The role of a glucagon/cAMP-dependent protein kinase–inducible coactivator PGC-1α signaling pathway is well characterized in hepatic gluconeogenesis. However, an opposing protein kinase B (PKB)/Akt-inducible corepressor signaling pathway is unknown. A previous report has demonstrated that small heterodimer partner–interacting leucine zipper protein (SMILE) regulates the nuclear receptors and transcriptional factors that control hepatic gluconeogenesis. Here, we show that hepatic SMILE expression was induced by feeding in normal mice but not in db/db and high-fat diet (HFD)-fed mice. Interestingly, SMILE expression was induced by insulin in mouse primary hepatocyte and liver. Hepatic SMILE expression was not altered by refeeding in liver-specific insulin receptor knockout (LIRKO) or PKB β-deficient (PKBβ−/−) mice. At the molecular level, SMILE inhibited hepatocyte nuclear factor 4–mediated transcriptional activity via direct competition with PGC-1α. Moreover, ablation of SMILE augmented gluconeogenesis and increased blood glucose levels in mice. Conversely, overexpression of SMILE reduced hepatic gluconeogenic gene expression and ameliorated hyperglycemia and glucose intolerance in db/db and HFD-fed mice. Therefore, SMILE is an insulin-inducible corepressor that suppresses hepatic gluconeogenesis. Small molecules that enhance SMILE expression would have potential for treating hyperglycemia in diabetes.
Insulin induces the insulin receptor tyrosine kinase–mediated activation of the phosphatidylinositol 3-kinase pathway that controls hepatic glucose production. Ablation of insulin signaling leads to the increased gluconeogenesis in type 2 diabetes (1–3). In the fed condition, insulin inhibits hepatic gluconeogenesis by downregulating the expression of PEPCK and glucose-6-phosphatase (G6Pase). This pathway involves phosphorylation of the forkhead transcription factor FOXO1 and CREBP (4–7) and recruitment of coactivators, including PGC-1α and CREB-regulated transcription coactivator 2 (8,9).
Small heterodimer partner–interacting leucine zipper protein (SMILE), including two alternative translation-derived isoforms, SMILE-L (CREBZF: long form of SMILE) and SMILE-S (Zhangfei: short form of SMILE), has been classified as a member of the CREB/ATF family of basic region-leucine zipper transcription factors. However, SMILE cannot bind to DNA as a homodimer (10–12). SMILE has also been reported to function as a coactivator of activating transcription factor 4 or as a corepressor of host cell factor-binding transcription factor (13,14). Previously, we have reported that SMILE is a corepressor of the estrogen receptor–related receptor γ, glucocorticoid receptor (GR), constitutive androstane receptor, hepatocyte nuclear factor 4α (HNF4α), and CREBH (15–17). A recent study demonstrated that SMILE activates tumor suppressor p53 and inhibits the function of bone morphogenetic protein 6 by interacting with Smads (18,19). However, the roles of SMILE in hepatic glucose metabolism still need to be clarified.
PGC-1α is a multifunctional transcriptional coactivator involved in diverse physiological metabolisms. In the liver, PGC-1α expression is induced in the fasting state and increases the expression of gluconeogenic genes via direct interaction with transcription factors including HNF4α, FOXO1, and GR (8,20). Overexpression of PGC-1α leads to the increased expression of G6Pase and PEPCK, key enzymes in the hepatic glucose production. Conversely, knockdown or knockout of PGC-1α results in the lower blood glucose levels as a result of reduced gluconeogenesis.
In this report, we show that SMILE is tightly regulated by the nutritional status, displaying an expression pattern opposite to that of PGC-1α. Whereas hepatic SMILE expression was elevated in the fed state relative to fasting, this regulation was lost in the absence of insulin signaling. Moreover, SMILE regulates hepatic gluconeogenesis via inhibition of PGC-1α–mediated gluconeogenic genes expression. Collectively, this study indicates that SMILE counteracts the stimulatory effect of PGC-1α on hepatic gluconeogenesis and plays an important role in insulin action on hepatic glucose metabolism.
