Chronic HMGCR/HMG-CoA reductase inhibitor treatment contributes to dysglycemia by upregulating hepatic gluconeogenesis through autophagy induction

Statins increase gluconeogenic enzyme expression in HepG2 cells

To evaluate whether hepatic gluconeogenesis is involved in the diabetogenic effects of statins, we tested the effects of rosuvastatin, fluvastatin, pravastatin, and atorvastatin on the expression of key enzymes involved in gluconeogenesis and glycolysis in HepG2 hepatocellular carcinoma cells (Fig. 1A). To validate the system, we treated HepG2 cells with insulin to confirm that insulin promotes glycolysis in these cells. Results of real-time quantitative reverse transcription-polymerase chain reaction (qRT-PCR) confirmed that insulin decreased expression of genes encoding the gluconeogenic enzymes G6PC (glucose-6-phosphatase, catalytic subunit) and PCK1 (phosphoenolpyruvate carboxykinase 1 [soluble]) and increased expression of genes encoding the glycolytic enzymes GCK (glucokinase [hexokinase 4]) and PKLR (pyruvate kinase, liver and RBC) (Fig. 1B). After 24-h treatment with rosuvastatin, fluvastatin, pravastatin, or atorvastatin (20 µM), mRNA levels of G6PC and PCK1 increased in response to each statin (Fig. 1C and Fig. 1D) and protein levels of these enzymes also increased (Fig. 1E), suggesting increased gluconeogenesis. In contrast, statins had little or no effect on mRNA levels of GCK and PKLR (Fig. 1F and Fig. 1G). These data indicate that statins specifically affect the expression of gluconeogenic enzymes in HepG2 cells.

Statins induce autophagic flux leading to enhanced expression of gluconeogenic enzymes

Because autophagy promotes gluconeogenesis in the liver,¹⁷ we tested whether statins increase autophagy in hepatocytes by transfecting HepG2 cells with a vector expressing the autophagy marker MAP1LC3A/LC3A (microtubule-associated protein 1 light chain 3 α) fused to green fluorescent protein (GFP). We found that statin treatment increased the number of GFP-LC3A fluorescent puncta representing autophagosomes in the cytosol (Fig. 2A and Fig. 2B). In contrast, puncta were barely discernible in control cells (Fig. 2A and Fig. 2B). To confirm this result, we evaluated LC3B-II expression by western blot analysis, which showed increased LC3B-II levels in statin-treated cells compared with controls, suggesting that statins promote autophagy in HepG2 cells (Fig. 2C). We treated cells with bafilomycin A₁ (Baf A1), a lysosomal blocker, to block autophagy and treated with statins to reveal whether increased LC3B with statin is due to autophagy induction or autophagy flow blockade. Statins additionally increased LC3B-II in cells pretreated with Baf A1, which suggests that statins induce autophagy rather than blocking autophagy flow (Fig. 2C). Then we transfected mRFP (monomeric red fluorescent protein)-GFP-LC3 tandem construct encoding LC3 fused to mRFP and GFP to HepG2 cells to evaluate autophagic flux.

The GFP protein is degraded in acidic conditions inside the lysosome, leading to loss of the green fluorescent signal whereas RFP is more stable in acidic condition, maintaining the red fluorescent signal. Therefore autophagosomes show the yellow fluorescent signal (merged signal of mRFP and GFP) and autolysosomes show only red signals (mRFP). The number of red and yellow puncta increased in HepG2 cells treated with statin, indicating that stains indeed induce autophagosome and autolysosome formation, representing an increase in autophagic flux (Fig. 2D and Fig. 2E). The possibility that the statins' effect is due to mere inhibition of lysosomal degradation is ruled out based on the result that Baf A1 increased number of yellow puncta without an increase in red puncta (Fig. 2D and Fig. 2E).

Blockage of autophagic flux attenuates the effect of statins on gluconeogenesis.

To see whether the gluconeogenic effect of statins is mediated by the autophagic process, we utilized a lysosomal inhibitor such as chloroquine (CQ) or Baf A1, with a statin treatment. Increased G6PC and PCK1 expression with statin treatment were attenuated by CQ (Fig. 3A and Fig. 3B) and glucose production was also decreased (Fig. 3C). Inhibition of the autophagic process with Baf A1 treatment resulted in attenuation of increased G6PC and PCK1 expression with statins (Fig. 3D and Fig. 3E). In addition glucose production was also decreased with Baf A1 (Fig. 3F).

