Properties of Xenopus Kv1.10 Channels Expressedin HEK293 Cells

Mark Fry,1 Robert A. Maue,1,2 Frances Moody-Corbett3

1 Department of Physiology, Dartmouth Medical School, Hanover, New Hampshire 03755

2 Department of Biochemistry, Dartmouth Medical School, Hanover, New Hampshire 03755

3 Division of Basic Medical Sciences, Memorial University of Newfoundland, St John’s,Newfoundland, Canada, A1B 3V6

Received 30 June 2003; accepted 19 November 2003

ABSTRACT: Voltage-gated K� channels play im-portant roles in shaping the characteristics of actionpotentials and electrical activity. In a previous study, weisolated cDNAs encoding several distinct K� channelisoforms, including a novel isoform (XKv1.10) expressedin Xenopus laevis spinal cord neurons and myocytes.Here, we report the biophysical characterization ofXKv1.10 expressed in transiently transfected HEK293cells. Whole cell patch clamp recordings revealed a volt-age-gated, rapidly activating and inactivating K� cur-rent. Interestingly, the rate of inactivation of XKv1.10channels showed apparent voltage dependence, withtime constants between 77.7–213.3 ms. The predictedprotein sequence of XKv1.10 does not appear to encodean N-terminal inactivating “ball and chain” domain,and instead these channels may inactivate via a C/P-typemechanism. Consistent with this, either increasing the

external concentration of K� or external application oftetraethylammonium caused a decrease in the rate ofinactivation. Pharmacologically, XKv1.10 K� channelswere sensitive to 4-aminopyridine and tetraethylammo-nium with apparent IC50 values of 68.5 �M and 17.1mM, respectively. When simulated action potentialswere used as a voltage command, XKv1.10 was similarto XKv1.4 in that it carried more repolarizing currentduring the action potential than XKv1.2. However, whileXKv1.4 was active during the interspike interval,XKv1.10 and XKv1.2 were not. Overall, the data suggestthat XKv1.10 channels make a unique contribution tothe developmental maturation of electrical signaling inXenopus laevis. © 2004 Wiley Periodicals, Inc. J Neurobiol 60:

227–235, 2004

Keywords: patch clamp; Kv1.10 potassium channel; Xe-nopus laevis; HEK 293; action potential

INTRODUCTION

Voltage-gated K� channels play a major role in de-termining electrical excitability. Tightly regulated ex-pression of K� currents has been shown to be impor-tant for normal maturation of neurons and skeletal

muscle cells (for review see Spitzer and Ribera,1998). For example, in Xenopus laevis spinal cordneurons in culture the differentiation of the delayedrectifier (IKv) is necessary for the transition from aslow, primarily Ca��-dependent action potential toa fast, predominantly Na�-dependent one (Barish,1986; O’Dowd et al., 1988). In addition, after thefirst day in culture, expression of an inactivatingK� current (IKA) leads to further shortening of theaction potential and limited ability of the neurons tofire repetitively (Ribera and Spitzer, 1990; Desar-menien et al., 1993). At early times in culture,

Correspondence to: M. Fry (wmfry@dartmouth.edu).Contract grant sponsor: NSERC.

© 2004 Wiley Periodicals, Inc.Published online 22 April 2004 in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/neu.20024

227

Xenopus skeletal muscle cells also express twovoltage-gated K� currents, IK and IIK. The IK showslittle inactivation with membrane depolarization,whereas IIK shows rapid inactivation, similar to IA

(Moody-Corbett and Gilbert, 1992a,b; Spruce andMoody, 1992; Linsdell and Moody, 1995). As themuscle cells mature in culture, expression of bothcurrents increases (Ribera and Spitzer, 1991;Moody-Corbett and Gilbert, 1992a; Linsdell andMoody, 1995; Chauhan-Patel and Spruce, 1998;Currie and Moody, 1999). Patch clamp experimentsusing cell attached recording techniques suggestthat IIK and IK are mediated by at least two differentstructural classes of K� channel (Ernsberger andSpitzer, 1995; Fry and Moody-Corbett, 1999). Inaddition, the inactivating properties of IIK fromwhole cell current recordings of skeletal muscle inculture suggest that this current may be a compositeof more than one channel subtype (Ribera andSpitzer, 1991; Moody-Corbett and Gilbert, 1992b).

