1521-0111/86/5/580–591$25.00 MOLECULAR PHARMACOLOGY Copyright ª 2014 by The American Society for Pharmacology and Experimental Therapeutics

http://dx.doi.org/10.1124/mol.114.092874 Mol Pharmacol 86:580–591, November 2014

Metformin Restores Intermediate-Conductance CalciumActivated K+ channel– and Small-Conductance CalciumActivated K+ channel–Mediated Vasodilatation Impaired by Advanced Glycation End Products in Rat Mesenteric Artery s Li-Mei Zhao, Yan Wang, Yong Yang, Rong Guo, Nan-Ping Wang, and Xiu-Ling Deng Department of Physiology and Pathophysiology (L.-M.Z., Y.W., Y.Y., R.G., X.-L.D.) and Cardiovascular Research Center (L.-M.Z., Y.W., N.-P.W., X.-L.D.), School of Medicine, Xi’an Jiaotong University, Xi’an, Shaanxi, China; and Key Laboratory of Cardiovascular Medicine Research, Harbin Medical University, Ministry of Education, Nangang District, Harbin, Heilongjiang, China (L.-M.Z.) Received March 24, 2014; accepted August 14, 2014

ABSTRACT The present study was designed to investigate the effect of metformin on the impairment of intermediate-conductance and small-conductance Ca21-activated potassium channels (IKCa and SKCa)–mediated relaxation in diabetes and the underlying mechanism. The endothelial vasodilatation function of mesenteric arteries was assessed with the use of wire myography. Expression levels of IK Ca and SK Ca and phosphorylated Thr 172 of AMPactivated protein kinase (AMPK) were measured using Western blot technology. The channel activity was observed using a whole-cell patch voltage clamp. Reactive oxygen species (ROS) were measured using dihydroethidium and 29,79-dichlorofluorescein diacetate. Metformin restored the impairment of IKCa- and SKCamediated vasodilatation in mesenteric arteries from streptozotocin-

Introduction Vascular diseases are the principal causes of morbidity and mortality in patients with diabetes mellitus. Loss of the modulatory role of the endothelium may be a critical and initiating factor in the development of diabetic vascular diseases (Hartge et al., 2006; Vlassara et al., 2008; Vanhoutte et al., 2009). Endothelium-dependent vasodilatation is an accessible parameter This work was supported by the National Nature Science Foundation of China [Grants 81370191, 81170137, and 81200097] and funding from Key Laboratory of Cardiovascular Medicine Research (Harbin Medical University), Ministry of Education [Grant 2013010]. dx.doi.org/10.1124/mol.114.092874. s This article has supplemental material available at molpharm.aspetjournals. org.

induced type 2 diabetic rats and that from normal rats incubated with advanced glycation end products (AGEs) for 3 hours. In cultured human umbilical vein endothelial cells (HUVECs), 1 mM metformin reversed AGE-induced increase of ROS and attenuated AGEand H2O2- induced downregulation of IKCa and SKCa after longterm incubation (.24 hours). Short-term treatment (3 hours) with 1 mM metformin reversed the decrease of IKCa and SKCa currents induced by AGE incubation for 3 hours without changing the channel expression or the AMPK activation in HUVECs. These results are the first to demonstrate that metformin restored IKCa- and SKCa-mediated vasodilatation impaired by AGEs in rat mesenteric artery, in which the upregulation of channel activity and protein expression is likely involved.

to probe endothelial function in different pathophysiological conditions. The endothelium-mediated dilation of arteries can be achieved by the release of relaxing factors, such as nitric oxide (NO), prostacyclin, and endothelium-derived hyperpolarizing factor (EDHF), from the endothelial cells. Findings from arteries of many species, including humans, indicate that endothelial intermediate-conductance Ca21-activated potassium channels (IKCa) and small-conductance Ca21-activated potassium channels (SKCa) play a pivotal role in mediating the EDHF effects in many small vascular beds (Busse et al., 2002; Dora et al., 2008; Félétou and Vanhoutte, 2009; Grgic et al., 2009). Studies indicate that the main IKCa isoform is the KCa3.1 subunit, and the SKCa isoform is KCa2.3 in endothelial cells (Eichler et al., 2003). In vivo studies in mice show that deficiency of KCa2.3

ABBREVIATIONS: Ab120335, 6,7-dihydro-4-hydroxy-3-(29-hydroxy[1,19-biphenyl]-4-yl)-6-oxo-thieno[2,3-b]pyridine-5-carbonitrile; ACh, acetylcholine; AGE, advanced glycation end product; AGE-BSA, advanced glycation end product–bovine serum albumin; AMPK, AMP-activated protein kinase; ANOVA, analysis of variance; COX, cyclooxygenase; DCF-DA, 29,79-dichlorofluorescein diacetate; DHE, dihydroethidium; DM, diabetes mellitus; EDHF, endothelium-derived hyperpolarizing factor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HbA1c, glycated hemoglobin; HUVEC, human umbilical vein endothelial cell; IKCa, intermediate-conductance calcium-activated K1 channel; INDO, indomethacin; I-V, current-voltage; KHS, Krebs-Henseleit solution; L-NAME, N-nitro-L-arginine; NO, nitric oxide; NOS, nitric oxide synthase; NS309, 6,7-dichloro-1H-indole-2,3-dione 3-oxime; p-AMPK, phosphorylated AMPK; PBS, phosphate-buffered saline; PE, phenylephrine; RAGE, receptor for AGEs; RIPA, RadioImmunoprecipitation Assay; ROS, reactive oxygen species; SKCa, small-conductance calcium-activated K1 channel; SNP, sodium nitroprusside; STZ, streptozotocin; T2DM, type 2 diabetes mellitus; TRAM-34, 1-[(2-chlorophenyl)diphenylmethyl]-1H-pyrazole; UCL-1684, 6,12,19,20,25,26hexahydro-5,27:13,18:21,24-trietheno-11,7-metheno-7H-dibenzo[b,m][1,5,12,16]tetraazacyclotricosine-5,13-diium. 580

Metformin Restores AGE-Induced Vasodilative Dysfunction

and/or KCa3.1 channels (Taylor et al., 2003; Si et al., 2006; Brähler et al., 2009) leads to elevated blood pressure. In addition, opening KCa2.3 and KCa3.1 channels decreases myogenic tone and increases acetylcholine (ACh)-induced relaxation in rat cremaster arterioles (Sheng et al., 2009). It has been reported that IKCa- and SKCa-mediated relaxation is impaired in resistance arteries of diabetic rats, which is related to the downregulation of these channels (Wigg et al., 2001; Burnham et al., 2006; Leo et al., 2011). Advanced glycation end products (AGEs) are a heterogeneous group of complex compounds that are formed irreversibly in serum and tissues via a chain of nonenzymatic chemical reactions (Vlassara et al., 2008). AGEs are thought to impair endothelium-dependent vasodilatation in diabetes (Gao et al., 2008; Sena et al., 2012). Our previous study found that AGEs impaired IKCa- and SKCa-mediated relaxation in rat mesenteric arteries via downregulation of KCa3.1 and KCa2.3 channels induced by oxidative stress (Zhao et al., 2014). Metformin is currently considered the first-line pharmacological therapy in type 2 diabetes mellitus (T2DM) (Consoli et al., 2004; Bennett et al., 2011), and showed a greater effect than sulfonylurea or insulin for any diabetes-related endpoint, all-cause mortality, and stroke (UK Prospective Diabetes Study (UKPDS) Group, 1998). Metformin provides more cardiovascular protection. A significant reduction was also observed regarding myocardial infarction at the end of the trial in the metformin group compared with the conventional group (UK Prospective Diabetes Study (UKPDS) Group, 1998). Metformin is associated with improvement in some markers of endothelial function, which may explain why it is associated with a decreased risk of cardiovascular disease in T2DM (de Jager et al., 2014). Both in vivo and in vitro metformin improved AChinduced relaxation in insulin-resistant rats, apparently through NO-dependent relaxation (Katakam et al., 2000). Metformin improved the vascular function in obese nondiabetic rats through reduction of reactive oxygen species (ROS) generation, modulation of membrane hyperpolarization, correction of the unbalanced prostanoids release, and increase in the sensitivity of the smooth muscle to NO (Lobato et al., 2012). But the effect of metformin on IKCa- and SKCa-mediated vasodilatation in small arteries impaired by AGEs remains unknown. In this study, we found that metformin restored IKCa- and SKCa-mediated vasodilatation impaired by AGEs in rat mesenteric artery, which was related to the upregulation of channel activity and protein expression.

