Biochem. J. (2014) 464, 281–289 (Printed in Great Britain)
281
doi:10.1042/BJ20141005
Hepatic serum- and glucocorticoid-regulated protein kinase 1 (SGK1) regulates insulin sensitivity in mice via extracellular-signal-regulated kinase 1/2 (ERK1/2) Hao Liu*1 , Junjie Yu*1 , Tingting Xia*, Yuzhong Xiao*, Qian Zhang*, Bin Liu*, Yajie Guo*, Jiali Deng*, Yalan Deng*, Shanghai Chen*, Aniko Naray-Fejes-Toth†, Geza Fejes-Toth† and Feifan Guo*2 *Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institute for Biological Sciences, Chinese Academy of Sciences, 294 Taiyuan Road, Shanghai 200031, China †Department of Physiology and Neurobiology, Geisel School of Medicine at Dartmouth, Lebanon, NH 03756-0001, U.S.A.
Insulin resistance is a major hallmark of metabolic syndromes, including Type 2 diabetes. Although numerous functions of SGK1 (serum- and glucocorticoid-regulated kinase 1) have been identified, a direct effect of SGK1 on insulin sensitivity has not been previously reported. In the present study, we generated liver-specific SGK1-knockout mice and found that these mice developed glucose intolerance and insulin resistance. We also found that insulin signalling is enhanced or impaired in Hep1-6 cells infected with adenoviruses expressing SGK1 (Ad-SGK1) or shRNA directed against the coding region of SGK1 (Ad-shSGK1) respectively. In addition, we determined that SGK1 inhibits ERK1/2 (extracellular-signal-regulated kinase 1/2) activity in
liver and Ad-shERK1/2-mediated inhibition of ERK1/2 reverses the attenuated insulin sensitivity in Ad-shSGK1 mice. Finally, we found that SGK1 functions are compromised under insulinresistant conditions and overexpression of SGK1 by Ad-SGK1 significantly ameliorates insulin resistance in both glucosaminetreated HepG2 cells and livers of db/db mice, a genetic model of insulin resistance.
INTRODUCTION
regulated kinase 1/2) [20] signalling pathway, which exerts significant effects on insulin signalling [21]. For these reasons, we hypothesized that SGK1 expressed in liver might play an important role in regulating insulin sensitivity. The aim of the present study was to investigate this possibility and elucidate underlying mechanisms.
SGK1 (serum- and glucocorticoid-regulated kinase 1) belongs to the family of serine/threonine kinases and is ubiquitously expressed in body tissues and organs, including liver, lung and brain [1]. The Sgk1 coding region was originally isolated from rat mammary tumour cells, where it functions as an immediate early gene that is transcriptionally stimulated by serum and glucocorticoids [2–4] via activation of GR (glucocorticoid receptor), RXR (retinoid X receptor), PPARγ (peroxisomeproliferator-activated receptor γ ) and NF-κB (nuclear factor κB) [5]. Various functions of SGK1 have been identified [5–13]. Previous studies have also shown that SGK1 is involved in the control of blood glucose levels, a process that is maintained by optimal glucose metabolism and insulin function. For example, SGK1 up-regulates the activity of GLUT1 (glucose transporter 1) by trafficking the transporter to the plasma membrane [14]. Deletion of SGK1 in mice results in glucose intolerance [15], and SGK1 mediates glucocorticoid inhibition of insulin secretion [16]. It is unclear, however, whether SGK1 also plays a role in the regulation of insulin sensitivity, another important aspect in determining blood glucose levels [17]. An epidemiological survey showed that SGK variability affects diabetes risk in European populations [18,19], and SGK1 is reported to negatively regulate the ERK1/2 (extracellular-signal-
Key words: db/db mice, extracellular-signal-regulated kinase 1/2 (ERK1/2), glucose, hepatocyte, insulin signalling, liver, serumand glucocorticoid-regulated kinase 1 (SGK1).