Of the K� channels having six transmembranedomains (Kv1 to Kv9, eag, erg, elk, KQT, SK, andSLO), at least 14 main classes have been identified todate (for review see Coetzee et al., 1999). Of these,homologues of Kv 1.1, Kv1.2, Kv1.4, Kv1.10, Kv2.1,Kv2.2, Kv3.1 KvLQT1 (Ribera, 1990; Ribera andNguyen, 1993; Burger and Ribera, 1996; Sanguinettiet al., 1996; Gurantz et al., 2000; Fry et al., 2001;Kerschbaum et al., 2002), Kv4.3 (Lautermilch andSpitzer, unpublished clone accession #U89265), andSLO (Kukuljan et al., unpublished clone accession#AF274053) have been isolated from Xenopus. Whilethe biophysical properties of some of these channelshave been examined, the functional properties of thenovel Kv1 subfamily member XKv1.10 (Fry et al.,2001) have not been reported. Therefore, the purposeof this study was to examine the biophysical proper-ties of XKv1.10 when transiently expressed inHEK293 cells. Interestingly, although XKv1.10 doesnot appear to contain an N-terminal “ball and chain”inactivation domain (Zagotta et al., 1990; Tseng-Crank et al., 1993; Comer et al., 1994), it does exhibitsignificant inactivation during depolarizing voltagesteps. Furthermore, the inactivation of the channeldisplays apparent voltage dependence and is sensitiveto either increasing concentration of external K� orthe presence of external tetraethylammonium. Finally,when compared to XKv1.2 and XKv1.4 during actionpotential-like changes in membrane potential, the re-sults suggest XKv1.10 may play a unique role indetermining the properties of excitable cells duringXenopus development.

METHODS

cDNA Clones

The open reading frames of XKv1.2, XKv1.4, and XKv1.10cDNAs were amplified from embryonic Xenopus skeletalmuscle cDNA using the polymerase chain reaction, as pre-viously described (Fry et al., 2001). The PCR products wereinitially subcloned into the pCR2.1 vector (Invitrogen,Carlsbad, CA), then excised and subcloned into thepCDNA3 expression vector (Invitrogen). DNA sequencingand restriction mapping confirmed the identity of the con-structs. The pGL plasmid encoding green fluorescent pro-tein (GFP) was obtained from Gibco BRL (Carlsbad, CA).

Cell Culture and Transfection

HEK293 cells were obtained from American Type CultureCollection (Manassas, VA) and maintained at 37°C and 5%CO2 in Dulbecco’s minimal essential medium (DMEM;Gibco BRL) supplemented with 10% fetal calf serum, pen-icillin (100 units/mL), streptomycin (100 �g/mL), and 2mM glutamine. Cells were plated on 35 mm plastic culturedishes and allowed to reach 70–90% confluency beforetransfection. Transient transfections were carried out using4 �g of LipofectAMINE and 2 �L of Plus reagent (GibcoBRL), 1 �g of one of the three K� channel pCDNA3constructs, and 0.5-1 �g of pGL diluted in 200 �L DMEM.The cells were incubated overnight before they were re-moved using 0.05% trypsin/0.53 mM EDTA (Cellgro,Herndon, VA) and replated at lower density for electrophys-iological recording.

Electrophysiological Recording andData Analysis

The functional properties of XKv1.2, XKv1.4, and XKv1.10in HEK293 cells were assessed by whole cell patch clamprecording (Hamill et al., 1981) 24–36 h after transfection.Whole cell current recordings were made using an EPC-9patch clamp amplifier (Instrutech, Greenvale, NY). Stimu-lation and recording were controlled by a Macintosh 8600with Pulse software. Data were filtered at 7 kHz and ac-quired at 50 kHz, except for experiments examining time-dependent recovery from inactivation, where data were fil-tered at 2 kHz and acquired at 5 kHz. Cells were examinedon a Zeiss IM 35 equipped with fluorescence optics. Onlyfluorescent cells exhibiting simple morphology and a lack ofcontact with other cells were used for recording. The extra-cellular recording solution contained (in mM): 143 NaCl, 2KCl, 1.2 CaCl2, 1 MgCl2, 10 glucose, 10 N-2-hydroxyeth-ylpiperazine-N�-2-ethanesulfonic acid (HEPES) (pH 7.4 ad-justed with NaOH). In some solutions Na� was replacedwith equimolar K� or tetraethylammonium ions (TEA).4-Aminopyridine (4AP) was added to the external recordingsolution (final concentration between 1 �m and 3 mM) insome experiments. Solutions were exchanged in some ex-periments using a syringe driven system. The patch elec-

228 Fry et al.

trode contained (in mM): 140 KCl, 5 ethylene-glycol-bis-amino-ethyl-tetraacetic acid (EGTA), 5 MgCl2, and 10HEPES (pH 7.4 adjusted with NaOH). Electrodes werefabricated from borosilicate glass (World Precision Instru-ments, Sarasota, FL) and had resistances of 1.5–2.1 M�when filled with the patch electrode solution. Capacitivetransients and series resistance were minimized via thesoftware-controlled amplifier (series resistance compensa-tion was between 60–85%). Currents were not leak sub-tracted unless otherwise noted. Data were analyzed usingPulseFit (Instrutech) and Origin v6.1 (OriginLab Corp,Northampton, MA) software. Mean values are reported as� standard error of the mean.