Materials and Methods Animals. Experimental protocols were approved by the Institutional Animal Care and Use Committee of Xi’an Jiaotong University and conformed with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. Male Sprague-Dawley rats (173 6 18 g) were housed under the following conditions: 22 6 2°C, humidity of 55 6 5%, and a 12/12-hour light/dark cycle. They were randomly divided into control animals (n 5 6) and diabetic animals (n 5 14). Diabetes was induced by a single intraperitoneal injection of 30 mg·kg21 streptozotocin (STZ) after 8 weeks on a modified high-fat and high-glucose diet (10% grease, 20% sucrose, 1% bile salt, and 2.5% cholesterol) (Wang et al., 2010; Su et al., 2011). Blood glucose was measured using a One Touch Sure Step Glucose Meter (LifeScan Inc., Milpitas, CA) 1 week after STZ injection. Animals with .11.1 mmol·l21 blood glucose (n 5 12) were randomly divided into the diabetes mellitus (DM) group (n 5 6) and the metformin group (n 5 6).

581

Animals in the DM and metformin groups were fed a high-fat and high-glucose diet for another 4 weeks after STZ injection, then a standard rat chow for an additional 2 months. Rats in the metformin group were treated with metformin (300 mg·kg21·d21 intragastric administration) after their blood glucose increased for 3 months. Control animals were fed the standard rat chow and injected with the drug vehicle citric acid buffer. Preparation of Advanced Glycation End Product–Bovine Serum Albumin. Advanced glycation end product–bovine serum albumin (AGE-BSA) was prepared as described previously (Zhao et al., 2012). In brief, BSA with low endotoxin was dissolved in 0.5 mM phosphate-buffered saline (PBS; pH 7.4) containing 1.6 mM D-glucose, sterilized by ultrafiltration, and then incubated at 37°C for 12 weeks, followed by dialysis against PBS for 24 hours to remove unincorporated glucose. Control nonglycated BSA was prepared with the same procedure without D-glucose inclusion. AGE-BSA content was estimated by fluorescence spectroscopy (excitation at 390 nm and emission at 460 nm) at a protein concentration of 1 mg·ml21, which indicated a 10-fold increase in fluorescence in AGE-BSA compared with control BSA. The prepared AGE-BSA and unmodified BSA were sterilized by ultrafiltration, respectively, and stored at 220°C until use. Endotoxin levels of each preparation were measured by limulus amebocyte lysate assay (E-Toxate; Sigma-Aldrich, St. Louis, MO) and were found to be ,0.8 EU·ml21. Isometric Force Measurements in Mesenteric Small Arteries. After the rats were anesthetized with sodium pentobarbital (50 mg·kg21, i.p.), the mesenteric arcade was isolated and immediately placed in ice-cold Krebs-Henseleit solution (KHS; pH 7.4) gassed with a mixture of 95% O2 and 5% CO2. The composition of KHS was as follows (in mM): NaCl 115, NaHCO3 25, KCl 4.6, NaH2PO4 1.2, MgCl2 1.2, CaCl2 2.5, and glucose 10.0. The third-order branches of the superior mesenteric artery (internal diameter ∼300 mm) were isolated, cleared of fat and connective tissue, and cut into 2-mm-long rings. The arterial rings were threaded onto two stainless steel wires (40 mm in diameter) and were mounted in 5-ml chambers of a multimyograph system (model 610M; Danish Myo Technology, Aarhus, Denmark) containing KHS continuously aerated with 95% O2 and 5% CO2 at 37°C for isometric force measurements. Tension signals were relayed to a PowerLab recording unit (ADInstruments Ltd., Aarhus, Denmark) and saved to Chart 7 for Windows software (ADInstruments Ltd.). After being mounted, the arteries were allowed to equilibrate for 20 minutes before normalization. The passive tension–internal circumference was determined by stretching to 90% of L100, where L100 is defined as the circumference of the relaxed artery exposed to a transmural pressure of 100 mm Hg, as previously described by McPherson (1992). The vessels of rats in the control, DM, and metformin groups were then allowed to equilibrate for at least 60 minutes, and with the bath solution changed every 15 minutes. After the equilibration, reactivity of the rings was checked thrice by administration of 60-mM KCl (achieved by substitution of NaCl in KHS with an equimolar concentration of KCl). To assess the integrity of the endothelium, mesenteric arteries were precontracted with phenylephrine (PE; 1 mM), and a high dose of ACh (10 mM) was used to relax the artery rings. ACh-induced relaxation was greater than 80% of the precontracted tone in all cases, indicating that the endothelium was functionally intact. Myograph Protocol. After further washouts, arteries were again precontracted with 1 mM PE. The effect of the treatment on relaxant responses was determined by cumulative concentration-response curves to ACh (0.1 nM to 10 mM) and sodium nitroprusside (SNP; 0.01 nM to 10 mM). In addition, responses to ACh were examined after 15-minute incubation with 100 mM N-nitro-L-arginine (L-NAME), a nonselective nitric oxide synthase (NOS) inhibitor, and 10 mM indomethacin (INDO), a blocker of cyclooxygenase (COX). The arteries were again washed and rested for 30 minutes in KHS before being incubated with 1-[(2-chlorophenyl)diphenylmethyl]-1H-pyrazole (TRAM34; 10 mM), a selective blocker of IKCa, or 6,12,19,20,25,26-hexahydro5,27:13,18:21,24-trietheno-11,7-metheno-7H-dibenzo[b,m][1,5,12,16]

582

Zhao et al.

