peroxisome proliferation-activated receptor γ coactivator 1α (PGC1α) is a master regulator of the gluconeogenic program in fasting and diabetic states (40). PGC1α was also predicted as an miR-214 target by PicTar and TargetScan. However, the protein levels of PGC1α were not changed in primary hepatocytes or the livers of mice when miR-214 was overexpressed (Fig. 4A). Based on the findings of a previous study demonstrating that miR-214 inhibited ATF4 expression (27), we tested the hypothesis that miR-214 regulates gluconeogenesis via targeting ATF4. As predicted, ATF4 expression was significantly suppressed by miR-214, as was expression of TRB3, the downstream target of ATF4 (Fig. 4B) (41). Gene silencing of ATF4 reduced Pck and G6p mRNA and protein levels, as well as glucose production, in primary hepatocytes (Fig. 4, C–F).

FIGURE 4.

ATF4 mediates the suppressive effects of miR-214 on gluconeogenesis. A–O, primary mouse hepatocytes (P.H., n = 6 dishes/condition) were transfected or infected with negative control (− miR-214) or miR-214 mimics (+ miR-214), negative control (− siATF4) or small interfering RNA for ATF4 (+ siATF4), control vector (− HA-ATF4), or HA-tagged ATF4 vector (+ HA-ATF4), Ad-GFP (− Ad-miR-214) or Ad-miR-214 (+ Ad-miR-214), Ad-GFP (−Ad-FOXO1) or Ad-FOXO1 (+ Ad-FOXO1), as indicated, for 48 h. A, male C57/B6J wild-type (WT) mice (n = 6 for each group) were injected with Ad-GFP (− Ad-miR-214) or Ad-miR-214 (+ Ad-miR-214) for 9 days. Means ± S.E. shown are representative of at least two independent in vitro or in vivo experiments, with the number of dishes/condition or mice included in each group in each experiment indicated. Statistical significance was determined by two-tailed Student's t test or one-way ANOVA followed by the SNK test for the effect of miR-214, siATF4, HA-ATF4, Ad-miR-214, or Ad-FOXO1 versus corresponding control (*, p < 0.05) or ATF4 overexpression or silence versus control in Ad-miR-214- or Ad-FOXO1-infected cells (†, p < 0.05). A, PGC1α protein (right, Western blot; left, quantitative measurements of PGC1α protein relative to actin); B, ATF4 and TRB3 protein (right, Western blot; left, quantitative measurements of ATF4 and TRB3 protein relative to actin); C and G, Atf4 mRNA; D and H, Pck and G6p mRNAs; E, I, and L, ATF4, PCK and Glc-6-P (G6P) protein (right, Western blot; left, quantitative measurements of ATF4, PCK, and Glc-6-P protein relative to actin); F, J, M, and O, glucose output assay; K, Atf4, Pck, and G6p mRNAs; N, Foxo1, Atf4, and Pck mRNAs.

We then investigated the effects of ATF4 overexpression on gluconeogenesis in primary hepatocytes. The plasmid pCMV-HA-ATF4 (HA-ATF4) induced higher Atf4, Pck, and G6p mRNA and protein expression, as well as glucose production (Fig. 4, G–J), in primary hepatocytes.

To further investigate whether miR-214 suppress gluconeogenesis through targeting ATF4, primary hepatocytes were cotransfected with miR-214 and Ad-ATF4. Overexpression of ATF4 significantly reversed the suppressive effect of miR-214 on PCK and Glc-6-P expression, as well as glucose production, in primary hepatocytes (Fig. 4, K–M).

Given that ATF4 interacts with FOXO1 to regulate the expression of PCK in the bones (30), we speculated that ATF4 could affect FOXO1 transcriptional activity to regulate the expression of PCK in primary hepatocytes. We found that forced expression of FOXO1 induced higher mRNA levels of Pck, which were significantly reduced by ATF4 knockdown (Fig. 4N). The stimulatory effect of FOXO1 on glucose production was also reversed by ATF4 knockdown (Fig. 4O).

miR-214 Suppresses Gluconeogenesis in an ATF4-dependent Manner in Vivo

To further investigate the physiology function of ATF4 on hepatic gluconeogenesis, LV-ATF4 KO were generated (Fig. 5, A and B). Liver lysates from LV-ATF4KO mice showed a noteworthy decrease in Atf4 mRNA and protein levels compared with conditional ATF4 knock-out mice (ATF4^f/f) mice (Fig. 5, C and D). LV-ATF4 KO mice exhibited decreased fed or fasting blood glucose levels (Fig. 5E). PCK, but not Glc-6-P, were significantly suppressed both in mRNA and protein levels (Fig. 5, F and H). Although there were no differences in glucose tolerance between the LV-ATF4 KO and ATF4^f/f mice, the PTT assay showed that hepatic ATF4 knock-out induced a lower rate of de novo glucose synthesis (Fig. 5G). We then injected LV-ATF4 KO mice with Ad-miR-214 to examine its effect on hepatic gluconeogenesis without ATF4. Result revealed that miR-214 could not cause a further decrease in fed or fasting blood glucose levels, nor did it further improve the pyruvate tolerance without ATF4 (Fig. 5, I and J).

