Involvement of pregnane X receptor in the suppression of carboxylesterases by metformin in vivo and in vitro, mediated by the activation of AMPK and JNK signaling pathway

European Journal of Pharmaceutical Sciences 102 (2017) 14–23

Contents lists available at ScienceDirect

European Journal of Pharmaceutical Sciences journal homepage: www.elsevier.com/locate/ejps

Involvement of pregnane X receptor in the suppression of carboxylesterases by metformin in vivo and in vitro, mediated by the activation of AMPK and JNK signaling pathway Enfang Shan, Zhu Zhu, Shuangcheng He, Dongbao Chu, Dinghao Ge, Yunran Zhan, Wei Liu, Jian Yang ⁎, Jing Xiong ⁎ Department of Pharmacology, Nanjing Medical University, Nanjing, Jiangsu 210029, China

a r t i c l e

i n f o

Article history: Received 2 August 2016 Received in revised form 16 February 2017 Accepted 20 February 2017 Available online 24 February 2017 Chemical compounds studied in this article: Metformin (PubChem CID: 4091) Para-nitrophenylacetate (PubChem CID: 6931101) Clopidogrel bisulfate (PubChem CID: 115366) Irinotecan hydrochloride (PubChem CID: 74990) Compound C (PubChem CID: 1e1524144) SP600125 (PubChem CID: 8515) Rhodamine 123 (PubChem CID: 65217)

a b s t r a c t Type 2 diabetes mellitus (T2D) is a complex metabolic disorder requiring polypharmacy treatment in clinic, with metformin being widely used antihyperglycemic drug. However, the mechanisms of metformin as a perpetrator inducing potential drug-drug interactions and adverse drug reactions are scarcely known to date. Carboxylesterases (CESs) are major hydrolytic enzymes highly expressed in the liver, including mouse carboxylesterase 1d (Ces1d) and Ces1e. In the present study, experiments are designed to investigate the effects and mechanisms of metformin on Ces1d and Ces1e in vivo and in vitro. In results, metformin suppresses the expression and activity of Ces1d and Ces1e in a dose- and time-dependent manner. The decreased expression of nuclear receptor PXR and its target gene P-gp indicates the involvements of PXR in the suppressed expression of carboxylesterases by metformin. Furthermore, metformin significantly suppresses the phosphorylation of AMPK and JNK, and the suppression of carboxylesterases induced by metformin is repeatedly abolished by AMPK inhibitor Compound C and JNK inhibitor SP600125. It implies that the activation of AMPK and JNK pathways mediates the suppression of carboxylesterases by metformin. The findings deserve further elucidation including clinical trials and have a potential to make contribution for the rational medication in the treatment of T2D patients. © 2017 Elsevier B.V. All rights reserved.

Keywords: Carboxylesterases Type 2 diabetes Pregnane X receptor P-glycoprotein AMP-activated protein kinase c-Jun N-terminal kinase

1. Introduction Type 2 diabetes mellitus (T2D) is a metabolic disorder, which is characterized by chronic hyperglycemia and insulin resistance, affecting a large amount of the global population with potentially serious health outcomes (Ginter & Simko, 2012). Since polypharmacy treatment is always required for the T2D patients, biotransformation characteristics and potential drug-drug interactions in patients deserve full elucidation (Breuker et al., 2017). Metformin, a historical antihyperglycemic drug, contributes to lifespan extension, treatment and prevention of sedentariness damages, insulin resistance, and obesity for T2D patients

⁎ Correspondence authors. E-mail addresses: [email protected] (J. Yang), [email protected] (J. Xiong).

http://dx.doi.org/10.1016/j.ejps.2017.02.031 0928-0987/© 2017 Elsevier B.V. All rights reserved.

(Senesi et al., 2016). Considering the extended clinical use of metformin (Loos et al., 2017; Li et al., 2016; Castillo-Quan & Blackwell, 2016), the risks of unwanted drug-drug interactions and drug adverse reactions associated with metformin-based polypharmacy treatment for patients should be thoroughly evaluated. However, previous studies predominantly focus on how other drugs may affect the pharmacokinetics of metformin as a victim (Zack et al., 2015; El Messaoudi et al., 2016), with limited data available regarding the influence of metformin as a perpetrator on the metabolism and clearance of other co-administered drugs. In all factors that may alter the hepatic capacity of drug metabolism, regulated expression of Drug-metabolizing enzymes (DMEs) and drug transporters contributes the most. Carboxylesterases (CESs) represent an important group of enzymes that have highly abundant expression in the liver. The drugs being hydrolyzed by CESs contain such functional