EXPERIMENTAL Animals and treatment
Leptin receptor-mutated (db/db) mice were obtained from Shanghai Laboratory Animal Co. Male C57BL/6J mice and albumin promoter-driven Cre transgenic mice (Alb-Cre mice) were purchased from Nanjing University (MARC, Nanjing, China). Mice with a floxed SGK1 allele (SGK1lox/lox ) were generated [22] and provided by Dr Geza Fejes-Toth and Dr Aniko Naray-Fejes-Toth (Dartmouth Medical School, Hanover, NH, U.S.A.). SGK1 floxed mice were crossed with Alb-Cre mice to generate AlbCre-SGK1lox/lox [SGK1 LKO (liver-specific knockout)] mice, all maintained in a C57BL/6J background. The sequences of primers used for checking deletion efficiency in LKO mice have been described previously [22]. Mice were maintained on a 12 h light/dark cycle at 25 ◦ C and provided free access to commercial rodent chow and tap water prior to initiation of the experiments. Examination of glucose and
Abbreviations: ERK1/2, extracellular-signal-regulated kinase1/2; G6Pase, glucose 6-phosphatase; GSK3β, glycogen synthase kinase 3β; GTT, glucose tolerance test; HOMA-IR, homoeostasis model assessment of insulin resistance; IR, insulin receptor; IRS1, insulin receptor substrate 1; ITT, insulin tolerance test; LKO, liver-specific knockout; NDRG1, N-Myc down-regulated gene 1; pfu, plaque-forming units; RT, reverse transcription; SGK1, serumand glucocorticoid-regulated kinase 1. 1 These authors contributed equally to this study. 2 To whom correspondence should be addressed (email
[email protected]). c The Authors Journal compilation c 2014 Biochemical Society
282
H. Liu and others
insulin-related parameters were conducted as described in the Supplementary material. These experiments were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee of Shanghai Institute for Nutritional Sciences, Chinese Academy of Sciences. Generation and administration of recombinant adenoviruses
Recombinant adenoviruses expressing mouse SGK1 were generated using the AdEasyTM Adenoviral Vector System (Qbiogene) according to the manufacturer’s instructions. The cDNA of mouse SGK1 was obtained by RT (reverse transcription)–PCR from a liver sample. Adenoviruses expressing scrambled (Ad-scrambled) or shRNA directed against the coding region of ERK1/2 (Ad-ERK1/2) and Ad-shSGK1 were generated using the BLOCK-iTTM Adenoviral RNAi Expression System (Invitrogen) according to manufacturer’s instructions. The scrambled sequence was 5 -TTCTCCGAACGTGTCACGT-3 . The shRNA sequence for mouse SGK1 was 5 -CACCGGGTACGAAGATGGATCAAGACGAATCTTGATCCATCTTCGTACCC-3 . The shRNA sequence for mouse ERK1/2 was 5 CACCGCAATGACCACATCTGCTACTCGAAAGTAGCAGATGTGGTCATTGC-3 . High-titre stocks of amplified recombinant adenoviruses were purified as described previously [23]. Viruses were diluted in PBS and administered at a dose of 1×107 pfu (plaque-forming units)/well in 12-well plates (detailed information of cell culture and treatment are described in the Supplementary material) or via tail vein injection using 5×108 pfu/mice. RT–PCR and Western blot analysis
Sgk1 mRNA levels were examined by RT–PCR as described previously [24]. The sequences of primers used for RT– PCR were: sense primer 5 -CTGCTCGAAGCACCCTTACC-3 and antisense primer 5 -TCCTGAGGATGGGACATTTTCA-3 . Western blot analysis was conducted as described in the Supplementary material. Statistics
All data are expressed as means + − S.E.M. Significant differences were assessed either by two-tailed Student’s t test or one-way ANOVA followed by the Student–Newman–Keuls test. P < 0.05 was considered statistically significant. RESULTS SGK1 LKO mice show glucose intolerance and insulin resistance
To evaluate the role of hepatic SGK1 in insulin sensitivity, we generated LKO mice. As described previously [22,25], we found that Sgk1 mRNA was almost completely absent in the liver of male LKO mice, but not other tissues examined (Figure 1A). mRNA levels of other isoforms of SGK were not affected (Supplementary Figure S1). Food intake, body weight and tissue mass (fat and liver) in LKO mice were similar to that in floxed mice (Supplementary Figure S2). Although the blood glucose and serum insulin of fed mice were not changed, LKO mice surprisingly demonstrated a significant elevation in fasting blood glucose and an increased tendency in fasting serum insulin compared with floxed mice (Figures 1B and 1C). As a result, LKO mice showed an increased HOMA-IR (homoeostasis model assessment of insulin resistance) index [26] (Figure 1D). In LKO c The Authors Journal compilation c 2014 Biochemical Society
