Brief Communication

Inhibition of phenylephrine induced hypertrophy in rat neonatalcardiomyocytes by the mitochondrial KATP channel opener diazoxide

Ying Xia, Venkatesh Rajapurohitam, Michael A. Cook, Morris Karmazyn*

Department of Physiology and Pharmacology, University of Western Ontario, Medical Sciences Building, London, Ont., Canada N6A 5C1

Received 31 March 2004; received in revised form 1 July 2004; accepted 7 July 2004

Available online 22 September 2004

Abstract

The effect of the putative mitochondrial KATP channel opener diazoxide (100 µM) was studied in terms of its ability to modulate thehypertrophic effect of 24 h treatment with the a1 adrenoceptor agonist phenylephrine (PE; 10 µM) in cultured neonatal rat ventricularmyocytes. PE on its own significantly increased cell size by 40%, 3H leucine incorporation by 37% and produced more than a threefoldelevation in both atrial natriuretic peptide and myosin light chain-2 expression. These effects were nearly completely prevented by diazoxidealthough the inhibitory effect of this agent was generally mitigated by the mitochondrial KATP channel antagonists 5-hydroxydecanoic acid(100 µM) and glibenclamide (50 µM). Although PE produced an early threefold elevation in MAP kinase activation this was generallyunaffected by diazoxide. PE also produced a greater than threefold increase in Na–H exchanger isoform 1 (NHE-1) expression which, wasprevented by diazoxide treatment. Our study therefore, demonstrates a potential antihypertrophic influence of mitochondrial KATP channelactivation which, is related to diminished NHE-1 expression. Mitochondrial KATP channel activation could represent an effective approach tominimize the myocardial hypertrophic process.© 2004 Elsevier Ltd. All rights reserved.

Keywords: Mitochondrial KATP channel; Diazoxide; 5-Hydroxydecanoic acid; Glibenclamide; Cardiomyocyte hypertrophy; NHE-1; MAPK

1. Introduction

There is considerable evidence implicating mitochondrialKATP channel opening in myocardial protection especiallyrelated to ischemic preconditioning [1,2]. Although the acuteprotective effects of mitochondrial KATP channel openinghave been well established, little is known concerning longterm effects of this treatment. The use of KATP channelopeners such as pinacidil has been shown to reduce hypertro-phy in patients with essential hypertension which has beenassumed to be due to the drug’s blood pressure loweringeffect [3]. Moreover, the mitochondrial KATP channel openernicorandil reduced myocardial remodeling in rats treatedwith the nitric oxide synthase inhibitor L-NAME [4]. Thecardioprotective effects of mitochondrial KATP channel acti-vation have been shown to be associated with diminishedintracellular Ca2+ accumulation which could result, in thelong term, in reduced calcium-dependent activation of intra-cellular hypertrophic factors. Whether mitochondrial KATP

channel opening produces a direct antihypertrophic influenceon the cardiomyocyte has, to our knowledge, not been stu-died. Accordingly, we examined the effect of the putativemitochondrial KATP channel opener diazoxide on the hyper-trophic effect of phenylephrine (PE) using cultured neonatalmyocytes in which extra cardiac effects of treatments couldbe precluded.

2. Material and methods

2.1. Primary neonatal cardiac myocytes culture

Myocytes were prepared from the ventricles of 4-day-oldSprague–Dawley rats as described in detail previously [5]. Inbrief the ventricles were excised, washed and cut into smallpieces in 15 ml Hanks’ balanced salt solution (HBSS; Invi-trogen, Burlington, Ont., Canada), then digested in 60 ml ofHBSS containing 800 U collagenase (Worthington Bioche-mical Corporation, Lakewood, NJ, USA)/ventricle. The di-gestion was performed in a circulating water bath to keep thereaction temperature at 37° C. The digestion was terminated

* Corresponding author. Tel.: +1-519-661-3872; fax: +1-519-661-3827.E-mail address: mkarm@uwo.ca (M. Karmazyn).