To better understand the effect of statins on autophagy, we transfected HepG2 cells with short hairpin RNAs (shRNAs) against the gene encoding BECN1, which plays an important role in autophagy induction.²² After confirming the knockdown of BECN1 (Fig. 3G), we observed statin-dependent increases in G6PC and PCK1 mRNA levels in transfected cells (Fig. 3H and Fig. 3I). These data support the role of autophagy in statin-induced gluconeogenesis.

ATG7 is necessary for statin-induced gluconeogenesis in the liver

To further evaluate the role of autophagy in regulating the expression of gluconeogenic enzymes, ex vivo studies were performed with primary hepatocytes isolated from liver-specific Atg7 (autophagy-related 7)-deficient mice. Deficiency of Atg7 in the liver was confirmed by qRT-PCR (Fig. 4A), the cultured hepatocytes were treated with statins for 24 h. Our results showed that the statin-induced increase in expression of the genes encoding gluconeogenic enzymes (G6pc and Pck1) was blocked in Atg7-deficient hepatocytes (Fig. 4B). In addition, glucose output did not increase in response to statins in Atg7-deficient hepatocytes (Fig. 4C). Collectively, these data suggest that ATG7, the core autophagy regulator, is required for statin-induced gluconeogenesis in the liver, supporting the role of autophagy in statin-induced hepatic gluconeogenesis.

Statins increase hepatic gluconeogenesis, leading to hepatic insulin resistance in high-fat diet-fed mice

To evaluate the in vivo effects of statins on glucose homeostasis, beginning at 5 wk of age mice were fed a high-fat diet supplemented with rosuvastatin, fluvastatin, pravastatin, or atorvastatin for 16 wk. Mean body weight significantly increased in statin-treated mice compared with untreated mice (Fig. 5A), independent of food intake (Fig. 5B). We measured fasting blood glucose levels and performed the oral glucose tolerance test and insulin tolerance test at wk 15 and pyruvate tolerance test at wk 16. Our results showed that fasting blood glucose levels were higher in mice treated with pravastatin or atorvastatin compared with mice treated with rosuvastatin or fluvastatin (Fig. 5C). Results of the oral glucose tolerance test did not differ between pravastatin- and atorvastatin-treated mice and control mice (Fig. 5D), indicating that pancreatic β-cell function was not impaired at 15 wk. However, blood glucose levels failed to decrease upon insulin treatment in statin-treated mice, indicating insulin resistance (Fig. 5E), and area under the curve for the insulin tolerance test differed significantly between the control group and the statin-treated groups (Fig. S1). To evaluate the possibility that statins induce hepatic insulin resistance by increasing hepatic glucose production, we performed the pyruvate tolerance test, which showed significantly elevated blood glucose levels in pravastatin- and atorvastatin-treated mice over 3 h (Fig. 5F and Fig. S1: area under the curve). These data suggest that pravastatin and atorvastatin increase blood glucose levels in vivo, at least in part, by stimulating hepatic gluconeogenesis.

To determine whether statins function as HMGCR inhibitors under our experimental conditions, we measured serum levels of cholesterol, triglycerides, and free fatty acids in the high-fat diet-fed mice. Serum cholesterol levels were only marginally decreased by rosuvastatin and fluvastatin (Fig. 5G); however, serum triglyceride and free fatty acid levels were significantly decreased in the statin-treated groups compared with the control group (Fig. 5H and Fig. 5I). These data suggest that under conditions in which statins increase hepatic gluconeogenesis, they function as HMGCR inhibitors within hepatocytes, and these statin-induced metabolic changes may be related to the inhibition of endogenous cholesterol synthesis.

Statins increase hepatic gluconeogenesis and autophagy in high-fat diet-fed mice

To determine whether statins increase hepatic gluconeogenesis and autophagy in vivo, expression of key gluconeogenic enzymes in the livers of high-fat diet-fed mice was evaluated by qRT-PCR. Consistent with in vitro results, statin treatment caused a significant increase in the expression of hepatic gluconeogenic genes (G6pc and Pck1) in mice (Fig. 6A and Fig. 6B). However, expression of glycolytic genes (Gck and Pklr) was not affected by statins (Fig. 6C and Fig. 6D).

To determine whether statins induce autophagy in mouse livers in vivo, electron microscopy analysis was performed. Transmission electron microscopy analysis revealed prominent vacuolization and autophagosomes in the hepatocytes of statin-treated mice (Fig. 7A). Autophagic vacuoles are increased in statin treated mouse livers (Fig. 7B). Collectively, these data demonstrate that statin treatment leads to insulin resistance by increasing gluconeogenesis, which is tightly coupled to autophagy.

