Levels of miR-22-3p, a highly abundant hepatic microRNA, are abnormally increased in mouse models of insulin resistance and type 2 diabetes, yet its contribution to deregulated hepatic metabolism under diseased states is not well understood. Here, we unravel a novel link between elevated hepatic miR-22-3p expression and impaired gluconeogenesis in diabetic db/db mice via the regulation of Tcf7 (transcription factor 7). Our data demonstrate that miR-22-3p binds to the 3′ untranslated region of TCF7 and downregulates it, and this microRNA-mediated regulation of TCF7 increases the expression of enzymes of the gluconeogenic pathway in HepG2 cells. Small interfering RNA–mediated knockdown of TCF7 in HepG2 cells also causes similar upregulation of gluconeogenic genes. Furthermore, in vivo silencing of miR-22-3p by antagomiR administration lowered random as well as fasting glucose levels in diabetic mice. miR-22-3p antagonism improved glucose tolerance and insulin sensitivity. Importantly, the hepatic Tcf7 levels were restored along with reduced hepatic glucose output, which was also reflected by the decreased expression of gluconeogenic genes. Our results support a critical role for miR-22-3p and its target, Tcf7, in the pathogenesis of diabetes by upregulating gluconeogenesis. Moreover, targeting the miR-22/Tcf7/Wnt axis might hold therapeutic potential for the treatment of altered hepatic physiology during insulin resistance and type 2 diabetes.
MicroRNAs (miRNAs) are small noncoding RNAs that regulate gene expression by post-transcriptional silencing (1). miRNAs are essential components of the regulatory circuits involved in almost every aspect of biology from development to physiology and disease (2,3). Aberrant miRNA expression underlies the pathogenesis of several diseases, including metabolic disorders such as obesity and type 2 diabetes (4–10). Previously, we had reported an altered hepatic miRNA signature for obese diabetic db/db mice (11). Targets to these altered miRNAs mapped onto the Wnt signaling pathway, and we further observed that many components of this pathway were downregulated in the diabetic liver. Besides the well-known functions of Wnt signaling in early development and organogenesis, it plays a role in postnatal liver growth and regeneration, and in the establishment of hepatic zonation (12). However, the Wnt pathway has not been explored extensively in the context of hepatic metabolism. Evidence suggests that Wnt/β-catenin signaling regulates hepatic glucose production and is important for bile acid synthesis in mice (13,14). miR-22-3p (also mentioned as miR-22), an abundantly expressed hepatic miRNA, is elevated in diabetic mouse liver, but its metabolic implications have not been studied. We had reported (11) that miR-22 was predicted to target more than one gene of the Wnt signaling pathway, suggesting that it might be a critical modulator of this pathway.
Metabolic disorders such as obesity and type 2 diabetes are characterized by deregulated glucose metabolism (15). One such deregulation is increased hepatic glucose production due to impaired gluconeogenesis, which contributes to elevated blood glucose levels (16–18). In subjects with type 2 diabetes, the rates of gluconeogenesis and gluconeogenic flux are elevated and strongly correlate with fasting glucose levels (19). Even in fed states, gluconeogenesis is inadequately repressed, hence, causing higher glucose levels in the postprandial phase as well (20). Such metabolic abnormalities are also observed in prediabetic subjects, suggesting that defects in the gluconeogenic pathway occur early and might be a major contributing factor in the development and progression of type 2 diabetes (21).
We therefore sought to explore the link between miR-22–mediated alterations in the Wnt signaling pathway and dysregulated hepatic glucose metabolism. Here, we show that miR-22 targets Tcf7 (transcription factor 7), an important effector molecule in the Wnt pathway, and, in doing so, regulates gluconeogenesis in vitro and in vivo. Furthermore, we provide evidence for therapeutic targeting of miR-22 to treat insulin resistance and type 2 diabetes.