tetraazacyclotricosine-5,13-diium (UCL-1684; 0.1 mM), a selective blocker of SKCa, in the continuous presence of INDO and L-NAME, and the ACh concentration-response curve was repeated. In addition, the cumulative concentration-response curves to 6,7-dichloro-1Hindole-2,3-dione 3-oxime (NS309; 0.1 nM to 1 mM), an IKCa and SKCa opener, of the arteries incubated with or without 10 mM TRAM-34 or 0.1 mM UCL-1684 were also recorded. In a subset of experiments, the vessels from normal rats underwent the same process, except that the vessels were equilibrated for 3 hours in KHS without or with 200 mg·ml21 BSA, 200 mg·ml21 AGE-BSA, or 1 mM metformin (preincubation for 30 minutes) plus 200 mg·ml21 AGE-BSA after normalization. Besides the responses to ACh, the NS309 (0.1 nM to ∼1 mM) concentration-response curves of the vessels incubated with or without 10 mM TRAM-34 or 0.1 mM UCL-1684 were also recorded. Cell Culture and Treatments. Human umbilical vein endothelial cells (HUVECs) (CRL-1730) were purchased from American Type Culture Collection (Manassas, VA). Cells were cultured in Dulbecco’s modified Eagle’s medium/F12 (Gibco, Manassas, VA) containing 10% fetal bovine serum (Invitrogen, Carlsbad, CA), penicillin (100 U·ml21), and streptomycin (100 mg·ml21) and incubated at 37°C in a humidified atmosphere with 5% CO2. The medium was changed twice weekly. Cells at 80% confluence were starved for 24 hours in Dulbecco’s modified Eagle’s medium/F12 containing 0.1% fetal bovine serum before being treated with different interventions. Western Blot Analysis. HUVECs incubated with a different stimulus were lysed with ice-cold modified RIPA (Radio-Immunoprecipitation Assay) Lysis Buffer (60 mM Tris-HCl, 0.25% SDS, 1 mM sodium fluoride, 1 mM sodium orthovanadate, 10 mg·ml21 aprotinin and leupeptin). The lysates were then centrifuged at 12,000 g for 15 minutes at 4°C. After transferring the supernatant to a fresh icecold tube, the protein concentration was determined with BCA protein assay. Equal concentrations of proteins were mixed with SDS sample buffer and denatured at 100°C for 10 minutes. The samples were separated on an SDS–10% polyacrylamide gel, then transferred to a PVDF (polyvinylidene difluoride) membrane at 300 mA for 2 hours in a transfer buffer containing 20 mM Tris, 150 mM glycine, and 20% methanol. The membranes were blocked with 5% nonfat dried milk in Tris-buffered saline/Tween 20 (0.1% Tween 20) for 1 hour. After blocking, the blots were incubated in primary antibody for KCa3.1 (1:300), KCa2.3 (1:200), AMP-activated protein kinase (AMPK) (1:800), or phosphorylated AMPK (p-AMPK) Thr172 (1:800) at 4°C overnight, and then incubated with horseradish peroxidase–conjugated secondary antibodies (1:5000) for 1 hour at room temperature. The bound antibodies were detected with an enhanced chemiluminescence detection system (Amersham, Piscataway, NJ) and quantified by densitometry using a Chemigenius Bio-imaging System (Syngene, Cambridge, UK). Based on the results from our preliminary experiments, we found that stimulation with 200 mg·ml21 BSA or 200 mg·ml21 AGE-BSA for 48 hours did not affect glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression. Therefore, after stripping, the membrane was then reprobed with a loading control antibody GAPDH (1:5000) to normalize for the amount of proteins. The ratio of band intensity to GAPDH was obtained to quantify the relative protein expression level. Electrophysiology. Membrane current was recorded in cultured HUVECs with the whole-cell patch-clamp technique as described previously (Wang et al., 2013; Wu et al., 2013). The detached HUVECs were placed into a 0.2-ml cell chamber mounted on an inverted microscope, allowed to settle to the bottom (∼20 minutes), and then superfused with Tyrode’s solution (pH 7.4), which contained the following (mM): 136 NaCl, 5.4 KCl, 1.0 MgCl2, 1.8 CaCl2, 10 glucose, and 10 HEPES. Borosilicate glass electrodes (1.2-mm outer diameter) were pulled with a Brown-Flaming puller (model P-97; Sutter Instrument Co., Novato, CA), and had a resistance of 2–3 MV when filled with pipette solution, which contained the following (mM): 120 K-aspartate, 20 KCl, 1.0 MgCl 2 , 5.0 EGTA, 5 Na-phosphocreatine, 5 Mg-ATP, 0.1 GTP, 10 HEPES, and 500 nM free Ca21 with pH adjusted to 7.2 with KOH. The tip potentials were zeroed before the pipette contacted the

cell. After a gigaohm seal was obtained by negative suction, the cell membrane was ruptured by a gentle suction to establish whole-cell configuration. Membrane currents were recorded with an EPC-9 patchclamp amplifier and Pulse software (Heka Elektronik, Lambrecht, Germany). Command pulses were generated by a 12-bit digital-to-analog converter controlled by Pulse software. All current recording experiments were conducted at room temperature (22–23°C). Intracellular ROS Detection. Intracellular ROS generation was measured with the fluorescent probes of dihydroethidium (DHE) and 29,79-dichlorofluorescein diacetate (DCF-DA). HUVECs grown on coverslips were stimulated with 200 mg·ml21 BSA, 200 mg·ml21 AGE-BSA, or 200 mg·ml21 AGE-BSA plus 1 mM metformin for 24 hours. Serumdeprived cells were washed with PBS, loaded with 5 mM DHE or 4 mM DCF-DA, and incubated for 30 minutes at 37°C. Differential interference contrast images were obtained using fluorescence microscopy. Reagents. UCL-1684, TRAM-34, NS309, metformin, BSA, DHE, DCF-DA, and other chemicals used in this study were obtained from Sigma-Aldrich. Anti-KCa3.1 and anti-KCa2.3 antibodies were from Alomone Laboratories Ltd. (Jerusalem, Israel). All drugs were dissolved in distilled water, with the exception of INDO, NS309, and UCL-1684, which were dissolved in dimethylsulfoxide. Statistical Analysis. Results are presented as means 6 S.E.M., and n represents the number of animals or the number of assays. Concentration-response curves from rat isolated mesenteric arteries were made with Prism version 5.0 software (GraphPad Software, San Diego, CA) and other figures with SigmaPlot 10.0 software (Systat Software, San Jose, CA). Percentage values of relaxation to ACh or SNP were measured as a percentage of precontraction to PE. In this study, IKCa- and SKCa-mediated relaxation refers to the relaxation which can be blocked by the selective blockers, UCL-1684 or TRAM-34. Percentage values of relaxation of IKCa- and SKCa-mediated relaxation were calculated as the difference in value between the percentage values of relaxation before and after incubation with 0.1 mM UCL-1684 or 10 mM TRAM-34. Group percentage values of relaxation were compared via one-way analysis of variance (ANOVA) with Tukey’s post test, whereas one-way ANOVA with Bonferroni’s post test was used for other results. P , 0.05 was considered statistically significant. All statistical analyses were made with Prism version 5.0 software.

Results Characteristics of Diabetic Rat Model. Table 1 displays the characteristics of a diabetic rat model after 3 months receiving STZ or vehicle injection. Compared with control, the diabetic rats exhibited decreased body weight and increased blood glucose and glycated hemoglobin (HbA1c) level. Treatment with metformin partly reversed the increasing of blood glucose and HbA1c level but further decreased the body weight of diabetic rats. Influences of Metformin on IKCa- and SKCa-Mediated Relaxation in Small Arteries from Diabetic Rats. To examine whether endothelium-dependent relaxation in small arteries was damaged in diabetic rats, the thirdorder mesenteric arteries were isolated and their isometric force TABLE 1 Body weight, blood glucose, and glycated hemoglobin levels of rats in control, DM, and metformin groups Values are the mean 6 S.E.M. and were analyzed by one-way ANOVA with Bonferroni’s post test (n = 6 rats).

Body weight (g) Blood glucose (mM) HbA1c (%)

Control

DM

Metformin

515 6 15 6.33 6 1.33 9.35 6 0.81

351 6 12** 23.13 6 1.36** 18.57 6 0.53**

278.3+7.2## 13.37 6 6.33## 15.06 6 0.86##

**P , 0.01 versus control rats;

##

P , 0.01 versus DM rats.

Metformin Restores AGE-Induced Vasodilative Dysfunction

to 10210∼1025 M ACh was measured. Figure 1A illustrates the typical records of concentration-dependent relaxations induced by ACh in PE-precontracted mesenteric arteries from control rats, diabetic rats (DM), and metformin-treated diabetic rats. As shown in Fig. 1C, diabetes significantly reduced the relaxation in response to ACh in mesenteric arteries (n 5 6, P , 0.01 vs. control), whereas metformin treatment restored the adverse condition (n 5 6, P , 0.05, P , 0.01 vs. DM). However, the relaxation in response to SNP was not affected (Fig. 1B). Incubation with 100 mM L-NAME and 10 mM INDO to block NOS and COX depressed the relaxant response to ACh in arteries from all rats, but the relaxation still decreased in mesenteric arteries of diabetic rats compared with control rats, and treatment with metformin partly reversed this decrease (Fig. 1D; n 5 6, P , 0.01 vs. control, P , 0.01 vs. DM). These results demonstrate that metformin restored the impaired EDHF-mediated relaxation in small arteries of diabetic rats.

