Ranolazine Increases �-Cell Survival and Improves GlucoseHomeostasis in Low-Dose Streptozotocin-Induced Diabetesin Mice

Yun Ning, Wei Zhen, Zhuo Fu, Jenny Jiang, Dongmin Liu, Luiz Belardinelli,and Arvinder K. DhallaDepartment of Biology, Gilead Sciences, Palo Alto, California (Y.N., J.J, L.B., A.K.D.); and Department of Human Nutrition,Foods and Exercise, Virginia Polytechnic Institute and State University, Blacksburg, Virginia (W.Z., Z.F., D.L.)

Received October 22, 2010; accepted January 10, 2011

ABSTRACTIn addition to its anti-ischemic and antianginal effects, rano-lazine has been shown to lower hemoglobin A1c (HbA1c) inpatients with coronary artery disease and diabetes. Thepresent study was undertaken to test the hypothesis thatranolazine lowers HbA1c because of improved glucose ho-meostasis in an animal model. Diabetes in mice was inducedby giving multiple low doses of streptozotocin. Ranolazinewas given twice daily via an oral gavage (20 mg/kg) for 8weeks. Fasting plasma glucose levels were significantlylower in the ranolazine-treated group (187 � 19 mg/dl) com-pared with the vehicle group (273 � 23 mg/dl) at 8 weeks.HbA1c was 5.8 � 0.4% in the vehicle group and 4.5 � 0.2%in the ranolazine-treated group (p � 0.05). Glucose disposalduring the oral glucose tolerance test (OGTT) and insulintolerance test were not different between the two groups;

however, during OGTT, peak insulin levels were significantly(p � 0.05) higher in ranolazine-treated mice. Mice treatedwith ranolazine had healthier islet morphology and signifi-cantly (p � 0.01) higher �-cell mass (69 � 2% per islet) thanthe vehicle group (50 � 5% per islet) as determined fromhematoxylin and eosin staining. The number of apoptoticcells was significantly (p � 0.05) less in the pancreas of theranolazine-treated group (14 � 2% per islet) compared withthe vehicle group (24 � 4% per islet). In addition, ranolazineincreased glucose-stimulated insulin secretion in rat andhuman islets in a glucose-dependent manner. These datasuggest that ranolazine may be a novel antidiabetic agentthat causes �-cell preservation and enhances insulin secre-tion in a glucose-dependent manner in diabetic mice.

IntroductionType 2 diabetes is a major public health problem affecting

more than 250 million people worldwide. The underlyingcauses of type 2 diabetes are a combination of impairment ininsulin-mediated glucose disposal (insulin resistance) and adefect in insulin secretion from pancreatic �-cells (Wajchen-berg, 2007). Type 2 diabetes is also a risk factor for cardio-vascular disease and frequently coexists with cardiovascularcomorbidities, including dyslipidemia and hypertension, andthus is a significant predictor of cardiovascular mortality(Cannon, 2008). It is reported that the presence of diabetes inpatients with coronary artery disease (CAD) doubles theirrisk for adverse cardiovascular events (Pepine et al., 2003;Bakris et al., 2004). The treatment options for diabetic pa-

tients with CAD currently have several safety concerns. Sev-eral clinical trials have shown that patients with CADtreated with �-blockers and calcium channel blockers haveconsistently and significantly higher rates of newly diag-nosed diabetes because of the prodiabetic effects of thesemedications (Pepine and Cooper-Dehoff, 2004; Wofford andKing, 2004; Dahlof et al., 2005).

Ranolazine is a novel anti-ischemic and antianginal drugthat reduces angina frequency and improves exercise perfor-mance in patients with chronic stable angina, an insidiousmanifestation of CAD (Chaitman et al., 2004; Stone et al.,2006). Ranolazine reduces myocardial ischemia by improvingsodium-calcium homeostasis via inhibition of the late phaseof the inward sodium current (late INa) during cardiac repo-larization. In addition to its antianginal effects, ranolazinehas been shown to reduce HbA1c in patients with CAD anddiabetes in two clinical studies (Timmis et al., 2006; Scirica etal., 2007; Chisholm et al., 2010). In the CARISA (Combina-

Article, publication date, and citation information can be found athttp://jpet.aspetjournals.org.

doi:10.1124/jpet.110.176396.

ABBREVIATIONS: CAD, coronary artery disease; HbA1c, hemoglobin A1c; FPG, fasting plasma glucose; OGTT, oral glucose tolerance test; PBS,phosphate-buffered saline; H&E, hematoxylin and eosin; GSIS, glucose-stimulated insulin secretion; FITC, fluorescein isothiocyanate; KRB,Krebs-Ringer-bicarbonate; GLP-1, glucagon-like peptide-1.

0022-3565/11/3371-50–58$20.00THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 337, No. 1Copyright © 2011 by The American Society for Pharmacology and Experimental Therapeutics 176396/3675363JPET 337:50–58, 2011 Printed in U.S.A.

50

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from

tion Assessment of Ranolazine In Stable Angina)trial (Tim-mis et al., 2006), ranolazine treatment for 12 weeks signifi-cantly reduced HbA1c levels by 0.7 � 0.18% regardless ofconcomitant insulin and/or oral hypoglycemic therapy. In theMERLIN-TIMI 36 (Metabolic Efficiency with Ranolazine forLess Ischemia in Non–ST-Elevation Acute Coronary Syn-drome Thrombolysis in Myocardial Infarction 36) trial, a0.6 � 0.2% reduction in HbA1c was observed in patientstreated with ranolazine in addition to standard antidiabetictherapy. Furthermore, in patients with moderate/severe hy-perglycemia, ranolazine treatment was associated with areduction (25.7 mg/dl) in fasting plasma glucose (FPG) after4 months of treatment (Scirica et al., 2007; Chisholm et al.,2010). However, no preclinical data have been reported, dem-onstrating either the antidiabetic properties of ranolazine orthe pharmacological basis for reduction of HbA1c observed inclinical trials.

