999

Modulation of an Inactivating Human CardiacK' Channel by Protein Kinase C

Katherine T. Murray, Stephen A. Fahrig, Karen K. Deal, Sunny S. Po, Ning Ning Hu,Dirk J. Snyders, Michael M. Tamkun, Paul B. Bennett

Abstract The transient outward current (ITO) is an importantrepolarizing component of the cardiac action potential. In nativecardiac myocytes, ITO is modulated after activation of proteinkinase C, although the molecular nature of this effect is not wellunderstood. A channel recently cloned from human ventricularmyocardium (KvL.4, HK1) produces a rapidly inactivating K'current, which has phenotypic similarities to the 4-aminopyri-dine-sensitive component of ITO- Therefore, we examinedwhether this recombinant channel was also modulated by proteinkinase C activation by investigating the effects of the diacylglyc-erol analogue phorbol 12-myristate 13-acetate (PMA) on Kvl.4K' current expressed in Xenopus oocytes. At a concentration of10 nmol/L, PMA caused a biphasic response with an initialincrease (14+4%, mean+SEM) in current, which peaked in 14minutes. This was followed by a significant reduction (40+11%)in the current within 30 minutes. There was no significant changein cell membrane electrical capacitance with 10 nmol/L PMA

(1±1% decline in 30 minutes), demonstrating that loss of cellmembrane surface area did not explain the reduction in K'current, although cell capacitance did decrease when using ahigher concentration of PMA (81 nmol/L). The inactive stereo-isomer, 4a-PMA, had no effect on Kvl.4 current, whereaspreincubation with the protein kinase inhibitor staurosporine orprotein kinase C-selective chelerythrine prevented the effects ofPMA. When purified from a stably transfected mammalian cellline by using immunoprecipitation, the channel protein wasreadily phosphorylated in vitro by purified protein kinase C.These results indicate that human Kvl.4-induced current ismodulated by protein kinase C activation and suggest a role fordirect K' channel phosphorylation as the molecular mechanismof this effect. (Circ Res. 1994;75:999-1005.)Key Words * K' channels * transient outward current -

protein kinase C * heart

R apidly inactivating or A-type K' currents playan important role in repolarization in a widevariety of excitable cells. In cardiac myocytes,

the transient outward current (ITO) is an A-type K'current thought to be largely responsible for earlyrepolarization or phase 1 of the cardiac action poten-tial.1,2 Although its importance varies among tissuetypes and species, a large component of ITO has beendemonstrated in atrial as well as ventricular myocytesobtained from human heart.3-5 Protein kinase C playsan important role in modulating ion channel function inmany types of cells.6 Recent advances in the cloning ofvoltage-gated K' channels7'8 provide a unique opportu-nity to study molecular mechanisms of kinase effects onthe function of channels derived from human tissue.Previous studies have shown that agents such as a-ad-renergic receptor agonists, which stimulate phosphati-dylinositol turnover and the subsequent activation ofprotein kinase C, suppress ITO, with a resultant increasein action potential duration.9-12 However, investigationof this current in native myocytes is complicated by thepresence of multiple current components.2"13

Received April 12, 1994; accepted August 25, 1994.From the Departments of Pharmacology (K.T.M., K.K.D.,

S.S.P., N.N.H., D.J.S., M.M.T., P.B.B.), Medicine (K.T.M., S.A.F.,S.S.P., D.T.S., P.B.B.), and Molecular Physiology and Biophysics(M.M.T.), Vanderbilt University Medical Center, Nashville, Tenn.

Previously published as preliminary results in abstract form(Circulation. 1992;86[suppl I]:I-26; Biophys J. 1993;64:A313).Correspondence to Katherine T. Murray, MD, Vanderbilt Uni-

versity Medical Center, Department of Pharmacology, 560 Medi-cal Research Bldg, Nashville, TN 37232-6602.© 1994 American Heart Association, Inc.

To explore the role of protein phosphorylation in K'channel function at the molecular level, we examinedwhether protein kinase C activation could result in bothmodulation of channel activity and phosphorylation ofpurified channel protein when using an inactivating K'channel cloned from human cardiac tissue.7,8 ThecDNA encoding this channel was originally termed HK1and is designated Kvl.4 according to a proposed nomen-clature for vertebrate K' channel genes.14 The aminoacid sequence contains multiple potential phosphoryla-tion sites for protein kinase C,15,16 suggesting that thismay be a mechanism of channel modulation.

Materials and MethodsExperimental PreparationThe synthesis ofRNA from Kvl.4 cDNA and preparation of

Xenopus oocytes have been previously described.817 Afterdisaggregation, the cells were placed in a physiological solution(ND-96, see below) and incubated at room temperature for 1to 24 hours until cytoplasmic RNA injection. By use ofRNase-free water, 5'-capped cRNA was diluted so that K'current expressed in 24 to 48 hours was <5 ,uA when 30 to 35nL was injected. The oocytes were then kept at 18°C for 1 to 4days until experimentation. Each batch of cells was screenedfor endogenous Cl- currents by using depolarizing voltage-clamp steps (range, -110 to +40 mV). Oocytes having endog-enous currents at these voltages, which were >1% of ex-pressed currents, were not used.

