Expression, Localization, and Pharmacological Role of Kv7Potassium Channels in Skeletal Muscle Proliferation,Differentiation, and Survival after Myotoxic Insults□S

Fabio Arturo Iannotti, Elisabetta Panza, Vincenzo Barrese, Davide Viggiano,Maria Virginia Soldovieri, and Maurizio TaglialatelaDivision of Pharmacology, Department of Neuroscience, University of Naples Federico II, Naples, Italy (F.A.I., E.P., V.B., M.T.);and Department of Health Sciences, University of Molise, Campobasso, Italy (F.A.I., D.V., M.V.S., M.T.)

Received October 19, 2009; accepted December 16, 2009

ABSTRACTChanges in the expression of potassium channels regulateskeletal muscle development. The purpose of this study was toinvestigate the expression profile and pharmacological role ofKv7 voltage-gated potassium channels in skeletal muscle dif-ferentiation, proliferation, and survival after myotoxic insults.Transcripts for all Kv7 genes (Kv7.1–Kv7.5) were detected bypolymerase chain reaction (PCR) and/or real-time PCR in mu-rine C2C12 myoblasts; Kv7.1, Kv7.3, and Kv7.4 transcripts wereup-regulated after myotube formation. Western blot experi-ments confirmed Kv7.2, Kv7.3, and Kv7.4 subunit expression,and the up-regulation of Kv7.3 and Kv7.4 subunits during invitro differentiation. In adult skeletal muscles from mice andhumans, Kv7.2 and Kv7.3 immunoreactivity was mainly local-

ized at the level of intracellular striations positioned betweenankyrinG-positive triads, whereas that of Kv7.4 subunits waslargely restricted to the sarcolemmal membrane. In C2C12 cells,retigabine (10 �M), a specific activator of neuronally expressedKv7.2 to Kv7.5 subunits, reduced proliferation, acceleratedmyogenin expression, and inhibited the myotoxic effect of me-vastatin (IC50 � 7 �M); all these effects of retigabine wereprevented by the Kv7 channel blocker 10,10-bis(4-pyridinyl-methyl)-9(10H)-anthracenone (XE-991) (10 �M). These datacollectively highlight neural Kv7 channels as significant phar-macological targets to regulate skeletal muscle proliferation,differentiation, and myotoxic effects of drugs.

Skeletal muscle development is a multifactorial, highlycontrolled process involving the coordinated regulation of alarge number of genes that allows proliferating myoblasts towithdraw from the cell cycle, to fuse in ordered arrays oflarge multinucleated myotubes, and to further differentiateinto mature muscle fibers (Walsh and Perlman, 1997).

Changes in the expression of potassium (K�) channelsseem to be associated with the myoblast to myotube transi-tion (Cooper, 2001). After earlier reports on distinct voltage-sensitive K� channels being developmentally regulated in

vivo (Lesage et al., 1992), K� currents carried by ether-a-go-go (Kv10.1) (Bijlenga et al., 1998), and KIR2.1 (Wielandand Gong, 1995; Fischer-Lougheed et al., 2001) channelshave been shown to play a specific role in determining themembrane hyperpolarization that occurs during the earlyphases of myoblast commitment and fusion in vitro. Thisphenomenon is believed to allow significant Ca2� influxthrough T-type voltage-gated Ca2� channels, thereby trig-gering the increase in intracellular Ca2� concentration([Ca2�]i) necessary for myoblast fusion (Bijlenga et al., 2000).Other K� channels, such as intermediate-conductance Ca2�-activated K� channels, seem to be down-regulated duringmyogenesis, although they do not play a significant role inmyoblast proliferation (Fioretti et al., 2005).

Voltage-dependent K� channels belonging to the Kv7 sub-class have been implicated in resting membrane potentialand electrical excitability control in many cell types (Miceli etal., 2008). The Kv7 family consists of five members (Kv7.1–

This work was supported by ERA-Net for Research Programs on RareDisease (E-Rare 2007) Telethon Fondazione ONLUS, Italy [Grant GGP07125];Regione Molise; and the Italian Ministry of Education, University, and Re-search (PRIN 2007); and Italian Ministry of Health (Progetto Doping 2005).(M.T.).

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

doi:10.1124/jpet.109.162800.□S The online version of this article (available at http://jpet.aspetjournals.org)

contains supplemental material.

ABBREVIATIONS: XE-991, 10,10-bis(4-pyridinylmethyl)-9(10H)-anthracenone; IKM, M-current; GM, growth medium; DM, differentiation medium;MyHC, myosin heavy chain; AnkG, ankyrin-G; MEV, mevastatin; FBS, fetal bovine serum; CHO, Chinese hamster ovary; PCR, polymerase chainreaction; RT-PCR, reverse transcription-polymerase chain reaction; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; PBS,phosphate-buffered saline; OD, optical density.

0022-3565/10/3323-811–820$20.00THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 332, No. 3Copyright © 2010 by The American Society for Pharmacology and Experimental Therapeutics 162800/3565299JPET 332:811–820, 2010 Printed in U.S.A.

811

http://jpet.aspetjournals.org/content/suppl/2009/12/29/jpet.109.162800.DC1Supplemental material to this article can be found at:

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

Kv7.5), each showing distinct tissue distribution and subcel-lular localization, as well as biophysical, pharmacological,and pathophysiological properties. Although Kv7.1 is ex-pressed mainly in cardiac muscle, gastrointestinal epithelia,and inner ear, Kv7.2 and Kv7.3 are expressed in the centraland peripheral nervous system (Cooper et al., 2000). How-ever, Kv7.4 subunits seem to be expressed mainly in thecochlea and central auditory pathways (Kharkovets et al.,2000), whereas Kv7.5 transcripts have been detected in sev-eral brain regions and in sympathetic ganglia (Lerche et al.,2000; Schroeder et al., 2000). Because Kv7.2 to Kv7.5 werefirst discovered in neurons, they are currently identified asneural Kv7 genes (Brown and Passmore, 2009).

Kv7 channels give rise to outwardly rectifying voltage-dependent K� currents with an heterogeneous functionalrole. Kv7.1 subunits form the molecular basis of IKs, a cardiaccurrent involved in action potential repolarization (Sangui-netti et al., 1996), whereas Kv7.2 and Kv7.3 subunits (possi-bly in association with Kv7.4 and Kv7.5 at specific neuronalsites) can form homomeric or heteromeric K� channels un-derlying the M-current (IKM) (Wang et al., 1998). IKM is aslowly activating/deactivating and noninactivating sub-threshold current that regulates neuronal excitability, func-tioning as a brake for repetitive action potential firing(Brown and Adams, 1980). In humans, mutations in theKv7.1 gene are responsible for one form of long QT syndrome,whereas mutations targeting Kv7.4 cause a rare form ofnonsyndromic autosomal dominant hearing loss; gene defectsin Kv7.2 or Kv7.3 are responsible for benign familial neonatalseizures, an inherited epilepsy of the newborn (Miceli et al.,2008; Brown and Passmore, 2009).

In addition to neural tissue, Kv7.4 and Kv7.5 expression hasalso been described in smooth muscle cells from several vascu-lar beds, where they control reactivity to vasopressors (Yeung etal., 2007; Mackie et al., 2008). In addition, Kv7.1 (Tsevi et al.,2005) and Kv7.5 (Lerche et al., 2000; Schroeder et al., 2000)transcripts have been detected in adult skeletal muscle;changes in Kv7.1 and Kv7.5 transcript levels occur during pro-liferation and differentiation of rat L6E9 myoblasts in vitro(Roura-Ferrer et al., 2008). However, no data are yet availableon other Kv7 family members, and the functional role of Kv7.1and Kv7.5 in skeletal muscle is still poorly defined.

In this study, with use of murine C2C12 cells as an experi-mental model for in vitro skeletal myogenesis, we investigatedthe expression pattern of Kv7.1 to Kv7.5 transcripts and pro-teins during skeletal muscle differentiation. Moreover, the ex-pression of Kv7 subunits in adult skeletal muscle from mouseand humans, and their possible role in C2C12 cell proliferation,differentiation, and protection from drug-induced myotoxicity,were also investigated. To this aim, Kv7 channel blockers (XE-991) and activators (retigabine) (Miceli et al., 2008) were used.The results obtained indicate that neural Kv7 channels regulateskeletal muscle cell proliferation, differentiation, and responseto myotoxic stimuli, thus highlighting novel potential therapeu-tic opportunities for Kv7 channel modulators in skeletal muscledisorders including drug-induced myopathies.