Therefore, the objectives of the present study were 1) toreproduce the clinical findings with ranolazine in an animalmodel, 2) to correlate the HbA1c-lowering effect with glucoselevels, and 3) to understand the mechanisms underlying theantidiabetic effects of ranolazine. We determined the effectsof an 8-week treatment with ranolazine in a diabetic mousemodel induced with multiple low doses of streptozotocin. Wemonitored body weight and FPG weekly and HbA1c every 4weeks during 8 weeks of ranolazine treatment followed byoral glucose tolerance test (OGTT) and insulin tolerance testat the end of the treatment. Pancreatic islet �-cell mass andapoptosis rate were also determined. Our data show thatlong-term treatment with ranolazine in STZ-induced diabeticmice improves FPG and HbA1c levels and the insulin re-sponse during an OGTT. Moreover, STZ mice treated withranolazine had healthier islet morphology and significantlyhigher �-cell mass compared with the vehicle-treated group.The number of apoptotic islet cells was also significantly lessin pancreas from ranolazine-treated mice. In addition, weinvestigated the acute effect of ranolazine on GSIS in isolatedislets. Our data show that ranolazine increases GSIS in bothrat and human-pancreatic islets in a concentration- and glu-cose-dependent manner. Together, these data suggest thatlong-term treatment with ranolazine results in �-cell preser-vation and enhances insulin secretion, which could explainthe improvement in glucose and HbA1c levels in diabeticmice.

Materials and MethodsAnimals. All experimental procedures were performed under a

protocol approved by the Institutional Animal Care and Use Com-mittee of Gilead Palo Alto Inc. (Palo Alto) and in accordance with therecommendations set forth in the Guide for the Care and Use ofLaboratory Animals published by the National Research Council.Five-week-old male C57BL/6J mice were purchased from The Jack-son Laboratory (Bar Harbor, ME). Animals were housed five per cagein a room maintained on a 12-h light/dark cycle (lights on from 6:00AM to 6:00 PM) under constant temperature (22–25°C) with accessto food and water ad libitum.

Generation of Multiple Low-Doses of Streptozotocin Dia-betic Animal Model. After an initial acclimation period (7 days),mice were injected with streptozotocin (40 mg/kg i.p., dissolved inice- 0.025 M sodium citrate-buffered solution, pH 4.5, freshly maderight before injection) for 5 consecutive days. Fasting blood glucoselevels were monitored 2 days after STZ treatment. The diabetic mice

were then divided into STZ � vehicle and STZ � ranolazine groups(10 mice/group) based on body weight and blood glucose levels. Ageand sex-matched nondiabetic mice were used as “normal” controls(n � 3). For the next 8 weeks, mice were given either vehicle orranolazine (20 mg/kg in water p.o., twice daily). Plasma concentra-tions of ranolazine after an oral dose of 20 mg/kg in mice at 15, 30,and 60 min were determined to be 9.1 � 3.3, 7.4 � 3.0, and 6.1 � 2.8�M, respectively, which are within the clinically therapeutic concen-trations (5–10 �M) of the drug. Body weight and fasting blood glu-cose levels were monitored once a week. HbA1c levels were measuredusing a DCA 2000� clinical analyzer (Siemens Healthcare Diag-nostics, Deerfield, IL) at 0, 4, and 8 weeks of treatment. At the endof 8-week treatment, the OGTT and insulin tolerance test wereperformed.

Oral Glucose Tolerance Test. Mice were fasted for 4 h (8:00AM–12:00 PM) before the OGTT. At 0 min, a drop of blood (5 �l) wastaken via a tail nick before giving the oral glucose load. Mice werethen given a glucose load (2 g/kg) in distilled water. Blood samples(40 �l) were obtained at 15, 30, 60, 90, 120,180, and 240 min afterglucose load for the determination of blood glucose with a glucom-eter. Insulin levels were measured using a rat/mouse insulin en-zyme-linked immunosorbent assay kit for each sample. At the end ofthe experiment, the animals were euthanized, and the pancreasesfrom all mice were collected for morphological and immunohisto-chemical analysis.

Insulin Tolerance Test. Mice were fasted for 4 h (8:00 AM–12:00PM) and then given an intraperitoneal injection of insulin (1 U/kg) indistilled water. Blood samples were taken at 0, 15, 30, 60, 90, and120 min. Blood glucose levels were determined using a glucometer(Roche Diagnostics, IN).

Islet Morphometry. Pancreases were fixed in 10% formalin over-night and then embedded in paraffin. Sections (7 �m) were obtainedthroughout the pancreas and were deparaffinized and rehydratedsequentially in xylene, xylene/ethanol, and ethanol (100, 95, 80, 70,and 50%), and then the sections were placed in distilled water for 10min. Pancreatic sections were stained with hematoxylin and eosin(H&E) using standard protocols.

Immunohistochemistry. Pancreases from age and sex-matchedmice were stained as normal control. A series of pancreatic sections(7 �m thickness, n � 4) from normal, STZ � vehicle, and STZ �ranolazine groups were prepared, mounted on glass slides, andstained for insulin, glucagon, and apoptosis using fluorescent anti-bodies. In brief, the tissue sections were blocked with 10% donkeyserum in PBS buffer for 30 min and then washed with PBS. For thedetection of insulin and glucagon, a cocktail of primary antibodies(guinea pig anti-insulin, mouse antiglucagon) was added to the slidesand incubated at 4°C overnight. Slides were washed with PBS andthen incubated with a cocktail of secondary antibodies to the respec-tive primaries (donkey anti-guinea pig Cy3, donkey anti-mouseFITC) for 30 min at room temperature. Staining was preserved, andnuclei were identified by adding a drop of Vectashield mountingmedium (Vector Laboratories, Burlingame, CA) with 4�,6-diamidino-2-phenylindole to each tissue section before placing coverslip. Usingthis procedure, insulin is stained red, whereas glucagon appearsgreen and nuclei are stained blue. Similar procedure was used todetect apoptotic �-cells as determined by sequential labeling of thesections with activated caspase-3 antibody and donkey anti-rabbitFITC secondary antibody. All sections were viewed under fluorescentmicroscope, and the stained areas were digitally photographed at amagnification of 20. The images taken at different magnificationwere normalized using the standard ruler grade (S1 Finder Grati-cule, 68040; Electron Microscopy Science, Hatfield, PA). Analyses ofislet areas and entire section areas were performed using ImageJsoftware (National Institutes of Health, Bethesda, MD). At least foursections from each animal (n � 3 mice in normal group; n � 6 micefor STZ � vehicle and STZ � ranolazine groups) were analyzed.