583

To determine the role of IKCa and SKCa in impaired EDHFtype relaxation in diabetes, we examined responses to ACh in the presence of TRAM-34 (IKCa blocker) or UCL-1684 (SKCa blocker) in addition to NOS and COX inhibition (L-NAME 1 INDO). Compared with arteries of control rats, IKCa- and SKCa-mediated relaxation decreased in arteries of DM rats (Fig. 1, E and G; n 5 6, P , 0.01 vs. control). Metformin treatment partly reversed the decrease of IKCa- and SKCa-mediated relaxation (Fig. 1, F and H; n 5 6, P , 0.05, P , 0.01 vs. DM). Effect of Metformin on IKCa- and SKCa-Mediated Relaxation in Small Arteries Incubated with AGEs. We then examined the role of metformin in AGE-induced impairment of EDHF-type relaxation in small arteries. The thirdorder mesenteric arteries from normal rats were incubated without or with 200 mg·ml21 BSA, 200 mg·ml21 AGE-BSA, or 1 mM metformin plus 200 mg·ml21 AGE-BSA for 3 hours. BSA incubation did not affect the response of arteries to ACh, but AGE-BSA incubation attenuated it. Metformin preincubation

Fig. 1. Effect of metformin on IKCa- and SKCa-mediated ACh-induced relaxation in small arteries of diabetic rats. (A) Typical records showing concentration-dependent relaxations induced by ACh in PE-precontracted mesenteric arteries from control rats, diabetic rats (DM), and metformintreated diabetic rats (dots indicate ACh concentration from 10210 to 1025 M). (B–H) Cumulative concentration-response curves for SNP (B) and ACh in the absence (C) or presence of L-NAME+INDO (D), L-NAME+INDO+TRAM-34 (E and F), and L-NAME+INDO+UCL-1684 (G and H) in endotheliumintact mesenteric arteries isolated from control rats, diabetic rats, and metformin-treated diabetic rats. In each group of experiments, mesenteric arteries were precontracted with 1 mM PE. Results are shown as the mean 6 S.E.M. n = 6 rats, one-way ANOVA with Tukey’s post test. **P , 0.01 vs. control; #P , 0.05, ##P , 0.01 vs. DM).

584

Zhao et al.

inhibited the adverse effect of AGEs (Fig. 2B; n 5 6, P , 0.01 vs. BSA; P , 0.05, P , 0.01 vs. AGE-BSA). However, the relaxation in response to SNP was not affected by AGE-BSA incubation (Fig. 2A). When incubating arteries with L-NAME and INDO, AGE-BSA also decreased the relaxant response to ACh, and metformin restored the adverse condition (Fig. 2C; n 5 6, P , 0.05, P , 0.01 vs. BSA; P , 0.05, P , 0.01 vs. AGE-BSA). Effects of metformin on AGE-induced impairment of IKCaand SKCa-mediated relaxation in the third-order mesenteric arteries from normal rats were further investigated. As shown in Fig. 2 (D and G), there was no difference between the control and BSA groups. Compared with incubation with BSA, 3-hour incubation with 200 mg·ml21 AGE-BSA reduced both IKCa- and SKCa-mediated relaxation in response to ACh (Fig. 2, E and H; n 5 6, P , 0.05, P , 0.01 vs. BSA). Metformin treatment also incompletely reversed this damage (Fig. 2, F and I; P , 0.05, P , 0.01 vs. AGE-BSA). To confirm this conclusion, NS309, an opener of both IKCa and SKCa channels, was then used to induce vasorelaxation in

small arteries. Figure 3 displays the cumulative concentrationresponse curves to NS309 in endothelium-intact mesenteric arteries from normal rats. There was no difference in the relaxation in response to NS309 between control and BSAincubated arteries (Fig. 3, A and D). AGE-BSA incubation significantly reduced the relaxant response to NS309 in arteries treated without or with 10 mM TRAM-34 (Fig. 3B; n 5 6, P , 0.05, P , 0.01 vs. BSA) or 0.1 mM UCL-1684 (Fig. 3E; n 5 6, P , 0.01 vs. BSA). The adverse effects were partially revived by preincubation with 1 mM metformin (Fig. 3, C and F; n 5 6, P , 0.05, P , 0.01 vs. AGE-BSA). These results further confirm that metformin restored IKCa- and SKCa-mediated relaxation in small arteries damaged by AGEs. Effect of Metformin on IKCa and SKCa Channel Expression in HUVECs Incubated with AGEs. According to our previous study, AGEs reduced the expression of KCa3.1 and KCa2.3 channels in dose-dependent and time-dependent manners (Zhao et al., 2014). We therefore investigated whether the effect of metformin on IKCa- and SKCa-mediated relaxation

Fig. 2. Effect of metformin on IKCa- and SKCa-mediated ACh-induced relaxation in small arteries incubated with AGEs. (A–I) Cumulative concentrationresponse curves for SNP (A) and ACh in the absence (B) or presence of L-NAME+INDO (C), L-NAME+INDO+TRAM-34 (D–F), and L-NAME+INDO+UCL1684 (G–I) in endothelium-intact mesenteric arteries isolated from normal rats incubated without (control) or with 200 mg·ml21 BSA (BSA), 200 mg·ml21 AGE-BSA (AGE-BSA), or 200 mg·ml21 AGE-BSA plus 1 mM metformin for 3 hours. In each group of experiments, mesenteric arteries were precontracted with 1 mM PE. (n = 6 rats, one-way ANOVA with Tukey’s post test, *P , 0.05 and **P , 0.01 vs. BSA; #P , 0.05 and ##P , 0.01 vs. AGE-BSA).

Metformin Restores AGE-Induced Vasodilative Dysfunction

585

Fig. 3. Effect of metformin on IKCa- and SKCa-mediated NS309-induced relaxation in small arteries incubated with AGEs. (A–F) Cumulative concentration-response curves for NS309 in the absence or presence of TRAM-34 (A–C) or UCL-1684 (D–F) in endothelium-intact mesenteric arteries from normal rats incubated without (control) or with 200 mg·ml21 BSA (BSA), 200 mg·ml21 AGE-BSA (AGE-BSA), or 200 mg·ml21 AGE-BSA plus 1 mM metformin for 3 hours. In each group of experiments, mesenteric arteries were precontracted with 1 mM PE. (n = 6 rats, one-way ANOVA with Tukey’s post test, *P , 0.05 and **P , 0.01 vs. BSA; #P , 0.05 and ##P , 0.01 versus AGE-BSA).

is related to its regulation of KCa3.1 and KCa2.3 expression. HUVECs were cultured in the medium without (control) or with 200 mg·ml21 BSA, 200 mg·ml21 AGE-BSA, and AGE-BSA plus a different dose of metformin (1026, 1027, 1028 M) for 48 hours to determine the change in the expression level of KCa3.1 and KCa2.3 channels by Western blot. Incubation with a different dose of metformin and 200 mg·ml21 BSA for 48 hours did not affect the protein expression of KCa3.1 and KCa2.3 channels in HUVECs, whereas treatment with 200 mg·ml21 AGE-BSA for 48 hours decreased the protein expression of KCa3.1 and KCa2.3 by 35.52 and 49.48%, respectively (Fig. 4, A– D; n 5 5, P , 0.01 vs. BSA). Incubation with 200 mg·ml21 AGEBSA plus 1026, 1027, or 1028 M metformin for 48 hours attenuated the decrease of protein level, and the difference was significant at 1026 and 1027 M (Fig. 4, C and D; n 5 5, P , 0.01 vs. AGE-BSA). These results demonstrated that metformin restored the downregulation of IKCa and SKCa channels induced by AGEs. We also observed the effect of short-term incubation of metformin on KCa channel expression in HUVECs. The results showed that incubation with 200 mg·ml21 BSA, 200 mg·ml21 AGE-BSA, and metformin (1026 M) plus AGE-BSA for 3 hours did not change the expression of either KCa3.1 or KCa2.3 channels significantly in cultured HUVECs (shown in Supplemental Fig. 1; n 5 5, P . 0.05 vs. control), suggesting that regulation of KCa channel expression is not involved in the effect of short-term incubation with AGE-BSA and metformin. Effect of Metformin on IKCa and SKCa Currents Inhibited by AGEs in Cultured HUVECs. We then investigated whether short-term incubation with AGE-BSA and metformin influences KCa channel activity. Figure 5A shows

the membrane currents recorded in HUVECs with 300-ms voltage steps to between 2100 mV and 160 mV from a holding potential of 270 mV (inset). The membrane current with weak inward rectification at positive potentials was sensitive to the inhibition by IKCa channel blocker TRAM-34 (1 mM) and SKCa channel blocker UCL-1684 (0.1 mM) in a representative cell incubated with 200 mg·ml21 BSA for 3 hours (Fig. 5A, upper panel). Similar results were obtained in the other four cells. These results suggest that both IKCa and SKCa channels are predominantly expressed in cultured HUVECs. To determine whether the membrane currents were affected by AGEs, the cells were incubated with the medium containing 200 mg·ml21 AGE-BSA for 3 hours. The membrane current was significantly decreased in cells treated with AGE-BSA, and the sensitivity to TRAM-34 (1 mM) or UCL-1684 (0.1 mM) was also weakened by AGE-BSA (Fig. 5A, middle panel). Pretreatment with metformin (1 mM) partly reversed the inhibitory effects of AGE-BSA on the currents (Fig. 5A, lower panel). Figure 5B illustrates the mean values of current-voltage (I-V) relationships of TRAM34–sensitive current obtained by digital subtraction of the current before TRAM-34 from the current after TRAM-34 application in cells with 200 mg·ml21 BSA (control) or 200 mg·ml21 AGE-BSA and 1 mM metformin plus AGE-BSA. The I-V curves of TRAM-34–sensitive current showed a weak inward rectification. The current density at 260 to 160 mV was smaller in cells treated with AGE-BSA (n 5 5, P , 0.05 or P , 0.01 vs. control). The decreased current by AGE-BSA was partly antagonized by coincubation of metformin and AGEBSA (n 5 5, P , 0.05 or P , 0.01 vs. AGE-BSA alone). These results indicate that the decrease of IKCa current by AGE-BSA is upregulated by metformin in cultured HUVECs.