Materials and MethodsCell Culture and Reagents

Murine C2C12 myoblasts were propagated in a growth medium(GM) composed of Dulbecco’s modified Eagle’s medium supple-

mented with 10% fetal bovine serum (FBS), 50 U/ml penicillin plus50 �g/ml streptomycin, and 1% L-glutamine (Invitrogen, Milan, It-aly), in a humidified atmosphere of 95% air/5% CO2 at 37°C. Expo-sure of proliferating C2C12 cells for 24 to 72 h to a lower (from 10% to0.1%) FBS concentration, plus the addition of 5 �g/ml insulin and 5�g/ml transferrin (differentiation medium, DM), induced their dif-ferentiation into myotubes, as revealed by phase-contrast lighttransmission analysis showing an increased cell size and an elon-gated fiber-like morphology (Supplemental Fig. 1A, a and b). Confo-cal immunofluorescence analysis revealed the induction in C2C12

myotubes of myosin heavy chain (MyHC), a marker of late skeletalmuscle cells maturation (Bennett and Tonks, 1997) (SupplementalFig. 1A, c and d), a result also confirmed by Western blot analysis(Supplemental Fig. 1B). Staining with chromomycin A3, a fluores-cent DNA-binding antibiotic labeling cell nuclei, confirmed in vitromyotube formation, as multiple nuclei were evident in a single fiber(Supplemental Fig. 1A, c and d).

Chinese hamster ovary cells (CHO cells) were grown in 60-mmplastic Petri dishes in Dulbecco’s modified Eagle’s medium contain-ing 10% FBS, nonessential amino acids (0.1 mM), penicillin (50U/ml), and streptomycin (100 �g/ml) in a humidified atmosphere at95% O2/5% CO2 at 37°C. After plating, the cells were transfected onthe next day with plasmids encoding Kv7 cDNAs by use of Lipo-fectamine 2000 (Invitrogen), a plasmid encoding for the EnhancedGreen Fluorescent Protein (Clontech, Mountain View, CA) was usedas a transfection marker, and total cDNA in the transfection mixturewas kept constant at 4 �g.

RNA Extraction, Semiquantitative RT-PCR, andQuantitative Real-Time PCR

Total RNA was isolated from C2C12 myoblasts and myotubes orfrom native tissues by use of the TRI-Reagent (Sigma-Aldrich, Mi-lan, Italy), reacted with DNase-I (1U/�l; Sigma-Aldrich) for 15 minat room temperature, followed by spectrophotometric quantification.Final preparation of RNA was considered DNA- and protein-free ifthe ratio between readings at 260/280 nm was �1.7. Isolated mRNAwas reverse-transcribed by use of MuLV high-capacity reverse tran-scriptase (50 U/�l; Applied Biosystems, Monza, Italy) in buffer con-taining 4 mM dNTP, Random Primers, 1 �l of RNase Inhibitor at37°C for 120 min; cDNA obtained was amplified in reverse transcrip-tion-polymerase chain reaction (RT-PCR) using PCR gold buffer, alsocontaining 2 mM MgCl2, 0.8 mM dNTP mix, 0.001 mM forward andreverse primers (see Supplemental Table 1 for primer sequences),and 1 U/�l Amplitaq Gold (Applied Biosystems). The protocol usedfor PCR amplification was the following: denaturation at 95°C for 1min, annealing at 54 to 60°C (template-dependent) for 1 min, andelongation at 72°C for 1 min (35 cycles, one every 13 min). To test theability of the Kv7 primers to specifically recognize the mRNA targetagainst which they were designed, RT-PCR experiments were per-formed with use of cDNA templates from adult mouse brain andheart mRNAs (Supplemental Fig. 2). Neural Kv7.2 to Kv7.5 tran-scripts were effectively detected in brain tissue (Supplemental Fig. 2,left). A very faint band corresponding to Kv7.1 mRNA could also beidentified, consistent with previous results (Ohya et al., 2003; Yeunget al., 2007). Moreover, in mouse heart, only Kv7.1 and Kv7.4 tran-scripts (the latter of presumed vascular origin) showed significantexpression (Supplemental Fig. 2).

Quantitative real-time PCR was carried out in a 7500 fast real-time PCR system (Applied Biosystems) with the Kv7 primers by useof SYBR Green detection. Samples were amplified simultaneously ina triplicate in one-assay run, and the ct (threshold cycle) value foreach experimental group was determined. Data normalization wasperformed by using as a control the ct from S16, a constitutivelyexpressed ribosomal protein; differences in mRNA content betweengroups were calculated as normalized values by use of the 2��ct

formula.

812 Iannotti et al.

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

Cell Proliferation

To assess C2C12 cell proliferation, myoblasts were trypsinized andseeded onto 24-well plastic plates at a 2 � 103 cells/cm2 density. Afteradhesion, the cells were synchronized in DM for 30 h; cells were thenwashed twice with phosphate-buffered saline (PBS), and 1 �Ci/ml[methyl-3H]thymidine (GE Healthcare, Milan, Italy) was added infresh GM or DM medium containing the drug(s) of interest. After24 h, the cells were fixed in cold methanol, washed three times inice-cold 10% trichloroacetic acid, and solubilized in 1% SDS and 0.3%NaOH (Roura-Ferrer et al., 2008). The radioactivity in each samplewas determined by liquid scintillation counting in a LS/5000/CEsystem (Beckman Coulter, Milan, Italy).

Cell Viability

C2C12 cells were seeded at 2 � 103 cells/cm2 density in 24-wellplastic plates. One day after plating, mevastatin (alone or in thepresence of Kv7 channel modulators) was added to the culture me-dium for the indicated times (24, 48, and 72 h). Cell viability wasevaluated with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazo-lium bromide (MTT; 10 mg/ml; Sigma-Aldrich) reduction assay, andformazan salts formation upon MTT reduction by mitochondria ofliving cells was detected spectrophotometrically at 595 nm.

Confocal Immunofluorescence Analysis

Cells. C2C12 cells, plated on poly(L-lysine)-coated coverslips, werewashed three times with PBS and incubated at room temperature(20–22°C) with freshly made paraformaldehyde (4% w/v) for 10 min.The cells were then incubated for 5 min with 0.1 M glycine, washedin PBS, and incubated for 1 h with anti-mouse monoclonal anti-MyHC (dilution, 1:500; Millipore, Vimodrone—Milan, Italy). Subse-quently, the cells were incubated at room temperature for 1 h with afluorescent secondary antibody (The Jackson Laboratory, Bar Har-bor, ME) anti-mouse IgG conjugated to CY3 for anti-MyHC, diluted1:200 in PBS containing 10% (v/v) FBS and 0.1% Triton X-100.Chromomycin A3 was used as nuclear marker. Coverslips weremounted in SlowFade (Invitrogen).