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

groups as carboxylic acid ester, amide and thioester, and occupy about 20% of therapeutic agents in clinic (Xiao et al., 2012). There are two major CESs expressed in the liver, human carboxylesterase 1 (CES1) and carboxylesterase 2 (CES2) (Xiong et al., 2014a). Pregnane X receptor (PXR) is one of the most important nuclear receptors regulating the transcriptional expression of DMEs and drug transporters, including Cytochrome P450 3A4 (CYP3A4), multidrug resistance 1 (MDR1)-encoded P-glycoprotein (P-gp), as well as CESs (Luo et al., 2017; Pondugula et al., 2015). Upon activation, PXR is translocated from the cytoplasm to the nuclear compartment, heterodimerizes with the retinoid X receptor (RXR) and upregulates the transcription by binding to the response element in the promoter of target DMEs and drug transporters (Gu et al., 2006). A number of pathways including AMP-activated protein kinase (AMPK) and c-Jun N-terminal kinase (JNK) are involved in the regulation of PXR and its target genes (Kumari et al., 2015; Krausova et al., 2011). As a known activator of AMPK, metformin appears to exert its pharmacological actions both AMPK-dependently and -independently (Lee et al., 2012; Do et al., 2013). It is also reported that AMPK interacts with transcription factors apart from PXR (Sozio et al., 2011) and disturbs the expression of PXR-target genes (Wang et al., 2015). In addition, involvement of the JNK pathway in the actions of metformin has also been recently reported, as the activation of JNK pathway is involved in metformin-induced apoptosis in lung cancer cells (Wu et al., 2011). As mouse carboxylesterase 1d (Ces1d) and Ces1e strongly crossreact with human CES1 and CES2 respectively (Xiao et al., 2012), experiments are designed to investigate the effect of metformin on the expression and activity of mouse hepatic Ces1d and Ces1e both in vivo and in vitro. Based on the above information, nuclear receptor PXR, AMPK and JNK signaling pathways are hypothesized to mediate the regulation of CESs by metformin. The findings could make a great contribution to move a step in guiding clinical rational drug use for the T2D patients and deserve further investigations including clinical trials. 2. Materials and methods 2.1. Materials Metformin hydrochloride (MET) was purchased from Aladdin (Shanghai, China). Para-nitrophenylacetate (PNPA), clopidogrel bisulfate, irinotecan hydrochloride, AMPK inhibitor 6-[4-(2-piperidin-1ylethoxy) phenyl]-3-pyridin-4-ylpyrazolo [1,5-a] pyrimidine (Compound C), JNK inhibitor SP600125 and Rhodamine 123 (Rho123) were from Sigma (St. Louis, MO, USA). Dulbecco's modified Eagle's medium (DMEM) was from Invitrogen (Carlsbad, CA, USA); and fetal bovine serum (FBS) was from Hyclone Laboratories (Logan, UT, USA). Mouse Ces1d or Ces1e was detected by antibody against human CES1 or human CES2, kindly provided by Dr. Bingfang Yan (Xiong et al., 2014b). Antibody against mouse PXR was also donated by Dr. Yan. Antibodies against pAMPK, AMPK, pJNK, JNK and β-actin were from Bioworld (St. Louis Park, USA), antibody against P-gp was from Abcam (Cambridge, UK). The goat anti-rabbit or anti-mouse IgG conjugated with horseradish peroxidase was from Pierce Chemical (Pierce, Rockford, IL, USA). All other reagents were of analytical grade and commercially available. All the experiments for this study were conducted in compliance with the principles of Good Laboratory Practice (GLP) standards for Nonclinical Laboratory Studies. 2.2. Animal experiments The use of animals was approved by IACUC (Institutional Animal Care and Use Committee) of Nanjing Medical University. To demonstrate the in vivo effects of metformin on the expression of mouse Ces1d and Ces1e in the liver, male Institute of Cancer Research (ICR) mice (20–25 g) obtained from Yangzhou University (Yangzhou, China) were housed individually in animal cages in a room with

15

controlled temperature (22 ± 1 °C), humidity (50 ± 10%), and lighting (lights on from 8:00 to 20:00). After 1 week of acclimatization, metformin was dissolved in distilled water and administered to mice by oral gavage (0, 100 or 300 mg/kg/day for 3 days, or 300 mg/kg/day for 0, 1 or 3 days). The dosage for mice is converted from human dosage using body surface area normalization (Reagan-Shaw et al., 2008), and refers to previous studies (Krausova et al., 2011; Pham et al., 2017; Xu et al., 2015). 24 h after the last administration of drug or vehicle, mice were anesthetized by injection with urethane (1 g/kg body weight). The liver was perfused with PBS through the portal vein to remove blood and frozen at − 80 °C for preparing S9 fractions. Every effort was made to minimize animal suffering and to reduce the number of animals used for the experiments. 2.3. Preparation of S9 fractions The frozen livers were thawed in homogenization buffer (50 mM Tris–HCl, pH 7.4, 150 mM KCl and 2 mM EDTA) and then homogenized with six passes of Teflon pestle driven by a Wharton stirrer. The homogenates were centrifuged at 10,000 ×g for 20 min at 4 °C. The S9 fractions of liver (supernatant) were assayed for the hydrolysis of PNPA and for protein expression. 2.4. Primary mouse hepatocyte culture and treatment Mouse hepatocytes were isolated from livers of male ICR mice here referred to as modification of the two-step perfusion method, described previously (Xiong et al., 2014a). Isolated hepatocytes were suspended in the DMEM supplemented with 10% FBS, seeded into collagen-coated six-well plates and maintained at 37 °C for 4 h to allow attachment. After being continuously cultured for one day with a change of fresh medium, primary mouse hepatocytes were treated with MET (0, 0.125, 0.25 or 0.5 mM) for 24 h. 2.5. Enzyme activity assay The overall hydrolytic activity was tested using standard substrate PNPA by spectrophotometer. After the treatment with metformin (0, 0.125, 0.25, and 0.5 mM), cells were rinsed with D-Hank's and harvested in 80 μL of 100 mM potassium phosphate buffer (pH 7.4). The cell suspension was sonicated by a sonifier (Nanjing, China). Then the supernatants were obtained by precipitation using centrifugation at 12,000 ×g for 15 min at 4 °C, and assayed for overall hydrolytic activity toward PNPA as described previously (Xiong et al., 2014a). A sample tube (1 mL) contained 10 μg of cell lysates in 100 mM potassium phosphate buffer (pH 7.4), 10 μL PNPA and substrate (1 mM) at 37 °C. The addition of PNPA was used to initiate the reactions and the hydrolytic rate was recorded from an increase in absorbance at 400 nm. The extinction coefficient (E400) was determined to be 13 mM−1 cm−1. Several controls were conducted including incubation without proteins. 2.6. Cytotoxicity and morphologic assay Primary mouse hepatocytes were seeded into 96-well plates at the density of 5000 cells each well. After continued culture for one day, the medium was replaced with fresh medium with or without 0.5 mM metformin for 12 h. After that, the cells were washed with DMEM twice and treated with different concentrations of clopidogrel (0, 10, 30 or 100 μM) for 30 h or irinotecan (0, 10, 30 or 100 μM) for 24 h. The medium was replaced with fresh medium containing 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium (MTT) at a final concentration of 0.5 mg/mL. After 4 h incubation at 37 °C, the medium was gently decanted, and dimethyl sulfoxide was added to dissolve formazan product. The optical density (OD) was determined at 570 nm, and the final OD values were expressed by subtracting the