mice, GTT (glucose tolerance test) and ITT (insulin tolerance test) demonstrated impaired glucose clearance and insulin sensitivity (Figure 1E). On the basis of these results, we next examined the effects of SGK1 deletion on phosphorylation of the IR (insulin receptor) (Tyr1150 /Tyr1151 ), IRS1 (IR substrate 1) (Tyr608 ), Akt (Ser473 ) and GSK3β (glycogen synthase kinase 3β) (Ser9 ), which are major components of insulin signalling [27] following insulin stimulation. Although phosphorylation of IR remained constant, phosphorylation of IRS1, Akt and GSK3β were significantly attenuated in the liver of LKO mice, following insulin stimulation (Figure 1F). In contrast, insulin signalling was not altered in skeletal muscle and adipose tissue of LKO mice (Supplementary Figure S3). No significant differences were observed in food intake, body weight, blood glucose and insulin levels, HOMA-IR, GTT or ITT between Alb-Cre and floxed mice (Supplementary Figure S4). We also examined the expression of genes related to glucose metabolism in LKO mice and found that only mRNA levels of G6Pase (glucose 6-phosphatase), a gluconeogenesis-related gene [27], were significantly induced in the liver of LKO mice compared with control mice (Supplementary Figure S5A). Consistent with these results, knockdown of SGK1 expression via adenovirus-expressing shRNA directed against the coding region of SGK1 (Ad-shSGK1) or overexpressing SGK1 via adenovirusexpressing SGK1 (Ad-SGK1) increased or decreased expression of G6Pase and glucose production respectively, in mouse primary hepatocytes compared with control cells (Supplementary Figures S5B and S5C). Insulin signalling is inhibited by SGK1 knockdown or activated by SGK1 overexpression in vitro
The contribution of SGK1 to insulin signalling in liver was examined by comparing Hep1-6 cells infected with Ad-shSGK1 versus cells infected with scrambled adenovirus (Ad-scrambled). Consistent with our in vivo observations (Figure 1F), insulinstimulated phosphorylation of IRS1, Akt and GSK3β were inhibited in Ad-shSGK1 cells when endogenous SGK1 protein levels were reduced compared with control cells (Figure 2A). Opposite effects were observed in SGK1-overexpressing Hep1-6 cells infected with Ad-SGK1 compared with control cells infected with control GFP (Ad-GFP) (Figure 2B). The in vivo effect of SGK1 on insulin sensitivity was further investigated in male C57BL/6J wild-type (WT) mice via adenovirus-mediated knockdown or overexpression of SGK1. Aligning well with the results of LKO mice, we found that knockdown of hepatic SGK1 by Ad-shSGK1 impaired insulin sensitivity, whereas overexpression of SGK1 by Ad-SGK1 improved insulin sensitivity in these mice compared with control mice (Supplementary Figures S6 and S7). Food intake and body weight were not affected by either Ad-shSGK1 or Ad-SGK1 (Supplementary Figure S8). ERK1/2 inhibition mediates the promoting effects of SGK1 on insulin signalling in vitro
Previous studies have proposed SGK1 as a negative regulator of the Raf/MEK (mitogen-activated protein kinase/ERK kinase)/ERK1/2 cascade [20], and ERK1/2 signalling has been implicated in the regulation of insulin sensitivity [21], suggesting that inhibition of ERK1/2 may mediate the promoting effects of SGK1 on insulin sensitivity. We found that ERK1/2 phosphorylation was enhanced or suppressed in Hep1-6 cells
Hepatic SGK1 regulates insulin sensitivity
Figure 1
283
LKO mice show glucose intolerance and insulin resistance
PCR products of Sgk1 mRNA in liver, skeletal muscle (muscle), white adipose tissue (WAT) (A), fed or fasting blood glucose (B) and serum insulin (C) levels, HOMA-IR index (D), GTTs and ITTs (E) and insulin signalling in liver before (- Ins) and after ( + Ins) insulin stimulation (2 units/kg) for 3 min (F) were examined in 3-month-old male floxed (SGK1lox/lox ) or LKO (Alb/SGK1lox/lox ) mice. Data were obtained with mice described above (n = 14 mice per group) and are presented as means + − S.E.M. Statistical significance was calculated using the two-tailed Student’s t test for the effects of the LKO compared with the floxed group (*P < 0.05). (F) p-IR (Tyr1150 /Tyr1151 ), p-IRS1 (Tyr608 ), p-Akt (Ser473 ), p-GSK3β (Ser9 ) [left-hand side, Western blot; right-hand side, quantitative measurements of p-IR, p-IRS1, p-Akt and p-GSK3β against their total (t) protein or actin].