Cell culture and drug treatments

The primary hepatocytes and hepatocellular carcinoma HepG2 cell lines were cultured in Dulbecco's modified Eagle's medium (Thermo Scientific, SH30243.01) containing 10% fetal bovine serum (Thermo Scientific, SH30071.03), 100 U/ml penicillin, and 100 µg/ml streptomycin (Thermo scientific, SV30010) in a 5% CO₂ incubator at 37°C. The statin drugs rosuvastatin (Sigma-Aldrich, SML1264), fluvastatin (Sigma-Aldrich, Y0001090), pravastatin (Sigma-Aldrich, P4498), and atorvastatin (Sigma-Aldrich, PZ0001) were dissolved in dimethyl sulfoxide before dilution in the culture medium. In all experiments the final statin concentration was 20 µM, and final dimethyl sulfoxide concentration was ≤ 0.1%. Chloroquine (Sigma-Aldrich, C6628) and bafilomycin A₁ (Sigma-Aldrich, B1793) were dissolved in distilled water before treatments. The final concentration was 50 uM and 20 nM each, respectively.

Analysis of autophagy by confocal microscopy

HepG2 cells were transfected with the expression vector GFP-LC3 and mRFP-GFP-LC3 using Lipofectamine 2000 (Invitrogen, 11668–027) for 48 h. The cells were then treated with statins for 24 h and fixed in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 10 min. After fixation, HepG2 cells were washed in phosphate-buffered saline (PBS; Amresco, E404–200TABS) 3 times for 5 min and then observed using LSM 700 and LSM780 confocal microscopes (Zeiss, Gottingen, Germany).

RNA extraction and qRT-PCR

Total RNA was extracted from HepG2 cells using the RNeasy Mini Kit (Qiagen, 74104) and from primary hepatocytes using TRIzol (Invitrogen, 15596–018). Reverse transcription was carried out with 2 μg total RNA using the QuantiTect Reverse Transcription kit (Qiagen, 205311) according to the manufacturer's instruction. Expression of target genes G6PC, PCK1, GCK, and PKLR was analyzed by qPCR using SYBR Premix Ex Taq (Clontech, RR420A) and gene-specific primers designed from sequences submitted to the NCBI nucleotide sequence database. Amplification was carried out using the Takara Thermal Cycler Dice® Real-Time system (Otsu, Shiga, Japan) and the following cycling conditions: 40 cycles of 95°C for 5 sec, 58°C for 10 sec, and 72°C for 20 sec. All reactions were performed in triplicate, and target gene expression was normalized to that of the internal control glyceraldehyde 3-phosphate dehydrogenase.

Immunoblotting

Cells were lysed in buffer consisting of 50 mM Tris-HCl, pH 8.0, 5 mM EDTA, 150 mM NaCl, 0.5% sodium deoxycholate (Sigma-Aldrich, D6750), 1% Nonidet P-40 (Sigma-Aldrich, 74385), 0.1% sodium dodecyl sulfate (Sigma-Aldrich, L3771), 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium fluoride, 1 mM sodium orthovanadate, and protease inhibitor cocktail (Roche, 11 836 153 001). Equivalent amounts of each protein extract were separated on 10% polyacrylamide gels and electrophoretically transferred onto polyvinylidene fluoride membrane (Millipore, IPVH00010). After blocking, the membranes were incubated with primary antibodies against PCK1 (Santa Cruz Biotechnology, sc-32879), G6PC (Santa Cruz Biotechnology, sc-25840), FOXO1 (Cell Signaling Technology, 2880), phospho-FOXO1 (S256; Cell Signaling Technology, 9461), AKT (Cell Signaling Technology, 9272), phospho-AKT (S473; Cell Signaling Technology, #9271), LC3B (Sigma-Aldrich, L7543), ACTB (Sigma-Aldrich, A1978), BECN1 (Santa Cruz Biotechnology, sc-11427), MTOR (Santa Cruz Biotechnology, sc-8319), phospho-MTOR (S2448; Santa Cruz Biotechnology, sc-101738), RPS6KB (Santa Cruz Biotechnology, sc-230), and phospho-RPS6KB (T389; Santa Cruz Biotechnology, sc-11759) followed by horseradish peroxidase-conjugated IgG (Santa Cruz Biotechnology, sc-2371) and anti-rabbit IgG (Santa Cruz Biotechnology, sc-2030). The blots were developed using an enhanced chemiluminescent detection kit.

Glucose output assay

Glucose output from HepG2 cells was quantified using a colorimetric glucose assay kit (BioVision, K686–100) according to the manufacturer's instructions. Briefly, HepG2 cells were treated with or without statin for 24 h. The conditioned medium was then collected and incubated with the reaction mix for 30 min at room temperature. Absorbance at 450 nm was measured in a 96-well plate reader (Molecular Devices, Sunnyvale, CA).