Journal of Molecular and Cellular Cardiology 37 (2004) 1063–1067

www.elsevier.com/locate/yjmcc

0022-2828/$ - see front matter © 2004 Elsevier Ltd. All rights reserved.doi:10.1016/j.yjmcc.2004.07.002

by adding the same volume of 20% fetal bovine serum(FBS). The cells were sorted by a cell strainer to removeundigested particles and then centrifuged at 600 × g for 5 minat 4° C. The cell pellet was resuspended in a plating mediumcontaining 10% FBS, 0.1 mM bromodeoxyuridine, and waspreplated in tissue culture flasks for two times at 20 min toreduce contaminating non myocytes after which the cellswere transferred into Primaria™ cell culture dishes (BectonDickinson Labware, Mississauga, Ont., Canada) and cultu-red for 48 h. The medium was replaced with a serum freemaintenance medium and incubated for another 24 h beforebeing used for study. Approximately 95% of cells preparedby this method demonstrated sarcomeric myosin heavy chainstaining, indicating relatively low non myocyte contamina-tion [6].

2.2. Experimental protocol

Cells were first serum-starved for 24 h after which thecells were treated for a further 24 h with PE (10 µM). Toassess the effect of mitochondrial KATP channel modulators,the cells were first treated with either diazoxide (100 µM),5-hydroxydecanoate (5-HD; 100 µM) or glibenclamide(50 µM) or a combination of these drugs as described inSection 3. All drugs were added 30 min before PE adminis-tration. Mitochondrial KATP modulators were also studied ontheir own by administering each for 24 h in cells not treatedwith PE. All parameters were measured 24 h after PE addi-tion except for MAPK activation which, was studied usingcells treated with PE for 10 min. All drugs were purchasedfrom Sigma (Oakville, Ont., Canada).

2.3. Cell area measurement

The cells were plated at a density of 1 × 106 cells/6 cmdish to obtain individually plated cells. At the end of thetreatment period the cells were washed twice with PBS afterwhich they were viewed using a Leica DMIL inverted mi-croscope equipped with a Polaroid digital camera. Eightrandom photographs were taken from each sample and atleast 40 individual cell surface area measurements weremade from each photograph using Mocha software.

2.4. Real time PCR

Myocytes were plated at 6 × 106 cells/6 cm dish. Afterwashing twice with PBS, RNA was isolated by adding 1 mlTrizol reagent (Invitrogen) to each dish. Five micrograms oftotal RNA were applied for reverse transcription by Supers-cript II reverse transcriptase (Invitrogen). One microliterfrom the 20 µl cDNA product was used for each PCR reac-tion. Real time PCR was performed with a DNA EngineOpticon Real Time System (MJ Research, Waltham, MA,USA) with a SYBR Green JumpStart Taq ReadyMix kit(Sigma) according to the manufacturer’s instructions. Re-sults for gene expression are presented as a ratio to 18s rRNA(18s) as the reference gene.

2.5. Incorporation of 3H leucine

Cells were plated at 1 × 106 cells per well in 24 welldishes. 3H leucine (2 µCi/ml) was added to each well justafter the last treatment. Twenty-four hours later, the cellswere washed three times with ice-cold PBS, and incubatedwith 5% TCA on ice for 20 min. After the cells were washedtwice with ice-cold 5% TCA, they were solubilized in 100 µl0.5 N NaOH for 30 min on ice, and then neutralized by 100 µl0.5 M HCl. This was then transferred to 10 ml scintillationsolution for radioactivity determination.

2.6. Western blot analysis for MAP kinases

Cells were plated at a concentration of 6 × 106 cells/6 cmdish. After washing twice with PBS the cells were scrapedinto 100 µl lysis buffer (20 mM Tris, 150 mM NaCl, 1%Triton X 100, 10% glycerol, 2 mM EDTA, 2 mM EGTA,50 mM NaF, 200 µM Na3VO4, 10 mM Na4P2O7, 40 mMb-glycerophosphate, 10 µg/ml leupeptin, 1 µM pepstatin A,1 mM PMSF, and 1 µM colyculin A). The lysate was trans-ferred to 1.5 ml Eppendorf tubes, homogenized and thencentrifuged at 10,000 × g for 5 min at 4° C. The supernatantwas transferred to a fresh tube and the protein concentrationwas assayed by Bradford Protein Assay Kit (Bio Rad, Mis-sissauga, Ont., Canada). Thirty micrograms protein wereloaded in 10% SDS PAGE, and transferred to nylon mem-brane (Amersham, Little Chalfont Buckinghamshire, UK).The membranes were blocked in 5% dry milk for 3 h, pri-mary antibody for 2 h, secondary antibody for 1 h, and thendetected by ECL reagent (Amersham). All antibodies werepurchased from Cell Signaling (Beverly, MA, USA) andused at a 1:1000 dilution.