16

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

background reading (no seeded cells). Morphologic changes were detected under microscope before MTT assay. 2.7. Western blotting Cells used for phospho-lysates were lysed with lysis buffer enriched with protease and phosphatase inhibitors cocktail. Protein concentrations were determined with BCA protein assay based on the albumin standard (Pears, Rockford, IL, USA). Equal amounts of protein were separated on a 10% SDS polyacrylamide gel and transferred electrophoretically onto polyvinylidene difluoride membranes (Bio-Rad Laboratories, Hercules, CA, USA). The membranes were blocked with 5% non-fat milk in Tris-buffered saline/0.1% Tween 20 for 1 h, subsequently blotted with respective primary antibodies overnight, and then blotted with horseradish peroxidase-conjugated secondary antibody for 1 h. The protein bands were visualized with enhanced chemiluminescence detection system. Protein levels were quantified by density analysis using Image J software (NIH), and expressed as fold change of interest protein/β-actin. 2.8. Immunofluorescence analysis Cells were seeded at 2 × 105 cells/well on glasses bottom dishes. After treatment with saline or metformin (0, 0.125, 0.25, or 0.5 mM), cells were rinsed with PBS and fixed with 4% paraformaldehyde for 10 min. Permeabilization was performed in PBS with 0.3% Triton X100 for 10 min. After blocking for 2 h with 5% bovine serum albumin, the cells were incubated with anti-PXR primary antibody (1:400) at 4 °C overnight. After washing with PBS, FITC-conjugated secondary antibody (Bioworld, St. Louis Park, USA) was added for 1 h in the dark. Nuclei were stained with 4′,6′-diamidino-2-phenylindole (DAPI; Bioworld, St. Louis Park, USA), and the cells were detected through using a fluorescent microscope (Olympus, Tokyo, Japan). 2.9. Intracellular rhodamine 123 accumulation assay Efflux activity of P-gp was evaluated by intracellular accumulation assay of a well-established substrate for P-gp, Rho123, as described (Xiong et al., 2014a). After treatment with saline or metformin (0, 0.125, 0.25, and 0.5 mM), cells were rinsed with PBS and resuspended in DMEM (10% FBS) supplemented with 5 μg/mL Rho123 to allow incubation at 37 °C, in a humidified atmosphere of 5% CO2 for 1 h in the dark. Following, the cells were rinsed twice with ice-cold PBS and visualized immediately by using a fluorescent microscope (Olympus, Tokyo, Japan). 2.10. Statistical analysis The experimental results are presented as mean ± SEM from at least three independent experiments. The multiple comparisons were performed by a one-way ANOVA with Duncan's multiple range test. The differences were considered to be statistically significant when p b 0.05. 3. Results 3.1. Metformin suppresses the expression and hydrolytic activity of mouse hepatic Ces1d and Ces1e in vivo First of all, the effects of metformin on mouse Ces1d and Ces1e in vivo are determined by administering metformin in different doses (0, 100 or 300 mg/kg) once daily for three consecutive days by oral gavage. It is shown that metformin decreases the expression of Ces1d and Ces1e in a concentration-dependent manner, with lowest expression level by 300 mg/kg metformin (Fig. 1A). Then, metformin is administered to mice with 300 mg/kg for different days (0, 1 or 3 days). The results