Tissue Slices. Human muscle preparations were obtained fromthe biopsy of the brachial biceps of a 30-year-old male, after informedconsent approved by the local Ethical Committee was given. Animalmuscle preparations derived from adult male C57/BL6 mice (50–60g) housed under constant conditions of temperature (22 � 1°C) andrelative humidity (50%) with a regular light-dark schedule (lights onfrom 7:00 AM to 7:00 PM) and free access to food and water. Allexperiments were conducted in accordance with the European Com-munity Council Directive (86/609/EEC) and were approved by theEthics Committees of the Institutions where the experiments wereperformed. All efforts were taken to minimize animal suffering andto use the minimal number of animals necessary to obtain reliableresults. Animals were deeply anesthetized, perfused with saline (10ml) followed by 50 ml of freshly made paraformaldehyde (4% w/v inPBS). From these animals, soleus muscles were removed and post-fixed in the same fixative for 2 h at 4°C, and washed with 0.1 Mglycine for 5 to 10 min. Mouse or fixed human tissue was cut with acryostat into 20-�m-thick sections that were mounted onto pre-treated Superfrost slides, stored at �80°C until further processing.Mouse and human sections were incubated overnight at 4°C with thefollowing primary antibodies: a) rabbit-polyclonal anti-Kv7.2 (dilu-tion, 1:100; a gift from Dr. Holger Lerche) (Luisi et al., 2009); b)rabbit polyclonal anti-Kv7.3 (dilution, 1:500; a gift from Dr. AlvaroVillarroel); c) mouse monoclonal anti Kv7.4 (dilution, 1:100; Neu-romab, Davis, CA); d) mouse monoclonal anti-ankyrin-G (AnkG;dilution, 1:50; Abcam plc, Cambridge, UK). The primary antibodieswere diluted in PBS containing 10% (v/v) FBS and 0.1% Triton X-100and incubated overnight at 4°C. Slices were then washed in PBS andincubated at room temperature for 1 h with anti-mouse IgG conju-gated to CY3 for Kv7.4, or anti-rabbit IgG conjugated to CY3 forKv7.2 or Kv7.3 (The Jackson Laboratory), or anti-mouse IgG conju-

gated to CY5 for AnkG; secondary antibodies were diluted 1:100 inPBS containing 10% (v/v) FBS and 0.1% Triton X-100. ChromomycinA3 was used as a nuclear marker. Slices were allowed to air drybefore mounting in SlowFade (Invitrogen).

Cells and tissues were analyzed by use of a Zeiss LSM 510 Metaargon/krypton laser scanning confocal microscope (Carl Zeiss, Jena,Germany). Images were acquired using the multitrack system toavoid cross talk among channels. The excitation/emission settingswere as follows: 430/505–550 nm for chromomycin, 543/560–615 nmfor CY3, and 633/650 nm for CY5. Images were confocally capturedby use of a 63� oil immersion objective (PlanApochromat; numericalaperture 1.4), with a maximal confocal zoom factor of 3, fixed boxsizes of 512 � 512 pixels, and pinhole below 1 Airy unit. Each imagewas acquired four times, and the signal was averaged to improve thesignal to noise ratio.

Western Blot

Cells were washed two times in cold PBS and lysed with lysissolution (150 mM NaCl, 1 mM EDTA, pH 7.4, 10 mM Tris-HCl, pH8, 1% SDS, and protease inhibitors). Lysates (100–150 �g) wereboiled 5 min in Laemmli SDS loading buffer and separated by 8%SDS-polyacrylamide gel electrophoresis. Filters were incubatedovernight at 4°C with the following antibodies: a) rabbit anti-Kv7.2(dilution, 1:2000; Dr. H. Lerche) (Luisi et al., 2009); b) rabbit anti-Kv7.3 (dilution, 1:1000; Dr. Alvaro Villarroel); c) mouse anti Kv7.4(dilution, 1:500; Neuromab, Davis, CA); d) mouse anti-MyHC (dilu-tion, 1:1000; Millipore). An anti--tubulin antibody (dilution, 1:5000;Sigma-Aldrich) was used to check for equal protein loading. Reactivebands were detected by chemiluminescence (ECL-plus; GE Health-care). Images were analyzed on a ChemiDoc station with Quantity-one software (Bio-Rad, Segrate, Italy).

Materials

N-(2-Amino-4-(4-fluorobenzylamino)-phenyl)-carbamic acid ethylester (retigabine) was from ASTA Medica (Radebeul, Germany) orValeant Pharmaceuticals (Aliso Viejo, CA); XE-991 was from TocrisBioscience (Bristol, UK). Fura-2 (Calbiochem-Inalco, Milan, Italy)was kept as 1 mM stock solutions in dimethyl sulfoxide at �20°C. Allthe remaining material was from Sigma-Aldrich.

Statistics

Data are expressed as means� S.E.M. of the given number ofexperiments (n). Data sets were compared by use of matched Stu-dent’s t tests or, if necessary, with one-way analysis of variance,followed by the Newman-Keul test. Statistically significant differ-ences were accepted when p was �0.05.

ResultsExpression of Kv7 K� Channel Transcripts and Pro-

teins during C2C12 Cell Differentiation. To investigatethe relative changes in the expression of transcripts encodingfor distinct classes of Na�, Ca2�, and K� channels in in vitrodifferentiating C2C12 cells, RT-PCR experiments were per-formed with use of gene-selective primers (Supplemental Ta-ble 1). For each gene, its expression was quantified as rela-tive to that of the constitutive ribosomal protein S16. Thedata obtained (shown in Table 1) revealed that, although themRNAs from most of the ion channel genes presently testedcould not be detected at either myoblast or myotube stage,other mRNAs, such as those encoding for Kir2.1, Kir2.2, andKv10.1, were already expressed at the myoblast stage.Among Kv7 members, the present RT-PCR results revealedthat, in addition to Kv7.1 and Kv7.5 transcripts whose skel-etal muscle expression has been described previously (Roura-Ferrer et al., 2008), those encoding for Kv7.2, Kv7.3, and

Kv7 Expression and Function in Skeletal Muscle 813

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

Kv7.4 were also detected in C2C12 myoblasts cultured in GM(Table 1 and Fig. 1A).

C2C12 cell incubation for 72 h in DM, an experimentalcondition promoting myotube formation (Supplemental Fig.1), induced the novel expression of transcripts encoding forother ion channels such as Cav1.1, Cav3.1, Nav1.4, Nav1.9,and, among voltage-gated K� channels, Kv1.5, Kv3.1, andKv3.4. In addition, the expression levels of the majority of ionchannel transcripts detected at the myoblast stage did notincrease during the differentiation process, rather showing insome cases a substantial decrease (KCa3.1, Kv10.1). By con-trast, time course experiments revealed that the abundanceof the mRNAs encoding for Kv7.1, Kv7.3, and Kv7.4, but notfor Kv7.2 and Kv7.5, showed a time-dependent increase after48 h of cell exposure to DM, reaching a plateau at 72 h (Table1 and Fig. 1A). The fact that Kv7 genes appear as the only theion channel subclass whose mRNAs can be detected in C2C12

myoblasts, and whose transcript levels are increased uponmyotube formation, suggests their potential role during skel-etal myogenesis.

Because the RT-PCR technique does not allow one to quan-tify in absolute terms the extent of gene expression changes,quantitative real-time PCR experiments, using a relative-comparative method, were performed. This technique, bycomparing the cycle-dependent increase in the relative fluo-rescence of the amplified product, allows one to define, foreach experimental condition, a cycle threshold (ct). This pa-rameter is then normalized to the ct value of a constitutivegene (S16) whose expression does not change during theexperimental procedures, thus defining a �ct value (Fig. 1B).Comparison of �ct values obtained in nondifferentiated ver-sus differentiated C2C12 cells (��ct), revealed that Kv7.1,Kv7.3, and Kv7.4 mRNA expression was increased 15-, 10-,and 25-fold after 72 h of myoblast exposure to DM, respec-tively; real-time PCR experiments, similarly to semiquanti-tative RT-PCR, did not reveal significant changes in Kv7.2and Kv7.5 mRNA abundance during C2C12 cell differentia-tion in vitro (data not shown).

To evaluate the expression of the proteins encoded by theKv7.2, Kv7.3, and Kv7.4 mRNAs herewith described for thefirst time, Western blot analysis using subunit-specific anti-bodies was performed in cell lysates from C2C12 myoblastsand myotubes (72 h of DM exposure) (Fig. 2). Antibody spec-ificity was tested in parallel Western blot experiments by use

of membrane fractions from control CHO cells and from CHOcells transfected with the respective Kv7 cDNA. Each anti-Kv7 antibody revealed a single subunit-specific band only inCHO cells transfected with the corresponding cDNA; un-transfected CHO cells did not show any specific signal. InC2C12 cells (Fig. 2A, right), anti-Kv7.2, -Kv7.3, and -Kv7.4antibodies recognized a single band of 95, 97, and 77 kDa,respectively; the size of these bands was in accordance withthe predicted molecular mass of the recognized proteins, andidentical to those identified in transfected CHO cells. More-over, C2C12 cell differentiation significantly increased theintensities of the bands corresponding to Kv7.3 and Kv7.4,but not that of Kv7.2. Figure 2B quantifies the pooled resultsof the densitometric analysis of Kv7-specific bands in nondif-ferentiated and differentiated C2C12 cells. These results sug-gest that, in addition to Kv7.1 and Kv7.5 subunits (Roura-Ferrer et al., 2008), Kv7.2, Kv7.3, and Kv7.4 subunits are alsoexpressed in C2C12 cells, with Kv7.3, and more so Kv7.4,being up-regulated during myotube formation in vitro.