Research Design and Methods
All animal experiments were performed according to the guidelines of the Institutional Animal Ethical Committee. Male diabetic (C57BLKs-db/db) and nondiabetic (C57BLKs-db/+) mice 10–12 weeks of age were maintained at the Council of Scientific and Industrial Research-Central Drug Research Institute (Lucknow, India) on a 12-h light-dark cycle. The mice had ad libitum access to food and water. Wild-type male mice 10–12 weeks of age (C57BL/6J; 25 ± 1.6 g, n = 7) were injected intravenously (once on 2 alternate days) with either the scramble or an miR-22 mimic (5 mg/kg body wt; Ambion; Life Technologies, Carlsbad, CA) using Invivofectamine (Life Technologies). Another set of such mice (n = 6) was injected with the scramble or the miR-22 antagomiR (antagomiR-22 [or Ant-22]) (40 mg/kg body wt; GE Dharmacon, Lafayette, CO) for 3 consecutive days. Random and fasting glucose levels were measured, and an oral glucose tolerance test (OGTT) was performed as described below.
Cell Culture and Transfections
HepG2 or Hepa 1–6 cells were reverse transfected with the mature miR-22 mimic (40 nmol/L) with or without its hairpin inhibitor (40 nmol/L) (GE Dharmacon), and control cells were transfected with scramble. Endogenous miR-22 was knocked down by transfection with the miR-22 inhibitor (10–50 nmol/L, 48 h). HepG2 cells were reverse transfected with TCF7 small interfering RNA (siRNA) (20 nmol/L, 48 h; GE Dharmacon) using Lipofectamine RNAiMax (Life Technologies). HepG2 cells transfected with the scramble or miR-22 were serum starved for 12 h and incubated in either the absence (fasted) or presence of insulin (refed; 50 nmol/L). After 4 h, total RNA was isolated, and levels of PCK1 (phosphoenol pyruvate carboxykinase 1) and G6PC (glucose 6-phosphatase) were assessed by quantitative real-time PCR (qRT-PCR).
Cell or tissue samples were lysed in radioimmunoprecipitation assay lysis buffer containing protease inhibitors (Calbiochem, Darmstadt, Germany). Protein samples (30 μg) were analyzed by Western blotting for TCF7 (Sigma-Aldrich, St. Louis, MO) and PCK1 (Abcam, Cambridge, U.K.). Vinculin, GAPDH, or Hsc70 (Santa Cruz Biotechnology, Dallas, TX) was used as a loading control. Scramble or miR-22 mimic transfected cells were incubated in the absence or presence of insulin (50 nmol/L, 20 min). Cells were lysed, and 30 μg protein was subjected to Western blot analyses to evaluate for the levels of insulin signaling intermediates, insulin receptor, phospho-insulin receptor, pAkt, and Akt (Cell Signaling Technology, Danvers, MA). β-Actin was taken as the loading control.
Cloning, Mutagenesis, and Luciferase Assays
Luciferase reporter constructs were generated by cloning the 3′ untranslated region (UTR) of the human TCF7 gene using primers (Table 1) spanning the binding sites for miR-22, downstream of Renilla luciferase in psiCheck2 vector (Promega Corporation, Madison, WI), and mutations in the binding site were generated using specific primers (Table 1) and a site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA). HEK293 cells were cotransfected with either the wild-type or mutated 3′ UTR reporter plasmids along with miR-22 mimics and/or miR-22 hairpin inhibitors. A dual luciferase assay (Promega Corporation) was performed after 48 h of transfection, and luminescence was measured on an Infinite M200 Pro Multimode Reader (TECAN, Männedorf, Switzerland). Renilla luciferase values were normalized to those of firefly luciferase.
RNA Isolation and qRT-PCR
Total RNA was isolated from cells transfected with either the scramble or miR-22 and/or its inhibitor (48 h) or TCF7 siRNA (48 h) or miR-22 together with a TCF7 overexpression vector (500 ng, 36 h; OriGene, Rockville, MD) using Invitrogen TriZol (Life Technologies). For animal tissue samples, total RNA was isolated using the Ambion mirVana miRNA isolation kit (Life Technologies). RNA (2 μg) was reverse transcribed using random hexamers, and the expression levels of genes (TCF7, PCK1, G6PC, FBP1 [fructose 1–6 bisphosphatase], PC [pyruvate carboxylase], TBP [TATA box binding protein]) were measured using specific primers (Table 1) and Applied Biosystems SYBR green Master Mix (Life Technologies) on an Applied Biosystems StepOne Plus Real Time PCR (Life Technologies). TBP was taken as the normalization control. Relative levels of miR-22 were quantified using stem loop primers, as described previously (11), normalized to U6 or sno234, and data were analyzed by the 2-ΔΔCt method. For quantification of endogenous miR-22 levels and after overexpression, serial dilutions of small synthetic RNA (21 nucleotides; Systems Biosciences, Mountain View, CA) were reverse transcribed and assayed by qRT-PCR. Total RNA (250 ng) from scramble and miR-22–overexpressed HepG2 cells were also subjected to qRT-PCR. A standard curve was generated for the synthetic RNA and was used to quantify endogenous levels of miR-22 in HepG2 cells.