Electrophysiological MeasurementsOocytes were voltage-clamped by using the two-microelec-

trode voltage-clamp technique as previously described.818 Mi-croelectrodes were pulled from borosilicate glass (RadnotiGlass Technology, Inc) with tip resistances of 0.3 to 0.8 MQ for

by guest on August 1, 2016http://circres.ahajournals.org/Downloaded from

1000 Circulation Research Vol 75, No 6 December 1994

the current-passing electrode and 2 to 7 Mfl for the voltage-measuring electrode. Recordings were obtained using anAxoclamp amplifier (Axon Instruments, Inc) with data sam-pled at 5 kHz and filtered at 1 kHz. Voltage commandpotentials were generated using a 12-bit digital-to-analogconverter (Tecmar) controlled by customized PCLAMP software(Axon Instruments). Because of the slow kinetics of Kvl.4recovery from inactivation,8 voltage-clamp pulses were deliv-ered every 15 seconds unless otherwise noted. The holdingpotential was -120 mV to allow full recovery from inactiva-tion. Current recordings were obtained initially every 30seconds by using a pulse to +40 mV for 10 minutes afterelectrode impalement to ensure cell stability; K' current wasalso monitored in this way after the addition of a test com-pound such as phorbol 12-myristate 13-acetate (PMA) toobserve changes in current amplitude. The clamp protocolsconducted under control conditions were repeated afterchanges observed in current amplitude after the administra-tion of PMA approached steady state (.1% change in peakcurrent amplitude per minute for 5 minutes). Cell membraneelectrical capacitance calculations were made by recording thecapacitive transient associated with a small voltage step (-120to -110 mV), during which Kvl.4 current was not activated.Integration of the leak-corrected transient yielded the charge(Q) transferred during the voltage step (V), from whichcapacitance (C) was calculated: C=Q/V. All experiments wereperformed at room temperature (22°C±2°C).

SolutionsOocytes were continuously superfused at a rate of 0.5 to 1

mL/min with a physiological solution (ND-96) containing(mmol/L) NaCl 96, KCl 2, CaCl2 1.8, MgCl2 1.0, and HEPES 5(pH 7.50). This solution was also used for incubation beforecRNA injection. PMA was purchased from Sigma ChemicalCo, 4a-PMA and chelerythrine were from LC Services, andstaurosporine was from Kyowa Hakko.

Data AnalysisAnalysis was performed with custom programs that were

designed to read and analyze PCLAMP data files. K' currentsderived from Kvl.4 were activated at voltages more positivethan -60 mV. Therefore, leak subtraction was performed byobtaining the least-squares fit to the passive linear leak currentobtained during voltage steps from -120 to -80 mV. The timecourse of macroscopic current inactivation was fitted with anexponential function by using a nonlinear least-squares algo-rithm.8 Measurements of peak current amplitude after theadministration of a test compound were normalized to predrugvalues by using an average of five test pulses (over 2.5 minutes)immediately before drug superfusion. Data are expressed asmean±SEM. Student's paired t test was used to comparevoltage-dependent and kinetic properties of Kvl.4 currentbefore and after the administration of PMA; the effects ofdifferent concentrations of PMA on current amplitude werecompared by using an unpaired t test. The effect of drugadministration over time was analyzed by one-way ANOVA,and comparison of the effects of different PMA concentrationson capacitance over time was performed by using ANOVA forrepeated measures. A value of P<.05 was used to reject thenull hypothesis.

Phosphorylation by Protein Kinase CThe ability of the K' channel protein to serve as substrate

for phosphorylation by protein kinase C was tested by using amammalian cell line transfected with Kvl.4 as describedpreviously.19 A parallel transfection with Kvl.1 cDNA20'21 wasperformed to obtain control cells. Culture methods have beenpreviously described.19,21 Confluent dishes of KvL.4- andKvl.1-transfected cells were induced with 4 gmol/L dexameth-asone for 18 hours. Each 60-mm dish was solubilized with 3

kept at 4°C throughout the solubilization process. The cellularextract was incubated with 8 ,uL of affinity-purified Kvl.4polyclonal antisera21 (see below) for 2 hours at room temper-ature on a rocking platform. Packed protein A beads (2 ,uL)preadsorbed with phosphate-buffered saline and 0.2% IgG-free bovine serum albumin were added to the extract andincubated for an additional 2 hours at room temperature. Thebeads were sedimented at 900g for 10 seconds and washed.22

Protein kinase C phosphorylation was carried out by using a

modification of previously published protocols.23'24 Briefly,washed beads were resuspended in a 100-,uL reaction volumecontaining 50 mmol/L Tris (pH 7.5), 10 mmol/L MgCl2, 1mmol/L EGTA, 1.5 mmol/L CaCl, 60 ,mol/L phosphatidylser-ine (Sigma), 1.61 ,umol/L diolein (Sigma), 1.8 ,Ci ['y_2P]ATP(0.45 Ci/mmol), and 15.8 nmol/L rat brain protein kinase C(Calbiochem). Phosphatidylserine and diolein were added as a

sonicated micelle mixture. The reaction mixtures were incu-bated at 30°C for 25 minutes with frequent vortexing. The beadswere then pelleted and washed quickly with 1 mL of 0.1%sodium dodecyl sulfate (SDS), 0.1% (wt/vol) Triton X-100, 300mmol/L NaCl, and 50 mmol/L Tris (pH 7.5). The Kvl.4 proteinwas then eluted at 100°C in SDS-polyacrylamide gel samplebuffer22 and analyzed by electrophoresis and autoradiography.