show that metformin also decreases the expression of Ces1d and Ces1e in a time-dependent manner, with lowest expression levels after administration lasting for three days (Fig. 1B). Next, whether the decreases of Ces1d and Ces1e expression could be translated into the decreases of the hydrolytic activity is further investigated. As shown in Fig. 1C and D, comparable with the protein expression of CESs, the hepatic overall hydrolytic activity has also been decreased in a dose- and time-dependent manner. The data suggest that metformin significantly suppresses the expression and hydrolytic activity of mouse hepatic Ces1d and Ces1e in vivo. 3.2. Metformin represses the expression of mouse Ces1d and Ces1e in vitro. Following, whether metformin affects the expression of mouse Ces1d and Ces1e in primary mouse hepatocytes is investigated. Primary mouse hepatocytes are treated with metformin (0, 0.125, 0.25 or 0.5 mM) for 24 h. The protein expression of Ces1d and Ces1e is detected by Western blot analysis. As shown in Fig. 2A, metformin consistently decreases Ces1d and Ces1e protein levels. On the basis of significant suppression on Ces1d and Ces1e expression caused by metformin in a concentration-dependent manner, we further examine whether metformin has effect on Ces1d and Ces1e expression in a time-dependent manner. Primary mouse hepatocytes are treated with metformin (0.5 mM) for corresponding time periods (0, 3, 6, 12, 24 or 48 h). In results, the decreases of Ces1d and Ces1e protein levels start after the treatment of metformin (0.5 mM) for 12 h, and reach the lowest levels after 48 h. It supports a time-dependent manner in suppressing CESs expression by metformin (Fig. 2B). The results imply that metformin represses the Ces1d and Ces1e protein expression in primary mouse hepatocytes in vitro. 3.3. Metformin attenuates the hydrolytic activity of Ces1d and Ces1e which is consistent with the decreased protein levels of Ces1d and Ces1e in vitro. We further determine whether the altered expression of Ces1d and Ces1e translates into the alteration of the hepatic hydrolytic activity by using standard substrate PNPA. Primary mouse hepatocytes are treated with metformin (0, 0.125, 0.25, 0.5 mM) for 24 h, and cell lysates are prepared. Consistent with the decreases in mouse Ces1d and Ces1e protein levels, the hydrolysis of PNPA is significantly decreased (Fig. 3A). Since the hydrolysis of PNPA represents the activity of numerous hepatic esterases more than CESs, further experiments are designed to detect the functional alterations of respective carboxylesterase by using specific substrates, clopidogrel (antiplatelet drug) for Ces1d and irinotecan (anticancer agent) for Ces1e (Shang et al., 2016; Yang et al., 2011; Timsit & Negishi, 2007; Ling et al., 2014). The hydrolysis of clopidogrel produces metabolites which is less toxic than the parent drug, whereas the hydrolysis of irinotecan leads to a higher toxic product than the parent drug. Primary mouse hepatocytes are first treated with metformin (0.5 mM) for 24 h, washed twice, and then treated with irinotecan for 24 h or clopidogrel for 30 h at various concentrations. After that, cell viability is measured by MTT assay, following the morphologic detection under microscope. As shown in Fig. 3B and C, the cells pretreated with metformin alone causes no difference in the cell viability (the concentration of either clopidogrel or irinotecan is 0 μM), compared with the ones non-pretreated with metformin. However, cells pretreated with metformin (0.5 mM) followed by clopidogrel (100 μM) or irinotecan (100 μM) shows statistical and morphological changes at corresponding concentration compared with the cells non-pretreated with metformin (Fig. 3B and C). Pretreatment with metformin increases the cytotoxicity of clopidogrel in primary mouse hepatocytes, while decreases the cytotoxicity of irinotecan. Consistently, when exposed to 100 μM clopidogrel for 30 h, cells pretreated with metformin were round, isolated, and aggregated; while the cells without metformin pretreatment were spread, with

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

17

Fig. 1. Metformin reduces Ces1d and Ces1e expression and the hydrolytic activity of the liver in vivo. (A) and (C). Metformin reduces the expression of Ces1d and Ces1e and the hepatic hydrolytic activity in a dose-dependent manner. Metformin (0, 100 and 300 mg/kg) was dissolved in distilled water and administered to mice by oral gavage. The liver was isolated and prepared for the experiments. (A). The expression of Ces1d and Ces1e was determined by Western blot. (C). The overall hydrolytic activity of the liver was detected using PNPA. (B) and (D). Metformin reduces the expression of Ces1d and Ces1e and the hepatic hydrolytic activity in a time-dependent manner. Mice were administered 300 mg/kg metformin for 0, 1 or 3 days and the expression (B) and activity of Ces1d and Ces1e (C) were determined. Data are expressed as mean ± SEM (n = 6). *p b 0.05, **p b 0.01 indicate a statistically significant result compared to vehicle-treated mice.

well-extended projects (Fig. 3C). On the contrary, when exposed to 100 μM irinotecan for 24 h, cells without metformin pretreatment were isolated and shrank, whereas cells pretreated with metformin

were morphologically normal (Fig. 3B). It suggests that altered expression of CESs by metformin has important pharmacological and toxicological consequences.

Fig. 2. Metformin represses the expression of Ces1d and Ces1e in primary mouse hepatocytes. (A). Metformin suppresses both Ces1d and Ces1e expression in a concentration-dependent manner. Primary mouse hepatocytes were treated with metformin (0, 0.125, 0.25 or 0.5 mM) for 24 h. (B). Metformin suppresses Ces1d and Ces1e expression in a time-dependent manner. Primary mouse hepatocytes were treated with 0.5 mM metformin for 0, 3, 6, 12, 24 or 48 h. Total protein extracted from hepatocytes was subjected to Western blot analysis for Ces1d and Ces1e expression. Experiments described in the figure were repeated independently at least three times, and data are expressed as mean ± SEM. *p b 0.05, **p b 0.01,***p b 0.001 compared with control.

18

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

Fig. 3. Metformin reduces the hydrolytic activity in primary mouse hepatocytes. (A). Primary mouse hepatocytes were treated with metformin (0, 0.125, 0.25 or 0.5 mM) for 24 h and the overall hydrolytic activity was determined. (B) and (D). Morphological analyses. The cells were treated with or without metformin 0.5 mM, followed by the treatment with 100 mM clopidogrel, or 100 mM irinotecan in full medium. After additional 30 h or 24 h incubation, images were taken under bright field. (C) and (E). Cell viability assay. Primary mouse hepatocytes were seeded into 96-well plates. Twenty-four hours later, the cells were treated with or without metformin (0.5 mM) for 12 h, followed by the treatment with clopidogrel, or irinotecan at various concentrations. After additional 30 h or 24 h incubation, cell viability was tested by MTT assay. All the experiments were repeated independently at least three times, and data are expressed as mean ± SEM. *p b 0.05, **p b 0.01 versus control (without metformin and clopidogrel or irinotecan); #p b 0.05 versus corresponding non-treatment with metformin.

3.4. The downregulation of PXR expression is involved in the suppressed expression of Ces1d and Ces1d by metformin.

(Fig. 4C). The data suggest the involvement of PXR in the suppression of Ces1d and Ces1e expression caused by metformin.