Isolation and Culture of Pancreatic Islets. Pancreatic isletswere isolated from 8- to 12-week-old rats as described previously

Antidiabetic Effects of Ranolazine 51

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from

(Yang et al., 2004). In brief, animals were anesthetized with carbondioxide and euthanized. The abdomen was opened, and the commonbile duct was cannulated followed by infusion of 10 ml of collagenasesolution (0.5 mg/ml collagenase P in Hanks’ balanced salt solutionsupplemented with 1% bovine serum albumin and 0.0375% NaCO3).Afterward, the pancreases were removed and incubated in a waterbath at 37°C for 20 min with gentle shaking. Digested pancreaseswere then filtered and washed, and the islets were separated fromexocrine tissue by centrifugation over a discontinuous Histopaque1077 (Sigma-Aldrich, St. Louis, MO) gradient.

Human islets were obtained through the National Institutes ofHealth-supported Islet Cell Resource Centers and the Islet Distribu-tion Program at the Juvenile Diabetes Research Foundation, both ofwhich are administrated by Administrative and Bioinformatics Co-ordinating Center. The islet purity was 80 to 90%, and viability was80 to 97%. Before the experiment, the islets were maintained inCMRL-1066 medium (Mediatech, Manassas, VA) containing 10%FBS (HyClone Laboratories, Logan, UT).

Insulin Secretion. Isolated pancreatic islets were maintained incomplete RPMI 1640 medium overnight (Yang et al., 2004). Insulinsecretion assay was performed essentially as described previously(Liu et al., 2006). In brief, before the experiment, islets were prein-cubated in Krebs-Ringer-bicarbonate (KRB) buffer containing 3 mMglucose for 30 min, after which islets were washed and incubated intriplicate in 24-well plate (50 islets/well), in oxygenated KRB bufferwith 3 or 20 mM glucose in the presence of various concentrations ofranolazine or vehicle for 60 min at 37°C. In some experiments, isletswere also incubated with glucagon-like peptide-1 (GLP-1) (Sigma-Aldrich), which served as a positive control. Insulin in experimentalsamples was measured by an enzyme-linked immunosorbent assaykit (Mercodia, Winston-Salem, NC). Our preliminary experimentsshow that exposure of the islets to the compounds for 1 h had noeffect on total insulin or protein content. Therefore, all insulin se-cretion data in the present study are presented as insulin secretionper well.

Chemicals and Biological Reagents. Ranolazine is a pipera-zine derivative chemically described as 1-piperazineacetamide [N-(2,6-dimethylphenyl)-4-[2-hydroxy-3-(2-methoxyphenoxy)pro-pyl]-, (�)-]. Ranolazine was synthesized at Gilead Palo Alto Inc.For in vivo study, ranolazine was dissolved in 0.025 N HCl, pH4.0, and the solution was filtered with a 0.2 �M sterilizing filter(Millipore, Billerica, MA). Other agents were purchased from thefollowing sources: streptozotocin (S0130), human insulin (19278,275 U/ml), glucose (G-7528), and GLP-1 (G9416; Sigma-Aldrich);rat/mouse insulin enzyme-linked immunosorbent assay kit(EZRMI-13K; Millipore); Accu-Chek active diabetic test strips andglucometer (Roche Diagnostics); and HbA1c measure system(5035C, DCA systems; Siemens Healthcare Diagnostics). The an-tibodies were purchased from following sources: 10% donkey se-rum (Jackson ImmunoResearch Laboratories, West Grove, PA),guinea pig anti-insulin (Sigma I-8510, 1:500; Sigma-Aldrich),mouse anti-glucagon (Sigma G-2654, 1:100; Sigma-Aldrich), don-key anti-guinea pig Cy3 (706-165-148, 1:500; Jackson ImmunoRe-search Laboratories), donkey anti-mouse FITC (715-095-150,1:200; Jackson ImmunoResearch Laboratories), rabbit caspase-3antibody (1:50; BD Biosciences), and donkey anti-rabbit FITCsecondary antibody (1:100; Jackson ImmunoResearch Laborato-ries). For ex vivo experiments, stock solution of ranolazine wasmade in dimethyl sulfoxide and immediately frozen at 80°Cbefore use and then diluted with KRB buffer (129 mM NaCl, 4.8mM KCl, 1.2 mM MgSO4, 1.2 mM KH2PO4, 2.5 mM CaCl2, 5 mMNaHCO3, 0.1% bovine serum albumin, 10 mM HEPES, pH 7.4) tomake various concentrations.

Data Analysis. Data are reported as mean � S.E.M. wheredesignated. Analyses of metabolic parameters were performedusing a two-tailed, unpaired Student’s t test or analysis of vari-ance followed by Bonferroni’s test. p value �0.05 is consideredstatistically significant.

ResultsRanolazine Lowers FPG and HbA1c Levels in Dia-

betic Mice. The time course of the effect of ranolazine onbody weight, FPG, and HbA1c is shown in Fig. 1. Ranolazineeffectively prevented body weight loss during the week 1after STZ injection. After this period, mice in both groupsshowed similar trends in body weight gain through 8-weektreatment (STZ � vehicle: 23 � 0.3 g versus STZ � ranola-zine: 24 � 0.7 g) (Fig. 1A). FPG increased significantly withtime in both STZ � vehicle and STZ � ranolazine groupsafter STZ injection (STZ � vehicle group: baseline 108 � 3 to342 � 29 mg/dl at week 4, STZ � ranolazine group: baseline115 � 2 to 264 � 30 mg/dl at week 4), demonstrating thatmice in both groups developed diabetes. However, fromweeks 6 to 8, FPG was significantly lower in mice treatedwith ranolazine than that of the mice in the vehicle group

Fig. 1. Body weight (A), FPG (B), and HbA1c (C) in STZ-induced diabeticmice treated with vehicle or ranolazine (20 mg/kg p.o., twice daily) for 8weeks. Animals were fasted for 4 h before FPG and HbA1c measurement.Data are expressed as mean � S.E.M. �, p � 0.05 versus STZ � vehiclegroup. B signifies baseline.