Western Blot Analysis of Kvl.4 Antibody Binding toL-Cell MembranesThe production and affinity purification of the Kvl.4 anti-

sera as well as methods for Western blot analysis have beenpreviously described in detail.2' Confluent flasks of Kvl.4- and

sham-transfected L cells were induced with 4 ,umol/L dexa-methasone for 24 hours and membrane fractions isolated.After SDS-polyacrylamide gel electrophoresis (PAGE) andtransfer to nitrocellulose, incubation was performed with a

1:400 dilution of Kvl.4 antisera for 2.5 hours followed by a

1:7500 dilution of horseradish peroxidase-conjugated goatanti-rabbit IgG. Detection was achieved with an enhancedchemiluminescence kit.

ResultsIn experiments with both 10 and 81 nmol/L (50

mg/mL) PMA, marked suppression of Kvl.4 currentwas observed over time. Fig 1 demonstrates outward K'currents in Xenopus oocytes injected with Kvl.4 cRNAduring step depolarizations before (Fig 1A) and after(Fig 1B) the addition of 10 nmol/L PMA. After steadystate was achieved after drug exposure, peak currentamplitude at +40 mV was reduced by 78% in thisexperiment. In the experiments with 10 nmol/L PMA,this reduction was significant after 30 minutes of expo-

sure to PMA, averaging 40±11% (n=10). At a higherPMA concentration of 81 nmol/L, the reduction incurrent was 54±7% (n-11) for the same time period,although this apparent increased effect was not statisti-cally significant. The current-voltage relation for eachexperiment was normalized to peak current amplitudeat +50 mV under control conditions. Averaged valuesfor the entire group of experiments are shown in Fig 1C.It is apparent that Kvl.4 current was reduced by thephorbol ester at all test potentials examined.To characterize further the nature of this effect, we

examined whether PMA affected the voltage depen-dence of channel gating. A standard two-pulse protocolwas used to assess the voltage dependence of steady-state channel availability. Fig 2 shows averaged valuesfor normalized inactivation curves before and after theadministration of PMA. There was no significant change

mL of detergent extraction buffer.22 Cellular extracts were in the midpoint of the inactivation curve after exposure

by guest on August 1, 2016http://circres.ahajournals.org/Downloaded from

Murray et al Modulation of Inactivating Human K' Channel 1001

A o +40

-50-120-

a

100 mS

C1.0 -

K zz W

°: c 0.5-

W.: +U W.

CONTROL

-100 -50 0 50 100

TEST PULSE ImVIFIG 1. Tracings (A and B) and graph (C) showing suppressionof Kvl.4 current by phorbol 12-myristate 13-acetate (PMA). A,Outward K+ currents during step depolarizations in Xenopusoocytes injected with Kvl.4 cRNA are shown (pulse protocol ininset). B, Currents from the same cell as in panel A are shownafter exposure to 10 nmol/L PMA. C, Data from each experimentwere normalized to peak current amplitude at +50 mV undercontrol conditions. Values from the group of experiments wereaveraged to obtain the peak current-voltage relations(mean+SEM). Error bars are shown where they exceed symbolsize.

to PMA at a concentration of 10 nmol/L (-55±1 mVbefore and -55+±2 mV after PMA, n=6). The timecourse of macroscopic K+ current decline was examinedto obtain an estimate of the apparent rate of inactiva-tion. The time constant was 35±2 milliseconds at +40mV under baseline conditions and did not change afterthe administration of PMA (38±2 milliseconds, n=9).Finally, a two-pulse protocol was used to measurerecovery from K+ current inactivation at -120 mV. Fig3 demonstrates that PMA did not alter the time con-stant of this process from the control values (3.8±0.2seconds at baseline and 3.9±0.2 seconds after PMA,n=4). When a higher concentration of PMA (81

nmol/L) was used, there was also no apparent change inthe voltage dependence of gating as reflected by thetime course of inactivation, the steady-state inactivationcurve, and recovery from inactivation.The time course of effect was examined in more detail

by recording current at +40 mV every 30 secondsduring the addition of PMA as described above. In Fig4A, a representative experiment is shown to demon-strate the time course of the effect of PMA on peakKv1.4 current by using data normalized to an averagedvalue just before drug exposure. Initially PMA caused a