FIGURE 5.

Suppressive effects of miR-214 on gluconeogenesis are abrogated in liver-specific ATF4 knock-out (LV-ATF4 KO) mice. A and B, generation of LV-ATF4 KO mice. C–J, 8–10-week old male C57/B6J LV-ATF4 KO mice (+ LV-ATF4 KO) or control mice (− LV-ATF4 KO) or LV-ATF4 KO mice injected with Ad-GFP (− Ad-miR-214) or Ad-miR-214 (+ Ad-miR-214) for 9 days were used for all of the assays (n = 5∼7 for each group). Means ± S.E. shown are representative of at least two independent in vivo experiments, with the number of mice included in each group in each experiment indicated. Statistical significance was determined by two-tailed Student's t test or one-way ANOVA followed by the SNK test for the effect of LV-ATF4 KO versus control mice (*, p < 0.05), or miR-214 overexpression versus control in LV-ATF4 KO mice (†, p < 0.05). A and B, diagrams of ATF4 loxP mice generation strategy and genotyping of ATF4 loxP alleles; C, Atf4 mRNA; D, nuclear ATF4 protein (top, Western blot; bottom, quantitative measurements of ATF4 protein relative to LAMIN B1); E and I, fed and fasting blood glucose levels; F, Atf4, Pck, and G6p mRNAs; G, GTTs and PTTs; H, PCK and Glc-6-P (G6P) protein (left, Western blot; right, quantitative measurements of PCK and Glc-6-P protein relative to actin); J, PTTs.

LV-miR-214 TG Mice Exhibit Suppressed Gluconeogenesis with Lower Expression of ATF4 in Liver

To further confirm the specific effect of miR-214 on hepatic gluconeogenesis, LV-miR-214 TG mice were constructed. The expression of miR-214 was up-regulated by 1.7-fold in the livers of LV-miR-214 TG mice compared with that of control mice (Fig. 6A). We observed no difference in body weight and liver-to-body-weight ratio between LV-miR-214 TG and control mice (Fig. 6, B and C). LV-miR-214 TG mice exhibited significantly decreased fed and fasting blood glucose levels compared with control mice (Fig. 6D). Additionally, LV-miR-214 TG mice showed improved glucose and pyruvate tolerance (Fig. 6E). ATF4, PCK, and Glc-6-P protein levels were significantly reduced in their livers compared with control mice (Fig. 6F).

FIGURE 6.

Gluconeogenesis is suppressed in liver-specific miR-214 transgenic (LV-miR-214 TG) mice. A–G, 10-week-old male C57/B6J control mice (− LV-miR-214 TG) or LV-miR-214 TG mice (+ LV-miR-214 TG) were used for all of the analyse (n = 8–10 for each group). Means ± S.E. shown are representative of at least two independent in vivo experiments, with the number of mice included in each group in each experiment indicated. Statistical significance was determined by two-tailed Student's t test for the effect of LV-miR-214 TG versus control mice (*, p < 0.05). A, miR-214, Atf4, Pck, and G6p mRNAs; B, body weight; C, liver weight to body weight ratio; D, fed and fasting blood glucose levels; E, GTTs and PTTs; F, ATF4, PCK, and Glc-6-P (G6P) protein (left, Western blot; right, quantitative measurements of ATF4, PCK, and Glc-6-P protein relative to actin).

DISCUSSION

The transcriptional regulators of hepatic gluconeogenesis are considered potential therapeutic targets for the treatment of T2D (4). Recently, microRNAs, such as miR-29a-c, miR-23a, and miR-33, have attracted attention as regulators of glucose metabolism (11, 12, 42). In this study, we found that the miR-214 was down-regulated in the livers of fasted, HFD-fed, and db/db mice; and adenovirus-mediated overexpression of miR-214 reduced fasting blood glucose levels in C57BL/6J and HFD-fed mice. Furthermore, LV-miR-214 TG mice exhibited lower fasting blood glucose levels and decreased expression of PCK and Glc-6-P. These findings suggest a pivotal role for miR-214 as a regulator of gluconeogenesis in the fasting or diabetic state.

For the first time, we discovered that miR-214 is down-regulated in mouse liver under fasting or diabetic states. This conclusion is drawn based on results obtained using two different controls, 18 S and U6. Moreover, blocking cAMP-PKA pathway with H89, a widely used PKA inhibitor (34), could reverse the effect of glucagon on the expression of miR-214. This result revealed that expression of miR-214 was indeed regulated by the classical glucagon-GR-PKA pathway, although the direct transcription activators of miR-214 require further investigation. Interestingly, our results further indicated that glucagon decreased miR-214 levels most likely via affecting the maturation of miR-214.