52 Ning et al.

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from

(STZ � vehicle: 273 � 23 mg/dl versus STZ � ranolazine:188 � 20 mg/dl, p � 0.05) (Fig. 1B), suggesting that ranola-zine slows the progression of diabetes. HbA1c levels alsoincreased significantly in STZ-induced diabetic mice atweeks 4 and 8 compared with baseline; however, HbA1c levelsin STZ � ranolazine group were significantly lower thanthose in STZ � vehicle group after 4- and 8-week treatment(at week 8 STZ � vehicle: 5.8 � 0.4% versus STZ � ranola-zine: 4.6 � 0.2%, p � 0.05) (Fig. 1C), consistent with theobservation in FPG.

Long-Term Treatment of Ranolazine ImprovesRapid Insulin Secretion. To investigate the effect of long-term treatment with ranolazine on glucose disposal, an oralglucose tolerance test was performed in STZ � vehicle andSTZ � ranolazine groups after 8 weeks of treatment. Overall,glucose levels were lower in ranolazine-treated groups com-pared with vehicle-treated group (Fig. 2A); however, no dif-ference was observed in plasma concentration–time of glu-cose curves (i.e., area under the curve for glucose disposal)between the two groups (Fig. 2B).

In fasting state, insulin is secreted constantly at a lowbasal level. Postprandial in vivo insulin secretion is biphasicwith a rapid and transient (within 10 min) burst followed bya second phase that is characterized by sustained flat orgradual decrease that could last 2 to 3 h (Bratanova-Toch-kova et al., 2002). Baseline insulin levels were similar inSTZ � vehicle (0.25 � 0.04 ng/ml) and STZ � ranolazine(0.3 � 0.05 ng/ml) groups. Insulin levels in both groupsincreased after glucose injection; however, mice treatedwith ranolazine had significantly (p � 0.05) higher insulinlevels (0.54 � 0.06 ng/ml) compared with that of STZ �vehicle (0.34 � 0.05 ng/ml) at 15 min (Fig. 2C).

Ranolazine Has No Effect on Insulin Sensitivity inDiabetic Mice. To determine the long-term effect of ranola-zine on insulin sensitivity, insulin tolerance test was per-formed in STZ � vehicle and STZ � ranolazine groups after8 weeks of treatment. Mice were fasted for 4 h and then givenan intraperitoneal injection of insulin (1 U/kg). As shown inFig. 3A, glucose levels decreased in response to insulin injec-tion in both groups, and there was no difference in glucosedisposal curves between the two groups (Fig. 3B), suggestingthat ranolazine treatment has no significant effect on insulinsensitivity.

Ranolazine Preserves Islet Morphology and �-CellMass in Diabetic Mice. To further explore the mechanismunderlying the beneficial effect of ranolazine on FPG andHbA1c, islet morphology was examined using both H&E andimmunostaining. Pancreases from age- and sex-matched nor-mal mice were also stained as normal control for comparisonwith STZ � vehicle and STZ � ranolazine groups. As shownin Fig. 4A, STZ treatment severely decreased �-cell mass anddisrupted the islet architecture. The clear round islet bound-ary was destroyed, and islet shrinkage was observed inSTZ � vehicle group compared with healthy islets of thenormal group. Treatment with ranolazine partially pre-vented the shrinkage of the islets. This result was furtherconfirmed by fluorescence staining of insulin-expressing�-cells and glucagon-expressing �-cells (Fig. 4B). The per-centages of total islet (insulin and glucagon) area in STZ �vehicle and STZ � ranolazine were 0.21 � 0.02 and 0.30 �0.03%, respectively (p � 0.01) (Fig. 5A). In STZ � vehiclegroup, insulin-positive area (red staining) was significantly

decreased (50 � 4.6% per islet), whereas glucagon-positivearea (green staining) was increased (50 � 2.1% per islet)compared with the normal group. Ranolazine treatment sig-nificantly increased insulin-positive area (69 � 2.4%, p �0.05) compared with STZ � vehicle group, suggesting partialpreservation of functional �-cell mass in pancreas.

Ranolazine Reduces Apoptosis in Pancreatic Isletsin Diabetic Mice. To determine the role of apoptosis inSTZ-induced pancreatic morphological changes, we assessed

Fig. 2. Glucose (A) and AUC of glucose disposal (B) and insulin (C) levelsduring OGTT in STZ-induced diabetic mice treated with vehicle or rano-lazine for 8 weeks. Mice were fasted for 4 h before the experiment.Glucose was given at 2 g/kg b.wt. p.o. in distilled water. Data are ex-pressed as mean � S.E.M. �, p � 0.05 versus STZ � vehicle group.

Antidiabetic Effects of Ranolazine 53

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from

pancreatic histological sections for the presence of apoptoticcells. Activated caspase-3, a well recognized apoptosismarker, was measured using immunohistochemistry andpresented as percentage of caspase-3 positive area per islet.As shown in Fig. 6, A and B, caspase-3-positive area (24 �3.5%) was significantly increased in STZ � vehicle groupcompared with the normal group (6 � 0.8%). Ranolazinetreatment significantly reduced caspase-3-positive area to14 � 1.6% in diabetic mice (p � 0.05 versus STZ � vehiclegroup).