BECN1 RNA interference

Electroporation of shRNA-expressing plasmids in HepG2 cells was performed using the Neon® transfection system (Invitrogen, MPK10096) according to the manufacturer's protocol. Briefly, trypsinized HepG2 cells (1 × 10⁶ cells) were washed in PBS and then resuspended in Neon Resuspension Buffer R. The cell suspension was mixed with 2 µg shRNA against BECN1 (Santa Cruz Biotechnology, sc-29797-SH) or a scrambled shRNA sequence (Santa Cruz Biotechnology, sc-108060) as a negative control and pulsed twice at 1200 V for 50 msec. After electroporation, cells were quickly seeded into 6-well plates and grown in culture medium for further experiments. Successful inhibition of BECN1 expression was verified by western blot analysis.

Animals

Four-wk-old male C57BL/6J mice were housed under controlled conditions (21°C ± 2°C, 60% ± 10% humidity, 12-h light/12-h dark cycle) with ad libitum access to food and water. After 1 wk, the mice were divided into 5 groups according to treatment (untreated control, n = 9 ; rosuvastatin, n = 7 ; fluvastatin, n = 7 ; pravastatin, n = 11 ; atorvastatin, n = 11 ). Beginning at 5 wk of age, all mice were fed a high-fat diet that included 45% lipids (Research Diets, Inc., D12451). The food given to each treatment group was supplemented with 0.01% (w/w) of the appropriate statin. Food intake and body weight of the mice were evaluated 2 times a wk at the same time of day. Fasting blood glucose level was measured weekly in the evening after an 8-h fast. After 16 wk, the mice were anesthetized with zolazepam and tiletamine (Zoletil, 50 mg/kg; Virbac France GTIN: 03597132126045), and blood was collected by cardiac puncture. Primary hepatocytes were isolated from male liver-specific atg7-deficient mice⁴¹ and wild-type mice (9 wk old) using a previously described method.⁴⁸ Atg7f/f mice were crossed with albumin promoter-driven Cre mice to generate liver-specific atg7-deficient mice⁴¹ (atg7f/f;alb-Cre mice). The animal protocol was approved by the institutional animal care and use committee at Yonsei University College of Medicine.

Oral glucose tolerance, insulin tolerance, and pyruvate tolerance tests

To perform the oral glucose tolerance test, 40% glucose (2 g/kg body weight) was administered via oral gavage after a 6-h fast. Blood was collected from the tail vein at 0, 30, 60, 90, and 120 min after glucose administration. To assay insulin tolerance, fasting glucose was measured 4 h after fasting, and then mice were intraperitoneally injected with 0.75 U/kg human insulin-R (Sigma-Aldrich, I9278) dissolved in PBS. Blood glucose was measured at 15, 30, 60, 90, and 120 min after injection. To assay pyruvate tolerance, mice were intraperitoneally injected with 2 g/kg sodium pyruvate (Sigma-Aldrich, P2256) dissolved in PBS after an 18-h fast. Blood was collected from the tail vein before pyruvate injection (0 h) and at 15, 30, 60, 90, and 120 min after injection. Glucose levels were determined using an Accu-Chek Performa® glucometer (Boehringer-Mannheim, Indianapolis, IN).

Plasma glucose, cholesterol, triglyceride, and free fatty acid measurement

Blood was collected in microcentrifuge tubes and centrifuged to obtain serum, which was divided into aliquots and stored at -80°C for subsequent assays. Serum glucose, cholesterol, triglyceride, and free fatty acid levels were measured with the respective assay kits (BioAssay Systems, EBGL-100 for glucose, ECCH-100 for cholesterol, ETGA-200 for triglyceride and EFFA-100 for free fatty acid) according to the manufacturer's instructions.

Transmission electron microscopy

Autophagic vacuoles in the liver were visualized by transmission electron microscopy. Glutaraldehyde-fixed mouse liver tissues were postfixed in 2% osmium tetroxide, dehydrated in graded alcohol, and flat embedded in Epon 812 (Electron Microscopy Sciences, 100503–876). Ultrathin tissue sections (300 nm) were stained with uranyl acetate and lead citrate and examined with an electron microscope (JEM-1011, JEOL/MegaView III, Olympus, Tokyo, Japan).

Statistical analysis

Data are presented as mean ± standard error of the mean. Groups were compared using the Student t test or one-way analysis of variance followed by the Dunnett multiple comparison test, where appropriate; P < 0.05 was considered significant. Data analysis was carried out using Prism 5.0 software (GraphPad Software, La Jolla, CA).