Cignal Reporter Assay
HepG2 cells were transfected with the scramble or miR-22 and/or its inhibitor (48 h), TCF7 siRNA (48 h), or miR-22 together with a TCF7 overexpression vector, and Wnt signaling was assessed using the TCF/LEF Reporter Assay Kit (CCS-018L; QIAGEN). Luciferase activity was measured after 48 h. Renilla reporter plasmid served as the normalization control.
Glucose Production Assay and Lactate Measurements
HepG2 cells were transfected with the scramble or miR-22 for 48 h. Prior to the termination of incubation, cells were either serum starved for 2 h and then incubated in glucose-free media for another 3 h for the estimation of extracellular and intracellular lactate levels using an L-Lactate Assay Kit (Cayman Chemical, Ann Arbor, MI) or were incubated in glucose production media (DMEM glucose-free, 20 mmol/L lactate, 2 mmol/L pyruvate, and 0.5% BSA) for 4 h for glucose content in the media using an Amplex Red Glucose/Glucose Oxidase Assay Kit (Life Technologies). Data were normalized to the protein content.
AntagomiRs were custom synthesized by GE Dharmacon, as described in a previous study (22). The sequence for the antagomiR-22 used was 5′-ascsaguucuucaacuggcagscsusus-Chol-3′, where the lowercase letters represent 2′-O methyl–modified nucleotides and subscript “s” represents a phosphorothioate linkage and a 3′-cholesterol moiety linked through a hydroxyprolinol linkage. An antagomiR-scramble (5′-gsascuccacucuucuagaausasascs-Chol-3′) was used as a negative control.
The db/db mice (n = 6) were injected with anatgomiR-22 or antagomiR-scramble (Ant-Scr) for 3 consecutive days via the tail vein at a dose of 40 mg/kg body wt. The db/+ mice were injected with saline or Ant-Scr. Random blood glucose levels were measured (ACCU-CHEK; Roche, Mannhein, Germany), and OGTT, pyruvate tolerance test (PTT), and insulin tolerance test (ITT) were performed. Mice were given normal access to food before euthanasia (9 days after the first antagomiR injection). Liver tissue samples were isolated, and levels of miR-22, Tcf7, and gluconeogenic genes were assessed by Western blot and qRT-PCR. Adipose and skeletal muscle tissue samples were also collected to evaluate the levels of miR-22.
OGTT and PTT
OGTT and PTT were performed on overnight fasted mice (12 h). Fasting glucose levels were measured, and mice were administered a bolus of glucose orally at a dose of 2 g/kg body wt for the OGTT or were given an intraperitoneal injection of sodium pyruvate at a dose of 2 g/kg body wt for the PTT. Blood glucose measurements were made after 30, 60, 90, and 120 min.
ITTs were conducted on mice fasted for 5 h. Mice received intraperitoneal injections of insulin (5 units/kg for db/db mice and 0.75 units/kg for db/+ mice), and subsequently blood glucose levels were determined at 0, 15, 30, and 60 min.
Blood samples were collected from mice prior to sacrifice. Random insulin levels were estimated by an insulin ELISA kit according to the manufacturer protocol (Calbiotech, Spring Valley, CA). Lipid analysis was performed using a bioanalyzer (DTN-410-K; DIALAB, Vienna, Austria). Random triglyceride, total cholesterol, LDL, HDL, alanine aminotransferase (ALT), and aspartate aminotransferase (AST) levels were determined using kits (DIALAB) according to the company protocol.
Alpha DigiDoc 1201 software (Alpha Innotech Corporation, San Leandro, CA) was used to evaluate protein expression by densitometric analysis.