Ranolazine Increases GSIS. We also investigatedwhether ranolazine has an effect on GSIS in isolated islets.Rat pancreatic islets were stimulated with 20 mM glucose inthe absence or presence of various concentrations of ranola-zine for 1 h. As expected, an increase in glucose from 3 to 20mM caused a 2.5-fold increase in insulin release (Fig. 7).Acute static incubation of rat pancreatic islets with variousconcentrations of ranolazine (1 nM–1 �M) in the presence of20 mM glucose caused a further 1.4- to 2-fold increase inGSIS in a concentration-dependent manner, with a maximaleffect observed at 1 �M (Fig. 7). Similar effects of ranolazinewere observed qualitatively in human pancreatic islets. Asshown in Fig. 8A, ranolazine greatly potentiated GSIS inhuman isolated islets, although the maximal increase inglucose-induced insulin release was observed at 5 �M rano-lazine. To determine whether this effect of ranolazine isglucose-dependent, human islets were incubated with vari-ous concentrations of glucose in the absence or presence of 1�M ranolazine. As shown in Fig. 8B, ranolazine had no effecton basal insulin secretion but significantly augmented GSISunder high glucose concentrations. The magnitude of theranolazine-induced GSIS was approximately 15% higher

than that achieved by 50 nM incretin hormone glucagon-like-peptide-1(7–36) amide (Fig. 8C); however, that difference didnot reach a statistical significance.

DiscussionRanolazine is a novel anti-ischemic and antianginal drug

that increases exercise duration and time to angina episodesin patients with chronic angina pectoris. Timmis et al. (2006)and Morrow and colleagues (Scirica et al., 2007) have shownpreviously that ranolazine also has beneficial metabolic ef-fects in diabetic subjects as detected by significant reductionsof HbA1c in clinical trials. However, the mechanism of thisHbA1c-lowering effect remains unknown.

One objective of the present study was to reproduce the an-tidiabetic effect of ranolazine, observed in humans, in an animal

Fig. 4. Representative pancreatic islets shown by H&E staining(A) from two different animals and fluorescent staining (B) from normalmice and STZ-induced diabetic mice treated with vehicle or ranolazine for8 weeks. Red stains for insulin-expressing �-cells (cysteine; 20); greenstains for glucagon-expressing �-cells (FITC; 20).

Fig. 3. Glucose levels (A) and AUC of glucose disposal (B) during insulintolerance test in STZ-induced diabetic mice treated with vehicle or rano-lazine for 8 weeks. Mice were fasted for 4 h before the experiment. Insulinwas given at 1 U/kg b.wt. in distilled water intraperitoneally. Data areexpressed as mean � S.E.M.

54 Ning et al.

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from

model that can then be used to study the mechanism of theantidiabetic effects of ranolazine. By use of a diabetic mousemodel induced with multiple low doses of STZ, which causesmild to moderate levels of diabetes as a result of destruction ofpancreatic �-cells, we found that 8-week treatment with rano-lazine reduces the severity of diabetes as evidenced by loweringof FPG and HbA1c and improving pancreatic �-cell morphologyand mass and islet cell survival.

STZ is a widely used diabetogenic agent that produces aselective toxic effect on pancreatic �-cells and induces diabe-tes mellitus in most laboratory animals (LeDoux et al., 1986).High doses of STZ induce rapid and complete insulin defi-ciency resembling type 1 diabetes with ketosis. However,

lower doses of STZ, which cause a partial destruction of �-cellmass, can be used to produce a mild insulin deficiency resem-bling type 2 diabetes without a tendency to cause ketosis(Portha et al., 1989). STZ given in multiple low doses canactivate immune mechanisms in animals. It induces inflam-mation of the islets by recruitment of mononucleates andmacrophages from extra-islet area with subsequent destruc-tion of islet �-cells within a few days (Papaccio et al., 1991,1993, 2000). Numerous doses of STZ have been reported toinduce diabetes in various non-human primate species, andmultiple low doses of 40 mg/kg have been reported to induce�-cell destruction in mice (Takamura et al., 1999; Sun et al.,2005). Results from the present study are consistent withprevious findings in the literature (Takamura et al., 1999;Sun et al., 2005). Mice treated with 40 mg/kg STZ for 5 daysgradually developed diabetes within 14 days, showing signif-icant increase in FPG levels (Fig. 1B) and HbA1c (Fig. 1C).Histological analysis of the pancreas using either H&E (Fig.4A) or fluorescent staining (Fig. 4B) showed the collapse ofislets. The total number of islets in whole pancreas (Fig. 5A)and percentage of �-cells in each islet (Fig. 5B) significantlydeclined after treatment with STZ. Islet apoptosis rate wassignificantly higher in STZ mice compared with healthy con-trols (Fig. 6). Therefore, we have reproduced a stable diabeticmouse model to investigate the effects of ranolazine in vivo.

In the current study, we report that long-term treatmentwith ranolazine resulted in significant improvement in gly-cemic control. Fasting plasma glucose levels represent themost common means of assessing metabolic control in diabe-tes. Ranolazine-treated STZ mice had significantly lowerFPG at 6 weeks compared with the vehicle-treated mice, and

Fig. 5. Quantification of total islet area (A), insulin-expressing �-cells (B), and glucagon-expressing �-cells (C) in pancreas from normal mice andSTZ-induced diabetic mice treated with vehicle or ranolazine for 8 weeks. Normal group, n � 3, STZ � Vehicle and STZ � ranolazine groups, n � 6. Dataare expressed as mean � S.E.M. �, p � 0.05; ��, p � 0.01; ���, p � 0.001.

Fig. 6. A, fluorescent staining (20) of activated caspase-3 in pancreas from normal, STZ � vehicle, and STZ � ranolazine groups. B, apoptosis ratein islets from normal mice and STZ-induced diabetic mice treated with vehicle or ranolazine for 8 weeks. Normal group, n � 3, STZ � Vehicle andSTZ � ranolazine groups, n � 6. Data are expressed as mean � S.E.M. �, p � 0.05; ��, p � 0.01.

Fig. 7. GSIS in rat pancreatic islets. Freshly isolated rat islets werestimulated with various concentrations of ranolazine (R) in the presenceof 20 mM glucose (Glu) in KRB buffer for 60 min at 37°C. Data areexpressed as mean � S.E.M. of 12 experiments in duplicate each. �, p �0.05 versus 3 mM glucose-treated islets; †, p � 0.05 versus 20 mMglucose-alone-treated islets.