586

Zhao et al.

Fig. 4. Effect of metformin on downregulation of K Ca 3.1 and K Ca 2.3 channels induced by AGEs in cultured HUVECs. (A) Representative Western blot images and quantitative analysis of KCa3.1 channel protein levels in HUVECs treated without (control) or with different doses of metformin (1028∼1026 M) for 48 hours. (B) Representative Western blot images and quantitative analysis of KCa2.3 channel protein levels in HUVECs treated with the interventions as described in (A). (C) Representative Western blot images and quantitative analysis of KCa3.1 channel protein levels in HUVECs treated without (control) or with 200 mg·ml21 BSA (BSA), 200 mg·ml21 AGE-BSA (AGE-BSA), and 200 mg·ml21 AGE-BSA plus different doses of metformin (1028∼1026 M) for 48 hours. (D) Representative Western blot images and quantitative analysis of KCa2.3 channel protein levels in HUVECs treated with the interventions as described in (C). (n = 5, one-way ANOVA with Bonferroni’s post test, **P , 0.01 vs. BSA; ##P , 0.01 vs. AGE-BSA).

Figure 5C displays the I-V relationships of mean values of UCL-1684–sensitive current obtained by digitally subtracting the current before UCL-1684 from the current after UCL1684 application in cells with 200 mg·ml21 BSA (control) or 200 mg·ml21 AGE-BSA and metformin (1 mM) plus AGE-BSA. The current density at 110 to 160 mV was remarkably decreased in cells treated with AGE-BSA (n 5 5, P , 0.05 or P , 0.01 vs. control) and almost fully restored by metformin plus AGE-BSA (n 5 5, P , 0.05 or P , 0.01 vs. AGE-BSA alone). These results indicate that the inhibition of SKCa current by AGE-BSA is upregulated by metformin in cultured HUVECs. Involvement of ROS in the Effect of Metformin on Regulation of KCa Channels in HUVECs. We further explored whether intracellular ROS generation detected by DCF-DA and DHE was influenced by metformin. Compared with unstimulated and BSA-treated cells, ROS activity in HUVECs treated by 200 mg·ml21 AGE-BSA for 24 hours increased remarkably as indicated by fluorescence intensity of both fluorescent probes (Fig. 6, A–C; n 5 8, P , 0.01 vs. BSA). Metformin (1 mM) treatment reversed this condition (Fig. 6, A–C; n 5 8, P , 0.01 vs. AGE-BSA). To mimic the impairment of ROS, we used 40 mM H2O2 to treat HUVECs preincubated with or without different doses of

metformin (1026, 1027, 1028 M) for 48 hours. Incubation with 40 mM H2O2 for 48 hours reduced the expression of the KCa2.3 channel by 64.08% and expression of the KCa3.1 channel by 55.98%, respectively (Fig. 6, D and E; n 5 5, P , 0.01 vs. control). Incubation with 40 mM H2O2 plus 1026, 1027, or 1028 M metformin for 48 hours attenuated the decreasing of protein level, and the difference was significant at 1026 and 1027 M (Fig. 6, D and E; n 5 5, P , 0.05 or P , 0.01 vs. AGEBSA). These results demonstrated that metformin restored the downregulation of IKCa and SKCa channels induced by H2O2, which suggested that inhibiting the impairment of ROS on HUVECs is likely related to the beneficial effect of metformin on regulation of KCa channels. Effect of Metformin on AMPK Activation under AGEInduced Pathologic Conditions. To evaluate whether the beneficial effect of metformin on regulation of KCa channels in HUVECs is related to the activation of the AMPK pathway, we performed a Western blot analysis with a specific antibody against phosphorylated Thr172 of AMPK, which is reported to be essential for the AMPK activity, in HUVECs treated with 200 mg·ml21 BSA, 200 mg·ml21 AGE-BSA, AGE-BSA plus 1026∼1023 M metformin, 6,7-dihydro-4-hydroxy-3-(29-hydroxy [1,19-biphenyl]-4-yl)-6-oxo-thieno[2,3-b]pyridine-5-carbonitrile (Ab120335), or Compound C for 3 hours. As shown in Fig. 7A,

Metformin Restores AGE-Induced Vasodilative Dysfunction

587

Fig. 5. Effect of metformin on IKCa and SKCa currents inhibited by AGEs in cultured HUVECs. (A) Membrane currents recorded with the protocol as shown in the inset in cells treated with BSA (200 mg·ml21), AGE-BSA (200 mg·ml21), and AGE-BSA plus metformin (1 mM) for 3 hours before and after application of TRAM-34 (1 mM) and UCL-1684 (0.1 mM). (B) I-V relationships of the mean values of TRAM-34–sensitive current obtained by digital subtraction of the current before TRAM-34 from the current after TRAM-34 application (n = 5 for each groups). (C) I-V relationships of the mean values of UCL-1684–sensitive current obtained by digital subtraction of the current before UCL-1684 from the current after UCL-1684 application (n = 5 for each groups).

the relative level of p-AMPK Thr172 was reduced in cells incubated with AGE-BSA (n 5 5, P , 0.05 vs. BSA), which was antagonized by coincubation of 10 mM Ab120335, an AMPK agonist (n 5 5, P , 0.05 vs. AGE-BSA), but not 1026 M metformin, the dose we used in this study in vitro. Only more than 1024 M metformin significantly reversed the decrease of p-AMPK level by AGE-BSA (Fig. 7B; n 5 5, P , 0.05, P , 0.01 vs. AGE-BSA), and the effect of 1023 M metformin was countered by 10 mM Compound C (Fig. 7B; n 5 5, P , 0.05 vs. AGE-BSA plus 1023 M metformin), a specific AMPK inhibitor. These results indicated that high concentrations of metformin are required to activate AMPK under AGE-induced pathologic conditions.

Discussion In the present study, we found that 1) long-term treatment with metformin improved the impairment of IKCa- and SKCamediated relaxation in the third-order mesenteric arteries

from rats with STZ-induced diabetes, in which restoring IKCa and SKCa channel expression downregulated by AGEs through inhibiting ROS increase in endothelial cells, at least in part, was involved; and 2) short-term treatment with metformin reversed the damage of IKCa- and SKCa-mediated relaxation induced by short-term AGE incubation in small mesenteric arteries through activation of IKCa and SKCa channels without changing the channel expression. Metformin is one of the most commonly used therapeutic drugs for the treatment of T2DM. It not only lowers blood glucose concentration but also inhibits adipose tissue lipolysis, reduces circulating levels of free fatty acids, decreases formation of AGEs, and improves insulin sensitivity (Kirpichnikov et al., 2002). In addition, metformin improves lipid profiles and lowers blood pressure in both patients and animals with impaired glucose tolerance and T2DM (Verma et al., 1994; Kirpichnikov et al., 2002). It has been shown that metformin also has vasculoprotective effects (Katakam et al., 2000; Mather et al., 2001; Kirpichnikov et al., 2002; Iida et al.,

588

Zhao et al.