Expression of Kv7.2, Kv7.3, and Kv7.4 K� ChannelTranscripts and Proteins in Adult Skeletal Musclefrom Mice and Humans. In mice skeletal muscle tissue,mRNAs encoding for all Kv7 subunits could be detected byRT-PCR (Fig. 3A). Furthermore, confocal immunofluores-cence experiments on mouse skeletal muscle cryosectionsrevealed that Kv7.2 and Kv7.3 immunofluorescence wasmainly associated with intracellular striations; in addition,Kv7.2, but not Kv7.3, immunofluorescence also seemed to beexpressed along the sarcolemmal membrane. By contrast,Kv7.4 immunoreactivity seemed to be exclusively localized atthe level of the plasma membrane (Fig. 3B). To gain furtherinsight into the subcellular localization of Kv7.2, Kv7.3, andKv7.4 subunits, experiments were performed with use ofantibodies against AnkG, a scaffold protein localized withinthe sarcoplasmic reticulum in adult skeletal muscle fibers(Kordeli et al., 1995, 1998). In these experiments, the nuclearmarker chromomycin A3 was also used. In longitudinal sec-tions, AnkG immunofluorescence was associated with boththe subplasmalemmal region of the sarcomere and its intra-cellular striations, labeling two bands per sarcomere. Each ofthese bands is likely to correspond to a triad running on eachside of a Z-line; within each triad, AnkG has been shown to beconcentrated along plasmalemmal T-tubules (Flucher et al.,1990). A similar distribution pattern has been shown for

TABLE 1Ion channel genes expressed in C2C12 myoblasts and myotubes

Gene Myoblasts Myotubes Gene Myoblasts Myotubes

Cav1.1 � � Kv2.2 � �Cav1.2 � � Kv3.1 � �Cav1.3 � � Kv3.4 � �Cav3.1 � � Kv4.1 � �Cav3.2 � � Kv7.1 � 1Cav3.3 � � Kv7.2 � �Nav1.4 � � Kv7.3 � 1Nav1.5 � � Kv7.4 � 1Nav1.9 � � Kv7.5 � �Kir2.1 � � Kv10.1 � 2Kir2.2 � � Kv10.2 � �Kv1.1 � � Kv11.1 � �Kv1.4 � � KCa1.1 � �Kv1.5 � � KCa2.2 � �Kv1.7 � � KCa2.3 � �Kv2.1 � � KCa3.1 � �

�, not detected; �, detected; 2, down-regulated in myotubes; 1, up-regulated in myotubes.

814 Iannotti et al.

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

ryanodine receptors-1 (RyR-1) (Salanova et al., 2002). In thesame sections, Kv7.2- and Kv7.3-positive striations were thinand nonoverlapping with those recognized by anti-AnkG an-tibodies, being instead positioned between each pair of thickAnkG-labeled triads; these results suggest the possible ex-pression of Kv7.2 and Kv7.3 subunits in a sarcomeric patternat the level of the Z-line. Whereas Kv7.2 and Kv7.3 exhibitedintracytoplasmic immunoreactivity, Kv7.4 immunostainingdid not colocalize with AnkG at the cytoplasmic level, beingdetected only along the plasmalemmal membrane, as also re-vealed by the subsarcolemmal position of the myotube nuclei.

In human skeletal muscle, Kv7.2, Kv7.3, and Kv7.4 mRNAswere detected by RT-PCR (Fig. 3C). Moreover, an expressionpattern similar to that described previously in mouse muscletissue was observed with confocal immunohistochemistry(Fig. 3D); whereas Kv7.2 and Kv7.3 immunoreactivityseemed to be prevalently associated with intracytoplasmic

striations, sarcolemmal labeling was revealed instead by theanti-Kv7.4 antibody.

Activation of Kv7 Channels Regulates C2C12 CellProliferation and Differentiation In Vitro. To investi-gate the possible functional role of Kv7 channels in skeletalmuscle proliferation and differentiation, C2C12 cells wereexposed to retigabine, a compound acting as a selective acti-vator of neural Kv7 channels, but unable to modulate cardiacKv7.1 channels (Miceli et al., 2008), and to XE-991, a Kv7channel blocker that only poorly disciminates among chan-nels formed by different Kv7 subunits (Wang et al., 1998).Proliferation was measured by [3H]thymidine incorporation,whereas differentiation was quantified by the transcriptslevels of myogenin, a muscle-specific transcription factor re-quired for myotube formation (Bennett and Tonks, 1997).Retigabine (10 �M) reduced C2C12 myoblast proliferation tovalues similar to those observed in cells exposed to DM. The

B

Am

RN

Aexp

ress

ion

(2- ΔΔct

)

Kv7.1 Kv7.3 Kv7.40

0,01

0,02

0,03

0,04

0,05

0,06 myoblastsmyotubes

*

*

*

S16(bp)

100200300

Kv7.1 Kv7.2

500400

Kv7.3 Kv7.4 Kv7.5

1

2

GM 24 48 72

mR

NA

exp

ress

ion

(OD

Kv7

.1/

OD

S1

6)

* *

GM 24 48 72

time (h) in DM

0,5

1,0

1,5

mR

NA

exp

ress

ion

(no

rmali

zed

OD

)

1

2

3

GM 24 48 72

mR

NA

exp

ress

ion

(OD

Kv7

.2/

OD

S1

6)

1

2

GM 24 48 72

* *

mR

NA

exp

ress

ion

(OD

Kv7

.3/

OD

S1

6)

1

2

GM 24 48 72

* *

mR

NA

exp

ress

ion

(OD

Kv7

.4/

OD

S1

6)

1

2

GM 24 48 72

mR

NA

exp

ress

ion

(OD

Kv7

.5/

OD

S1

6)

100200300400 700

600

100200300400

100200300400

100200300400

Fig. 1. Expression of Kv7 transcripts dur-ing C2C12 cells differentiation. A, timecourse of Kv7 mRNA expression duringC2C12 cell in vitro differentiation. C2C12myoblasts were cultured in GM or ex-posed to DM for 24, 48, and 72 h. TotalmRNA was extracted and retrotrans-cribed, and the resulting cDNA was am-plified with Kv7-selective primers. Foreach gene, the top row shows agarose gelelectrophoresis of the RT-PCR productsat each time point (molecular mass of thestandards is shown on the left), whereasthe quantifications of the OD relative tothat of the reference gene S16 are re-ported in the bottom row. Each point isthe mean � S.E.M. of three separate de-terminations. �, p � 0.05 versus C2C12myoblasts (GM). Amplicon size was: 143bp for Kv7.1; 490 bp for Kv7.2; 238 bp forKv7.3; 100 bp for Kv7.4; 180 bp for Kv7.5.B, quantification of transcripts for Kv7members by use of quantitative real-timePCR. Data are expressed as 2-�ct rela-tive to S16, as described in Materials andMethods. �, myoblasts (GM); f, myotubes(DM for 72 h). Each bar is the mean �S.E.M. of at least four separate determina-tions. �, p � 0.05 versus myoblasts.

Kv7 Expression and Function in Skeletal Muscle 815

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

Kv7 blocker XE-991 (10 �M) failed to interfere with C2C12

cell proliferation, although it prevented the inhibitory effectexerted by retigabine (Fig. 4A). In RT-PCR experiments,retigabine (10 �M) increased the mRNA expression level formyogenin after 24 and 48 h of cell exposure to DM, an effectprevented by XE-991 (10 �M) (Fig. 4B); similar results werealso obtained with real-time PCR experiments (data notshown). These results suggest that the reduced C2C12 myo-blast proliferation caused by retigabine may be attributed, atleast in part, to an enhanced or anticipated differentiation ofmyoblasts into myotubes triggered by the pharmacologicalactivation of neural Kv7 channels.