significant increase in the current, averaging 14±4%within 14±2 minutes. In some oocytes, this increase wassubstantial with a maximum increase of 42%. In exper-iments using 81 nmol/L PMA, the current increaseoccurred earlier (11±1 minutes) but was similar inextent (+ 15±5%). Of note, variability was observed inthe time course of the effect of PMA so that the increaseand subsequent decrease of current were not aligned intime in all experiments. In two experiments using 10nmol/L PMA, the current peaked considerably later(:'23 minutes after the addition of drug); in theseexperiments, the time course of suppression was alsodelayed, with K+ current amplitude still above baselinevalues (by 11% and 28%) at 30 minutes. When thesetwo experiments were excluded, the decrease in peakcurrent at 30 minutes in the remaining experiments(n=8) averaged 54%.To examine the possibility that the K+ current reduc-

tion caused by PMA was due to nonspecific internaliza-tion of channel-containing oocyte plasma membrane25rather than modulation of Kv1.4 channel function,capacitance was measured during the time course ofexperiments as an indicator of cell membrane surfacearea.26 Fig 4A also demonstrates that significant sup-pression of the current was typically accompanied by a

minimal effect on capacitance. In the experimentshown, there was a 51% reduction in peak current at=45 minutes of PMA exposure with only a 2% declinein capacitance. The phorbol ester caused a concentra-tion- and time-related effect on capacitance in oocytes,as shown in Fig 4B. At a concentration of 10 nmol/L,there was essentially no change in capacitance after 30minutes of superfusion with PMA (1±1% decrease),although it subsequently declined over time (14±7%decrease at 50 minutes). When 81 nmol/L PMA was

used, capacitance decreased significantly over 30 min-

a0

PMA

=L a

50 msec

FIG 2. Graph (left) and tracings (right) show-ing voltage dependence of Kvl.4 current in-activation. Left, A standard two-pulse protocol(inset) was used to assess the voltage depen-dence of steady-state channel availability (in-activation curve) by using initial 2-second pre-pulses varying from -90 to -25 mV followedby a test pulse to +20 mV. Data were normal-ized to peak current amplitude by using aprepulse of -90 mV at baseline, and aver-aged values (mean±SEM) are presented.The line through the data represents the bestnonlinear least-squares fit of a Boltzmannfunction. The half-inactivation voltage before(o) and after (a) the administration of phorbol12-myristate 13-acetate was -55±1 and-55±2 mV, respectively. Right, Selected cur-rent tracings from a representative experi-ment are shown.

CONTROL1.0

ocz0

L)

U-

z

W

C.)

0.5

0 -

-120 -90 -60 -30 0

PRE-PULSE POTENTIAL ImnV

by guest on August 1, 2016http://circres.ahajournals.org/Downloaded from

1002 Circulation Research Vol 75, No 6 December 1994

CONTROL1.0

czz WO

c 0.50 0

0

0 5 10 15 20

RECOVERY TIME Is)

utes (22+7%). However, this degree of reduction wasconsiderably less than the total current suppression(54±7%) at the same time point.

Fig 4C and 4D show results from additional experi-ments performed to determine whether the effects ofPMA resulted from protein kinase C stimulation rather

FIG 3. Graph (left) and tracings (right)showing recovery of Kvl.4 current from in-activation. Left, Recovery of Kvl.4 currentfrom inactivation before (o) and after (a) theadministration of phorbol 12-myristate 13-acetate (PMA) was measured by using atwo-pulse protocol with an initial prepulse topermit development of inactivation (+30 mVfor 500 milliseconds) followed by a secondtest pulse (+30 my) at different time points(recovery time) after the prepulse. r indi-cates time constant. The holding potentialwas -120 mV. Data were normalized to themaximum current value under control con-ditions (recovery time, 20 seconds) andaveraged. Right, Selected current tracingsobtained at recovery times of 0.2, 0.5, 5, 15,and 20 seconds during an individual exper-iment are shown.

a

50 msec

than a nonspecific effect of the phorbol ester on channelfunction. The inactive stereoisomer 4a-PMA had essen-tially no effect on Kvl.4 current (Fig 4C), with a 3±f-5%increase at 30 minutes (n=4). In another set of exper-iments, oocytes were preincubated in either the proteinkinase inhibitor staurosporine (3 ,umol/L, 20 to 25

B

Wm 1.0 -

z

°0.U

0.0

C1.5 -

-JI-r z 1.0-zW

C + 0.5-L. X

_i

z0

0

PMA l o nM

o currenta capacitance

W

z

CLOK0)

1.0-

0.5 -

0.0 -0 25 50 75 100

TIME Iminl D1.5 -

4a-PMA 10 nM

45

z 1.0

o cz:

cCC+ 0.5-

0Y

FIG 4. Graphs showing timecourse of the effects of phorbol1 2-myristate 13-acetate (PMA). A,The time course of the effect ofPMA on peak K' current (o) at+40 mV and capacitance (a) isillustrated for an individual exper-iment. Data are expressed nor-malized to predrug values; eitheran averaged value from five testpulses (current) or a single value(capacitance) immediately before

------ drug superfusion was used. Timezero corresponds to the time of

* oocyte impalement, and the ar-row indicates the start of drug

PMAinfusion. PMA caused a biphasic

PMA response with an initial 15% in-0 10 nM crease in current within 18 min-* 81 nM utes in this experiment, followed

by marked suppression (51%)without significant change (2%

0 10 20 30 40 50 decline) in membrane capaci-TIME Iminl tance at -45 minutes. B, Aver-

aged values for membrane ca-pacitance in Xenopus oocytes areshown as a function of time afterthe addition of two different con-