Since our previous study has indicated that the decreases of PXR is involved in the decreases of mouse Ces1d and Ces1e (Xiong et al., 2014a; Shang et al., 2016), the roles of PXR in the decreases of Ces1d and Ces1e expression induced by metformin are further explored in the following study. Primary mouse hepatocytes are treated with metformin (0, 0.125, 0.25 or 0.5 mM) for 24 h and then harvested for determining the PXR expression by Western blot analysis. As shown in Fig. 4A, metformin inhibits PXR expression at protein levels in a concentration-dependent manner, and the most significant repression by metformin is shown when the concentration is 0.5 mM. Thereafter, metformin (0.5 mM) is added into primary mouse hepatocytes and incubated for different period (0, 3, 6, 12, 24 or 48 h) in the following. The data show that PXR protein expression is suppressed by metformin in a time-dependent manner, starting from 12 h (Fig. 4B). To further confirm the decreased PXR expression by metformin, immunofluorescence analysis is performed by using PXR specific antibody to validate the results obtained with qualitative Western blot analysis. The expression of PXR regulated by metformin is consistent with the expression of CESs

3.5. Metformin decreases the expression and efflux activity of PXR target gene P-gp. To further confirm the involvement of PXR in the suppression of CESs induced by metformin, we thus detect the influence of metformin on the expression and activity of P-gp, a well-established target gene of PXR. Primary mouse hepatocytes are treated with various concentrations of metformin (0, 0.125, 0.25 or 0.5 mM) for 24 h and then harvested for determining the P-gp expression by Western blot analysis. In results, the P-gp protein levels are significantly decreased by metformin in a concentration-dependent manner, comparable with the suppressed expression of PXR and CESs (Fig. 5A). Moreover, metformin (0.5 mM) represses the P-gp expression in a time-dependent manner (Fig. 5B). The efflux activity of P-gp is determined by the intracellular accumulation of Rhodamine 123 (Rho123), which is a classical transporting substrate of P-gp. It shows that the treatment of metformin significantly increases intracellular Rho123 accumulation, indicating inhibited efflux activity of P-pg by metformin, which is consistent with the suppressed

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

19

Fig. 4. Effects of metformin on the expression of PXR in primary mouse hepatocytes. (A). Metformin suppressed PXR expression in a concentration-dependent manner. Primary mouse hepatocytes were treated with metformin (0, 0.125, 0.25 and 0.5 mM) for 24 h. Whole cell lysates were prepared for the detection of PXR expression. (B). Metformin suppressed PXR expression in a time-dependent manner. Primary mouse hepatocytes were treated with 0.5 mM metformin for 0, 3, 6, 12, 24 or 48 h. Total protein extracted from hepatocytes was subjected to Western blot analysis of PXR expression. The relative protein level was expressed as PXR/β-actin arbitrary units. Experiments described in the figure were repeated independently at least three times, and data are expressed as mean ± SEM. *p b 0.05 compared with control. (C). Primary mouse hepatocytes were treated with metformin (0, 0.125, 0.25 or 0.5 mM) for 24 h. Samples were stained by anti-PXR antibody and DAPI, then prepared for confocal microscopy analysis (200×). Images shown in this figure were representative of three independent experiments.

20

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

Fig. 5. Effects of metformin on the expression and efflux activity of P-gp in primary mouse hepatocytes. (A). Metformin suppressed P-gp expression in a concentration-dependent manner. Cells were treated with metformin (0, 0.125, 0.25 or 0.5 mM) for 24 h. (B). Metformin suppressed P-gp expression in a time-dependent manner. Primary mouse hepatocytes were treated with 0.5 mM metformin for 0, 3, 6, 12, 24 or 48 h. Whole cell lysates were prepared and equal amounts of total protein were used for the Western analysis. Experiments described in this figure were repeated independently at least for three times and data are expressed as mean ± SEM. *p b 0.05, **p b 0.01 compared with saline-treated group. (C). Metformin suppressed the efflux activity of P-gp. After treated with metformin (0, 0.125, 0.25, or 0.5 mM) for 24 h, cells were incubated with 5 μg/mL Rho123 in DMEM for 1.5 h in the dark. Then cells were prepared for confocal microscopy analysis (200×). Images shown in this figure were representative of three independent experiments.

expression of P-gp. These data suggest that not only the expression of PXR, but also its activity as a transcription factor is attenuated by metformin.

3.6. Metformin significantly decreases the expression of PXR and its target gene P-gp in vivo. Furthermore, the regulation of PXR and its target gene P-gp by metformin is confirmed in the in vivo experiment. Different doses of metformin (0, 100 or 300 mg/kg) are administered to mice once daily for three consecutive days by oral gavage. In comparison with control group, both PXR and its target gene P-gp protein expression significantly decrease in a dose concentration-dependent manner in vivo (Fig. 6A and C). Next, metformin (300 mg/kg) is administered to mice for different days (0, 1 or 3 days). It is confirmed that metformin (300 mg/kg) significantly suppresses the expression of both PXR and P-gp for 1 day, and more obviously when the administration lasting for 3 days, indicating a time-dependent manner (Fig. 6B and D). The data imply that the downregulation PXR is associated with the suppression of Ces1d and Ces1e induced by metformin in vivo.

3.7. The suppression of Ces1d and Ces1e by metformin is dependent on the activation of both AMPK and JNK pathway. On the above basis, underlying mechanisms are investigated by focusing on the intracellular signaling transduction. Since metformin has been reported to activate AMPK and JNK signaling pathways, specific inhibitors are applied and the activation status of respective signaling pathway is analyzed. As shown in Fig. 7A, metformin (0.5 mM) repeatedly suppresses the protein expression of Ces1d and Ces1e, and the repression by metformin is partially restored by compound C (a classical inhibitor of AMPK pathway) and SP600125 (a established inhibitor of JNK pathway) (Fig. 7A). To further confirm the involvement of the pathways, phospho-lysates are prepared and Western blot experiments are performed to detect the activation status of the respective pathway using specific antibodies. Data show that metformin increases the phosphorylation of AMPK and JNK in a time-dependent manner. The activation of both AMPK and JNK pathways by metformin starts after the incubation lasting for 2 h and extends after 10 h incubation without any obvious attenuation (Fig. 7B). Collectively, the results suggest that metformin represses Ces1d and Ces1e expression at least partially mediated by the activation of AMPK and JNK signaling pathways.