Research Design and Methods
Four-week-old male C57BL/6 mice (for the DIO model) purchased from Orient (Seongnam, South Korea) were fed a high-fat diet (HFD) (60 kcal% fat diet, D12492; Research Diets, Inc.) for 12 weeks (12:12 h light:dark cycle). Male 8- to 12-week-old C57BL/6 and db/db mice were provided with a standard rodent diet. Livers of LIRKO and PKBβ−/− mice were provided by S.B. Biddinger (Harvard Medical School, Boston, MA) and B.A. Hemmings (Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland) as previously described (21). Adenoviruses GFP and SMILE were delivered by tail vein injection on the 7th day. Seven days later, protein and total RNA were extracted from liver for Western blot and quantitative PCR (qPCR) analyses, respectively. For measurement of fasting blood glucose levels, animals were fasted for 4 h with free access to water. For activation of the insulin signaling pathway, either PBS (for control) or insulin (0.5 unit/kg body wt i.p.) was injected for 1–12 h before the collection of liver for further analyses. All procedures were performed in a specific pathogen-free facility at the Korea University College of Medicine, based on the protocols that were approved by the Korea University Institutional Animal Care and Use Committee.
Glucose, Insulin, and Pyruvate Tolerance Tests
For the glucose or pyruvate tolerance tests, mice were fasted for 16 h and then injected with 2 g/kg i.p. glucose or 1–1.5 g/kg i.p. pyruvate (on day 6 post–adenoviral injection). Plasma glucose levels were measured using blood drawn from the tail vein at designated time points using an automatic glucose monitor (One Touch, LifeScan Ltd., Milpitas, CA). For the insulin tolerance test, mice were fasted for 4 h and then injected with 0.75–1 unit/kg i.p. insulin on day 6 after adenoviral injection.
Plasma insulin was measured using plasma collected at time of sacrifice by using mouse insulin ELISA kit (Shibayagi Co., Ltd., Ishihara, Japan).
Liver tissue (30 mg) was homogenized in 0.1 mol/L ice-cold Citrate buffer (pH 4.2) plus NaF, and glycogen was measured using Enzychrom Glycogen assay kit (Bioassay Systems, Hayward, CA) according to the manufacturer’s instructions.
Glucose Output Assay
Glucose production from primary hepatocytes was measured according to the manufacturer's protocol, using a colorimetric glucose oxidase assay (Sigma). Primary hepatocytes were seeded and cultured for 24 h, the cells were washed three times with phosphate-buffered saline, and the medium was then replaced with glucose production buffer (glucose-free DMEM, pH 7.4, containing 20 mmol/L sodium lactate, 1 mmol/L sodium pyruvate, and 15 mmol/L HEPES without phenol red). The glucose assays were performed in triplicate.
Insulin was purchased from Sigma-Aldrich (St. Louis, MO). Forskolin was purchased from Calbiochem and dissolved in the recommended solvents.
Cell Culture and Transient Transfection Assay
α-Mouse liver 12 (AML12) cells were maintained in DMEM/F-12 (Invitrogen), supplemented with 10% FBS (Cambrex Bio Science Walkersville, Inc., Walkersville, MD) and antibiotics (Invitrogen). Cells were split in 24-well plates (2–8 × 104 cells/well) at the day before transfection. Transient transfections were performed using the Lipofectamine 2000 transfection reagent (Invitrogen) according to the manufacturer's instructions. Total DNA used in each transfection was adjusted to 0.8 μg/well by adding appropriate amount of empty vector, and cytomegalovirus-β-galactosidase plasmids were cotransfected as an internal control. The cells were harvested 48 h after the transfection, and luciferase activity was measured and normalized to β-galactosidase activity.
Culture of Primary Hepatocytes
Primary hepatocytes were prepared from C57BL/6 mice by the collagenase perfusion method as previously described (22). After attachment, cells were infected with adenoviruses for 18 h. Subsequently, cells were maintained in the serum-free medium 199 media (Mediatech) overnight and treated with 10 μmol/L forskolin and/or 100 nmol/L insulin.
Adenoviruses expressing unspecific short hairpin (sh)RNA, shSMILE, control GFP, SMILE, HNF4α, PGC-1α, SREBP-1c, and dominant-negative SREBP-1c (dn-SREBP-1c) were described previously (17,22–24). All viruses were purified by using CsCl2 or the Adeno-X Maxi Purification kit (Clontech, Mountain View, CA).