--- centrations of PMA (10 nmol/L [o]and 81 nmol/L [ml). Data are nor-

PMA 10 nM malized to the value obtained justbefore the administration of PMA

.staurospoine3MM

(time zero); there was a signfi-o staurosporine 3 pJM cant difference between the two* chelerythrine 20 pM curves (*P<.05). C, The inactive

stereoisomer, 4a-PMA, had no ef-0 15 30 45 fect on Kv1.4 current. Current am-TIME Imin) plitude from each experiment

(n=4) was normalized to predrugvalues as described in panel A. D,After pretreatment with either theprotein kinase inhibitor staurospo-rine (3 ,umol/L) or chelerythrine (20,umol/L), PMA had no significanteffect on Kvl.4 current, with a6±7% increase or 9±9% decline at30 minutes, respectively.

A

0 15 30

TIME Iminl

by guest on August 1, 2016http://circres.ahajournals.org/Downloaded from

Murray et al Modulation of Inactivating Human K' Channel 1003

1 2 3

116kDa- -_97- f _

66 -

43 -

SW

ea

4 5

S Kvl.4

FIG 5. In vitro phosphorylation of Kvl.4 channel protein bypurified protein kinase C. Following exposure to 4 gmol/Ldexamethasone for 18 hours, L cells transfected with eitherKvl.4 (lane 1) or Kvl.1 (lane 3) cDNA or nontransfected L cells(lane 2) were detergent-extracted and incubated with Kv1.4-specific antisera as described in "Materials and Methods." Theaffinity-purified protein was then incubated with protein kinase Cand [92_PIATP and analyzed by sodium dodecyl sulfate-poly-acrylamide gel electrophoresis and autoradiography. Westernanalysis was also performed by using L cells transfected witheither vector containing Kvl.4 cDNA (lane 4) or no DNA (sham,lane 5) and Kvl.4 antisera. Molecular weight markers are shownon the left. The 85-kD band common to lanes 1 through 3represents autophosphorylation of protein kinase C. A bandrepresenting phosphorylated protein is present in lane 1 but notin lanes 2 and 3; this band migrates with the same electropho-retic mobility as Kv1.4, as indicated by the arrow. These resultsconfirm that the phosphorylated protein in lane 1 is the Kvl.4channel.

minutes) or chelerythrine (20 gmol/L, 40 to 45 min-utes), a more specific inhibitor of protein kinase C,27before the addition of PMA. As shown in Fig 4D, PMAhad no significant effect on outward K' current after 30minutes in the presence of either staurosporine or

chelerythrine (n=4 each).Channel protein was obtained from L cells expressing

Kvl.4 current to determine whether protein kinase Ccould phosphorylate the channel directly. Immunopre-cipitation with Kvl.4-specific antisera was performed byuse of L cells transfected with either Kv1.4 (Fig 5, lane 1)or Kvl.1 (lane 3) cDNA or nontransfected L cells (lane2), followed by incubation of the affinity-purified proteinwith [/y2-P]ATP and purified protein kinase C. Thephosphorylated protein was analyzed by SDS-PAGE andautoradiography; results are shown in Fig 5. A band at 85kD was observed in lanes 1 through 3 and representsautophosphorylation of protein kinase C. The bandpresent in lane 1 appears to represent phosphorylatedKv1.4 protein, since it is specific to the Kvl.4-expressingcells. Lanes 4 and 5 in Fig 5 demonstrate the results ofWestern blot analysis with Kvl.4 antisera of L cellstransfected with either Kvl.4 cDNA or the vector lackingchannel DNA (sham transfection), respectively. Thearrow indicates that the electrophoretic mobility of theKv1.4 channel protein in lane 4 is identical to thephosphorylated protein in lane 1. These findings confirmthat the Kvl.4 channel protein is a substrate for phos-phorylation in vitro by protein kinase C.

DiscussionOur results demonstrate that the phorbol ester PMA

modulates Kvl.4 current expressed in Xenopus oocytes.The bimodal nature of the effect of PMA is intriguing,and its significance is unclear at the present time,

although several explanations are possible. First, pro-tein kinase C activation can lead to a dual physiologicalresponse (with an initial stimulation and subsequentreduction) that is due to negative feedback control ofone or more steps in the cell-signaling process.8 Forexample, a biphasic effect of phorbol ester has beendescribed on B-lymphocyte activation, with an initialenhancement and subsequent suppression of prolifera-tion.29 In addition, it was recently reported that cardiacL-type Ca21 channels are modified in a bimodal fashionby protein kinase C both in mammalian ventricularmyocytes3031 and after the expression of these channelsin Xenopus oocytes.32,33