All bars represent the mean ± SD, and data were analyzed using a Student two-tailed t test.
miR-22 Levels Are Elevated in the Livers of Diabetic Mice
In a previous study (11), using an miRNA microarray, we had reported an altered miRNA signature in the db/db mouse liver. Thirteen miRNAs were altered, and pathway analyses of their predicted targets revealed over-representation of the Wnt signaling pathway. miR-22 was one of the significantly upregulated miRNAs, and its predicted targets mapped onto more than one point on the Wnt signaling pathway. Additionally, sequencing data from different studies (8,23) suggest that miR-22 is highly abundant in the adult liver. Though other studies (9,10) have also confirmed the increased hepatic expression of miR-22 under diabetic conditions, the metabolic consequences of this have not yet been studied. The high abundance of miR-22 in the liver and the evidence that it is dysregulated in diabetes prompted us to hypothesize that it could be important in regulating hepatic metabolic pathways, derangements of which contribute to the pathogenesis of type 2 diabetes. The evaluation of miR-22 levels by qRT-PCR demonstrated it to be significantly upregulated (by 2.5-fold) in the db/db mouse liver (Fig. 1A).
TCF7 Is a Direct and Conserved Target of miR-22
Using a pathway-specific PCR array, we had demonstrated in a previous study (11) that several components of the Wnt pathway, including Tcf7, are downregulated in the diabetic liver. This was validated for Tcf7 by qRT-PCR, and, as shown in Fig. 1B, hepatic levels of Tcf7 are significantly inhibited in db/db mice. Interestingly, Tcf7 protein levels are significantly decreased in livers of db/db mice (Fig. 1C). Using two different computational programs, TargetScan (http://www.targetscan.org/) and miRanda (http://microRNA.org/), we identified Tcf7 as a predicted target of miR-22, and various genetic association studies (24–27) across populations have shown that polymorphisms in genes of the TCF family are strongly associated with diabetes susceptibility. An increase in hepatic miR-22 levels accompanied by decreased Tcf7 levels suggested that it could be a direct target of miR-22. The TCF7 3′ UTR harbors two binding sites for miR-22; the site spanning 875–899 (human) and 958–979 (mouse) nucleotides (http://microRNA.org) is conserved across species and is shown in Fig. 1D. Overexpression of miR-22 in HepG2 cells led to an 200-fold increase over its endogenous levels (14,000 copies/cell), and this significantly decreased the mRNA and protein levels of TCF7 that were prevented by the miR-22 inhibitor (Fig. 1E and F). miR-22 inhibited Tcf7 protein levels in Hepa 1–6 cells too (Fig. 1F). The inhibition of endogenous miR-22 levels using an miR-22 inhibitor significantly upregulated TCF7 protein levels (Fig. 1G). miR-22 decreased the luciferase activity of the reporter plasmid containing the TCF7 3′ UTR (Fig. 1H). This decrease in the luciferase activity by miR-22 was prevented, both in the presence of the miR-22 inhibitor and in mutations in the conserved miR-22 binding site on the TCF7 3′ UTR (Fig. 1H). Mutation in the other miR-22 binding site did not restore the luciferase activity (data not shown). All of these data suggest that miR-22 binds to the 3′ UTR of TCF7, and this interaction specifically regulates TCF7 levels within the cell.
miR-22 Regulates Wnt Signaling and the Expression of Gluconeogenic Genes Through the Repression of TCF7
Since TCF7 is a major effector molecule of the Wnt pathway, we sought to assess the effect of miR-22 on the activity of this pathway. In the absence of miR-22, the TCF/LEF (lymphoid enhancer factor) reporter demonstrated significant activity of the luciferase reporter driven by the TCF/LEF response element. However, miR-22 decreased this activity, and this was prevented in the presence of the miR-22 inhibitor (Fig. 2A). A similar reduction in Wnt activity was also observed in the presence of TCF7 siRNA (Fig. 2A). Interestingly, when TCF7 was overexpressed in the presence of miR-22, the miR-22–mediated decrease was significantly abrogated, suggesting a mediatory role of TCF7 in the miR-22 effects. Results until now have demonstrated that miR-22 regulates the Wnt signaling pathway by binding to the TCF7 3′ UTR and regulating its levels.