Fig. 6. Involvement of ROS in the effect of metformin on regulation of KCa channels in HUVECs. (A) Image showing ROS generation detected by DCF-DA (green) and DHE (red) in cultured HUVECs treated without (control) or with 200 mg·ml21 BSA, 200 mg·ml 2 1 AGE-BSA, or 200 mg·ml 2 1 AGE-BSA plus 1 mM metformin for 24 hours. Magnification = 200. (B and C) quantitative analysis of DCF-DA (B) and DHE (C) positive staining in HUVECs treated with the interventions as described in (A) (n = 8, oneway ANOVA with Bonferroni’s post test, **P , 0.01 vs. BSA; ##P , 0.01 vs. AGE-BSA). (D and E) Representative Western blot images and quantitative analysis of KCa3.1 channel (D) and KCa2.3 channel (E) protein levels in HUVECs treated without (control) or with 40 mM H2O2, 40 mM H2O2 plus different doses of metformin (1028∼1026 M) for 48 hours (n = 5, one-way ANOVA with Bonferroni’s post test, **P , 0.01 vs. control; #P , 0.05 and ##P , 0.01 vs. H2O2).

2003), which are largely independent of its well known antihyperglycemic action. In overweight T2DM patients, metformin use is associated with a decrease in macrovascular morbidity and mortality (UK Prospective Diabetes Study (UKPDS) Group, 1998). Metformin improved the vascular function in obese nondiabetic rats (Lobato et al., 2012) and insulin-resistant rats (Katakam et al., 2000). In the present study, we used STZ-treated rats with a 3-month high-fat and high-glucose diet as an animal model of diabetes. The diabetes model was characterized by an increase of fasting blood glucose, serum cholesterol, glycosylated hemoglobin, blood serum insulin, and mean artery pressure, as well as reduced body weight, as previously described (Wang et al., 2010; Su et al., 2011), which was similar to the state of patients with T2DM. Our data showed that, after 3-month treatment with metformin, the impaired ACh-induced endothelium-dependent relaxation, especially the EDHF-type relaxation, in small mesenteric arteries was reversed (Fig. 1, C and D), which further confirmed its vasculoprotective effect.

Despite the large number of studies published, it remains unclear which mechanisms underlie the vasculoprotective actions of metformin. In resistance arteries, the contribution of NO appears to be less important than in conduit vessels, and EDHF appears to play a major role in regulating tissue blood flow (Félétou and Vanhoutte, 2009). Findings from IKCa- and/or SKCa-deficient mice indicate that endothelial IKCa and SKCa are indeed fundamental components of the EDHF-signaling pathway in vivo (Grgic et al., 2009). There are reports indicating that impaired EDHF-mediated relaxations in small mesenteric arteries in animal models of diabetes involve endothelial KCa channels (McPherson, 1992; Morikawa et al., 2005). In the present study, using selective blockers of IKCa and SKCa, TRAM-34 and UCL-1684, we found that metformin restored the impaired ACh-induced EDHFmediated vasodilation which was related to its beneficial effect on IKCa- and SKCa-mediated relaxation (Fig. 1, E–H). AGE is one of the key pathogenic factors of diabetic vascular complications (Yamagishi et al., 2005; Rahman et al., 2007;

Metformin Restores AGE-Induced Vasodilative Dysfunction

589

Fig. 7. Effect of metformin on the p-AMPK level under AGE-induced pathologic conditions. (A) Representative Western blot images and relative levels of p-AMPK Thr172 (to total AMPK) in HUVECs treated without (control) or with 200 mg·ml21 BSA, 200 mg·ml21 AGE-BSA, and AGE-BSA plus 10 mM Ab120335 or 1 mM metformin for 3 hours. (B) Representative Western blot images and relative levels of p-AMPK Thr172 (to total AMPK) in HUVECs treated with 200 mg·ml21 BSA, 200 mg·ml21 AGE-BSA, AGE-BSA plus different doses of metformin (1025∼1023 M), or 1023 M metformin and 10 mM Compound C for 3 hours. (n = 5, one-way ANOVA with Bonferroni’s post test, *P , 0.05 vs. BSA; #P , 0.05, ##P , 0.01 vs. AGE-BSA; &P , 0.05 vs. AGE-BSA+1023 M metformin).

Gao et al., 2008; Negre-Salvayre et al., 2009). Our previous study showed that short-term exposure to AGE-BSA impaired EDHF-mediated relaxations in rat mesenteric arteries (Zhao et al., 2014). In this study, we also used the same method to mimic the direct effect of AGEs and observe the influence of metformin on AGE-induced impairment in rat mesenteric arteries. The results showed that direct incubation of isolated mesenteric arteries with metformin for 3 hours significantly reversed AGE-impaired IKCa- and SKCa-mediated AChinduced EDHF-type relaxation (Fig. 2). NS309 is a positive modulator of both SKCa and IKCa channels with no effect on big-conductance Ca21-activated potassium channels (Si et al., 2006). In high concentrations, NS309 has been shown to inhibit the activity of the voltage-dependent calcium channels causing smooth muscle relaxation (Morimura et al., 2006). In the present study, NS309 was applied to confirm the effect of metformin on IKCa- and SKCa-mediated vasodilatation (Fig. 3), and the maximum concentration was 1 mM, which is well below the IC50 (10.6 6 1.1 mM) for inhibiting voltagedependent calcium channels previously found in tissue preparations. We did not observe the direct effect of metformin on vasodilation of rat mesenteric arteries. Previous doseresponse experiments have demonstrated that, in rat tail and mesenteric arteries, metformin-induced vasodilation only occurred at very high concentrations (.10 mM) (Chen et al., 1997; Katakam et al., 2000). Metformin is a guanidine compound that is structurally related to aminoguanidine, and has been reported to have a potential effect on the inhibition of glycation reactions (Tanaka et al., 1997; Ruggiero-Lopez et al., 1999; Kiho et al., 2005; Brown et al., 2006). Metformin therapy reduced the atherogenic AGE molecules in the serum of women with polycystic ovary syndrome (Diamanti-Kandarakis et al., 2007). Our data showed that metformin attenuated the increase of HbA1c in diabetic rats (Table 1), the level of which is positively correlated with the serum level of AGEs (Jakus et al., 2014). In vitro treatment of arterioles with 1 mM metformin also significantly reversed the impaired IKCa- and

SKCa-mediated relaxation of isolated mesenteric arteries induced by direct AGE-BSA incubation (Figs. 2 and 3), suggesting that there should be other mechanisms besides its inhibition of AGE generation. The results from our Western blot analysis showed that 3-hour incubation with 200 mg·ml21 AGE-BSA or 1 mM metformin plus AGE-BSA did not change the expression of either KCa3.1 or KCa2.3 channels (Supplemental Fig. 1). Patch voltage-clamp analysis showed that both TRAM-34–sensitive current (IKCa channel current) and UCL-1684–sensitive current (SKCa channel current) were decreased by 3-hour incubation with AGE-BSA (200 mg·ml21) and reversed by coincubation with metformin (1 mM) and AGE-BSA (Fig. 5). These evidences suggest that the shortterm effect of AGEs and metformin on vascular relaxation is not related to the alteration of KCa channel expression, but to the change of channel activity. However, 48-hour incubation with metformin reversed AGE-induced downregulation of KCa3.1 and KCa2.3 (Fig. 4), which is in line with our previous study (Zhao et al., 2014). These results indicated that alterations of the KCa channel both in protein expression and channel activity may be responsible for the beneficial effect of metformin on IKCa- and SKCa-mediated relaxation in vivo. However, we cannot exclude other mechanisms concerning the inhibitory effect of metformin on receptor for AGEs (RAGE) and downstream processes (Ouslimani et al., 2007; Ishibashi et al., 2012a,b). In endothelial cells, the interaction of AGEs with RAGE can activate complex signaling pathways, causing increased generation of ROS (Goldin et al., 2006). Enhanced generation of ROS and changes in antioxidant capacity are thought to be the etiology of chronic diabetic complications (Matheus et al., 2013). ROS have been proposed to be regulators of K1 channel function in various tissues (Kourie, 1998). In our previous study, diabetic rats treated with the antioxidant alpha lipoic acid, which can quench ROS, showed the restored IKCa- and SKCa-mediated relaxation in mesenteric arteries (Zhao et al., 2013). Alpha lipoic acid also attenuated AGE-induced impairment of IKCa- and SKCa-mediated relaxation in mesenteric