2.7. Statistical analysis

All values in the figures and text are presented asmean ± S.E.M. Sample size per experiment is indicated inSection 3. Data were analyzed by one-way ANOVA withP < 0.05 considered to represent significant differencesbetween groups.

3. Results and discussion

Fig. 1 summarizes the effect of diazoxide on variousparameters associated with the hypertrophic response fol-lowing PE addition. In all cases diazoxide prevented thehypertrophic phenotype as assessed by cell size (panel A), 3Hleucine incorporation (panel B) as well as both ANP andmyosin light chain-2 (MLC-2) expression (panels C and D).In cells not treated with PE, neither diazoxide, 5-HD orglibenclamide had any direct effects on any parameter stu-died after 24 h treatment (not shown). In addition, as shownin Fig. 1 neither 5-HD or glibenclamide had any effect ontheir own in PE-treated cells although 5-HD blunted the

1064 Y. Xia et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 1063–1067

PE-induced upregulation of MLC-2 (panel D). However,both 5-HD and glibenclamide either partially or completelyprevented the diazoxide-induced inhibition of hypertrophy interms of all parameters studied (Fig. 1).

As shown in Fig. 2, we also determined whether diazoxidemodulates expression of Na–H exchanger isoform 1 (NHE-1). Although NHE-1 is not considered as a classical hyper-trophic marker it is known to be upregulated in hypertrophyand has been proposed by various investigators as a keymodulator of the hypertrophic response to various agonists(reviewed in [7,8]). The ability of diazoxide to significantlyattenuate the upregulation of NHE-1 expression by PE likelyfurther attests to its antihypertrophic properties. Although5-HD partially reversed the effect of diazoxide, as for MLC

expression we did observe a direct blunting effect of thisagent as well as glibenclamide with regards to NHE-1 upre-gulation. The reason for the direct effect of 5-HD is uncertainbut may reflect its non specific actions not related to mito-chondrial KATP channel inhibition particularly since the ef-fect was not shared with glibenclamide, or it may reflectsome intrinsic KATP channel opening ability by this agent.Indeed, it should be noted that both diazoxide and 5-HD havebeen shown to exert effects unrelated to the mitochondrialKATP channel [9] and hence more direct evidence for mito-chondrial KATP channel involvement in cardiomyocyte hy-pertrophy is required.

As shown in Fig. 3, ERK phosphorylation was markedlyelevated 10 min after PE addition. Diazoxide failed to mar-

Fig. 1. Effect of the mitochondrial KATP channel opener diazoxide in the absence or presence of the mitochondrial KATP channel blockers 5-HD orglibenclamide on PE-induced hypertrophy by cell area (A), 3H leucine incorporation (B), ANP expression (C) and expression of MLC (D). Values indicatemean ± S.E. with n = 6 for all groups. * P < 0.05 from control (no treatment); # P < 0.05 from PE alone group.

1065Y. Xia et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 1063–1067

kedly alter ERK activation although in the presence of dia-zoxide without either 5-HD or glibenclamide values were notsignificantly different from control. No effect of diazoxide oneither p38 or JNK phosphorylation was seen (not shown).

The ability of mitochondrial KATP channel openers toprotect the ischemic and reperfused myocardium has beendemonstrated by numerous investigators [1,2]. The mecha-nism of protection afforded by this group of agents is notcompletely understood but the consequence of mitochondrialKATP channel activation is associated with modulation ofvarious cell signaling systems as well as inhibition of intra-cellular Ca2+ accumulation (reviewed in [2]). While protec-tion by mitochondrial KATP channel opening is well esta-blished, little is known concerning the potential longer termeffects of mitochondrial KATP channel opening. Our resultsreveal a potent antihypertrophic effect of diazoxide in termsof inhibition of increased cell area, upregulation of the hyper-trophic genes ANP and MLC-2 as well as diminution ofincreased leucine incorporation. The ability of diazoxide toinhibit the hypertrophic responses was reversed by the puta-tive mitochondrial KATP channel antagonists 5-HD and gli-benclamide with relatively identical efficacies.