Kv7 Channel Activation by Retigabine Prevents Sta-tin-Induced C2C12 Cell Toxicity. Skeletal muscle toxicityis a serious although uncommon side effect of therapy with3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitorsof the statin class. Because statin-induced myotoxicity in-

volves an inhibition of myotube formation in vitro (Baba etal., 2008), an effect opposite to that prompted by neural Kv7channel activation in the present experiments, the conse-quences of the pharmacological manipulation of Kv7 chan-nels on C2C12 myoblast viability after exposure to mevastatin(MEV) were evaluated. When proliferating C2C12 myoblastswere exposed to MEV (0.1–100 �M), a dose- and time-depen-dent reduction in cell viability was observed. The IC50 forMEV-induced toxicity was time-dependent, �20 �M after24 h, �6 �M after 48 h, and �2 �M after 72 h of drugexposure (Fig. 5A). Compared with myoblasts, differentiatedC2C12 myotubes seemed considerably less sensitive to statin-induced toxicity; in fact, the percentage of survival afterexposure to 3 �M MEV for 48 h was 57.5 � 4.4 (n 8) and93 � 6% (n 4) in myoblasts and myotubes, respectively (p �0.05). Even more dramatic differences were observed whenMEV concentrations were increased to 100 �M (cell survivalwas 16 � 0.5%, n 6, and 75 � 4%, n 5, in C2C12 myoblastsand myotubes, respectively; p � 0.05). The potential myopro-tective actions of Kv7 modulators was assessed upon simul-taneous incubation of proliferating C2C12 cells with retigab-ine, XE-991, and retigabine � XE-991 together with 3 �MMEV for 48 h, a value close to the calculated EC50 for thestatin. Retigabine (1–100 �M) dose-dependently and fullyprotected C2C12 cells from MEV-induced cytotoxicity; theIC50 for retigabine-induced myoprotection was �7 �M (Fig.5B). At each time point (24, 48, and 72 h), XE-991 (10 �M;Fig. 5C) counteracted the protective effect of retigabine onMEV-induced toxicity, strongly suggesting that the Kv7-ac-tivating potential of the drug was responsible for this phar-macological action.

DiscussionExpression and Modulation of KV7 Channel Mem-

bers in Developing and Adult Skeletal Muscle. Changesin K� channels function exert a complex regulation of cellfate, depending on cell type and differentiation stage, as wellas on the specific channel subtype. Indeed, in neurons, anincreased activity of plasma membrane K� channels is one ofthe earliest events triggered by neurodegenerative stimuli; inmost of these cases, the inhibition of K� efflux prevents celldeath (Yu, 2003). Enhanced K� channel function decreases[K�]i, causing cell shrinkage and cytosolic acidification, acti-vates some key enzymes involved in apoptosis, such ascaspases and endonucleases, and enhances the formation ofcytochrome c-dependent apoptosomes (Yu, 2003). By con-trast, in skeletal muscle cells, the early phases of myotubeformation depend on the coordinated expression of K� chan-nels of the delayed-rectifier (Kv10.1) (Bijlenga et al., 1998)and inward-rectifier (KIR2.1) (Wieland and Gong, 1995; Fi-scher-Lougheed et al., 2001) families. This phenomenonleads to a marked hyperpolarization of the cell membranepotential, a necessary prerequisite for Ca2�-dependent myo-blast acquisition of fusion competence (Bijlenga et al., 2000).

To identify additional ion channels that might participatein the early phases of skeletal muscle development, an ex-tensive analysis of the changes in the expression levels of anarray of ion channel transcripts during in vitro C2C12 myo-blast differentiation into myotubes was carried out in thepresent study. The results obtained reveal that, in addition toKv7.1 and Kv7.5 transcripts, whose expression in skeletal

Kv7.2

Kv7.3

Kv7.4

αα-tubulin

MW(kDa)m

yobla

sts

myo

tubesA

CHO/Kv7.2CHO

95

77

97

55

CHO/Kv7.3CHO

CHO/Kv7.4CHO

B myoblasts

myotubes

Kv7

.x p

rote

inexp

ress

ion

(no

rmali

zed

tom

yo

bla

sts)

0

1

2

3

Kv7.2 Kv7.4Kv7.3

*

*

Fig. 2. Expression of Kv7 subunits during C2C12 cell differentiation.A, Western blots on total cell lysates from CHO (left) and C2C12 cells(right); Kv7 subunits were detected by use of anti-Kv7-specific antibodies.C2C12 myoblasts were grown in GM, whereas myotubes were obtainedupon C2C12 cell exposure to DM for 72 h. The gels at the bottom of eachimage show the expression, on the same filters, of the intracellularprotein -tubulin, used as an internal standard for equal protein loading.B, quantification of the data shown in A. For each experiment, the OD ofthe Kv7.x bands was obtained by densitometric analysis; this value wasdivided by that of each respective -tubulin band intensity. Data areexpressed after normalization to the ratio OD Kv7.x/OD -tubulin inmyoblasts. �, myoblasts (GM); f, myotubes (DM for 72 h). Each datapoint is the mean � S.E.M calculated from four separate experiments.Asterisks denote values significantly different (p � 0.05) from C2C12myoblasts.

816 Iannotti et al.

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

muscle tissue had been reported previously (Lerche et al.,2000; Schroeder et al., 2000; Tsevi et al., 2005), Kv7.2, Kv7.3,and Kv7.4 mRNAs are also present in skeletal muscle cells.Moreover, Kv7.1, Kv7.3, and Kv7.4 transcripts, but not Kv7.2and Kv7.5, were up-regulated upon myotube formation, ar-guing in favor of a potential role for their protein productsduring myogenesis. Roura-Ferrer et al. (2008) recently inves-tigated the expression of Kv7.1 and Kv7.5 transcripts andproteins in differentiating rat L6E9 myoblasts; the presentresults confirm that Kv7.5 mRNAs did not increase duringmyotube formation in vitro, and reveal a previously un-no-ticed increase in Kv7.1 mRNA.

The expression of Kv7.2, Kv7.3, and Kv7.4 proteins inC2C12 myoblasts, as well as the up-regulation of Kv7.3 andKv7.4 (but not of Kv7.2) in C2C12 myotubes, was also con-firmed by Western blot experiments. The subcellular local-ization of these proteins in adult skeletal muscle from miceand humans was studied by immunohistochemistry. In lon-gitudinal sections from mouse muscle, Kv7.2 subunitsseemed to be located along the plasma membrane and withinthe cytoplasm, at the boundary between sarcomeres, flankedon both sides by AnkG-positive triads. This region seems tocorrespond to the Z-line, a multiprotein complex playing anintegral role in maintaining striated muscle structure andfunction, and showing a crucial role in the pathogenesis ofhuman skeletal muscle myopathies (Sheikh et al., 2007). Asimilar Z-line distribution to Kv7.2 has been also observed forKv7.3, whereas Kv7.4 expression seems restricted to theplasma membrane. It is noteworthy that Kv7.2 and Kv7.3subunits localize in intracellular regions where InsP3R andthe scaffolding protein Homer have been identified (Salanova

et al., 2002). Although the precise functional and structuralinter-relationships between these Z-line and sarcoplasmicreticulum proteins is still debated (Volpe et al., 2004), itseems plausible that they may regulate the function of thesarcoplasmic reticulum, thereby controlling skeletal muscleCa2� homeostasis and excitation-contraction coupling.