RESULTS

Biophysical Properties of XKv1.10

While we recently reported the expression of a novelK� channel gene (XKv1.10) in excitable tissue ofdeveloping Xenopus laevis (Fry et al., 2001), thebiophysical properties of this isoform were unknown.Whole cell patch clamp recording from HEK293 cellsexpressing XKv1.10 K� channel cDNA revealed K�

currents, which rapidly activated and incompletelyinactivated in response to depolarizing voltage pulses(Fig. 1). These currents were not observed in untrans-fected HEK293 cells (data not shown). The XKv1.10current activated at approximately �30 mV, andreached half maximum at �11.3 � 0.8 mV (k � 10.8� 0.7, n � 17) [Fig. 1(A,C)]. The XKv1.10 currentwas sensitive to membrane holding potential, exhib-iting steady state inactivation with a voltage of half-inactivation of �37.8 � 0.6 mV (k � 7.7 � 0.5, n� 16) [Fig. 1(B,C)]. The reversal potential of thecurrent carried by the XKv1.10 channel was investi-gated by examining tail currents. The ionic currentreversed at �81.8 � 1.1 mV (n � 11) in 2 mMexternal K�, �72.0 � 1.4 mV (n � 5) in 5 mMexternal K�, and �24.0 � 1.0 mV (n � 5) in 50 mMK� (data not shown). The observed shifts in reversalpotential closely followed those predicted by theNernst equation, indicating a strongly K�-selectivechannel. Together, these data confirm that XKv1.10encodes a voltage-gated K� channel.

Activation of XKv1.10 was rapid and voltage-dependent. The 10–90% rise times showed a markeddecrease with increasing voltage steps: from 14.2� 1.2 ms at �20 mV to 1.89 � 0.2 ms at �70 mV[Fig. 1(D)]. The inactivation of the XKv1.10 currentswere best fitted to a single exponential of the form I(t)� Io e (�t/�), where � is the time constant of decay(fitting to two � exponentials did not result in a sig-nificant decrease in the RMS error calculated by the

PulseFit software). The inactivation of XKv1.10showed apparent voltage dependence, with morerapid inactivation at more depolarized potentials [�� 213.8 � 29.8 ms at �10 mV to 77.7 � 7.9 ms at�70 mV, Fig. 1(E)]. Because the rate of inactivationof some Kv channels is sensitive to external K�

concentration (Lopez-Barneo et al., 1993; Marom andLevitan, 1994; Baukrowitz and Yellen, 1995; Ras-musson et al., 1995), we examined the effect of in-creasing the concentration of external K� onXKv1.10 inactivation. Increasing the extracellular K�

concentration from 2 to 50 mM caused a slight, butsignificant (p � 0.05, two population t test), increasein the time constant of XKv1.10 inactivation (�), from101 � 8 to 160 � 23 ms at �50 mV [n � 4, Fig.1(F)].

Because other related inactivating channels(rKv1.3, XKv1.4, rKv1.4, and mKv1.7) have beenshown to exhibit cumulative inactivation (Marom andLevitan, 1994; Roeper et al., 1997; Kalman et al.,1998; Kerschbaum et al., 2002), we determinedwhether the inactivating XKv1.10 exhibited similarbehavior. Cells expressing XKv1.10 were subjected toa series of 20 pulses (from �80 mV) to �40 mV,each 5 ms in duration, and separated by an interpulseinterval of varying durations (as described by Maromand Levitan, 1994). An interpulse interval of 10 s wassufficient to allow full recovery of the channels, as noinhibition of peak current was observed after 20pulses [Fig. 1(G)]. However shorter interpulse inter-vals resulted in significant inhibition (p � 0.05, twopopulation t test, n � 9). After 20 pulses with aninterpulse interval of either 1 s, 100 ms, 10 ms, or 5ms, peak currents were inhibited by 16.5 � 10.0, 31.3� 2.1, 32.9 � 2.7, and 35.0 � 2.5% respectively [n� 9, Fig. 1(G)]. We also tested whether cumulativeinactivation of XKv1.10 was sensitive to external K�

concentration or the holding potential of the interpulseinterval. When the extracellular K� concentration wasincreased to 100 mM, the cumulative inactivation thatoccurred was not significantly different from that ob-served in 2 mM external K�, independent of theinterpulse used (two population t test, n � 4). Specif-ically, after 20 pulses with an interpulse interval ofeither 10 s, 1 s, 100 ms, 10 ms, or 5 ms, peak currentswere inhibited by 2.8 � 2.7, 9.8 � 3.7, 36.7 � 5.5,39.7 � 5.6, and 39.8 � 5.9%, respectively. Similarly,using a holding potential of �100 mV instead of �80mV did not cause a significant change in the amountof cumulative inactivation independent of the inter-pulse interval (two population t test, n � 4). In par-ticular, after 20 pulses with an interpulse interval ofeither 10 s, 1 s, 100 ms, 10 ms, or 5 ms, peak currents

Properties of Kv1.10 K� Channels 229

were inhibited by 1.7 � 0.7, 13.3 � 3.1, 30.2 � 2.5,31.6 � 1.9, and 33.0 � 1.0%, respectively.