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

21

Fig. 6. Metformin decreases the expression of PXR and its target gene P-gp in vivo. (A) and (C). Metformin decreases the expression of PXR and P-gp in a dose-dependent manner. Different doses of metformin (0, 100, 300 mg/kg) were administered to mice by oral gavage. Total protein extracted from mice liver was subjected to Western blot analysis of PXR and P-gp expression. (B) and (C). Metformin decreases the expression of PXR and P-gp in a time-dependent manner. Metformin (300 mg/kg) was administered to mice for 0, 1 or 3 days, and the expression of PXR and P-gp was determined by Western blot analysis. Data are expressed as mean ± SEM (n = 6). *p b 0.05, **p b 0.01 indicate a statistically significant result compared to vehicle-treated animals.

4. Discussion As a widely used oral antihyperglycemic drug and a potential treatment for the cancer and aging (Senesi et al., 2016; Loos et al., 2017; Li et al., 2016; Castillo-Quan & Blackwell, 2016), coexposure of metformin with other remedies is highly frequent. Therefore, to elucidate the influence of metformin on the metabolism and clearance of other coadministered drugs is of great clinical importance. For the first time, we identify the suppression of Ces1d and Ces1e expression and activity by metformin both in vivo and in vitro (Figs. 1 and 2). Carboxylesterases are major enzymes involved in the hydrolytic metabolism of a wide range of drugs and other xenobiotics. The suppression of Ces1d and Ces1e by metformin indicates a clinical relevance that metformin probably inhibits the hydrolysis of many CESs-substrate drugs and subsequently affects their therapeutic actions as a perpetrator drug. When metformin is used in combination with enalapril which is a substrate activated by CES1 (Thomsen et al., 2014), potential drug-drug interactions may take place as metformin/enalapril has been previously identified to be the most common interacting pairs among hospitalized cardiac

patients (Sharma et al., 2014). Clinical pharmacokinetic drug-drug interaction studies with metformin and enalapril will need to be performed to further elucidate the potential clinical relevance of the presented in vivo and in vitro animal results. In a nonclinical level, the present study contributes to understand the underlying mechanisms of those drug-drug interactions and has a potential to guide the drug dosage adjustment in T2D patients with cardiovascular complications. When determining the alteration of carboxylesterase activity induced by metformin, we used esterase standard substrate PNPA (Fig. 3A) as well as specific substrates, clopidogrel for Ces1d and irinotecan for Ces1e. Since metformin suppresses the expression of Ces1d and Ces1e, it is speculated that metformin increases the toxicity of clopidogrel which is inactivated by Ces1d (Shi et al., 2006; Tang et al., 2006), while decreases the toxicity of irinotecan which is activated by Ces1e to produce a more toxic metabolite SN-38 (Wu et al., 2002; Mao et al., 2011). As expected, the increased toxicity of clopidogrel and decreased toxicity of irinotecan by the pretreatment of metformin provide evidence that metformin significantly attenuates the hydrolytic activity of Ces1d and Ces1e (Fig. 3B and C), which is comparable with

22

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

Fig. 7. AMPK and JNK pathways mediate the suppression of Ces1d and Ces1e induced by metformin. (A). The specific inhibitors of AMPK and JNK pathways abolished the suppression of Ces1d and Ces1e induced by metformin. Primary mouse hepatocytes were preincubated with compound C (20 μM) or Sp600125 (10 μM) for 1 h before the treatment with metformin (0.5 mM). Cells were then lysed and analyzed for Western blot against Ces1d and Ces1e. (B) and (C). Metformin suppressed the activation of AMPK and JNK pathways. Primary mouse hepatocytes were treated with metformin (0.5 mM) for indicate periods. Phospho-lysates were prepared for p-AMPK, AMPK, p-JNK, JNK and β-actin Western blot analysis. The relative protein level was expressed as p-AMPK/AMPK and p-JNK/JNK arbitrary units. Experiments described in the figure were repeated independently at least three times, and data are expressed as mean ± SEM. *p b 0.05, **p b 0.01 compared with control.

the suppressed expression by metformin (Fig. 2). The attenuation of CESs hydrolytic activity induced by metformin further support the drug-drug interactions in metformin-based medical combination. The mechanisms of the CESs suppression induced by metformin are further investigated in the present study by focusing on the nuclear receptor PXR. PXR regulates the transcriptional expression of drug metabolism-related genes such as one of the major DMEs CYP3A4 and efflux transporter P-gp, as well as the CESs family (Luo et al., 2017; Sivertsson et al., 2013; Shang et al., 2016; Yang et al., 2011). Therefore, PXR is largely responsible for the drug metabolism and the detoxification of environmental toxicants (Timsit & Negishi, 2007). In the present study, metformin suppresses the expression of PXR in a time- and dosedependent manner both in vivo and in vitro (Figs. 4, 6A and B), as demonstrated using Western blot and immunofluorescence microscopy analysis (Fig. 3C). The data indicate that PXR mediates the suppression of CESs by metformin, which is in consistence with previous findings that metformin dramatically suppresses PXR-mediated expression of CYP3A4 in human primary hepatocytes (Krausova et al., 2011). In addition, metformin also decreases the expression and activity of P-gp, a well-known target gene of PXR (Figs. 5, 6C and D). And the regulation of metformin on the expression of P-gp is supported by previous findings that metformin transcriptionally downregulates the expression of MDR1 (P-gp) (Ling et al., 2014).