Second, it is possible that after the initial stimulationof K' current activity, channel internalization resultedfrom PMA administration. In our experiments, weobserved a concentration- and time-related reduction incell membrane surface area as reflected by cell mem-brane capacitance; this reduction was significant after30 minutes only with a PMA concentration of >10nmol/L. These findings are consistent with a previousreport of downregulation of the Na+-ATPase in Xeno-pus oocytes after PMA administration; this downregu-lation was due to internalization of cell surface mem-brane and loss of ouabain binding sites.25 Aconcentration-related response was also observed inthat study accompanied by a change in capacitance andloss of cell membrane invaginations by electron micros-copy. In prior studies of cloned cardiac Ca' channelsexpressed in Xenopus oocytes, a significant reduction incurrent was associated with a drop in capacitance in onestudy33 and no change in capacitance in another.32Channel internalization is unlikely to explain our exper-imental results because the substantial depression ofcurrent that occurred at a time when capacitance (ie,membrane surface area) was unchanged. However, wecannot rule out the possibility of selective internaliza-tion of K' channel protein, especially if channels werenot evenly distributed on the cell surface membrane. Ofinterest, our findings of a biphasic effect of PMA do notfully agree with a recent report of the effects of proteinkinase C stimulation on RH10, a K' channel clonedfrom rat heart, which also corresponds to Kvl.4. Whenexpressed in Xenopus oocytes, RH10 current was sup-pressed by a higher concentration of PMA (100 nmol/L)over time without an initial increase observed.34 Thereason for this difference from our experimental resultsis not clear at the present time. Although cell capaci-tance was not measured in the other study, the concen-tration of PMA (100 nmol/L) used was similar to thehigher concentration (81 nmol/L) we used, which dem-onstrated an initial increase in current. Although the rat(RH10) and human (HK1) versions of Kvl.4 are 98%identical, three consensus sites for protein kinase Cphosphorylation present in the human channel aminoacid sequence are altered in the rat isoform (S67--T,S78->G, and S90->T). These differences could providea basis for explaining the discrepant effects of PMA onK' current derived from these channels, a hypothesisthat can be tested by using mutagenesis strategies.A third possible mechanism for the dual response in

Kvl.4 current that we observed is downregulation of thekinase as a result of persistent activation,28 as proposedfor the secondary suppression of cardiac L-type Ca'currents in neonatal rat ventricular myocytes.3' This

by guest on August 1, 2016http://circres.ahajournals.org/Downloaded from

1004 Circulation Research Vol 75, No 6 December 1994

process is typically associated with loss of measurableenzyme activity, most likely due to proteolysis.28,35 It isunlikely that downregulation accounts for our experi-mental findings, given the rapid time course (within 30minutes) of K' current suppression that was observed.Prior studies have demonstrated that decline in proteinkinase C activity after phorbol ester exposure typicallydid not occur for 2 to 3 hours.29'35'36 Moreover, the rateof enzyme loss was a function of the amount of phorbolester administered, occurring more slowly at a lowconcentration of PMA (10 nmol/L).29 At least twodistinct protein kinase C activities that depend on Ca21for maximal activation have been purified in Xenopusoocytes.37 After prolonged incubation with PMA, one ofthe Xenopus isozymes disappears rapidly within 30minutes; the second, more abundant isozyme is rela-tively unchanged for up to 2 hours. Therefore, it isconceivable that the biphasic response we observedcould have resulted from activation and subsequentdisappearance of this minor isozyme. The molecularnature of these and other protein kinase C isoforms inXenopus laevis oocytes is not well defined. Two Ca2'-dependent isozymes that are homologous to bovine a-and 8 forms of the kinase have been cloned from anoocyte cDNA library.38 Recent data indicate that boththese species are present in mammalian heart, where acomplex picture has emerged of multiple protein kinaseC isoforms differing in their subcellular distribution,response to activating factors such as Ca>2 and phorbolesters, and age-dependent expression.39Our results suggest that modulation of Kvl.4 current

by protein kinase C activation results from direct phos-phorylation of the channel protein itself. The aminoacid sequence of the channel has numerous potentialsites for phosphorylation by protein kinase C,15,16 al-though the exact site(s) responsible for the effects weobserved is unknown at the present time. It has beenreported that a Shaker K' channel could be phosphor-ylated by the kinase in the S4-S5 loop.40 We havemutated serines 461 in the S4 segment and 474 in theS4-S5 loop to alanines; however, these mutations re-sulted in loss of channel expression. Further studies arenecessary to determine conclusively whether directchannel phosphorylation is responsible for our experi-mental findings, and if so, which serine/threonine(s) ismodified.The physiological correlate(s) of Kvl.4 current in

native cardiac myocytes is not well understood. Inmammalian cardiac myocytes, ITO is modified by a-ad-renergic stimulation that is presumably due to proteinkinase C activation.9-12 In rabbit atrial cells, PMAcauses an increase in ITO41; in rat ventricular cells, thephorbol ester results in suppression of ITO.9 Kvl.4current is similar phenotypically to human ITO ,42 al-though recovery from inactivation is considerably slow-er.843 In this respect, it resembles ITO reported fromrabbit atrial myocytes.41 Nevertheless, coexpression ofKvl.4 with other mammalian Shaker homologues resultsin more rapid recovery from inactivation.43 This raisesthe possibility that the native channel exists as a heter-omultimer that could also include unique subunits, ashas been observed for Na+, Ca2+, and nicotinic acetyl-choline receptor channels. Thus, the heterogeneity ofnative cardiac A-type current phenotypes may reflectnot only the presence of distinct isoforms with different

kinetics and response to kinase activation but alsoheteromultimer formation. Nevertheless, the presentstudy demonstrates significant modulation of this im-portant K' channel and provides an initial step towardunraveling the complex nature of metabolic regulationof human cardiac K' channels.