Functional Role of KV7 Channel Members in SkeletalMuscle Proliferation and Differentiation. Retigabine isa novel IKM activator under clinical scrutiny for activityagainst hyperexcitability diseases, such as epilepsy, mi-graine, and chronic pain (Miceli et al., 2008). Retigabineactivates with different apparent affinity homomeric or het-eromeric channels formed by Kv7 neural subunits, but isunable to potentiate the currents carried by cardiac Kv7.1channels. Retigabine enhances neural Kv7 currents by caus-ing a hyperpolarizing shift of the voltage dependence of ac-tivation and/or by increasing the maximal amount of currentelicited by strong depolarizations (Miceli et al., 2008). In thepresent study, retigabine markedly reduced C2C12 myoblastproliferation and enhanced myotube formation, possibly be-cause the drug enhanced the contribution of Kv7 channels toC2C12 membrane potential control, thereby rendering morenegative the cell membrane potential, a phenomenon knownto facilitate skeletal muscle differentiation and to halt cellproliferation (Wieland and Gong, 1995; Bijlenga et al., 1998,2000; Fischer-Lougheed et al., 2001). Kv7.1, Kv7.3, and Kv7.4transcripts clearly seem to be up-regulated at a differentia-tion stage at which Kv10.1 transcripts show a decreased ex-pression (Table 1); the down-regulation of the Kv10.1 channelresults in a hyperpolarization to approximately �65 mV of themembrane potential of myoblasts as a result of the major re-

A CHuman muscle7.2 7.3 7.4

Mouse muscle7.57.1 7.2 7.3 7.4 (bp)(bp)

400300

300400

200200

100100

B D

KV7.3

KV7.2

AnkG MergeChromomycin

MergeAnkG Chromomycin

AnkG Chromomycin

KV7.3

KV7.2

MergeKV7.4 KV7.4

Fig. 3. Expression of Kv7 genes andproteins in adult skeletal musclefrom mice and humans. A, agarosegel electrophoresis of RT-PCR prod-ucts obtained from cDNA amplifica-tion of mouse muscle mRNA usingKv7-selective primers (Supplemen-tary Table 1). Amplicon sizes arereported in the legend to Fig. 1.B, confocal immunofluorescenceimages of mouse soleus muscle lon-gitudinal cryosections. Red, Kv7 im-munoreactivity (top, Kv7.2; middle,Kv7.3; bottom, Kv7.4); green, AnkGimmunoreactivity; blue, nucleiidentified with chromomycin A3.Scale bar, 10 �m for Kv7.2; 5 �m forKv7.3, and 5 �m for Kv7.4. Insets inthe Kv7.3 images show magnifica-tion, 3�. C, agarose gel electro-phoresis of RT-PCR products ob-tained from cDNA amplification ofhuman brachial biceps mRNA byuse of Kv7-selective primers. Ampli-con size: 115 bp for Kv7.2; 121 bp forKv7.3; and 192 bp for Kv7.4. D, con-focal immunofluorescence imagesof human brachial biceps musclecryosections. Kv7 immunoreactivityis shown in red, and nuclei are iden-tified in green with chromomycinA3. Scale bar, 5 �m for Kv7.2; 5 �mfor Kv7.3; and 10 �m for Kv7.4.

Kv7 Expression and Function in Skeletal Muscle 817

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

maining influence of KIR channels; this would permit an en-hanced impact of subthreshold Kv7 channels under moderatelydepolarizing conditions.

It should be mentioned that, in addition to activating neu-ral Kv7 channels, retigabine also affects other voltage- andligand-gated channels, but these pharmacological actions re-quire concentrations higher than those used in the presentstudy (Rundfeldt and Netzer, 2000). Moreover, the fact thatXE-991 fully counteracted retigabine-induced effects onC2C12 cell proliferation and differentiation strongly suggeststhat neuronal Kv7 channels act as molecular targets forretigabine. Although the pharmacological activation of neu-ral Kv7 channels by retigabine effectively interfered withC2C12 myoblast proliferation and differentiation, both phe-nomena were not affected by the Kv7 channel blocker XE-991. This result contrasts with that of Roura-Ferrer et al.(2008), showing that the IKM blocker linopiridine caused a60% reduction of C2C12 cell proliferation. However, in thisstudy, linopiridine was used at a concentration (100 �M) thatdoes not discriminate among voltage-gated potassium chan-nel subfamilies (Wang et al., 1998); therefore, further exper-iments are needed before this pharmacological effect may beattributed to a selective involvement of Kv7 channels.

Several plausible explanations might be given to interpretthe inability of XE-991 to interfere with C2C12 cell prolifer-ation and differentiation in the present experiments. First,Kv7 channels only contribute to a limited amount of outwardcurrent, and therefore their blockade does not significantlyinfluence C2C12 cell membrane potential. The fact that Kv7channels are not strongly active at the resting membranepotential, but only become so when activated by retigabine,has already been described in peripheral unmyelinated C-type nerve fibers (Lang et al., 2008) and in central myelin-ated axons (Rivera-Arconada and Lopez-Garcia, 2006; Ver-vaeke et al., 2006). Moreover, it should be emphasized thatXE-991 was used in the present experiments at 10 �M;although this concentration causes an almost complete blockof homomeric or heteromeric Kv7.2 or Kv7.3 channels (Wanget al., 1998), Kv7.4 (Søgaard et al., 2001), and, even more so,Kv7.5 (Yeung et al., 2008b) channels display a much lowersensitivity to blockade by XE-991. Therefore, the possibilityexists that the described insensitivity to XE-991 of C2C12 cellproliferation and differentiation might be ascribed to thepreferential involvement of Kv7.4 and Kv7.5 subunits. Al-though further experiments using antisense (Fischer-Lougheed et al., 2001) or RNA interference are needed toclarify the role of specific Kv7 subunits, the present resultsshowing a selective up-regulation of Kv7 mRNAs duringC2C12 cell differentiation argue in favor of a significant role ofneural Kv7 subunits in skeletal muscle maturation.

The Pharmacological Activation of Neural KV7Channels Protects Skeletal Muscle Cells from Statin-Induced Toxicity. Myopathy involving skeletal muscleweakness, tenderness, and pain, possibly associated toplasma creatine kinase elevation, is among the rare butpotentially severe toxicities associated with statins, widelyprescribed lipid-lowering agents acting as 3-hydroxy-3-meth-ylglutaryl coenzyme A reductase inhibitors (Sirvent et al.,2008). Despite the widespread concern over this clinicallyrelevant side effect, knowledge of its molecular pathogenesisis still a matter of debate; statins have been reported to affectmitochondrial function, thereby leading to a Ca2� leak thatdirectly interferes with the regulation of sarcoplasmic retic-ulum Ca2� cycling (Klopstock, 2008; Sirvent et al., 2008). Asa matter of fact, electron microscopy studies in skeletal mus-cles from statin-treated individuals have revealed ultrastruc-tural changes mostly located at the T-tubular level, even inthe absence of clinical and biochemical evidence of skeletalmuscle damage (Mohaupt et al., 2009). In consideration ofthe close proximity of retigabine-sensitive Kv7.2, Kv7.3, andKv7.4 subunits to key sites for [Ca2�]i handling in adultskeletal muscle fibers, the potential myoprotective effect ofretigabine on statin-induced C2C12 cell death was investi-gated. Mevastatin-induced myotoxicity was fully counter-acted by retigabine in a XE-991-sensitive manner, stronglysupporting the hypothesis that Kv7-opening actions werecritically involved in retigabine-induced myoprotection. Theapparent IC50 for retigabine-induced reversal of mevastatin-triggered myotoxicity was approximately 7 �M; although thisvalue is significantly higher than the EC50 reported for reti-gabine-induced activation of homomeric Kv7.2 or Kv7.3, orheteromeric Kv7.2/Kv7.3 or Kv7.3/Kv7.5 channels (0.3–2 �M),it is very close to that described for Kv7.4 homomeric chan-nels (5 �M) (Tatulian et al., 2001; Yeung et al., 2008a). Thisobservation suggests that Kv7.4 subunits may be the molec-

A

B

XE-991 RT RT+

XE-991

0.1% FBS (DM)

**

10% FBS (GM)M

yo

gen

in m

RN

A e

xp

ress

ion

(OD

myo

gen

in/

OD

S1

6)

Control (DM)

RT

RT + XE-991

XE-991

0

0,5

1

1,5

24 48 720

**

50

100

Pro

life

rati

ng

C

2C

12

myo

bla

sts

(%)

Control

Time (h)

Fig. 4. Effect of the pharmacological activation of Kv7 channels on C2C12cell proliferation and differentiation in vitro. A, for proliferation assays,C2C12 myoblast growth was synchronized by cell exposure to DM for 30 to36 h; after this period, cell proliferation was assessed by [3H]thymidineincorporation in cells grown for a further 24 h in DM (�) or in GM (withor without the indicated drugs; f). Retigabine (RT) or XE-991 were eachused at 10 �M. Each point is the mean � S.E.M. of three to four separatedeterminations performed in quadruplicate. �, p � 0.05 versus GM con-trols (no drug added). B, C2C12 cell differentiation was evaluated bymeasuring myogenin mRNA expression with RT-PCR during cell expo-sure to DM for the indicated times. When necessary, drugs were added attime 0. Myogenin expression was quantified densitometrically, and ODswere expressed relative to the reference gene S16. Each point is themean � S.E.M. of three separate determinations performed in triplicate.�, p � 0.05 versus C2C12 cells grown in DM (no drug added).