As a metabolic master switch in the modulation of energy metabolism, AMPK activation has been reported to play a central role in metformin's actions (Lee et al., 2012; Klein et al., 2016) as well as in the repressed transcriptional expression of PXR-targeted genes (Krausova et al., 2011; Ling et al., 2014). In the present study, we provide evidence that metformin suppresses the expression of CESs, which is significantly abolished by the inhibition of AMPK pathway with Compound C (Fig. 7A). It is investigated for the first time that AMPK pathway plays an important role in the repression of Ces1d and Ces1e by metformin (Fig. 7B). However, it still deserves full address on whether AMPK pathway mediates metformin-repressed CESs expression in a PXR-dependent or -independent way, which should be further elucidated in the following studies. In addition, as the involvement of JNK pathway in the action of metformin has been recently reported (Wu et al., 2011), JNK signaling is supposed to contribute to CESs expression repression induced by metformin. As shown in Fig. 7A and C, metformin (0.5 mM) significantly activates JNK phosphorylation, and the suppression of CESs expression is drastically abolished by the specific JNK inhibitor SP600125. The data are supportive for the involvement of JNK signaling in the regulation of CESs expression by metformin for the first time. There is one important limitation that should be acknowledged. Even though mouse Ces1d and Ces1e strongly cross-react with human

E. Shan et al. / European Journal of Pharmaceutical Sciences 102 (2017) 14–23

CES1 and CES2 respectively (Xiao et al., 2012), the expression of CESs is largely regulated in a species-dependent manner in some cases. In the present study, the findings of CESs suppression by metformin are demonstrated in mouse hepatocytes in vitro and mouse livers in vivo. Considering the species difference exists in the regulation of PXR and its target genes, the mouse cells and tissues employed in the present study may restrict the contribution of the findings on guiding the rational drug use in clinic. However, the mediation of PXR in metformin-suppressed CESs expression we found in the present study is comparable with previous findings which demonstrate the role of PXR using human primary hepatocytes (Krausova et al., 2011). It implies certain clinical significance of the present study and deserves further investigation including clinical trials in the future studies. In order to ensure the validity of quantitative Western blot, several measures are taken in detecting the protein expression of CESs, PXR and P-gp. In the in vivo experiments, S9 fractions of livers are prepared by accurately weighing the liver tissues and adding homogenization buffer. In the in vitro experiments, an equal number of cells are used for each group. Next, a protein assay is done to load equal amounts of protein in each group. Then, ponceau red staining of the PVDF membrane is applied after transfer. Finally, membranes are cut according to the prestained protein marker (Thermo Scientific) and incubated with primary antibodies against CESs, PXR or P-gp or control antibody against β-actin in the meantime. In addition, we also validate the suppression of PXR expression in primary mouse hepatocyte by performing immunofluorescence analysis. The results are compatible using different methods of Western blot and immunofluorescence. In conclusion, our results show that the antidiabetic agent metformin represses the expression and activity of Ces1d and Ces1e through the inhibition of PXR-mediated regulation. Furthermore, the suppression of CESs expression by metformin is mediated by the activation of AMPK and JNK pathways. The findings contribute to suggest a broader range of drug-drug interactions in metformin-based drug combination and should be acknowledged in the T2D drug therapy. Conflicts of interest The authors declare that there are no conflicts of interest. Acknowledgements The study was supported by the Natural Science Foundation of China (Nos. 81302855, 81573503, 81373443, 81173128), the Major Project of Jiangsu Provincial Department of Education (No. 13KJA310003), the Natural Science Foundation of Jiangsu Province, China (No. BK2012446). References Breuker, C., et al., 2017. Patients with diabetes are at high risk of serious medication errors at hospital: Interest of clinical pharmacist intervention to improve healthcare. Eur. J. Intern. Med. 38, 38–45. Castillo-Quan, J.I., Blackwell, T.K., 2016. Metformin: restraining nucleocytoplasmic shuttling to fight cancer and aging. Cell 167, 1670–1671. Do, M.T., et al., 2013. Metformin inhibits heme oxygenase-1 expression in cancer cells through inactivation of Raf-ERK-Nrf2 signaling and AMPK-independent pathways. Toxicol. Appl. Pharmacol. 271, 229–238. El Messaoudi, S., et al., 2016. The effect of dipyridamole on the pharmacokinetics of metformin: a randomized crossover study in healthy volunteers. Eur. J. Clin. Pharmacol. 72, 725–730. Ginter, E., Simko, V., 2012. Type 2 diabetes mellitus, pandemic in 21st century. Adv. Exp. Med. Biol. 771, 42–50. Gu, X., et al., 2006. Role of NF-kappaB in regulation of PXR-mediated gene expression: a mechanism for the suppression of cytochrome P-450 3A4 by proinflammatory agents. J. Biol. Chem. 281, 17882–17889. Klein, J.D., et al., 2016. Metformin, an AMPK activator, stimulates the phosphorylation of aquaporin 2 and urea transporter A1 in inner medullary collecting ducts. Am. J. Physiol. Ren. Physiol. 310, F1008–F1012.