Chromatin Immunoprecipitation Assay
The chromatin immunoprecipitation (ChIP) assay was performed as previously described (17). In brief, mouse livers (40 mg/sample) were fixed with 1% formaldehyde and sonicated. The soluble chromatin was then subjected to immunoprecipitation (IP) using anti-HA (cat. no. 2367; Cell Signaling Technology), anti-FLAG (2368; Cell Signaling Technology), anti–PGC-1α (sc-13067; Santa Cruz Biotechnology, Santa Cruz, CA) and anti-SMILE (sc-49329; Santa Cruz Biotechnology) followed by treatment with protein A-agarose/salmon sperm DNA (Upstate Biotechnology, Upstate, NY). DNA samples were quantified by quantitative real-time PCR using two pairs primers of the mouse Pepck promoter, forward (5′-GATGTAGACTCTGTCCTG-3′) and reverse (5′-GATTGCTCTGCTATGAGT-3′); mouse G6Pase promoter, forward (5′-CCTTGCCCCTGTTTTATATGCC-3′) and reverse (5′-CGTAAATCACCCTGAACATGTTTG-3′); or mouse PGC-1α promoter, forward (5′-GCCTATGAGCACGAAAGGCT-3′) and reverse (5′-GCGCTCTTCAATTGCTTTCT-3′).
Western Blot Analysis
Whole-cell extracts were prepared using RIPA buffer (Elpis-Biotech, Korea). Proteins from whole-cell lysates were separated by 10% SDS-PAGE and then transferred to nitrocellulose membranes. The membranes were probed with indicated antibodies. Immunoreactive proteins were visualized using an Amersham ECL kit (GE Healthcare, Piscataway, NJ), according to the manufacturer’s instructions.
Total RNA from primary hepatocytes or liver was extracted using an RNeasy mini-kit (Qiagen). All of the samples were combined for the experiment. cDNA was generated by Superscript II enzyme (Invitrogen) and analyzed by qPCR using an SYBR green PCR kit and a TP800 Thermal Cycler DICE Real Time system (Takara). All data were normalized to ribosomal L32 expression.
All values are expressed as means ± SEM. The significance between mean values was evaluated by a two-tailed unpaired Student t test.
Hepatic SMILE Gene Expression Is Elevated by Refeeding in Normal Mice but Not in Insulin-Resistant Mice
To investigate whether SMILE plays a role in glucose homeostasis, we monitored the effect of different nutritional conditions on hepatic SMILE gene expression in mouse liver. Ad libitum feeding increased SMILE expression in liver compared with 6-h-fasted mice (Fig. 1A). For further confirmation of the induction of SMILE expression in the refed condition, protein levels of SMILE were measured in the liver from fasted and refed mice. As expected, hepatic SMILE protein levels were increased by refeeding in a time-dependent manner, while PGC-1α protein levels were decreased (Fig. 1B). These results indicated that SMILE expression may be elevated by a signal induced by feeding. On the basis of SMILE gene expression in different nutritional conditions, we investigated whether insulin signaling is responsible for the modulation of hepatic SMILE gene expression in vivo. To confirm this, we examined whether SMILE expression is also increased by refeeding in insulin-resistant mouse models. qPCR analysis showed that the expression of hepatic SMILE was significantly higher in the refed conditions in normal mice but not in db/db mice (Fig. 1C). A similar pattern of SMILE protein levels was observed in response to fasting and refeeding conditions in normal versus db/db mice (Fig. 1D). Furthermore, SMILE mRNA (Fig. 1E) and protein levels (Fig. 1F) were decreased in the livers of HFD mice compared with those of normal chow diet mice. Collectively, these results demonstrate that hepatic expression of SMILE was tightly regulated by the nutritional condition in normal mice but not in insulin-resistant mouse models.