818 Iannotti et al.

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

ular targets responsible for the myoprotective actions of re-tigabine in the present experimental model. Among Kv7 tran-scripts, Kv7.4 clearly shows the largest increase after in vitrodifferentiation of C2C12 cells. The molecular mechanism(s)underlying retigabine-induced myoprotection are still underinvestigation; nevertheless, in view of the fact that, comparedwith myotubes, myoblasts seem significantly more sensitiveto mevastatin-induced myotoxic actions, an important roleseems to be attributable to the ability of retigabine to pro-mote C2C12 cell differentiation and withdrawal from the cellcycle, an effect opposite to that prompted by statins in thesame cells (Baba et al., 2008).

All together, the present results indicate that the pharma-cological activation of neural Kv7 subunits regulates skeletalmuscle cell proliferation, differentiation, and survival aftermyotoxic stimuli, thus revealing a novel role for these chan-nels as potential therapeutic targets in drug-induced myop-athies. Moreover, myoprotective effects triggered by thepharmacological activation of neural Kv7 subunits may be ofclinical relevance also in other skeletal muscle disorders. Theretigabine structural analog flupirtine, a nonopioid analgesicthat also acts as a neural Kv7 channel opener (Miceli et al.,2008), seems effective against different types of musculoskel-etal pains (Friedel and Fitton, 1993), and displays significantskeletal muscle-relaxing effects in humans (Lobisch et al.,1996). It remains to be investigated whether, in addition toits effects on neural sensory information processing, flupir-tine-induced opening of Kv7 channels at the skeletal musclelevel is also involved in these pharmacological effects.

Acknowledgments

We thank Prof. Andrea Graziani (University of Piemonte Orien-tale A. Avogadro, Novara, Italy) and Prof. Fabio Franciolini (Univer-

sity of Perugia, Perugia, Italy) for sharing C2C12 cells; Prof. ThomasJentsch (Leibniz-Institut fur Molekulare Pharmakologie; Berlin,Germany) for Kv7.2, Kv7.3, and Kv7.4 cDNAs; Profs. Holger Lercheand Snezana Maljevic (University of Ulm, Germany) and Prof. Mi-chael Schwake (Christian-Albrechts-University Kiel, Germany) foranti-Kv7.2 antibody; Prof. A. Villarroel (Consejo Superior de Inves-tigaciones Científicas, Universidad del Paìs Vasco, Bilbao, Spain) foranti-Kv7.3 antibody; Prof. Lucio Santoro (University of Naples Fed-erico II) for human skeletal muscle specimens; Prof. David Brown(University College, London, UK) for advice on the manuscript; andDrs. Vincenzo Romaniello (University of Naples Federico II) and Dr.Erika Di Zazzo (University of Molise, Campobasso, Italy) for techni-cal help. The monoclonal antibody Kv7.4 was developed by and/orobtained from the University of California Davis/National Institutesof Health NeuroMab Facility [supported by National Institutes ofHealth Grant U24-NS050606] and maintained by the Department ofNeurobiology, Physiology, and Behavior, College of Biological Sci-ences, University of California (Davis, CA).

ReferencesBaba TT, Nemoto TK, Miyazaki T, and Oida S (2008) Simvastatin suppresses the

differentiation of C2C12 myoblast cells via a Rac pathway. J Muscle Res Cell Motil29:127–134.

Bennett AM and Tonks NK (1997) Regulation of distinct stages of skeletal muscledifferentiation by mitogen-activated protein kinases. Science 278:1288–12891.

Bijlenga P, Liu JH, Espinos E, Haenggeli CA, Fischer-Lougheed J, Bader CR, andBernheim L (2000) T-type alpha 1H Ca2� channels are involved in Ca2� signalingduring terminal differentiation (fusion) of human myoblasts. Proc Natl Acad SciU S A 97:7627–7632.

Bijlenga P, Occhiodoro T, Liu JH, Bader CR, Bernheim L and Fischer-Lougheed J(1998) An ether -a-go-go K� current, Ih-eag, contributes to the hyperpolarizationof human fusion-competent myoblasts. J Physiol 512:317–323.

Brown DA and Adams PR (1980) Muscarinic suppression of a novel voltage-sensitiveK� current in a vertebrate neurone. Nature 283:673–676.

Brown DA and Passmore GM (2009) Neural KCNQ (Kv7) channels. Br J Pharmacol156:1185–1195.

Cooper E (2001) A new role for ion channels in myoblast fusion. J Cell Biol 153:F9–F12.

Cooper EC, Aldape KD, Abosch A, Barbaro NM, Berger MS, Peacock WS, Jan YN,and Jan LY (2000) Co-localization and coassembly of two human brain M-typepotassium channel subunits that are mutated in epilepsy. Proc Natl Acad SciU S A 97:4914–4919.

A B

Revers

al

of

MEV

-in

du

ced

toxic

ity (

%)

C

Cell v

iab

ilit

y

(OD

at

59

5 n

m)

0

0,2

24 h 48 h

GM

GM+MEV+RT+XE-991

GM+MEVGM+RT GM+MEV+RTGM+XE-991

72 h

*

Mevastatin ( M)

Cell v

iab

ilit

y

(norm

alize

d O

D a

t 5

95

nm

)

0,1 1 10 100

0,0

0,5

1,0

24h48h72h

1 10 100Retigabine ( M)

0

50

100

* * **

*

Fig. 5. Effect of Kv7 channel modulatorson mevastatin-induced cytotoxicity inC2C12 cells. A, dose- and time-dependenttoxicity exerted by mevastatin in C2C12myoblasts. MEV (0.1–100 �M) was addedto GM of proliferating C2C12 cells, andcell toxicity was assessed by the MTT re-duction assay 24 (f), 48 (F), or 72 h (�)later. Data are expressed as OD at 595nm, normalized to controls. In A to C,each point is the mean � S.E.M. of threeexperiments, each performed in quadru-plicate. The solid lines represent the fitsof the experimental data to the followingbinding isotherm: y max/(1 � X/IC50),where X is the MEV concentration. Fittedvalues for IC50 were �20 �M after 24 h ofexposure, �6 �M after 48 h, and �2 �Mafter 72 h of MEV incubation. B, reversalof MEV-induced toxicity by retigabine(RT). C2C12 myoblasts were exposed to 3�M MEV for 48 h in the presence of 1 to100 �M RT. The solid line represents thefit of the experimental data to the de-scribed binding isotherm. C, XE-991 pre-vents RT-induced protection from MEVtoxicity. Proliferating C2C12 cells were ex-posed to GM or GM added with MEV (3�M), RT (10 �M), XE-991 (10 �M), MEV(3 �M) � RT (10 �M), or MEV (3 �M) �RT (10 �M) � XE-991 (10 �M), as indi-cated, for 24, 48, and 72 h. At the end ofthese periods, mitochondrial function wasevaluated by the MTT reduction assay. �,p � 0.05 versus respective control (GMwith no drug added).

Kv7 Expression and Function in Skeletal Muscle 819

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from

Fioretti B, Pietrangelo T, Catacuzzeno L, and Franciolini F (2005) Intermediate-conductance Ca2�-activated K� channel is expressed in C2C12 myoblasts and isdownregulated during myogenesis. Am J Physiol Cell Physiol 289:C89–C96.

Fischer-Lougheed J, Liu JH, Espinos E, Mordasini D, Bader CR, Belin D, andBernheim L (2001) Human myoblast fusion requires expression of functionalinward rectifier Kir2.1 channels. J Cell Biol 153:677–686.

Flucher BE, Morton ME, Froehner SC, and Daniels MP (1990) Localization of thealpha 1 and alpha 2 subunits of the dihydropyridine receptor and ankyrin inskeletal muscle triads. Neuron 5:339–351.

Friedel HA and Fitton A (1993) Flupirtine. A review of its pharmacological proper-ties, and therapeutic efficacy in pain states. Drugs 45:548–569.