The liver is a unique tissue since it is capable of both glucose production and glucose consumption. While hepatic gluconeogenesis maintains normal blood glucose homeostasis, its uncontrolled activation significantly contributes to hyperglycemia during diabetes. Since miR-22 levels are elevated in the liver during diabetes, we evaluated the effect of miR-22–mediated regulation of TCF7 on the status of gluconeogenic gene expression. PCK1 and G6PC levels were significantly increased by approximately twofold and sevenfold, respectively (Fig. 2B), while FBP1 and PC were modestly increased in the presence of miR-22. This increase in the transcript levels of respective genes was prevented in the presence of the miR-22 inhibitor. Also, there was a significant increase in hepatic glucose output in the presence of miR-22 (Fig. 2C). However, cellular lactate content or lactate released into the medium did not show any significant alteration in the presence of miR-22 (Fig. 2D).
We subsequently sought to examine whether the effects of miR-22 on gluconeogenic genes (Fig. 2B) were mediated through inhibition of TCF7. TCF7 siRNA (20 nmol/L) significantly inhibited TCF7 at the transcript and protein levels (Fig. 2E) and upregulated the levels of rate-limiting genes of gluconeogenesis (i.e., PCK1 and G6PC) (Fig. 2F). Although miR-22 modestly increased FBP1 and PC expression levels (Fig. 2B), TCF7 silencing alone did not alter their transcript levels. This possibly suggests that the effects of miR-22 on gluconeogenesis mediated by TCF7 are primarily a result of PCK1 and G6PC regulation. This was further evident since TCF7 overexpression, at least in part, significantly abrogated the effects of miR-22 on PCK1 and G6PC (Fig. 2G). Taken together, these in vitro studies suggest that miR-22 plays a critical role in regulating the expression of important gluconeogenic genes. Since the gluconeogenic program is tightly regulated by insulin, we assessed insulin sensitivity in HepG2 cells in the presence of miR-22 and its inhibitor. Compared with scramble-transfected cells, there was no marked alteration in insulin-stimulated phosphorylation and the activation of two insulin-signaling intermediates, namely IR and Akt (Fig. 2H), in the presence of miR-22 or its inhibitor. These suggest that, although miR-22 affects the gluconeogenic pathway, this effect is independent of insulin signaling and presumably has a direct effect on gluconeogenic gene transcription. Insulin refeeding repressed fasting-induced expression of PCK1 and G6PC in both scramble- and miR-22–transfected cells, further suggesting that insulin signaling remains unaffected by miR-22 (Fig. 2I).
In Vivo Silencing of miR-22 Improves Circulatory Glucose and Insulin Levels
The schema of the in vivo experimental strategy is shown in Fig. 3A. OGTTs, ITTs, and PTTs were performed in the same animals, and tissue samples were collected for assessing the status of miR-22 and its targets. As shown in Fig. 3B, compared with db/db mice injected with Ant-Scr, the administration of antagomiR-22 resulted in a >90% inhibition of hepatic miR-22 levels in the livers of db/db mice, as assessed by qRT-PCR. miR-22 levels did not depict any alteration in the adipose tissue and skeletal muscle of db/db mice; however, systemic antagomiR-22 administration inhibited endogenous miRNA levels in both of these tissue samples. Compared with Ant-Scr–injected db/db mice, antagomiR-22 administration led to a significant lowering of random (measured after each injection) and fasting (on Day 5 from the day of the first injection) blood glucose levels (Fig. 3C and D). Also, miR-22 inhibition brought about a significant diminution in the random circulatory insulin levels (Fig. 3E).