AcknowledgmentsThis study was supported by the Charles E. Culpeper

Foundation and grants HL-02363, HL-49330, HL-47599, andHL-46681 from the National Institutes of Health. Dr Tamkunand Dr Bennett are Established Investigators of the AmericanHeart Association.

References1. Gintant GA, Cohen IS, Datyner NB, Kline RP. Time-dependent

outward currents in the heart. In: Fozzard HA, Haber E, JenningsRB, Kate AM, Morgan HE, eds. The Heart and CardiovascularSystem. 2nd ed. New York, NY: Raven Press Publishers; 1992:1121-1169.

2. Tseng G-N, Hoffman BF. Two components of transient outwardcurrent in canine ventricular myocytes. Circ Res. 1989;64:633-647.

3. Escande D, Coulombe A, Faivre JF, Deroubaix E, Coraboeuf E.Two types of transient outward currents in adult human atrialcells. Am J Physiol. 1987;252:H142-H148.

4. Shibata EF, Drury T, Refsum H, Aldrete V, Giles W. Contri-butions of a transient outward current to repolarization in humanatrium. Am J Physiol. 1989;257:H1773-H1781.

5. Nabauer M, Beuckelmann DJ, Erdmann E. Characteristics oftransient outward current in human ventricular myocytes frompatients with terminal heart failure. Circ Res. 1993;73:386-394.

6. Shearman MS, Sekiguchi K, Nishizuka Y. Modulation of ionchannel activity: a key function of the protein kinase C enzymefamily. Pharmacol Rev. 1989;41:211-237.

7. Tamkun MM, Knoth KM, Walbridge JA, Kroemer H, Roden DM,Glover DM. Molecular cloning and characterization of twovoltage-gated K' channel cDNAs from human ventricle. FASEB J.1991;5:331-337.

8. Po S, Snyders DJ, Baker R, Tamkun MM, Bennett PB. Functionalexpression of an inactivating potassium channel cloned fromhuman heart. Circ Res. 1992;71:732-736.

9. Apkon M, Nerbonne JM. a1-Adrenergic agonists selectivelysuppress voltage-dependent K' currents in rat ventricularmyocytes. Proc NatlAcad Sci USA. 1988;85:8756-8760.

10. Fedida D, Shimoni Y, Giles WR. A novel effect of norepinephrineon cardiac cells is mediated by a,-adrenoceptors. Am J Physiol.1989;256:H1500-H1504.

11. Fedida D, Shimoni Y, Giles WR. a-Adrenergic modulation of thetransient outward current in rabbit atrial myocytes. J Physiol(Lond). 1990;423:257-277.

12. Tohse N, Nakaya H, Hattori Y, Endou M, Kanno M. Inhibitoryeffect mediated by a,-adrenoceptors on transient outward currentin isolated rat ventricular cells. Pflugers Arch. 1990;415:575-581.

13. Nakayama T, Fozzard HA. Adrenergic modulation of the transientoutward current in isolated canine Purkinje cells. Circ Res. 1988;62:162-172.

14. Chandy GK. Simplified gene nomenclature. Nature. 1991;352:26.15. Kemp BE, Pearson RB. Protein kinase recognition sequence

motifs. Trends Biochem Sci. 1990;15:342-346.16. Graff JM, Stumpo DJ, Blackshear PJ. Characterization of the

phosphorylation sites in the chicken and bovine myristoylatedalanine-rich C kinase substrate protein, a prominent cellular sub-strate for protein kinase C. J Biol Chem. 1989;264:11912-11919.

17. Krieg PA, Melton DA. An enhancer responsible for activatingtranscription at the midblastula transition in Xenopus devel-opment. Proc Natl Acad Sci USA. 1987;84:2331-2335.

18. Kass RS, Bennett PB. Microelectrode voltage clamp: the cardiacPurkinje fiber. In: Smith TG, Lecar H, Redman S, Gage P, eds.Voltage and Patch Clamping With Microelectrodes. Bethesda, Md:Waverly Press; 1985:178-191.

19. Felipe A, Snyders DJ, Deal KK, Tamkun MM. Influence of clonedvoltage-gated K' channel expression of alanine transport, Rb+uptake, and cell volume. Am J Physiol. 1993;256:C1230-C1238.

20. Roberds SL, Tamkun MM. Cloning and tissue-specific expressionof five voltage-gated potassium channel cDNAs expressed in ratheart. Proc NatlAcad Sci USA. 1991;88:1798-1802.

by guest on August 1, 2016http://circres.ahajournals.org/Downloaded from

Murray et al Modulation of Inactivating Human K' Channel 1005

21. Deal KK, Lovinger DM, Tamkun MM. The brain Kv1.1 potassiumchannel: in vitro and in vivo studies on subunit assembly andposttranslational processing. J Neurosci. 1994;14:1666-1676.