23

Krausova, L., et al., 2011. Metformin suppresses pregnane X receptor (PXR)-regulated transactivation of CYP3A4 gene. Biochem. Pharmacol. 82, 1771–1780. Kumari, S., et al., 2015. Pregnane and xenobiotic receptor gene expression in liver cells is modulated by Ets-1 in synchrony with transcription factors Pax5, LEF-1 and c-Jun. Exp. Cell Res. 330, 398–411. Lee, J.O., et al., 2012. Metformin regulates glucose transporter 4 (GLUT4) translocation through AMP-activated protein kinase (AMPK)-mediated Cbl/CAP signaling in 3T3L1 preadipocyte cells. J. Biol. Chem. 287, 44121–44129. Li, K.L., et al., 2016. A multicenter double-blind phase II study of metformin with gefitinib as first-line therapy of locally advanced non-small-cell lung cancer. Clin. Lung Cancer http://dx.doi.org/10.1016/j.cllc.2016.12.003. Ling, S., et al., 2014. Metformin reverses multidrug resistance in human hepatocellular carcinoma Bel7402/5fluorouracil cells. Mol. Med. Rep. 10, 2891–2897. Loos, J.A., et al., 2017. Metformin exhibits preventive and therapeutic efficacy against experimental cystic echinococcosis. PLoS Negl. Trop. Dis. 11, e0005370. Luo, W., et al., 2017. Suppression of carboxylesterases by imatinib mediated by the downregulation of pregnane X receptor. Br. J. Pharmacol. http://dx.doi.org/10.1111/bph. 13731. Mao, Z., et al., 2011. Lipopolysaccharide down-regulates carbolesterases 1 and 2 and reduces hydrolysis activity in vitro and in vivo via p38MAPK-NF-kappaB pathway. Toxicol. Lett. 201, 213–220. Pham, H.T., et al., 2017. Effects of Averrhoa carambola L. (Oxalidaceae) juice mediated on hyperglycemia, hyperlipidemia, and its influence on regulatory protein expression in the injured kidneys of streptozotocin-induced diabetic mice. Am. J. Transl. Res. 9, 36–49. Pondugula, S.R., et al., 2015. Diindolylmethane, a naturally occurring compound, induces CYP3A4 and MDR1 gene expression by activating human PXR. Toxicol. Lett. 232, 580–589. Reagan-Shaw, S., et al., 2008. Dose translation from animal to human studies revisited. FASEB J. 22, 659–661. Senesi, P., et al., 2016. Metformin treatment prevents sedentariness related damages in mice. J. Diab. Res. 2016, 8274689. Shang, W., et al., 2016. Fluoxetine reduces CES1, CES2, and CYP3A4 expression through decreasing PXR and increasing DEC1 in HepG2 cells. Xenobiotica 46, 393–405. Sharma, S., et al., 2014. A study of potential drug-drug interactions among hospitalized cardiac patients in a teaching hospital in Western Nepal. Indian J Pharmacol. 46, 152–156. Shi, D., et al., 2006. Anti-influenza prodrug oseltamivir is activated by carboxylesterase human carboxylesterase 1, and the activation is inhibited by antiplatelet agent clopidogrel. J. Pharmacol. Exp. Ther. 319, 1477–1484. Sivertsson, L., et al., 2013. Induced CYP3A4 expression in confluent Huh7 hepatoma cells as a result of decreased cell proliferation and subsequent pregnane X receptor activation. Mol. Pharmacol. 83, 659–670. Sozio, M.S., et al., 2011. Activated AMPK inhibits PPAR-{alpha} and PPAR-{gamma} transcriptional activity in hepatoma cells. Am. J. Physiol. Gastrointest. Liver Physiol. 301, G739–G747. Tang, M., et al., 2006. Antiplatelet agents aspirin and clopidogrel are hydrolyzed by distinct carboxylesterases, and clopidogrel is transesterificated in the presence of ethyl alcohol. J. Pharmacol. Exp. Ther. 319, 1467–1476. Thomsen, R., et al., 2014. In vitro drug metabolism by human carboxylesterase 1: focus on angiotensin-converting enzyme inhibitors. Drug Metab. Dispos. 42, 126–133. Timsit, Y.E., Negishi, M., 2007. CAR and PXR: the xenobiotic-sensing receptors. Steroids 72, 231–246. Wang, Z., et al., 2015. Resveratrol induces AMPK-dependent MDR1 inhibition in colorectal cancer HCT116/L-OHP cells by preventing activation of NF-kappaB signaling and suppressing cAMP-responsive element transcriptional activity. Tumour Biol. 36, 9499–9510. Wu, M.H., et al., 2002. Irinotecan activation by human carboxylesterases in colorectal adenocarcinoma cells. Clin. Cancer Res. 8, 2696–2700. Wu, N., et al., 2011. Metformin induces apoptosis of lung cancer cells through activating JNK/p38 MAPK pathway and GADD153. Neoplasma 58, 482–490. Xiao, D., et al., 2012. Age-related inducibility of carboxylesterases by the antiepileptic agent phenobarbital and implications in drug metabolism and lipid accumulation. Biochem. Pharmacol. 84, 232–239. Xiong, J., et al., 2014a. Glucose dominates the regulation of carboxylesterases induced by lipopolysaccharide or interleukin-6 in primary mouse hepatocytes. Life Sci. 112, 41–48. Xiong, J., et al., 2014b. Fluoxetine suppresses AMP-activated protein kinase signaling pathway to promote hepatic lipid accumulation in primary mouse hepatocytes. Int. J. Biochem. Cell Biol. 54, 236–244. Xu, G., et al., 2015. Metformin ameliorates ionizing irradiation-induced long-term hematopoietic stem cell injury in mice. Free Radic. Biol. Med. 87, 15–25. Yang, J., et al., 2011. Byakangelicin induces cytochrome P450 3A4 expression via transactivation of pregnane X receptors in human hepatocytes. Br. J. Pharmacol. 162, 441–451. Zack, J., et al., 2015. Pharmacokinetic drug-drug interaction study of ranolazine and metformin in subjects with type 2 diabetes mellitus. Clin. Pharmacol. Drug Dev. 4, 121–129.