Silencing of miR-22 Improves Glucose, Pyruvate, and Insulin Tolerance In Vivo
AntagomiR-22–treated db/db mice demonstrated a significant improvement in glucose tolerance compared with scramble-injected db/db mice (Fig. 4A). Silencing miR-22 increased insulin sensitivity, as is evident by the ITT results, where Ant-22–treated diabetic mice displayed significantly enhanced clearance of basal glucose after insulin injection compared with Ant-Scr–injected mice at 60 min (Fig. 4B). AntagomiR-22–injected db/db mice also exhibited improved pyruvate tolerance (Fig. 4C). miR-22 antagonism in wild-type C57BL/6J mice did not have any effect on random glucose levels or glucose tolerance (Fig. 4D). Additionally, antagomiR-22–injected db/db mice also demonstrated moderate but significant improvement in circulatory triglycerides and HDL, LDL, and total cholesterol levels (Fig. 4E). Serum AST and ALT revealed that there was no hepatic toxicity associated with the antagomiR treatment (Fig. 4E). All of these data indicate that inhibition of miR-22 improves the circulatory metabolic parameters.
AntagomiR-22 Normalizes Hepatic Tcf7 Levels and Gluconeogenic Gene Expression In Vivo
Since, as demonstrated above, miR-22 antagonism alleviates hyperglycemia, improves insulin sensitivity, and decreases de novo glucose production, we further sought to evaluate its effect on Tcf7 and gluconeogenic gene expression. The inhibition of miR-22 restored hepatic Tcf7 levels in diabetic mice (Fig. 5A). This confirms that miR-22 also targets Tcf7 in vivo, and its antagonism normalizes hepatic Tcf7 levels.
There was significant improvement in the hepatic levels of Pck1, G6pc, Fbp1, and Pcx (pyruvate carboxylase). While the expression of all these enzymes was elevated in the livers of db/db mice compared with db/+ mice, in antagomir-22–injected db/db mice, the levels of the enzymes were significantly reduced (Fig. 5B). As in the transcript levels, hepatic protein levels of Pck1 also significantly improved in db/db mice treated with antagomiR-22 (Fig. 5C).
To evaluate gain of function in vivo, mature miR-22 mimics were injected into wild-type C57BL/6J mice. As shown in Fig. 5D, overexpression of miR-22 significantly increased both the random circulatory glucose levels (data presented are from the day just after the second injection) as well as the fasting circulatory glucose levels (data presented are from the second day after the second injection). Also, OGTT results revealed considerable impairment of glucose disposal. These in vivo gain-of-function studies prove that miR-22 alone is sufficient to induce the diabetic phenotype.
Our results, therefore, demonstrate that elevated levels of hepatic miR-22 in db/db mice target Tcf7 and, by doing so, modulate the levels of gluconeogenic genes and subsequently control circulatory glucose levels, insulin sensitivity, and de novo glucose production (Fig. 5E).
In the present study, we provide evidence in support of a critical role for the elevated expression of hepatic miR-22 in the pathogenesis of type 2 diabetes. miR-22 is abundantly expressed in the liver and is conserved across several vertebrate species. There are a few reports (28,29) that show altered hepatic expression of miR-22 during diseases. Interestingly, in concordance with our observation, increased hepatic expression of miR-22 has been reported in other mouse models of insulin resistance and type 2 diabetes (9,10), but its functional relevance has not been studied. We demonstrate that silencing miR-22 in db/db mice alleviates both random and fasting hyperglycemia. In addition, inhibiting miR-22 levels in vivo improves glucose tolerance and insulin sensitivity, suggesting a functional role of increased miR-22 levels in the pathophysiology of type 2 diabetes.