22. Tamkun MM, Fambrough DM. The (Na++K+)-ATPase of chicksensory neurons: studies on biosynthesis and intracellulartransport. J Biol Chem. 1986;261:1009-1019.

23. Murphy BJ, Catterall WA. Phosphorylation of purified rat brainNa+ channel reconstituted into phospholipid vesicles by proteinkinase C. J Biol Chem. 1992;267:16129-16134.

24. Kikkawa U, Takai Y, Minakuchi R, Inohara S, Nishizuka Y.Calcium-activated, phospholipid-dependent protein kinase fromrat brain: subcellular distribution, purification, and properties.J Biol Chem. 1982;257:13341-13348.

25. Vasilets LA, Schmalzing G, Madefessel K, Haase W, Schwarz W.Activation of protein kinase C by phorbol ester induces downreg-ulation of the Na+/K+-ATPase in oocytes of Xenopus laevis. JMembr Bio. 1990;118:131-142.

26. Penner R, Neher E. The patch-clamp technique in the study ofsecretion. Trends Neurosci. 1989;12:159-163.

27. Herbert JM, Augereau JM, Gleye J, Maffrand JP. Chelerythrine isa potent and specific inhibitor of protein kinase C. Biochem BiophysRes Commun. 1990;172:993-999.

28. Nishizuka Y. The molecular heterogeneity of protein kinase C andits implications for cellular regulation. Nature. 1988;334:661-665.

29. Mond JJ, Feuerstein N, June CH, Balapure AK, Glazer RI, With-erspoon K, Brunswick M. Bimodal effect of phorbol ester on B cellactivation. J Biol Chem. 1991;266:4458-4463.

30. Tseng GN, Boyden PA. Different effects of intracellular Ca andprotein kinase C on cardiac T and L Ca currents. Am J Physiol.1991;261:H364-H379.

31. Lacerda AE, Rampe D, Brown AM. Effects of protein kinase Cactivators on cardiac Ca2' channels. Nature. 1988;335:249-251.

32. Singer-Lahat D, Gershon E, Lotan H, Hullin R, Biel M, FlockerziV, Hofmann F, Dascal N. Modulation of cardiac Ca2+ channels inXenopus oocytes by protein kinase C. FEBS Lett. 1992;306:113-118.

33. Bourinet E, Fournier F, Lory P, Charnet P, Nargeot J. Proteinkinase C regulation of cardiac calcium channels expressed inXenopus oocytes. Pflugers Arch. 1992;421:247-255.

34. Okada H, Ishii K, Nunoki K, Abe T, Taira N. Modulation oftransient type K channel cloned from rat heart. Biochem BiophysRes Commun. 1992;189:430-436.

35. Young S, Parker PJ, Ullrich A, Stabel S. Down-regulation ofprotein kinase C is due to an increased rate of degradation.Biochem J. 1987;244:775-779.

36. Ballester R, Rosen OM. Fate of immunoprecipitable proteinkinase C in GH3 cells treated with phorbol 12-myristate 13-acetate.J Biol Chem. 1985;260:15194-15199.

37. Sahara S, Sato K, Aoto M, Ohnishi T, Kaise H, Koide H, Ogita K,Fukami Y. Characterization of protein kinase C in Xenopusoocytes. Biochem Biophys Res Commun. 1992;182:105-114.

38. Chen K, Peng Z, Lavu S, Kung H. Molecular cloning and sequenceanalysis of two distinct types of Xenopus laevis protein kinase C.Second Messengers Phosphoproteins. 1988 -1989;12:251-260.

39. Rybin VO, Steinberg SF. Protein kinase C isoform expressionand regulation in the developing rat heart. Circ Res. 1994;74:299-309.

40. Isacoff EY, Kimmerly W, Jan YN, Jan LY. Phosphorylation andmodulation of the Shaker K' channel by protein kinase C (PKC).Biophys J. 1992;61:A13. Abstract.

41. Fedida D, Giles WR. A phorbol ester increases the transientoutward current in rabbit atrial myocytes. J Physiol (Lond). 1989;415:106P. Abstract.

42. Fermini B, Wang Z, Duan D, Nattel S. Differences in ratedependence of transient outward current in rabbit and humanatrium. Am J PhysioL 1992;263:H1747-H1754.

43. Po S, Roberds S, Snyders DJ, Tamkun MM, Bennett PB. Het-eromultimeric assembly of human potassium channels: molec-ular basis of a transient outward current? Circ Res. 1993;72:1326-1336.

by guest on August 1, 2016http://circres.ahajournals.org/Downloaded from

BennettK T Murray, S A Fahrig, K K Deal, S S Po, N N Hu, D J Snyders, M M Tamkun and P B

Modulation of an inactivating human cardiac K+ channel by protein kinase C.

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1994 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.75.6.9991994;75:999-1005Circ Res. 

http://circres.ahajournals.org/content/75/6/999World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on August 1, 2016http://circres.ahajournals.org/Downloaded from