In order to compare cumulative inactivation ofXKv1.10 with that observed for Xenopus XKv1.4 andXKv1.2, cells transfected with K� channel cDNAwere subjected to a series of 150 ms pulses to �50mV from a holding potential of �80 mV every secondfor 20 s (as described by Kerschbaum et al., 2002).After 20 pulses, peak currents carried by the XKv1.10channel were only inhibited by 26.2 � 3.1% (n � 6).This frequency-dependent cumulative inactivation issignificantly less than that exhibited by XKv1.4 (87.3� 1.7%, n � 5), and significantly greater than thatexhibited by XKv1.2 (13.1 � 3.5%, n � 5) (p � 0.05,one-way ANOVA). Therefore, XKv1.10 displays cu-mulative inactivation properties different from bothXKv1.2 and XKv1.4 channels.

Recovery from inactivation in some Kv1 channels

has been shown to be affected by external K� (Tsengand Tseng-Crank, 1992; Demo and Yellen, 1991;Rasmusson et al., 1995; Levy and Deutsch, 1996),while recovery in other channels is independent ofexternal K� (Fedida et al., 1999). In order to inves-tigate the recovery of XKv1.10 from inactivation,HEK293 cells transfected with XKv1.10 cDNA weresubjected to a series of two-pulse protocols every 30 s(as described by Rasmusson et al., 1995). Briefly,

Figure 1 K� currents recorded from HEK293 cells trans-fected with XKv1.10 cDNA. (A) Family of current recordselicited by series of 1 s depolarizing steps at 30 s intervals(�70 to 70 mV in 10 mV increments) from holding poten-tial of �80 mV. (B) Family of current records showingsteady state inactivation of the current. The cell was held at�80 mV, at 30 s intervals subjected to a series of 1 sprepotentials (�90 to 50 mV in 10 mV increments) thenstepped to the test potential of 70 mV for 50 ms. Only thelast 75 ms are shown, for clarity. (C) Activation and steady-state inactivation conductance curves for XKv1.10 K�

channels. Data points representing the mean conductancesof 17 (activation) and 16 (inactivation) cells were fitted toBoltzmann distributions. (D) Voltage dependence of 10–90% rise time for step potentials �20 to 70 mV (meanvalues, n � 7 to 10). (E) Voltage dependence of the timeconstant of inactivation for step potentials �10 to 70 mVwas determined by fitting the decaying current with a singleexponential (mean values, n � 19 to 15). (F) External K�

slows rate of inactivation of XKv1.10 current. Cell was heldat �80 mV, and stepped to 50 mV in the presence of 2 mMexternal K�, or 50 mM external K� as indicated. As in-creasing external K� to 50 mM caused a reduction incurrent amplitude, currents have been normalized to facili-tate comparison. (G) Cumulative inactivation was deter-mined by holding a cell at �80 mV, and then subjecting thecell to 20 depolarizing steps to �40 mV at the variousinterpulse intervals indicated. Fractional current is ex-pressed as current elicited at the Ith pulse divided by currentelicited at the first pulse, Io. (H) Time-dependent recoverywas determined using a two pulse protocol in 2 mM (solidsquares; n � 9) and 100 mM (open squares; n � 4) externalK�: cells were held at �90 mV, then subjected to a pair ofdepolarizing pulses to �50 mV with interpulse intervalsranging from 400 ms to 10.4 s. Inset, example of time-dependent recovery from inactivation.

230 Fry et al.

cells were held at �90 mV, and depolarized by a pairof 800 ms duration pulses to �50 mV, separated by arepolarizing interpulse to interval �90 mV for vary-ing durations. Overall, the recovery in 2 mM externalK� did not differ from the recovery in100 mM exter-nal K� [Fig. 1(H)].

Sensitivity of XKv1.10 to TEA and 4AP

Sensitivity to the blocking agents 4AP and TEA var-ies between members of the Kv1 class of channel.Whole cell patch clamp analysis of embryonic Xeno-pus spinal cord neurons has detected the presence of4AP-sensitive and TEA-insensitive K� currents(Ribera and Spitzer, 1990), while studies of embry-onic skeletal muscle have revealed the presence of K�

currents that are reduced by both 4AP and TEA, albeitat relatively high concentrations (Moody-Corbett andGilbert, 1992b). Currents carried by XKv1.10 werefound to be sensitive to both of these agents, with anapparent IC50 of 68.5 �M for inhibition by 4AP [Fig.2(A,C)], and an apparent IC50 of 17.1 mM for TEA[Fig. 2(B,D)].

Externally applied TEA has been shown to slowthe rate of C-type inactivation, via a proposed “foot inthe door” mechanism (Choi et al., 1991) where theTEA prevents inactivation by binding to a site in theK� channel pore that normally becomes constrictedduring inactivation. Application of 20 mM externalTEA caused a small but significant decrease in therate of XKv1.10 inactivation [� � 145.5 � 15 ms incontrol, 215.2 � 20 ms in 20 mM TEA, n � 4, Fig.2(E,F)]. The block was reversible, with �70–80%recovery after washout [Fig. 2(E,F)].