590

Zhao et al.

arteries, and reversed AGE-induced downregulation of KCa3.1 and KCa2.3 channels in cultured HUVECs (Zhao et al., 2014). These previous studies confirmed the harmful role of ROS in IKCa- and SKCa-mediated relaxation. Metformin exhibits antioxidant properties that contribute to the vascular protective effects observed in several epidemiologic studies (UK Prospective Diabetes Study (UKPDS) Group, 1998). Treatment with metformin restores endothelial function through inhibiting endoplasmic reticulum stress and oxidative stress in obese diabetic mice (Cheang et al., 2014). In cultured bovine aortic endothelial cells, the intracellular antioxidant properties of metformin (10 mM) result in the inhibition of cell expression of both RAGE and lectin-like oxidized receptor 1, possibly through a modulation of redox-sensible nuclear factors such as NF-kB (Ouslimani et al., 2007). Our present data showed that metformin reversed the enhanced ROS generation induced by AGEs (Fig. 5, A–C) and the reduced expression of KCa3.1 and KCa2.3 by H2O2 (Fig. 6). These evidences indicated that the effect of metformin in this study might be related to inhibition of AGE-induced oxidation. Several reports (Zou et al., 2004; Towler and Hardie, 2007) have provided evidence that metformin (100 mM to 10 mM) activates AMPK. Our results also showed that only a high dose of metformin (.100 mM) significantly reversed the decrease of p-AMPK level by AGE-BSA (Fig. 7), indicating that high concentrations of metformin are required to activate AMPK under AGE-induced pathologic conditions. Therefore, it seems that the AMPK pathway is not involved in the observed effects of metformin at 1 mM in this study. In summary, the present study demonstrated that metformin treatment improved the vascular function in rats with STZ-induced diabetes by reversing AGE-induced impairment of IKCa- and SKCa-mediated relaxation. Our findings support the beneficial effects of metformin previously demonstrated in large intervention studies in type 2 diabetic patients. Authorship Contributions

Participated in research design: Zhao, Deng. Conducted experiments: Zhao, Y. Wang, Yang, Guo. Performed data analysis: Zhao, Y. Wang. Wrote or contributed to the writing of the manuscript: Zhao, N.-P. Wang, Deng. References Bennett WL, Maruthur NM, Singh S, Segal JB, Wilson LM, Chatterjee R, Marinopoulos SS, Puhan MA, Ranasinghe P, and Block L et al. (2011) Comparative effectiveness and safety of medications for type 2 diabetes: an update including new drugs and 2-drug combinations. Ann Intern Med 154:602–613. Brähler S, Kaistha A, Schmidt VJ, Wölfle SE, Busch C, Kaistha BP, Kacik M, Hasenau AL, Grgic I, and Si H et al. (2009) Genetic deficit of SK3 and IK1 channels disrupts the endothelium-derived hyperpolarizing factor vasodilator pathway and causes hypertension. Circulation 119:2323–2332. Brown BE, Mahroof FM, Cook NL, van Reyk DM, and Davies MJ (2006) Hydrazine compounds inhibit glycation of low-density lipoproteins and prevent the in vitro formation of model foam cells from glycolaldehyde-modified low-density lipoproteins. Diabetologia 49:775–783. Burnham MP, Johnson IT, and Weston AH (2006) Impaired small-conductance Ca21-activated K1 channel-dependent EDHF responses in Type II diabetic ZDF rats. Br J Pharmacol 148:434–441. Busse R, Edwards G, Félétou M, Fleming I, Vanhoutte PM, and Weston AH (2002) EDHF: bringing the concepts together. Trends Pharmacol Sci 23:374–380. Cheang WS, Tian XY, Wong WT, Lau CW, Lee SS, Chen ZY, Yao X, Wang N, and Huang Y (2014) Metformin protects endothelial function in diet-induced obese mice by inhibition of endoplasmic reticulum stress through 59 adenosine monophosphate-activated protein kinase-peroxisome proliferator-activated receptor d pathway. Arterioscler Thromb Vasc Biol 34:830–836. Chen XL, Panek K, and Rembold CM (1997) Metformin relaxes rat tail artery by repolarization and resultant decreases in Ca21 influx and intracellular. Ca21 J Hypertens 15:269–274. Consoli A, Gomis R, Halimi S, Home PD, Mehnert H, Strojek K, and Van Gaal LF (2004) Initiating oral glucose-lowering therapy with metformin in type 2 diabetic

patients: an evidence-based strategy to reduce the burden of late-developing diabetes complications. Diabetes Metab 30:509–516. de Jager J, Kooy A, Schalkwijk C, van der Kolk J, Lehert P, Bets D, Wulffelé MG, Donker AJ, and Stehouwer CD (2014) Long-term effects of metformin on endothelial function in type 2 diabetes: a randomized controlled trial. J Intern Med 275: 59–70. Diamanti-Kandarakis E, Alexandraki K, Piperi C, Aessopos A, Paterakis T, Katsikis I, and Panidis D (2007) Effect of metformin administration on plasma advanced glycation end product levels in women with polycystic ovary syndrome. Metabolism 56:129–134. Dora KA, Gallagher NT, McNeish A, and Garland CJ (2008) Modulation of endothelial cell KCa3.1 channels during endothelium-derived hyperpolarizing factor signaling in mesenteric resistance arteries. Circ Res 102:1247–1255. Eichler I, Wibawa J, Grgic I, Knorr A, Brakemeier S, Pries AR, Hoyer J, and Köhler R (2003) Selective blockade of endothelial Ca21-activated small- and intermediateconductance K1-channels suppresses EDHF-mediated vasodilation. Br J Pharmacol 138:594–601. Félétou M and Vanhoutte PM (2009) EDHF: an update. Clin Sci (Lond) 117:139–155. Gao X, Zhang H, Schmidt AM, and Zhang C (2008) AGE/RAGE produces endothelial dysfunction in coronary arterioles in type 2 diabetic mice. Am J Physiol Heart Circ Physiol 295:H491–H498. Goldin A, Beckman JA, Schmidt AM, and Creager MA (2006) Advanced glycation end products: sparking the development of diabetic vascular injury. Circulation 114: 597–605. Grgic I, Kaistha BP, Hoyer J, and Köhler R (2009) Endothelial Ca1-activated K1 channels in normal and impaired EDHF-dilator responses—relevance to cardiovascular pathologies and drug discovery. Br J Pharmacol 157:509–526. Hartge MM, Kintscher U, and Unger T (2006) Endothelial dysfunction and its role in diabetic vascular disease. Endocrinol Metab Clin North Am 35:551–560 viii–ix. Iida KT, Kawakami Y, Suzuki M, Shimano H, Toyoshima H, Sone H, Shimada K, Iwama Y, Watanabe Y, and Mokuno H et al. (2003) Effect of thiazolidinediones and metformin on LDL oxidation and aortic endothelium relaxation in diabetic GK rats. Am J Physiol Endocrinol Metab 284:E1125–E1130. Ishibashi Y, Matsui T, Takeuchi M, Yamagishi, and S (2012a) Metformin inhibits advanced glycation end products (AGEs)-induced renal tubular cell injury by suppressing reactive oxygen species generation via reducing receptor for AGEs (RAGE) expression. Horm Metab Res 44:891–895. Ishibashi Y, Matsui T, Takeuchi M, and Yamagishi S (2012b) Beneficial effects of metformin and irbesartan on advanced glycation end products (AGEs)-RAGEinduced proximal tubular cell injury. Pharmacol Res 65:297–302. Jakus V, Sándorová E, Kalninová J, and Krahulec B (2014) Monitoring of glycation, oxidative stress and inflammation in relation to the occurrence of vascular complications in patients with type 2 diabetes mellitus. Physiol Res 63:297–309. Katakam PV, Ujhelyi MR, Hoenig M, and Miller AW (2000) Metformin improves vascular function in insulin-resistant rats. Hypertension 35:108–112. Kiho T, Kato M, Usui S, and Hirano K (2005) Effect of buformin and metformin on formation of advanced glycation end products by methylglyoxal. Clin Chim Acta 358:139–145. Kirpichnikov D, McFarlane SI, and Sowers JR (2002) Metformin: an update. Ann Intern Med 137:25–33. Kourie JI (1998) Interaction of reactive oxygen species with ion transport mechanisms. Am J Physiol 275:C1–C24. Leo CH, Hart JL, and Woodman OL (2011) Impairment of both nitric oxide-mediated and EDHF-type relaxation in small mesenteric arteries from rats with streptozotocin-induced diabetes. Br J Pharmacol 162:365–377. Lobato NS, Filgueira FP, Hagihara GN, Akamine EH, Pariz JR, Tostes RC, Carvalho MH, and Fortes ZB (2012) Improvement of metabolic parameters and vascular function by metformin in obese non-diabetic rats. Life Sci 90:228–235. Mather KJ, Verma S, and Anderson TJ (2001) Improved endothelial function with metformin in type 2 diabetes mellitus. J Am Coll Cardiol 37:1344–1350. Matheus AS, Tannus LR, Cobas RA, Palma CC, Negrato CA, and Gomes MB (2013) Impact of diabetes on cardiovascular disease: an update. Int J Hypertens 2013: 653789 10.1155/2013/653789. McPherson GA (1992) Assessing vascular reactivity of arteries in the small vessel myograph. Clin Exp Pharmacol Physiol 19:815–825. Morikawa K, Matoba T, Kubota H, Hatanaka M, Fujiki T, Takahashi S, Takeshita A, and Shimokawa H (2005) Influence of diabetes mellitus, hypercholesterolemia, and their combination on EDHF-mediated responses in mice. J Cardiovasc Pharmacol 45:485–490. Morimura K, Yamamura H, Ohya S, and Imaizumi Y (2006) Voltage-dependent Ca21-channel block by openers of intermediate and small conductance Ca21activated K1 channels in urinary bladder smooth muscle cells. J Pharmacol Sci 100:237–241. Negre-Salvayre A, Salvayre R, Augé N, Pamplona R, and Portero-Otín M (2009) Hyperglycemia and glycation in diabetic complications. Antioxid Redox Signal 11: 3071–3109. Ouslimani N, Mahrouf M, Peynet J, Bonnefont-Rousselot D, Cosson C, Legrand A, and Beaudeux JL (2007) Metformin reduces endothelial cell expression of both the receptor for advanced glycation end products and lectin-like oxidized receptor 1. Metabolism 56:308–313. Rahman S, Rahman T, Ismail AA, and Rashid AR (2007) Diabetes-associated macrovasculopathy: pathophysiology and pathogenesis. Diabetes Obes Metab 9: 767–780. Ruggiero-Lopez D, Lecomte M, Moinet G, Patereau G, Lagarde M, and Wiernsperger N (1999) Reaction of metformin with dicarbonyl compounds. Possible implication in the inhibition of advanced glycation end product formation. Biochem Pharmacol 58:1765–1773. Sena CM, Matafome P, Crisóstomo J, Rodrigues L, Fernandes R, Pereira P, and Seiça RM (2012) Methylglyoxal promotes oxidative stress and endothelial dysfunction. Pharmacol Res 65:497–506.