Antidiabetic Effects of Ranolazine 55

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from

this effect persisted until the end of the experiment (8-weektreatment). HbA1c is a standard measure of long-term glyce-mic control and is used for monitoring the efficacy of theantidiabetic drugs (Derr et al., 2003). Reduction of HbA1c

with intensive hypoglycemic therapy is associated with areduction in complications of diabetes [UK Prospective Dia-betes Study (UKPDS) Group, 1998a,b; Stratton et al., 2000].We found that 8-week treatment with ranolazine at a dose of20 mg/kg twice a day significantly lowered HbA1c levels (atweek 8 STZ � vehicle: 5.8 � 0.4% versus STZ � ranolazine:4.6 � 0.2%, p � 0.05) in STZ-induced diabetic mice (Fig. 1C).This beneficial effect on glycemic control from ranolazinetreatment is similar to that seen with conventional dipepti-dyl peptidase-4 inhibitor sitagliptin (Mu et al., 2006), whichreduced HbA1c from 7.5 � 0.2 to 5.6 � 0.3% after 2 to 3months of treatment in STZ-induced diabetic mice on high-fat diet.

The deterioration of glycemic control over time in type 2diabetes is believed to be linked to the progressive loss of�-cell function and mass (Wajchenberg 2007). The majorfinding of this study is that long-term treatment of ranola-zine slows the progression of diabetes induced by STZ byincreasing GSIS and preserving �-cell mass in diabetic mice.Ranolazine treatment increased peak insulin secretion at 15min during an OGTT at 8 weeks (Fig. 2B), indicating someimprovement of first phase of insulin secretion from existing�-cells. Ranolazine treatment preserved �-cell mass and de-creased apoptosis rate (STZ � Vehicle: 23 � 3.6% versusSTZ � ranolazine: 14 � 1.5%) in the islets (Figs. 4, 5, and 6),suggesting that long-term treatment with ranolazine in STZ-induced diabetic mice ameliorates the progression of diabetesby targeting pancreatic �-cells. Thus, the protective effect ofranolazine in �-cell preservation may be the mechanism tomaintaining normal glycemia after STZ treatment. Similarfindings have been reported with exendin-4 and rosiglita-zone. Exendin-4 significantly reduced (4.5-fold) the numbersof apoptotic �-cells in mice administered both STZ and exen-din-4 compared with that in STZ-treated mice (Li et al.,2003). Rosiglitazone also improved glucose homeostasis bysignificantly decreasing apoptosis rate (percentage) of �-cellsin rosiglitazone-treated STZ-induced diabetic rat comparedwith that of nontreated STZ group (6.52 � 0.77 versus10.33 � 1.07%) (Liu and An, 2009). Other studies have shown

that STZ may increase production of inflammatory cytokineinterleukin-1� and nitric oxide (Corbett et al., 1992a,b; Kwonet al., 1994), which may contribute to STZ-induced �-cellapoptosis and diabetes. Exendin-4 has been reported to in-hibit interleukin-1� and nitric oxide production in islet cellsvia the cAMP/protein kinase A system (Kang et al., 2009).Whether ranolazine has any beneficial effect in regulation ofinflammatory cytokines and oxidative stress in vivo will beinvestigated in future studies.

Consistent with in vivo findings that ranolazine treatmentincreased transient insulin secretion during OGTT in dia-betic mice, data from our ex vivo studies demonstrate thatranolazine directly induces rapid glucose-dependent insulinsecretion from rat and human isolated islets as reportedpreviously (Dhalla et al., 2008) . Therefore, in addition to�-cell preservation, the effects of ranolazine on FPG andHbA1c in diabetic mice could also be due in part to improvedinsulin secretion from the existing residual �-cells. It hasbeen well characterized that glucose induces insulin secre-tion through glycolysis and mitochondrial oxidation in thecells, which increase intracellular ATP/ADP ratio, sequen-tially leading to closure of KATP channels, depolarization ofvoltage-gated L-type Ca2� channels on the plasma mem-brane, Ca2� influx, and activation of exocytosis of insulin-containing granules (Rutter 2001; Iezzi et al., 2004; Newsh-olme et al., 2005). Although the exact mechanism ofincreased insulin secretion by ranolazine is not known, rano-lazine significantly stimulated insulin secretion only in thepresence of high glucose but had minimal or no effect at lowglucose (Fig. 8B), suggesting that ranolazine does not in-crease insulin secretion via KATP channel-like sulfonylurea.Therefore, the risk of hypoglycemia (a major concern withsulfonylurea therapy) with ranolazine treatment may be verylow. Ranolazine inhibits INa and IKr in cardiac cells; the twoion currents have also been shown to play a role in insulinsecretion in �-cells. Inhibition of INa by ranolazine may in-crease the concentration of intracellular ATP by lowering theconsumption of ATP (Ding and Kitasato 1997), thereby in-hibiting IKATP and causing depolarization of the �-cell fol-lowed by activation of calcium channels, increase in intracel-lular calcium concentration, and consequent increase ininsulin release. Inhibition of IKr by ranolazine may decreaseoutward K� current, and this sustains the repolarization of

Fig. 8. GSIS in human pancreatic islets. Freshly isolated human islets were incubated with various concentrations of ranolazine (R) in the presenceof 20 mM glucose (Glu) (A), with 1.0 �M ranolazine (R) in the presence of indicated concentrations of glucose (B), or with 1.0 �M ranolazine (R), 50nM GLP-1 (GLP), or vehicle (C) in KRB buffer for 60 min at 37°C. Data are expressed as mean � S.E.M. of five (A) or nine (B) experiments in duplicateeach. �, p � 0.05 versus 3 mM glucose (A) or glucose alone (B)-treated islets; †, p � 0.05 versus 20 mM glucose alone-treated islets.

56 Ning et al.

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from

the �-cells that can result in activating calcium channels andincreasing in intracellular calcium concentration followed byan increase in insulin release (Rosati et al., 2000). Whetherthese properties of ranolazine play a role in its effect onglucose homeostasis remains to be determined. In addition tothe �-cell preservation and increased insulin secretion in thepancreas, ranolazine may also have effect on other organs,such as liver and skeletal muscle, which are currently underinvestigation.