Internal TEA can cause N-type inactivation to slowby competing with the inactivation particle for aninternal binding site (Choi et al., 1992). When weincluded 1 mM TEA in the internal recording solutionwe observed a reduction in the peak XKv1.10 currentwithout affecting inactivation kinetics [n � 3; Fig.2(G,H)]. In contrast, we observed a marked decreasein the inactivation of XKv1.4 when 1 mM TEA wasincluded in the internal recording solution (data notshown). These data suggest that block of XKv1.10with internally applied TEA does not interfere withbinding of an intracellular inactivation particle.

Activity of XKv1.10 during SimulatedAction Potentials

Although a large number of Kv1 K� channels havebeen cloned, in only a few cases have their individualroles been elucidated. In order to investigate the po-tential role(s) of XKv1.10 during electrical activity,

Figure 2 XKv1.10 is sensitive to block by 4AP and TEA.(A,B) Cells were held at �80 mV, and subjected to adepolarization to 50 mV every 30 s after application ofblocking agent. Current records shown represent steadystate response to the indicated concentration. (C,D) Doseresponse showing peak current inhibition by increasing con-centrations of 4AP (n � 3 to 6) and TEA (n � 4 to 5).Percent inhibition calculated by dividing peak current inpresence of blocking agent by peak current in the absence ofblocking agent. (E) Including 20 mM TEA in the externalrecording solution causes a reduction in current amplitudeand rate of inactivation (n � 4). (F) Currents in (E) havebeen normalized to facilitate comparison. (G) Including 1mM TEA in the internal recording eventually reduces thecurrent amplitude but does not affect the rate of inactivation(n � 3). (H) Currents in (G) have been normalized tofacilitate comparison.

Properties of Kv1.10 K� Channels 231

cells expressing either XKv1.2, XKv1.4, or XKv1.10were voltage clamped with a simulated action poten-tial protocol (SAPP). This protocol was intended toevoke K� currents more closely resembling currentsevoked during action potentials than currents evokedduring long, square wave depolarizing pulses. Forcomparison with SAPP, Figure 3[(A),left] shows cur-rents recorded from cells expressing XKv1.2,XKv1.4, or XKv1.10 in response to a square wavedepolarization to 70 mV. While the currents evokedfrom these three cells were of similar amplitude,XKv1.2 channels produced a slowly activating, non-inactivating current, whereas XKv1.4 and XKv1.10channels produced rapidly activating and more slowlyinactivating currents. In addition, all three channelisoforms exhibited different voltages of half-activa-tion (V1⁄2, act) and steady-state half-inactivation (V1⁄2,

inact): XKv1.2 V1⁄2, act � �2.3 � 0.8 mV (n � 4, datanot shown); XKv1.4 V1⁄2, act � �15.6 � 0.8 mV andV1⁄2, inact � �31.6 � 0.7 mV (n � 4, data not shown);XKv1.10 V1⁄2, act � �11.3 � 0.8 mV and V1⁄2, inact

� �37.8 � 0.8 mV [n � 17 and 16, Fig. 1(C)].However, in response to the SAPP, the amplitudes ofthe XKv1.2-mediated currents were significantly (p� 0.05, ANOVA) lower than the XKv1.4- andXKv1.10-mediated currents [Fig. 3(A), right]. Indeed,the peak SAPP evoked current mediated by XKv1.2was 1.4 � 0.3% (n � 6) of the current evoked by asquare wave depolarizing pulse, whereas the peakSAPP evoked current mediated by XKv1.4 andXKv1.10 was 10.8 � 2.0 (n � 5) and 13.8 � 1.8% (n� 6) of the current evoked by a square wave depo-larizing pulse, respectively [Fig. 3(B), upper]. In ad-dition, closer inspection of the SAPP evoked currentsrevealed that XKv1.4 currents were active betweendepolarizing spikes whereas XKv1.2 and XKv1.10currents were not [Fig. 3(B), lower]. These observa-tions suggest that XKv1.2, XKv1.4, and XKv1.10channels play different roles in the repolarization ofexcitable membranes during electrical activity.