Insulin Receptor/PKB Signaling Pathway Regulates SMILE Gene Expression During Refeeding
On the basis of different SMILE gene expression in normal mice and insulin-resistant mouse models, insulin signaling would be expected to modulate hepatic SMILE gene expression in vivo. To examine this hypothesis, we injected wild-type mice with insulin and measured hepatic SMILE gene expression levels. As expected, insulin injection led to a significant induction of hepatic SMILE mRNA and protein in wild-type mice (Fig. 2A and B). Therefore, we hypothesized that if insulin activates SMILE gene expression during refeeding condition, then liver-specific insulin receptor knockout (LIRKO) mice and protein kinase B (PKB) β-deficient (PKBβ−/−) mice should be resistant to these effects. To examine this hypothesis, we measured the expression levels of SMILE after fasting and refeeding in wild-type and LIRKO mice. LIRKO and PKBβ−/− mice exhibit elevated levels of plasma insulin, glucose, and hepatic gluconeogenic gene expression during refeeding compared with WT mice (21). Although hepatic SMILE gene expression was significantly increased by refeeding in wild-type mice, the hepatic SMILE mRNA and protein levels were not changed by fasting or refeeding in LIRKO mice (Fig. 2C and D). In contrast, the basal expression of hepatic PGC-1α was higher in LIRKO mice than in wild-type mice but not altered during fasting and refeeding. To further verify whether PKB signal pathway is involved in the induction of hepatic SMILE gene expression in vivo, we measured the expression levels of SMILE after fasting and refeeding in both PKBβ+/+ and PKBβ−/− mice. Consistent with SMILE gene expression pattern in LIRKO mice, the hepatic SMILE mRNA and protein levels were not altered by refeeding in PKBβ−/− mice compared with PKBβ+/+ mice (Fig. 2E and F). Taken together, these results demonstrate that insulin signaling via PKB is responsible for the induction of hepatic SMILE gene expression during refeeding.
Insulin Signaling Activates the SMILE Expression at the Transcriptional Level
To verify whether insulin activates SMILE expression at the transcriptional level, we performed transient transfection assay. Insulin treatment significantly activated SMILE promoter activity in a dose-dependent manner (Fig. 3A). We also found that insulin triggered induction of SMILE levels in a time-dependent manner (Fig. 3B). Consistent with the increased SMILE mRNA expression, exposure of mouse primary hepatocytes to insulin also increased the protein levels of SMILE in a time-dependent manner (Fig. 3C). Based on the ability of insulin to enhance the hepatic SMILE gene expression, we examined whether PKB increases SMILE gene expression using a constitutively active form of PKB adenovirus. Adenoviral overexpression of constitutively active PKB strongly increased SMILE mRNA and protein levels (Fig. 3D and E). Because insulin-induced protein modification such as phosphorylation often changes protein amounts by triggering their nuclear export (5–8), we examined the effects of insulin on subcellular localization of SMILE. Insulin treatment triggered increased accumulation of SMILE in the nucleus in a time-dependent manner, whereas the amounts of SMILE in the cytoplasm were not altered (Fig. 3F). Interestingly, forskolin treatment decreased the basal and insulin-stimulated SMILE gene expression (Fig. 3G). A previous report demonstrated that SREBP-1c increased SMILE gene expression in rat insulinoma cells (INS-1) (24). Consistent with that report, adenoviral overexpression of SREBP-1c dose dependently increased SMILE gene expression in mouse primary hepatocytes (Supplementary Fig. 1A). Moreover, insulin-induced SMILE expression was blocked by overexpression of dn-SREBP-1c (Supplementary Fig. 1B). Collectively, these results indicated that insulin/PKB signal pathway increases hepatic SMILE gene expression at the transcriptional level rather than at the level of protein modification.
SMILE Decreases HNF4α-Induced Gluconeogenic Gene Expression via Competing With PGC-1α
PGC-1α is induced by cAMP/cAMP-dependent protein kinase (PKA) signaling in hepatocytes and is strongly increased in the liver by fasting (25). Here, we found that hepatic PGC-1α was expressed in pattern opposite that of SMILE and was significantly elevated in both LIRKO and PKBβ−/− mice. Having seen the opposite expression pattern between SMILE and PGC-1α, we then examined the effect of SMILE on PGC-1α transcriptional activity. SMILE inhibited PGC-1α–mediated G6Pase and Pepck promoter activity (Fig. 4A). Moreover, PGC-1α–mediated G6Pase and Pepck expression and hepatic glucose production were significantly decreased by SMILE (Fig. 4B and C). A previous report demonstrated that PGC-1α activates G6Pase and Pepck gene expression by functional interaction with HNF4α (8). We have also reported that SMILE interacts with HNF4α and inhibits transcriptional activity of HNF4α (16). For exploration of whether PGC-1α–induced HNF4α transcriptional activity is repressed by SMILE, AML12 cells were cotransfected with HNF4α, PGC-1α, and SMILE expression vector together with gluconeogenic gene promoter-reporter constructs. HNF4α-mediated induction of G6Pase and PEPCK promoter activity was significantly decreased by SMILE but was recovered by coexpression of PGC-1α (Fig. 4D).