A limitation of our study is that we did not have a positivecontrol group; therefore, we can not directly compare themagnitude of the effect of ranolazine in this study with otherdrugs. Furthermore, the effects of ranolazine may be under-estimated in this study because the half-life of ranolazine inrodents is approximately 2 h; thus, dosing the mice twicedaily results in therapeutic levels of the drug (5–10 �M) for afew hours only (plasma concentrations of ranolazine at 15,30, and 60 min after oral dosing were: 9.1 � 3.3, 7.4 � 3.0,and 6.1 � 2.8 �M, respectively). To address this issue, afollow-up study, in which ranolazine concentration will bemaintained at therapeutic levels in animals by giving thedrug in diet, is planned.

In summary, the data from the present study replicate theantidiabetic effects of ranolazine seen in clinical trials. Weprovide evidence for the first time that ranolazine may exertan antidiabetic effect by protecting pancreatic �-cell massand improve insulin secretory function. Loss of functional�-cell mass is the key for deterioration of glycemic control inboth type 1 and type 2 diabetes. In this context, ranolazinehas the potential to become a novel agent that is capable oftreating patients with both angina and diabetes. However,more preclinical studies are needed to further characterizethe potential antidiabetic effect of ranolazine and to unravelthe molecular mechanisms underlying this effect.

Authorship Contributions

Participated in research design: Ning, Liu, Dhalla, and Belar-dinelli.

Conducted experiments: Ning, Zhen, Fu, Jiang, and Liu.Contributed new reagents or analytic tools: Ning and Jiang.Performed data analysis: Ning, Fu, and Jiang.Wrote or contributed to the writing of the manuscript: Ning, Liu,

Dhalla, and Belardinelli.

ReferencesBakris GL, Gaxiola E, Messerli FH, Mancia G, Erdine S, Cooper-DeHoff R, Pepine

CJ, and INVEST Investigators (2004) Clinical outcomes in the diabetes cohort ofthe INternational VErapamil SR-Trandolapril study. Hypertension 44:637–642.

Bratanova-Tochkova TK, Cheng H, Daniel S, Gunawardana S, Liu YJ, Mulvaney-Musa J, Schermerhorn T, Straub SG, Yajima H, and Sharp GW (2002) Triggeringand augmentation mechanisms, granule pools, and biphasic insulin secretion.Diabetes 51 (Suppl 1):S83–S90.

Cannon CP (2008) Mixed dyslipidemia, metabolic syndrome, diabetes mellitus, andcardiovascular disease: clinical implications. Am J Cardiol 102:5L–9L.

Chaitman BR, Skettino SL, Parker JO, Hanley P, Meluzin J, Kuch J, Pepine CJ,Wang W, Nelson JJ, Hebert DA, et al. (2004) Anti-ischemic effects and long-termsurvival during ranolazine monotherapy in patients with chronic severe angina.J Am Coll Cardiol 43:1375–1382.

Chisholm JW, Goldfine AB, Dhalla AK, Braunwald E, Morrow DA, Karwatowska-Prokopczuk E, and Belardinelli L (2010) Effect of ranolazine on A1C and glucoselevels in hyperglycemic patients with non-ST elevation acute coronary syndrome.Diabetes Care 33:1163–1168.

Corbett JA, Wang JL, Hughes JH, Wolf BA, Sweetland MA, Lancaster JR, Jr, andMcDaniel ML (1992a) Nitric oxide and cyclic GMP formation induced by interleu-kin 1 beta in islets of Langerhans. Evidence for an effector role of nitric oxide inislet dysfunction. Biochem J 287:229–235.

Corbett JA, Wang JL, Sweetland MA, Lancaster JR Jr, and McDaniel ML (1992b)Interleukin 1 beta induces the formation of nitric oxide by beta-cells purified fromrodent islets of Langerhans. Evidence for the beta-cell as a source and site of actionof nitric oxide. J Clin Invest 90:2384–2391.

Dahlof B, Sever PS, Poulter NR, Wedel H, Beevers DG, Caulfield M, Collins R,

Kjeldsen SE, Kristinsson A, McInnes GT, et al. (2005) Prevention of cardiovascularevents with an antihypertensive regimen of amlodipine adding perindopril asrequired versus atenolol adding bendroflumethiazide as required, in the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm (ASCOT-BPLA): a multicentre randomised controlled trial. Lancet 366:895–906.

Derr R, Garrett E, Stacy GA, and Saudek CD (2003) Is HbA(1c) affected by glycemicinstability? Diabetes Care 26:2728–2733.

Dhalla AK, Liu D, Santikul M, and Belardinelli L (2008) Ranolazine increasesglucose-stimulated insulin secretion in-vitro and in-vivo (abstract). J Am CollCardiol 51 (Suppl):A321.

Ding WG and Kitasato H (1997) Suppressing Na� influx induces an increase inintracellular ATP concentration in mouse pancreatic beta-cells. Jpn J Physiol47:299–306.

Iezzi M, Kouri G, Fukuda M, and Wollheim CB (2004) Synaptotagmin V and IXisoforms control Ca2�-dependent insulin exocytosis. J Cell Sci 117:3119–3127.

Kang JH, Chang SY, Jang HJ, Kim DB, Ryu GR, Ko SH, Jeong IK, Jo YH, and KimMJ (2009) Exendin-4 inhibits interleukin-1beta-induced iNOS expression at theprotein level, but not at the transcriptional and posttranscriptional levels, inRINm5F beta-cells. J Endocrinol 202:65–75.

Kwon NS, Lee SH, Choi CS, Kho T, and Lee HS (1994) Nitric oxide generation fromstreptozotocin. FASEB J 8:529–533.

LeDoux SP, Woodley SE, Patton NJ, and Wilson GL (1986) Mechanisms of nitro-sourea-induced beta-cell damage. Alterations in DNA. Diabetes 35:866–872.

Li Y, Hansotia T, Yusta B, Ris F, Halban PA, and Drucker DJ (2003) Glucagon-likepeptide-1 receptor signaling modulates beta cell apoptosis. J Biol Chem 278:471–478.

Liu D, Zhen W, Yang Z, Carter JD, Si H, and Reynolds KA (2006) Genistein acutelystimulates insulin secretion in pancreatic beta-cells through a cAMP-dependentprotein kinase pathway. Diabetes 55:1043–1050.