Glucagon levels are increased by fasting (43), and we found that miR-214 levels were suppressed by glucagon. Therefore, the increased glucagon levels might be the primary cause of the decreased miR-214 levels under fasting. However, regulatory mechanisms underlying HFD-suppressed miR-214 levels could be different. Although T2D is characterized by hyperglucagonemia (44–46), glucagon receptor expression was decreased in the livers of mice fed a HFD (47–49). These results suggest that HFD-suppressed miR-214 levels are unlikely to be mediated by stimulated glucagon signaling. HFD induces the higher levels of other hormones, such as insulin, leptin (50), and glucocorticoids (51), which have been shown to regulate gluconeogenesis (52–54). We speculate that the suppressed miR-214 expression by HFD might be mediated by other HFD-increased hormones. This possibility requires further investigation. Moreover, we found that the miR-214 inhibitor reduced the expression of miR-214 about 50%, which resulted in a 1.5-fold induction of Pck and 3-fold induction of G6p. However, the effect of glucagon (which reduces miR-214 by 50%) is to induce Pck and G6p >100-fold. These results suggested that miR-214 has evolved to finely control the expression of these genes, or the expression level of miR-214 has been strictly controlled.

Based on our findings, we further investigated the mechanisms underlying miR-214 regulation of gluconeogenesis. Previous studies reported that miR-214 targets ATF4 to inhibit bone formation (27) and that ATF4 plays a vital role in glucose metabolism (30, 55, 56). In our study, we showed that suppression of miR-214 gluconeogenesis is associated with lower levels of ATF4 protein. Moreover, in primary hepatocytes, ATF4 knockdown mimicked the suppressive effect of miR-214 on gluconeogenesis, whereas overexpression of ATF4 reversed this effect. Furthermore, gluconeogenesis and PCK expression were significantly suppressed in the livers of LV-ATF4 KO mice. The effects of miR-214 on gluconeogenesis were abrogated in these mice. Therefore, our results suggest that ATF4 is likely to play a key role in the regulation of gluconeogenesis by miR-214.

In a previous study, gluconeogenesis was impaired in the livers of Atf4^−/− mice, and this function of ATF4 occurred via its osteoblastic expression (56). Furthermore, the expression of gluconeogenic genes in primary hepatocytes prepared from Atf4^−/− mice was not significantly different from their expression in WT hepatocytes (56). We speculated that these results could be explained by compensatory adaptation to global ATF4 knockdown during development. Recently, Mendez-Lucas et al. (57) reported that the regulation of mitochondrial PCK expression requires recruiting ATF4 to the promoter of PCK. These results proved that ATF4 is a direct regulator of PCK expression and implied that ATF4 might be involved in the regulation of hepatic gluconeogenesis. Indeed, we proved that knockdown of ATF4 using siRNA significantly reduced glucose production and the expression of gluconeogenic genes such as PCK and Glc-6-P. Conversely, overexpression of ATF4 in primary hepatocytes increased glucose production and expression of PCK and Glc-6-P. Furthermore, LV-ATF4 KO mice exhibited decreased fed and fasting blood glucose levels and suppressed gluconeogenesis. Taken together, these results indicate that the expression of gluconeogenic genes is closely related to the expression of ATF4 in primary hepatocytes. Given that ATF4 can interact with transcription factor FOXO1 and promote its binding to the insulin-response element site of PCK (30), we further speculated that ATF4 is a coactivator of FOXO1 in the regulation of gluconeogenesis in primary hepatocytes. This speculation was supported by the observation that ATF4 knockdown partially reduced FOXO1-induced overexpression of PCK, as well as glucose production, in primary hepatocytes.

Given that overexpression of miR-214 resulted in significantly decreased expression of PCK and Glc-6-P and that liver-specific ATF4 deficiency suppressed the expression of PCK but not Glc-6-P, we speculated that another candidate, in addition to ATF4, also mediates miR-214-induced regulation of gluconeogenesis. However, we found that expression of PGC1α is not affected by miR-214 in hepatocytes in vitro and in vivo. Our previous work indicated that ATF4 overexpression inhibits the phosphorylation of AMP-activated protein kinase (58), which is a key regulator of gluconeogenesis (23). Therefore, miR-214 may also regulate gluconeogenesis via AMP-activated protein kinase. The involvement of AMP-activated protein kinase and other possible target genes in miR-214 regulation of hepatic gluconeogenesis requires further studies in the future.

In summary, this study reveals a pivotal role for miR-214 in the regulation of hepatic gluconeogenesis via targeting ATF4. Interestingly, forced expression of miR-214 alleviated hyperglycemia in HFD-induced diabetic mice. Our results indicate that the miR-214-ATF4 axis is a novel pathway for the regulation of hepatic gluconeogenesis, and they highlight the potential of miR-214 and ATF4 as therapeutic targets for the treatment of hyperglycemia in T2D.