miR-22 and Tcf7 demonstrate inverse patterns of expression in db/db mouse liver, and using HepG2 cells we validated TCF7 as a direct target of miR-22. Tcf7, in response to Wnt signals, associates with its most favorable coactivator, β-catenin, and regulates downstream target genes, primarily those involved in processes of cell proliferation, differentiation, and growth (30). Recently, Wnt signaling has emerged as an important regulator of hepatic metabolic pathways. Evidence suggests that Wnt/β-catenin signaling is important for bile acid metabolism, as β-catenin knockout mice accumulate hepatic cholesterol along with high serum bilirubin levels (13). Liu et al. (14) have shown that β-catenin alters serum glucose concentrations by regulating hepatic glucose production. Mice deficient in Sfrp1, a Wnt antagonist, demonstrate impaired glucose tolerance along with upregulation of gluconeogenesis (31). Single nucleotide polymorphisms in another member of this family, TCF7L2, confer increased risk toward development of diabetes in various ethnic groups including the Indian population (26,27,32). Interestingly, SNPs in the TCF7 gene represent a risk factor for type 1 diabetes but the role of TCF7 in type 2 diabetes is not known (24,25). These factors indicate that in addition to being critical in processes like development and differentiation, components of Wnt signaling are important in metabolic processes too. The observation that miR-22 antagonism normalizes hepatic levels of Tcf7 in diabetic mice along with improving metabolic parameters indicates an important role for Tcf7 in regulating hepatic glucose metabolism. Impaired hepatic gluconeogenesis is the principal cause of fasting and postprandial hyperglycemia in type 2 diabetes. Studies in rodents (33,34) indicate that increased hepatic glucose output is caused by increased transcription of two key gluconeogenic genes, Pck1 and G6pc, mediated by an integrated crosstalk among several transcription factors and cofactors (35). A few recent studies also suggest involvement of miRNAs in modulating gluconeogenesis (9,10,36,37). Our data show that overexpression of miR-22 and knockdown of TCF7 increases the levels PCK1 and G6PC. However, miR-22 does not interfere with insulin signaling suggesting that these miR-22 effects are independent of insulin action. miR-22 has been reported to target Glut1 (38); however, Glut1 levels are significantly low in insulin-responsive tissues, including the liver where it minimally participates only in basal glucose uptake (39). So, the improved glucose homeostasis that we observe does not seem to involve the role of Glut1. These data shed light on the mechanism by which augmented hepatic miR-22 might contribute to the pathogenesis of type 2 diabetes by impairing gluconeogenesis. Antagonism of miR-22 in db/db mice decreases hepatic expression of gluconeogenic genes (Pck1, G6pc, Fbp1, and Pcx), fasting glucose levels, and glucose output, suggesting a critical role of miR-22 inhibition in normalizing hepatic gluconeogenesis. Additionally, a moderate but favorable change in the lipid profile in antagomiR-22–treated diabetic mice suggests an encouraging effect of miR-22 antagonism. miR-22 has been shown to regulate cardiac hypertrophy and remodeling (29), and, since such cardiac complications are frequently encountered during diabetes, we think that, in addition to the beneficial effects of miR-22 antagonism that we present here, this might also be significant in potentially targeting diabetic cardiomyopathy.
While our results demonstrate that miR-22 targets Tcf7 and alters gluconeogenesis, the mechanism by which Tcf7 regulates hepatic gluconeogenesis is not known. Liu et al. (14) have shown that β-catenin binds with Foxo1, a master regulator of gluconeogenesis (40), instead of TCFs under starved conditions, resulting in the induction of the gluconeogenic program in mice. A reciprocal occupancy of the gluconeogenic gene promoters by Foxo1 and Tcf7l2 during fed and fasted states was shown by Oh et al. (41). Others have shown (42,43) increased interaction between β-catenin and members of the Foxo family under oxidative stress. These indicate a possible mechanism, but further investigations are required to explain the details of the mechanism by which miR-22–mediated repression of Tcf7 regulates gluconeogenesis.
The results from our current study, therefore, identify miR-22 as a novel metabolic regulator and provide evidence of its contribution in elevating hepatic gluconeogenic gene expression by targeting Tcf7. Our findings provide key information for designing appropriate therapeutic strategies for diabetes.
R.S. is currently affiliated with the Department of Hematology, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel.
Acknowledgments. The authors thank Dr. Rakesh K. Tyagi, Special Centre for Molecular Medicine, Jawaharlal Nehru University, New Delhi, India, for providing Hepa 1–6 cells for the experiments. The authors also thank Dr. Soumya Sinha Roy, Council of Scientific and Industrial Research-Institute of Genomics and Integrative Biology, New Delhi, India, for useful suggestions.
Funding. This work was supported by the Council of Scientific and Industrial Research, New Delhi, India (grant BSC0123). K.K. and S.V. were supported by fellowships from the University Grants Commission and the Indian Council of Medical Research, New Delhi, India.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. K.K. conceived, designed, and performed the experiments; analyzed the data; and wrote the article. S.V., A.M., and V.P.S. performed the experiments. R.S. performed the experiments and contributed reagents. A.K.S. contributed reagents. M.D. researched, conceived and designed the experiments, analyzed the data, and wrote the article. M.D. 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.