Next, to examine whether SMILE can affect the PGC-1α occupancy on the HNF4α binding region of the G6Pase and Pepck promoter in different nutritional conditions, we performed ChIP assays using fasted or refed mouse liver tissues. We observed increased occupancy of PGC-1α in the HNF4α binding region of the G6Pase and Pepck promoter in the fasted condition compared with the refed condition, whereas the occupancy of SMILE was significantly increased in the refed condition, which is completely opposite the occupancy of PGC-1α (Fig. 4E). Moreover, overexpression of SMILE significantly decreased PGC-1α occupancy in the HNF4α binding region of the G6Pase and Pepck promoter. In contrast, overexpression of PGC-1α decreased SMILE occupancy in the HNF4α binding region (Fig. 4F). These results suggest that SMILE competes with PGC-1α in forming a complex with HNF4α on G6Pase and Pepck promoter. Moreover, adenoviral overexpression of SMILE markedly inhibited HNF4α-mediated G6Pase and Pepck gene expression (Fig. 4G) or hepatic glucose production (Fig. 4H). We have also assessed the effect of SMILE on glycolytic gene expression. Overexpression of HNF4α increased the glucokinase (Gck) gene expression, whereas HNF4α-induced Gck gene expression was not affected by SMILE overexpression. Moreover, insulin-induced Gck expression was not altered by knockdown of SMILE (Supplementary Fig. 2A and B). These results indicate that SMILE specifically affects gluconeogenesis in the liver. A previous report showed that activation of FOXO1 by PGC-1α is required for gluconeogenic gene expression (20). We found that SMILE inhibits FOXO1-mediated hepatic gluconeogenic gene promoter activity or expression. Moreover, SMILE directly interacts with FOXO1 and competes with PGC-1α to inhibit FOXO1 transcriptional activity (Supplementary Fig. 3A–D). Taken together, these results demonstrate that SMILE regulates the HNF4α-mediated PEPCK and G6Pase gene expression via competing with PGC-1α and that stimulatory effect of PGC-1α on FOXO1 is also inhibited by SMILE.
Knockdown of SMILE Gene Expression Causes Hyperglycemia in Mice
The action of insulin in suppression of gluconeogenesis occurs rapidly via phosphorylation or dephosphorylation of its target molecules, whereas SMILE is regulated by insulin through gene transcription level. We assessed the effect of insulin-mediated SMILE gene expression on hepatic gluconeogenesis in early or delayed response to insulin. In the early response to insulin, knockdown of SMILE did not show any significant effect on insulin-mediated repression, whereas insulin-mediated repression of hepatic glucose output (Fig. 5A) and gluconeogenic gene expression (Fig. 5B) were significantly relieved by knockdown of SMILE in the delayed time. These results indicate that prolonged inhibition of hepatic gluconeogenesis by insulin occurs via induction of SMILE expression at the later time point. Next, we hypothesized that alteration in SMILE gene expression would affect hepatic glucose metabolism in vivo. To explore the action of hepatic SMILE in vivo, we injected mice with an adenovirus expressing (Ad-)shRNA against SMILE. As expected, knockdown of endogenous SMILE in the liver elicited a marked increase of fasting blood glucose levels in mice (Fig. 5C). However, plasma insulin levels were unaltered in Ad-shSMILE–injected mice compared with Ad-unspecific control–injected mice (Fig. 5D). Moreover, we confirmed a marked induction of hepatic gluconeogenic gene expression such as Pepck and G6Pase upon SMILE knockdown (Fig. 5E). Next, we further investigated whether ablation of endogenous SMILE affect glucose excursion, insulin sensitivity, and hepatic glucose production. During an intraperitoneal glucose tolerance test, glucose tolerance was significantly attenuated in Ad-shSMILE–injected mice compared with control mice (Fig. 5F). However, ablation of endogenous SMILE did not lead to the significant changes in insulin sensitivity as observed by insulin tolerance test in vivo (Fig. 5G). To examine the impact of ablation of endogenous SMILE on hepatic glucose production, we also performed a pyruvate tolerance test. Ablation of endogenous SMILE elevated the pyruvate-dependent increase of blood glucose levels (Fig. 5H).