DISCUSSION

XKv1.10 Encodes an Inactivating K�

Channel

This is the first report describing the biophysical prop-erties of the novel XKv1.10 K� channel. The currentsrecorded from HEK293 cells transiently expressingXKv1.10 exhibited rapid activation and inactivation.The observed inactivation was unexpected, as thepredicted protein did not contain a domain with ob-vious analogy to an inactivating “ball and chain”

domain (Zagotta et al., 1990; Tseng-Crank et al.,1993; Comer et al., 1994). While this in itself does noteliminate an N-type inactivation mechanism, a simplefunctional assessment of the role of the N-terminus ofXKv1.10 was precluded by the lack of functionalcurrents when three N-terminal deletion mutants weretested (unpublished observations). However, severalobservations suggest that the inactivation of XKv1.10is not mediated by an N-type mechanism. For exam-ple, the rate of N-type inactivation is insensitive to anelevated concentration of external K� (Demo andYellen, 1991; Tseng and Tseng-Crank, 1992; Ras-musson et al., 1995) and externally applied TEA(Choi et al., 1991), whereas the presence of 100 mMK� or 20 mM TEA in the external recording solutionslowed inactivation of XKv1.10. Furthermore, inter-nally applied TEA inhibits the rate of N-type inacti-vation by interfering with binding of the inactivationparticle (Choi et al., 1991), yet no change in inacti-vation kinetics was observed when TEA was appliedintracellularly to cells expressing XKv1.10. In addi-tion, the rate of recovery from N-type inactivation issensitive to external K� (Demo and Yellen, 1991;Tseng and Tseng-Crank, 1992; Rasmusson et al.,1995), but no such sensitivity was observed withXKv1.10. Lastly, N-type inactivation can be removedby treatment of the intracellular face of the channelwith trypsin (Hoshi et al., 1990), while preliminaryexperiments suggested that the rate of inactivation ofXKv1.10 was insensitive to internally applied trypsinunder conditions where the N-type inactivation ofXKv1.4 was inhibited by this treatment (data notshown). Together, these data suggest that inactivationof XKv1.10 may instead occur via a C/P-type mech-anism.

We determined whether XKv1.10 exhibited cumu-lative inactivation using paradigms similar to thoseused by Marom and Levitan (1994). The extent ofcumulative inactivation increased as the interpulseinterval decreased, up to an apparent maximum of�35–40% at interpulse intervals of 100 ms andshorter. This was unlike the biphasic pattern exhibitedby rKv1.3, where cumulative inactivation became lessprofound at interpulse intervals less than 600 ms. Thecumulative inactivation of XKv1.10 was also con-trasted with that of rKv1.3 (Marom and Levitan,1994) in that it was unaffected by hyperpolarization ofthe interpulse interval to �100 mV. Cumulative in-activation of XKv1.4 has been evaluated with aslightly different protocol (Kerschbaum et al., 2002).When XKv1.10 was evaluated using this protocol, itwas less sensitive to cumulative inactivation thanXKv1.4, but more sensitive than XKv1.2.

The rate of recovery of XKv1.10 from inactivation

232 Fry et al.

in low external K� was similar to that exhibited byother channels without N-type inactivation (Rasmus-son et al., 1995; Fedida et al., 1999). Recovery frominactivation of XKv1.10 was unaffected by increasingexternal K� concentration, similar to that observedfor some cases of C/P-type inactivation (Tseng andTseng-Crank, 1992; Fedida et al., 1999).

Previous results have shown that occupation ofposition �3 to the signature GYG motif by a nonaro-matic amino acid results in decreased sensitivity ofvoltage-gated K� channels to TEA (MacKinnon andYellen, 1990). Consistent with this observation,XKv1.10 possesses a serine residue in this positionand was sensitive to externally applied TEA with anapparent IC50 of 17.1 mM. When combined with theobservation that XKv1.10 was sensitive to 4AP withan apparent IC50 of 68.5 �M, the pharmacologicalcharacteristics of Kv1.10 are consistent with the re-ported sensitivities of other Kv1 channels (see Co-etzee et al., 1999, and references therein).

Contribution of XKv1.10 to the K�

Currents Observed in Excitable Cells ofEmbryonic Xenopus

The developmental expression, biophysical proper-ties, and sensitivity to TEA and 4AP suggest thatXKv1.10 channels could contribute to the increase in

inactivating K� current observed in Xenopus myo-cytes in culture (Ribera and Spitzer, 1991; Moody-Corbett and Gilbert, 1992a,b; Linsdell and Moody,1995; Chauhan-Patel and Spruce, 1998; Currie and

Figure 3 Differences in K� currents activated during asimulated action potential (SAPP). [(A), left] K� currentsevoked from HEK293 cells expressing either XKv1.2,XKv1.4, or XKv1.10. Cells were held at �80 mV andsubjected to a square wave depolarization to 70 mV, asshown at the bottom. [(A), right] K� currents evoked whenthe same cells were subjected to a SAPP: the cells were heldat �80 mV, ramped to �50 mV over 10 ms, ramped to 30mV over 2 ms, ramped to �100 mV over 1 ms, rampedback to �80 mV over 1 ms, then held at �80 mV for 2 msbefore repeating. The SAPP evoked current was leak sub-tracted with a P/5 routine. Note difference in the scales inthe left and right columns. [(B), upper] Relative amplitudeof peak SAPP evoked current was significantly lower forcells expressing XKv1.2 than for cells expressing XKv1.4or XKv1.10. Relative amplitude of peak SAPP evokedcurrent was determined by dividing the peak SAPP evokedcurrent by the peak current evoked by the square wavedepolarization to 70 mV. Each current record representsmean relative SAPP evoked currents from six, five, or sixcells expressing XKv1.2, XKv1.4, or XKv1.10, respec-tively. [(B), lower] Expansion of segment indicated in upperpanel. XKv1.4 was active between SAPP spikes whereasXKv1.2 and XKv1.10 were not. Dashed line represents zerocurrent.