Metformin Restores AGE-Induced Vasodilative Dysfunction Sheng JZ, Ella S, Davis MJ, Hill MA, and Braun AP (2009) Openers of SKCa and IKCa channels enhance agonist-evoked endothelial nitric oxide synthesis and arteriolar vasodilation. FASEB J 23:1138–1145. Si H, Heyken WT, Wölfle SE, Tysiac M, Schubert R, Grgic I, Vilianovich L, Giebing G, Maier T, and Gross V et al. (2006) Impaired endothelium-derived hyperpolarizing factor-mediated dilations and increased blood pressure in mice deficient of the intermediate-conductance Ca21-activated K1 channel. Circ Res 99:537–544. Su XL, Wang Y, Zhang W, Zhao LM, Li GR, and Deng XL (2011) Insulin-mediated upregulation of K(Ca)3.1 channels promotes cell migration and proliferation in rat vascular smooth muscle. J Mol Cell Cardiol 51:51–57. Tanaka Y, Iwamoto H, Onuma T, and Kawamori R (1997) Inhibitory effect of metformin on formation of advanced glycation end products. Curr Therap Res 58:693–697. Taylor MS, Bonev AD, Gross TP, Eckman DM, Brayden JE, Bond CT, Adelman JP, and Nelson MT (2003) Altered expression of small-conductance Ca21-activated K1 (SK3) channels modulates arterial tone and blood pressure. Circ Res 93:124–131. Towler MC and Hardie DG (2007) AMP-activated protein kinase in metabolic control and insulin signaling. Circ Res 100:328–341. UK Prospective Diabetes Study (UKPDS) Group (1998) Effect of intensive bloodglucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet 352:854–865. Vanhoutte PM, Shimokawa H, Tang EHC, and Feletou M (2009) Endothelial dysfunction and vascular disease. Acta Physiol (Oxf) 196:193–222. Verma S, Bhanot S, and McNeill JH (1994) Antihypertensive effects of metformin in fructose-fed hyperinsulinemic, hypertensive rats. J Pharmacol Exp Ther 271: 1334–1337. Vlassara H, Uribarri J, Cai W, and Striker G (2008) Advanced glycation end product homeostasis: exogenous oxidants and innate defenses. Ann N Y Acad Sci 1126: 46–52. Wang LP, Wang Y, Zhao LM, Li GR, and Deng XL (2013) Angiotensin II upregulates K(Ca)3.1 channels and stimulates cell proliferation in rat cardiac fibroblasts. Biochem Pharmacol 85:1486–1494.

591

Wang Y, Zhang HT, Su XL, Deng XL, Yuan BX, Zhang W, Wang XF, and Yang YB (2010) Experimental diabetes mellitus down-regulates large-conductance Ca21activated K1 channels in cerebral artery smooth muscle and alters functional conductance. Curr Neurovasc Res 7:75–84. Wigg SJ, Tare M, Tonta MA, O’Brien RC, Meredith IT, and Parkington HC (2001) Comparison of effects of diabetes mellitus on an EDHF-dependent and an EDHFindependent artery. Am J Physiol Heart Circ Physiol 281:H232–H240. Wu W, Sun HY, Deng XL, and Li GR (2013) EGFR tyrosine kinase regulates human small-conductance Ca21-activated K1 (hSKCa1) channels expressed in HEK-293 cells. Biochem J 452:121–129. Yamagishi S, Nakamura K, and Imaizumi T (2005) Advanced glycation end products (AGEs) and diabetic vascular complications. Curr Diabetes Rev 1:93–106. Zhao L, Wang Y, Ma X, Wang Y, and Deng X (2013) Oxidative stress impairs IKCaand SKCa-mediated vasodilatation in mesenteric arteries from diabetic rats. Nan Fang Yi Ke Da Xue Xue Bao 33:939–944. Zhao LM, Wang Y, Ma XZ, Wang NP, and Deng XL (2014) Advanced glycation end products impair K(Ca)3.1- and K(Ca)2.3-mediated vasodilatation via oxidative stress in rat mesenteric arteries. Pflugers Arch 466:307–317. Zhao LM, Zhang W, Wang LP, Li GR, and Deng XL (2012) Advanced glycation end products promote proliferation of cardiac fibroblasts by upregulation of KCa3.1 channels. Pflugers Arch 464:613–621. Zou MH, Kirkpatrick SS, Davis BJ, Nelson JS, Wiles WG, 4th, Schlattner U, Neumann D, Brownlee M, Freeman MB, and Goldman MH (2004) Activation of the AMP-activated protein kinase by the anti-diabetic drug metformin in vivo. Role of mitochondrial reactive nitrogen species. J Biol Chem 279:43940–43951.

Address correspondence to: Dr. Xiu-Ling Deng, Department of Physiology and Pathophysiology, School of Medicine, Xi’an Jiaotong University, 76 West Yanta Road, Xi’an, 710061, Shaanxi, China. E-mail: [email protected]