Liu GY and An ZM (2009) [Protective effect of rosiglitazone sodium on islet beta-cellof STZ induced diabetic rats through JNK pathway]. Sichuan Da Xue Xue Bao YiXue Ban 40:430–434.

Mu J, Woods J, Zhou YP, Roy RS, Li Z, Zycband E, Feng Y, Zhu L, Li C, Howard AD,et al. (2006) Chronic inhibition of dipeptidyl peptidase-4 with a sitagliptin analogpreserves pancreatic beta-cell mass and function in a rodent model of type 2diabetes. Diabetes 55:1695–1704.

Newsholme P, Brennan L, Rubi B, and Maechler P (2005) New insights into aminoacid metabolism, beta-cell function and diabetes. Clin Sci (Lond) 108:185–194.

Papaccio G, Frascatore S, Esposito V, and Pisanti FA (1991) Early macrophage infiltrationin mice treated with low-dose streptozocin decreases islet superoxide dismutase levels:prevention by silica pretreatment. Acta Anat (Basel) 142:141–146.

Papaccio G, Linn T, and Chieffi Baccari G (1993) Morphological observations onpancreatic islet blood vessels in low-dose streptozocin-treated mice. J Anat 182:45–53.

Papaccio G, Pisanti FA, Latronico MV, Ammendola E, and Galdieri M (2000) Mul-tiple low-dose and single high-dose treatments with streptozotocin do not generatenitric oxide. J Cell Biochem 77:82–91.

Pepine CJ and Cooper-Dehoff RM (2004) Cardiovascular therapies and risk fordevelopment of diabetes. J Am Coll Cardiol 44:509–512.

Pepine CJ, Handberg EM, Cooper-DeHoff RM, Marks RG, Kowey P, Messerli FH,Mancia G, Cangiano JL, Garcia-Barreto D, Keltai M, et al. (2003) A calciumantagonist vs a non-calcium antagonist hypertension treatment strategy for pa-tients with coronary artery disease. The International Verapamil-TrandolaprilStudy (INVEST): a randomized controlled trial. JAMA 290:2805–2816.

Portha B, Blondel O, Serradas P, McEvoy R, Giroix MH, Kergoat M, and Bailbe D(1989) The rat models of non-insulin dependent diabetes induced by neonatalstreptozotocin. Diabete Metab 15:61–75.

Rosati B, Marchetti P, Crociani O, Lecchi M, Lupi R, Arcangeli A, Olivotto M, andWanke E (2000) Glucose- and arginine-induced insulin secretion by human pan-creatic beta-cells: the role of HERG K(�) channels in firing and release. Faseb J14:2601–2610.

Rutter GA (2001) Nutrient-secretion coupling in the pancreatic islet beta-cell: recentadvances. Mol Aspects Med 22:247–284.

Scirica BM, Morrow DA, Hod H, Murphy SA, Belardinelli L, Hedgepeth CM, Mol-hoek P, Verheugt FW, Gersh BJ, McCabe CH, et al. (2007) Effect of ranolazine, anantianginal agent with novel electrophysiological properties, on the incidence ofarrhythmias in patients with non ST-segment elevation acute coronary syndrome:results from the Metabolic Efficiency With Ranolazine for Less Ischemia in NonST-Elevation Acute Coronary Syndrome Thrombolysis in Myocardial Infarction 36(MERLIN-TIMI 36) randomized controlled trial. Circulation 116:1647–1652.

Stone PH, Gratsiansky NA, Blokhin A, Huang IZ, Meng L, and ERICA Investigators(2006) Antianginal efficacy of ranolazine when added to treatment with amlodip-ine: the ERICA (Efficacy of Ranolazine in Chronic Angina) trial. J Am Coll Cardiol48:566–575.

Stratton IM, Adler AI, Neil HA, Matthews DR, Manley SE, Cull CA, Hadden D,Turner RC, and Holman RR (2000) Association of glycaemia with macrovascularand microvascular complications of type 2 diabetes (UKPDS 35): prospectiveobservational study. BMJ 321:405–412.

Sun N, Yang G, Zhao H, Savelkoul HF, and An L (2005) Multidose streptozotocininduction of diabetes in BALB/c mice induces a dominant oxidative macrophageand a conversion of TH1 to TH2 phenotypes during disease progression. MediatorsInflamm 2005:202–209.

Takamura T, Ando H, Nagai Y, Yamashita H, Nohara E, and Kobayashi K (1999)Pioglitazone prevents mice from multiple low-dose streptozotocin-induced insulitisand diabetes. Diabetes Res Clin Pract 44:107–114.

Timmis AD, Chaitman BR, and Crager M (2006) Effects of ranolazine on exercise toleranceand HbA1c in patients with chronic angina and diabetes. Eur Heart J 27:42–48.

UK Prospective Diabetes Study (UKPDS) Group (1998a) Effect of intensive blood-glucosecontrol with metformin on complications in overweight patients with type 2 diabetes(UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group. Lancet 352:854–865.

Antidiabetic Effects of Ranolazine 57

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from

UK Prospective Diabetes Study (UKPDS) Group. (1998b) Intensive blood-glucosecontrol with sulphonylureas or insulin compared with conventional treatment andrisk of complications in patients with type 2 diabetes (UKPDS 33). UK ProspectiveDiabetes Study (UKPDS) Group. Lancet 352:837–853.

Wajchenberg BL (2007) beta-cell failure in diabetes and preservation by clinicaltreatment. Endocr Rev 28:187–218.

Wofford MR and King DS (2004) Relationship between antihypertensive drugs andmetabolic syndrome. Metab Syndr Relat Disord 2:308–314.

Yang Z, Chen M, Ellett JD, Fialkow LB, Carter JD, and Nadler JL (2004) The novelanti-inflammatory agent lisofylline prevents autoimmune diabetic recurrence af-ter islet transplantation. Transplantation 77:55–60.

Address correspondence to: Dr. Arvinder K. Dhalla, Department of Biol-ogy, Gilead Palo Alto Inc., 1651 Page Mill Rd, Palo Alto, CA 94304. E-mail:arvinder.dhalla@gilead.com

58 Ning et al.

at ASPE

T Journals on Septem

ber 23, 2017jpet.aspetjournals.org

Dow

nloaded from