Properties of Kv1.10 K� Channels 233

Moody, 1999). Furthermore, the appearance ofXKv1.10 transcripts in Xenopus spinal cord coincideswith the time when action potential duration shortensin developing spinal cord neurons in culture (Barish,1986; O’Dowd et al., 1988; Ribera and Spitzer, 1990),and the potential of half-inactivation (�37.8 mV) forXKv1.10 is similar to the �40 mV reported for IKA inspinal cord neurons in culture (Ribera and Spitzer,1990), suggesting that XKv1.10 may contribute toIKA. Importantly, however, XKv1.10 may also formheteromultimers with other subunits, includingXKv1.4 (Fry et al., 2001; Kerschbaum et al., 2002)and/or � subunits (Lazaroff et al., 1999), and in thisway contribute to IKA and/or IIK.

Potential Contribution of XKv1.10 toElectrical Excitability

Inactivating A-currents, including those generated bythe Kv4 subfamily and perhaps Kv1.4, are thought toregulate excitability by determining the latency andultimately the rate of action potential firing. This mayoccur as a result of the subthreshold activation ofA-currents. In contrast, K� channels that activaterapidly at depolarized potentials, such as those of Kv3subfamily, contribute to the shortening of action po-tential duration (Moreno et al., 1995; Erisir et al.,1999; for review, see Rudy et al., 1999). Comparisonof the currents evoked by simulated action potentialsrevealed that there was proportionally more XKv1.4-and XKv1.10-mediated currents than XKv1.2-medi-ated current. This probably results from the slowerkinetics of XKv1.2 activation. However, while thefast activation of both the XKv1.4 and XKv1.10 chan-nels results in similar peak currents in response to theSAPP, XKv1.4 is activated at slightly more hyperpo-larized potentials than XKv1.10, which results in de-tectable XKv1.4-mediated currents, but not XKv1.10currents, between the simulated action potentials. Thissuggests that XKv1.10 may play a more prominentrole in shortening action potential duration than indetermining latency. Consistent with this, XKv1.10does not display significant frequency-dependent in-activation. Overall, the functional properties ofXKv1.10-mediated currents, especially in comparisonto XKv1.2- and XKv1.4-mediated currents, suggestthat this novel K� channel may play a distinct andimportant role in shaping electrical activity duringXenopus development.

The authors would like to thank G.D. Paterno and K.Mearow (Memorial University of Newfoundland) for assis-tance with the initial subcloning and expression of Kv1.10.The authors would also like to thank L.P. Henderson and

B.L. Jones (Dartmouth Medical School) for comments andsuggestions about the manuscript.

REFERENCES

Barish ME. 1986. Differentiation of voltage-gated potas-sium current and modulation of excitability in culturedamphibian spinal neurones. J Physiol 375:229–250.

Baukrowitz T, Yellen G. 1995. Modulation of K� currentby frequency and external [K�]: a tale of two inactivationmechanisms. Neuron 15:951–960.

Burger C, Ribera AB. 1996. Xenopus spinal neurons ex-press Kv2 potassium channel transcripts during embry-onic development. J Neurosci 16:1412–1421.

Chauhan-Patel R, Spruce AE. 1998. Differential regulationof potassium currents by FGF-1 and FGF-2 in embryonicXenopus laevis myocytes. J Physiol 512:109–118.

Choi KL, Aldrich RW, Yellen G. 1991. Tetraethylammo-nium blockade distinguishes two inactivation mecha-nisms in voltage-activated K� channels. Proc Natl AcadSci USA 88:5092–5095.

Coetzee WA, Amarillo Y, Chiu J, Chow A, Lau D, McCor-mack T, Moreno H, Nadal MS, Ozaita A, Pountney D, etal. 1999. Molecular diversity of K� channels. Ann NYAcad Sci 868:233–285.

Comer MB, Campbell DL, Rasmusson RL, Lamson DR,Morales MJ, Zhang Y, Strauss HC. 1994. Cloning andcharacterization of an Ito-like potassium channel fromferret ventricle. Am J Physiol 267:H1383–H1395.

Currie DA, Moody WJ. 1999. Time course of ion channeldevelopment in Xenopus muscle induced in vitro byactivin. Dev Biol 209:40–51.

Demo SD, Yellen G. 1991. The inactivation gate of theShaker K� channel behaves like an open-channelblocker. Neuron 7:743–753.

Desarmenien MG, Clendening B, Spitzer NC. 1993. In vivodevelopment of voltage-dependent ionic currents in em-bryonic Xenopus spinal neurons. J Neurosci 13:2575–2581.