by knocking down SGK1 using Ad-shSGK1 or overexpressing SGK1 using Ad-SGK1 respectively (Figures 3A and 3B). Similar results were obtained in vivo (Figures 3A and 3B). Furthermore, knocking down ERK1/2 using adenoviruses expressing shRNA specific for mouse ERK1/2 (Ad-shERK1/2) alone enhanced and also significantly reversed the suppressive effects of AdshSGK1 on insulin-stimulated phosphorylation of IRS1, Akt and GSK3β in Hep1-6 cells (Figure 3C). Phosphorylation of IR, however, was not affected by any of the treatments (Figure 3C).
Injection of Ad-shERK1/2 significantly lowered both fed and fasting glucose levels (Figure 4B); however, it had no effect on Ad-shSGK1-dependent increases in serum insulin levels (Figure 4C). The HOMA-IR index was decreased by AdshERK1/2 in Ad-shSGK1-treated mice (Figure 4D). Moreover, glucose clearance and insulin sensitivity, which were attenuated by knocking down SGK1, were also largely reversed by AdshERK1/2 (Figure 4E). Consistent with previous results [28–30], knock down ERK1/2 alone improved insulin sensitivity in mice (Figure 5).
Ad-shERK1/2 reverses attenuated insulin resistance in Ad-shSGK1 mice
Ad-SGK1 reverses insulin resistance in vitro and in vivo
To gain insights into the importance of ERK1/2 in mediating SGK1 knockdown induced insulin resistance in vivo, we injected Ad-shSGK1 mice with Ad-shERK1/2 or Ad-scrambled and examined whether knocking down ERK1/2 could ameliorate decreased insulin sensitivity in these mice. Functional validation of Ad-shSGK1 and Ad-shERK1/2 was demonstrated by the observed reduction in hepatic protein levels of SGK1 or ERK1/2 respectively compared with control mice (Figure 4A).
On the basis of the above results, we investigated whether SGK1 is involved in the regulation of insulin sensitivity under insulinresistant conditions. We examined expression levels of SGK1 in a cell culture model for insulin-resistance: HepG2 cells pre-treated with 18 mM glucosamine for 18 h, as described previously [31]. SGK1 functions were compromised significantly in these cells, as shown by the decreased levels of total and phosphorylated SGK1, as well as decreased phosphorylation of NDRG1 (N-Myc down-regulated gene 1), a specific downstream target of SGK1 c The Authors Journal compilation c 2014 Biochemical Society
284
Figure 2
H. Liu and others
Insulin signalling is inhibited by SGK1 knockdown or activated by SGK1 overexpression in vitro
(A and B) Hep1-6 cells were infected with Ad-shSGK1 ( + Ad-shSGK1) or Ad-scrambled (- Ad-shSGK1) for 72 h, or infected with Ad-SGK1 ( + Ad-SGK1) or Ad-GFP (- Ad-SGK1) for 48 h, followed with ( + Ins) or without (- Ins) insulin stimulation (10 nM) for 5 min. Results were obtained with at least three independent in vitro experiments and are presented as means + − S.E.M. Statistical significance was calculated using the two-tailed Student’s t test for the effects of Ad-SGK1 or Ad-shSGK1 compared with the control group (*P < 0.05). (A and B) p-IR (Tyr1150 /Tyr1151 ), p-IRS1 (Tyr608 ), p-Akt (Ser473 ), p-GSK3β (Ser9 ) and SGK1 protein [top, Western blot; bottom, quantitative measurements of p-IR, p-IRS1, p-Akt, p-GSK3β and SGK1 protein relative to their total (t) protein or actin].