Overexpression of SMILE in Normal Mice Improves Glucose Tolerance Without Markedly Affecting Fasting Blood Glucose Level
Based on the fact that ablation of SMILE increases blood glucose levels or hepatic gluconeogenesis, we suspected that induction of SMILE would have effects on hepatic glucose metabolism in normal mice. Adenovirus-mediated expression of SMILE did not elicit a significant change in overnight fasting blood glucose levels or liver glycogen contents (Fig. 6A and B). On the contrary, overexpression of SMILE significantly improved glucose tolerance and reduced gluconeogenesis as shown by improved pyruvate tolerance and reduced gluconeogenic gene expression (Fig. 6C–E). In addition, overexpression of SMILE did not cause significant induction of insulin signaling in the liver or in the muscle of normal chow-fed mice (Supplementary Fig. 4A and B). These results indicate that SMILE significantly improved glucose tolerance or pyruvate tolerance without affecting glycogen levels in normal chow diet–fed mice.
SMILE Decreases Gluconeogenic Gene Expression and Improves Hyperglycemia in Diabetic Mice Models
We next hypothesized that alteration in SMILE expression would affect hepatic glucose metabolism in diabetic mice. To further assess the functional consequences of SMILE-mediated repression in gluconeogenic gene expression and improvement of blood glucose level, we performed adenoviral overexpression of SMILE in db/db and HFD mice. The Ad-GFP– and Ad-SMILE–injected mice showed similar body weights (Fig. 7A), but overexpression of SMILE significantly reduced fasting blood glucose levels in db/db and HFD mice (Fig. 7B). Moreover, we observed a significant improvement of glucose tolerance and pyruvate tolerance without affecting insulin sensitivity in Ad-SMILE–injected db/db or HFD mice compared with Ad-GFP–injected db/db or HFD mice (Fig. 7C-E). Concomitantly, there was a significant reduction in mRNA levels of hepatic gluconeogenic gene, such as G6Pase, Pepck, and Pgc-1α, in Ad-SMILE–injected db/db or HFD mice compared with Ad-GFP–injected db/db or HFD mice (Fig. 7F and G). However, overexpression of SMILE did not lead to significant enhanced insulin signaling either in the liver or in the muscle of HFD mice (Supplementary Fig. 4C and D). Taken together, these results suggest that SMILE directly regulate hepatic glucose metabolism in diabetic mice. Therefore, activation of SMILE could be a new therapeutic approach to treat diabetes by suppressing gluconeogenesis and ameliorating hyperglycemia.
Insulin is the major hormone for the regulation of hepatic glucose metabolism by suppressing the expression of gluconeogenic enzyme genes (7,9,20). SMILE expression was increased in the refeeding condition but not in the fasting condition or insulin-resistant status, i.e., in HFD and db/db mice. Thus, it seemed that the insulin/PKB pathway plays a major role in the regulation of SMILE expression. In support of this hypothesis, we found that insulin increases hepatic expression of SMILE both in vitro and in vivo. However, the identity of the transcription factor that mediates activation of SMILE gene expression by insulin signaling is still elusive and requires further exploration.
One potential candidate transcription factor for mediating insulin signaling is SREBP-1c, a well-known factor in the transcriptional activation of lipogenesis. Indeed, SREBP-1c increased SMILE gene expression in rat insulinoma cells (INS-1) (24). In primary hepatocytes, adenoviral overexpression of SREBP-1c increased SMILE gene expression and insulin-induced SMILE expression was blocked by dn-SREBP-1c. These results suggest that insulin-mediated induction of SREBP-1c expression could increase SMILE gene expression in hepatocytes. Meanwhile, our previous study suggested that AMPK-induced SMILE decreases SREBP-1c gene expression (26). We speculated that the physiological induction of SMILE expression mainly represses gluconeogenesis under feeding conditions, while supraphysiological activation of SMILE expression could also reduce SREBP-1c expression and the subsequent lipogenic program. Alternatively, SMILE might be critical in fine-tuning the levels of fatty acid biosynthesis in the liver under feeding conditions. Reduced expression of SMILE in the livers of insulin resistance models could thus be associated with increased triacylglycerol synthesis in this setting. Taken together, insulin-mediated induction of SREBP-1c increases SMILE gene expression, which could decrease SREBP-1c expression via negative feedback mechanism. Under normal physiological conditions, SREBP-1c–mediated SMILE induction may inhibit the SREBP-1c gene expression to maintain lipid and glucose homeostasis.