Kharkovets T, Hardelin JP, Safieddine S, Schweizer M, El-Amraoui A, Petit C, andJentsch TJ (2000) KCNQ4, a K� channel mutated in a form of dominant deafness,is expressed in the inner ear and the central auditory pathway. Proc Natl Acad SciU S A 97:4333–4338.

Klopstock T (2008) Drug-induced myopathies. Curr Opin Neurol 21:590–595.Kordeli E, Lambert S, and Bennett V (1995) AnkyrinG. A new ankyrin gene with

neural-specific isoforms localized at the axonal initial segment and node of Ran-vier. J Biol Chem 270:2352–2359.

Kordeli E, Ludosky MA, Deprette C, Frappier T, and Cartaud J (1998) AnkyrinG isassociated with the postsynaptic membrane and the sarcoplasmic reticulum in theskeletal muscle fiber. J Cell Sci 111:2197–2207.

Lang PM, Fleckenstein J, Passmore GM, Brown DA, and Grafe P (2008) Retigabinereduces the excitability of unmyelinated peripheral human axons. Neuropharma-cology 54:1271–1278.

Lerche C, Scherer CR, Seebohm G, Derst C, Wei AD, Busch AE, and Steinmeyer K(2000) Molecular cloning and functional expression of KCNQ5, a potassium chan-nel subunit that may contribute to neuronal M-current diversity. J Biol Chem275:22395–22400.

Lesage F, Attali B, Lazdunski M, and Barhanin J (1992) Developmental expressionof voltage-sensitive K� channels in mouse skeletal muscle and C2C12 cells. FEBSLett 310:162–166.

Lobisch M, Schaffler K, Wauschkuhn H, and Nickel B (1996) Clinical pilot study ofthe myogenic effects of flupirtine in comparison to tetrazepam and placebo. Arz-neimittelforschung 46:293–298.

Luisi R, Panza E, Barrese V, Iannotti FA, Viggiano D, Secondo A, Canzoniero LM,Martire M, Annunziato L, and Taglialatela M (2009) Activation of pre-synapticM-type K� channels inhibits [3H]D-aspartate release by reducing Ca2� entrythrough P/Q-type voltage gated Ca2� channels. J Neurochem 109:168–181.

Mackie AR, Brueggemann LI, Henderson KK, Shiels AJ, Cribbs LL, Scrogin KE, andByron KL (2008) Vascular KCNQ potassium channels as novel targets for thecontrol of mesenteric artery constriction by vasopressin, based on studies in singlecells, pressurized arteries, and in vivo measurements of mesenteric vascularresistance. J Pharmacol Exp Ther 325:475–483.

Miceli F, Soldovieri MV, Martire M, and Taglialatela M (2008) Molecular pharma-cology and therapeutic potential of neuronal Kv7-modulating drugs. Curr OpinPharmacol 8:65–74.

Mohaupt MG, Karas RH, Babiychuk EB, Sanchez-Freire V, Monastyrskaya K, IyerL, Hoppeler H, Breil F, and Draeger A (2009) Association between statin-associated myopathy and skeletal muscle damage. CMAJ 7:E11–E18.

Ohya S, Sergeant GP, Greenwood IA, and Horowitz B (2003) Molecular variants ofKCNQ channels expressed in murine portal vein myocytes: a role in delayedrectifier current. Circ Res 92:1016–1023.

Rivera-Arconada I and Lopez-Garcia GA (2006) Retigabine-induced population pri-mary afferent hyperpolarisation in vitro. Neuropharmacology 51:756–763.

Roura-Ferrer M, Sole L, Martínez-Marmol R, Villalonga N, and Felipe A (2008)Skeletal muscle Kv7 (KCNQ) channels in myoblast differentiation and prolifera-tion. Biochem Biophys Res Commun 369:1094–1097.

Rundfeldt C and Netzer R (2000) Investigations into the mechanism of action of thenew anticonvulsant retigabine. Interaction with GABAergic and glutamatergicneurotransmission and with voltage gated ion channels. Arzneimittelforschung50:1063–1070.

Salanova M, Priori G, Barone V, Intravaia E, Flucher B, Ciruela F, McIlhinney RA,Parys JB, Mikoshiba K, and Sorrentino V (2002) Homer proteins and InsP(3)receptors co-localise in the longitudinal sarcoplasmic reticulum of skeletal musclefibres. Cell Calcium 32:193–200.

Sanguinetti MC, Curran ME, Zou A, Shen J, Spector PS, Atkinson DL, and KeatingMT (1996) Coassembly of K(V)LQT1 and minK (IsK) proteins to form cardiac I(Ks)potassium channel. Nature 384:80–83.

Schroeder BC, Hechenberger M, Weinreich F, Kubisch C, and Jentsch TJ (2000)KCNQ5, a novel potassium channel broadly expressed in brain, mediates M-typecurrents. J Biol Chem 275:24089–24095.

Sheikh F, Bang ML, Lange S, and Chen J (2007) “Z”eroing in on the role of Cypherin striated muscle function, signaling, and human disease. Trends Cardiovasc Med17:258–262.

Sirvent P, Mercier J, and Lacampagne A (2008) New insights into mechanisms ofstatin-associated myotoxicity. Curr Opin Pharmacol 8:333–338.

Søgaard R, Ljungstrøm T, Pedersen KA, Olesen SP, and Jensen BS (2001) KCNQ4channels expressed in mammalian cells: functional characteristics and pharma-cology. Am J Physiol Cell Physiol 280:C859–C866.

Tatulian L, Delmas P, Abogadie FC, and Brown DA (2001) Activation of expressedKCNQ potassium currents and native neuronal M-type potassium currents by theanti-convulsant drug retigabine. J Neurosci 21:5535–5545.

Tsevi I, Vicente R, Grande M, Lopez-Iglesias C, Figueras A, Capella G, Condom E,and Felipe A (2005) KCNQ1/KCNE1 channels during germ-cell differentiation inthe rat: expression associated with testis pathologies. J Cell Physiol 202:400–410.

Vervaeke K, Gu N, Agdestein C, Hu H, and Storm JF (2006) Kv7/KCNQ/M-channelsin rat glutamatergic hippocampal axons and their role in regulation of excitabilityand transmitter release. J Physiol 576:235–256.

Volpe P, Sandri C, Bortoloso E, Valle G, and Nori A (2004) Topology of Homer 1c andHomer 1a in C2C12 myotubes and transgenic skeletal muscle fibers. BiochemBiophys Res Commun 316:884–892.

Walsh K and Perlman H (1997) Cell cycle exit upon myogenic differentiation. CurrOpin Genet Dev 7:597–602.

Wang HS, Pan Z, Shi W, Brown BS, Wymore RS, Cohen IS, Dixon JE, and McKinnonD (1998) KCNQ2 and KCNQ3 potassium channel subunits: molecular correlates ofthe M-channel. Science 282:1890–1893.

Wieland SJ and Gong QH (1995) Modulation of a potassium conductance in devel-oping skeletal muscle. Am J Physiol 268:C490–C495.

Yeung S, Schwake M, Pucovsky V, and Greenwood IA (2008a) Bimodal effects of theKv7 channel activator retigabine on vascular K� currents. Br J Pharmacol 155:62–72.

Yeung SY, Lange W, Schwake M, and Greenwood IA (2008b) Expression profile andcharacterisation of a truncated KCNQ5 splice variant. Biochem Biophys Res Com-mun 371:741–746.

Yeung SY, Pucovsky V, Moffatt JD, Saldanha L, Schwake M, Ohya S, and Green-wood IA (2007) Molecular expression and pharmacological identification of a rolefor K(v)7 channels in murine vascular reactivity. Br J Pharmacol 151:758–770.

Yu SP (2003) Regulation and critical role of potassium homeostasis in apoptosis.Prog Neurobiol 70:363–386.

Address correspondence to: Dr. Maurizio Taglialatela, Department ofHealth Sciences, University of Molise. Via De Sanctis, 86100, Campobasso,Italy. E-mail: m.taglialatela@unimol.it

820 Iannotti et al.

at ASPE

T Journals on A

ugust 20, 2018jpet.aspetjournals.org

Dow

nloaded from