To our knowledge, this represents the first report docu-menting an antihypertrophic effect of mitochondrial KATP

channel inhibition although the mechanisms underlying thiseffect are uncertain. We considered the possibility that inhi-bition of MAP kinase activation could represent a possiblemechanism for this effect but diazoxide had minimal effectson p44/42 phosphorylation (shown) and no effect on eitherphosphorylated p38 or JNK levels. However, diazoxide po-tently inhibited NHE-1 upregulation which may contribute toits antihypertrophic effect, although whether this antihyper-trophic effect of diazoxide and NHE-1 inhibition represents acausal relationship needs to be determined. Indeed, the po-tential mechanisms for the antihypertrophic actions of dia-zoxide require substantial further studies. It also remains tobe determined whether mitochondrial KATP channel activa-tion modifies responses associated with cardiac remodelingsuch as those which may occur in the postischemic myocar-dium.

Acknowledgments

This study was supported by a grant from the CanadianInstitutes of Health Research. Dr. Karmazyn was a recipientof a Heart and Stroke Foundation of Ontario Career Investi-gator Award during the course of this study.

References

[1] Brayden JE. Functional roles of KATP channels in vascular smoothmuscle. Clin Exp Pharmacol Physiol 2002;29:312–6.

[2] Gross GJ, Peart JN. KATP channels and myocardial preconditioning:an update. Am J Physiol 2003;285:H921–H930.

[3] Steensgaard-Hansen F, Carlsen JE. Effects of long term treatmentwith pinacidil and nidefipine on left ventricular anatomy and functionin patients with mild to moderate systemic hypertension. Drugs 1988;36(Suppl 7):70–6.

Fig. 2. Effect of the mitochondrial KATP channel opener diazoxide in theabsence or presence of the mitochondrial KATP channel blockers 5-HD orglibenclamide on PE-induced upregulation of NHE-1 gene expression (de-picted as a ratio to 18s as the reference gene) determined by real time PCR.Values indicate mean ± S.E. with n = 6 for all groups. * P < 0.05 from control(no treatment).

Fig. 3. Effect of the mitochondrial KATP channel opener diazoxide in theabsence or presence of the mitochondrial KATP channel blockers 5-HD orglibenclamide on PE-induced activation of p44/42. Values indicatemean ± S.E. with n = 6 for all groups. * P < 0.05 from control (no treatment).Representative western blots for phosphorylated (pp44/42) and non phos-phorylated (p44/42) are shown below.

1066 Y. Xia et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 1063–1067

[4] Sanada S, Node K, Asanuma H, Ogita H, Takashima S, Minamino T, et al.Opening of the adenosine triphosphate-sensitive potassium channelattenuates cardiac remodeling induced by long-term inhibition of nitricoxide synthesis: role of 70-kDa S6 kinase and extracellular signal-regulated kinase. J Am Coll Cardiol 2002;40:991–7.

[5] Karmazyn M, Liu Q, Gan XT, Brix BJ, Fliegel L. Aldosteroneincreases NHE-1 expression and induces NHE-1-dependent hypertro-phy in neonatal rat ventricular myocytes. Hypertension 2003;42:1171–6.

[6] Rajapurohitam V, Gan XT, Kirshenbaum L, Karmazyn M. Theobesity-associated peptide leptin induces hypertrophy in neonatal ratventricular myocytes. Circ Res 2003;93:277–9.

[7] Cingolani HE. Na+/H+ exchange hyperactivity and myocardialhypertrophy: are they linked phenomena? Cardiovasc Res 1999;44:462–7.

[8] Karmazyn M. Role of sodium–hydrogen exchange in cardiac hyper-trophy and heart failure: a novel and promising therapeutic target.Basic Res Cardiol 2001;96:325–8.

[9] Hanley PJ, Mickel M, Loffler M, Brandt U, Daut J. KATP channel-independent targets of diazoxide and 5-hydroxydecanoate in the heart.J Physiol 2002;542:735–41.

1067Y. Xia et al. / Journal of Molecular and Cellular Cardiology 37 (2004) 1063–1067