[32], compared with control cells (Figure 6A). As predicted, overexpression of SGK1 partly reversed glucosamine-attenuated phosphorylation of Akt and GSK3β (Figure 6B). To validate further our hypothesis, we next examined the hepatic SGK1 levels in genetically diabetic db/db mice [33] and found that total SGK1 protein levels were higher in these mice compared with control mice (Figure 7A), possibly due to excessive circulating glucocorticoid [34]. Surprisingly, phosphorylated SGK1 and NDRG1 were significantly decreased in the livers of db/db mice compared with wild-type mice (Figure 7A), indicating a functional impairment of hepatic SGK1 in db/db mice. Overexpression of SGK1 by Ad-SGK1 significantly decreased the levels of fed and fasting blood glucose and serum insulin, reduced HOMA-IR and ameliorated glucose intolerance and insulin resistance in db/db mice compared with Ad-GFP injection (Figures 7B–7E). Consistent with a previous report [35], blood glucose levels were increased 15 min postinsulin injection in db/db mice during ITT. Possible reasons for this increase include the significantly impaired insulin sensitivity in db/db mice, which prolongs the time for insulin to decrease c The Authors Journal compilation c 2014 Biochemical Society
glucose levels, or the possibility that these mice are more susceptible to stress, such as insulin injection itself.
DISCUSSION
SGK1 is present in all metabolic tissues, including liver, skeletal muscle and adipose tissue [1]. Most of the information about SGK1 has been obtained using either cell lines or SGK1 global knockout mice [5]. Recent studies show that activation of SGK1 promotes T-helper cell differentiation [25,36,37], demonstrating a cell-type-specific effect of SGK1. The tissue-specific effect of SGK1, however, remains largely unknown. In the present study, we used LKO mice, as well as adenovirus-mediated overexpression or knockdown of SGK1 in hepatocyte cell lines or in mice, to demonstrate for the first time that hepatic SGK1 has beneficial effects on improving insulin sensitivity. This conclusion is supported by a previous study in mice showing that activation of global SGK1 with deoxycorticosterone acetate (a mineralocorticoid) increases glucose tolerance that had been
Hepatic SGK1 regulates insulin sensitivity
Figure 3
285
ERK1/2 inhibition mediates the promoting effects of SGK1 on insulin signalling in vitro
(A) Hep1-6 cells were infected with Ad-shSGK1 ( + Ad-shSGK1) or Ad-scrambled (- Ad-shSGK1) for 72 h or male C57BL/6J mice were infected with Ad-shSGK1 ( + Ad-shSGK1) or Ad-scrambled (- Ad-shSGK1) via tail vein injection, followed by examination of p-ERK1/2 in liver at day 11. (B) Hep1-6 cells were infected with Ad-SGK1 ( + Ad-SGK1) or Ad-GFP (- Ad-SGK1) for 48 h or male C57BL/6J mice were infected with Ad-SGK1 ( + Ad-SGK1) or Ad-GFP (- Ad-SGK1) via tail-vein injection, followed by examination of p-ERK1/2 in liver at day 12. (C) Hep1–6 cells were infected with Ad-shERK1/2 ( + Ad-shERK1/2) or Ad-scrambled (- Ad-shERK1/2) before being infected with Ad-SGK1 ( + Ad-SGK1) or Ad-GFP (- Ad-SGK1), followed with insulin stimulation (10 nM) for 5 min. Data were obtained with mice described above (n = 10–14 mice per group) or at least three independent in vitro experiments and are presented as means + − S.E.M. Statistical significance was calculated using the two-tailed Student’s t test for the effects of the Ad-SGK1 or Ad-shSGK1 compared with the control group (*P < 0.05) in (A) and (B); or using one-way ANOVA followed by the Student–Newman–Keuls (SNK) test for the effects of any group compared with the control group without Ad-SGK1 and Ad-shERK1/2 (*P