While SMILE gene expression was significantly increased by refeeding in normal mice, LIRKO and PKBβ−/− mice did not display induction of SMILE gene expression during refeeding. Thus, SMILE could be the nutritional status sensing nuclear corepressor by insulin signaling that represses gluconeogenesis in the fed status. Collectively, we suggest that enhanced expression of SMILE in the fed state is another critical mechanism by which insulin regulates hepatic gluconeogenesis. We have also shown that activation of AMPK by curcumin, a polyphenol compound, increases SMILE gene expression, which leads to the SMILE-mediated repression of CREBH transcriptional activity via competing with PGC-1α (17). These data suggest that pharmacological activation of SMILE expression can be used as a new therapeutic strategy to relieve hyperglycemia. Genetically altered animal models of SMILE are also needed to fully understand the physiological and pathological role of SMILE in hepatic gluconeogenesis.
Gluconeogenic enzyme gene expression is regulated by both transcription factors and coactivator, such as PGC-1α (8). In this study, SMILE inhibits HNF4α- or FOXO1-mediated hepatic gluconeogenic genes promoter activity or expression via competing with PGC-1α. In addition, previous report revealed that SMILE inhibits GR, which plays an important role in regulation of gluconeogenic genes expression such as PEPCK and G6Pase (16). Thus, SMILE could function as a universal transcriptional corepressor for PGC-1α–interacting transcription factors such as HNF4α, FOXO1, and GR. A previous report showed that insulin and PKB inhibit FOXO1 and PGC-1α activity by a phosphorylation-dependent mechanism (5,20,27). In the immediate early response of insulin, gluconeogenesis is regulated by classical phosphorylation mechanism. Here, we found the role of the delayed effect of insulin on the regulation of gluconeogenesis via induction of corepressor SMILE. These results suggest that SMILE has an important role in delayed action of insulin, and the induction of SMILE is another way by which insulin counteracts the action of PGC-1α in the regulation of hepatic glucose metabolism. In this study, we used the adenoviral approach to explore the regulatory role of SMILE in glucose homeostasis. Thus, genetically altered animal models of SMILE are also needed to understand the physiological and pathological role of SMILE in hepatic gluconeogenesis.
In the fasting condition, PKA/PGC-1α pathway increases hepatic gluconeogenesis. In the fed condition, the PKB/SMILE pathway turns off the PGC-1α–induced hepatic gluconeogenesis via a competitive binding mechanism (Fig. 8). Overall, our observations provide insight into a novel mechanism for insulin-mediated regulation in hepatic glucose metabolism that can be used as a novel therapeutic approach for the treatment of hyperglycemia in diabetes.
See accompanying article, p. 14.
Acknowledgments. This paper is dedicated to the memory of the late Dr. Richard W. Hanson (Case Western Reserve University) in recognition of his many great contributions to science.
Funding. This work was supported by a National Creative Research Initiatives grant through the National Research Foundation of Korea (NRF) (20110018305) funded by the Korean government (Ministry of Science, ICT and Future Planning). S.-H.K. is supported by the NRF (grant nos. NRF-2012M3A9B6055345, NRF-2015R1A5A1009024, and NRF-2015R1A2A1A01006687) and Korea University.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. J.-M.L., W.-Y.S., H.-S.C., and S.-H.K. designed most of the experiments. J.-M.L., W.-Y.S., H.-S.H., K.-J.O., Y.-S.L., D.-K.K., S.C., and B.H.C. performed the experiments. J.-M.L. wrote the manuscript. R.A.H., C.-H.L., S.-H.K., and H.-S.C. contributed to discussion and review and editing of the manuscript. S.-H.K. and H.-S.C. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.