Erisir A, Lau D, Rudy B, Leonard CS. 1999. Function ofspecific K(�) channels in sustained high-frequency firingof fast-spiking neocortical interneurons. J Neurophysiol82:2476–2489.

Ernsberger U, Spitzer NC. 1995. Convertible modes ofinactivation of potassium channels in Xenopus myocytesdifferentiating in vitro. J Physiol 484:313–329.

Fedida D, Maruoka ND, Lin S. 1999. Modulation of slowinactivation in human cardiac Kv1.5 channels by extra-and intracellular permanent cations. J Physiol 515:315–329.

Fry M, Moody-Corbett F. 1999. Localization of sodium andpotassium currents at sites of nerve-muscle contact inembryonic Xenopus muscle cells in culture. PflugersArch 437:895–902.

Fry M, Paterno G, Moody-Corbett F. 2001. Cloning andexpression of three K� channel cDNAs from Xenopusmuscle. Brain Res Mol Brain Res 90:135–148.

234 Fry et al.

Gurantz D, Lautermilch NJ, Watt SD, Spitzer NC. 2000.Sustained upregulation in embryonic spinal neurons of aKv3.1 potassium channel gene encoding a delayed recti-fier current. J Neurobiol 42:347–356.

Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ.1981. Improved patch-clamp techniques for high-resolu-tion current recording from cells and cell-free membranepatches. Pflugers Arch 391:85–100.

Hoshi T, Zagotta WN, Aldrich RW. 1990. Biophysical andmolecular mechanisms of Shaker potassium channel in-activation. Science 250:533–538.

Kalman K, Nguyen A, Tseng-Crank J, Dukes ID, ChandyG, Hustad CM, Copeland NG, Jenkins NA, Mohren-weiser H, Brandriff B, et al. 1998. Genomic organization,chromosomal localization, tissue distribution, and bio-physical characterization of a novel mammalian Shaker-related voltage-gated potassium channel, Kv1.7. J BiolChem 273:5851–5857.

Kerschbaum HH, Grissmer S, Engel E, Richter K, LehnerC, Jager HA. 2002. Shaker homologue encodes an A-typecurrent in Xenopus laevis. Brain Res 927:55–68.

Lazaroff MA, Hofmann AD, Ribera AB. 1999. Xenopusembryonic spinal neurons express potassium channel Kv-beta subunits. J Neurosci 19:10706–10715.

Levy DI, Deutsch C. 1996. A voltage-dependent role forK� in recovery from C-type inactivation. Biophys J71:3157–3166.

Linsdell P, Moody WJ. 1995. Electrical activity and calciuminflux regulate ion channel development in embryonicXenopus skeletal muscle. J Neurosci 15:4507–4514.

Lopez-Barneo J, Hoshi T, Heinemann SH, Aldrich RW.1993. Effects of external cations and mutations in thepore region on C-type inactivation of Shaker potassiumchannels. Receptors Channels. 1:61–71.

MacKinnon R, Yellen G. 1990. Mutations affecting TEAblockade and ion permeation in voltage-activated K�

channels. Science 250:276–279.Marom S, Levitan IB. 1994. State-dependent inactivation of

the Kv3 potassium channel. Biophys J 67:579–589.Moody-Corbett F, Gilbert R. 1992a. Two classes of potas-

sium currents in Xenopus muscle cells in young cultures.NeuroReport 3:319–322.

Moody-Corbett F, Gilbert R. 1992b. A K� current in Xe-nopus muscle cells which shows inactivation. NeuroRe-port 3:153–156.

Moreno H, Kentros C, Bueno E, Weiser M, Hernandez A,Vega-Saenz dM, Ponce A, Thornhill W, Rudy B. 1995.Thalamocortical projections have a K� channel that isphosphorylated and modulated by cAMP-dependent pro-tein kinase. J Neurosci 15:5486–5501.

O’Dowd DK, Ribera AB, Spitzer NC. 1988. Developmentof voltage-dependent calcium, sodium, and potassiumcurrents in Xenopus spinal neurons. J Neurosci 8:792–805.

Rasmusson RL, Morales MJ, Castellino RC, Zhang Y,Campbell DL, Strauss HC. 1995. C-type inactivationcontrols recovery in a fast inactivating cardiac K� chan-nel. Kv1.4 expressed in Xenopus oocytes. J Physiol 489:709–721.

Ribera AB. 1990. A potassium channel gene is expressed atneural induction. Neuron 5:691–701.

Ribera AB, Nguyen DA. 1993. Primary sensory neuronsexpress a Shaker-like potassium channel gene. J Neurosci13:4988–4996.

Ribera AB, Spitzer NC. 1990. Differentiation of IKA inamphibian spinal neurons. J Neurosci 10:1886–1891.

Ribera AB, Spitzer NC. 1991. Differentiation of delayedrectifier potassium current in embryonic amphibian myo-cytes. Dev Biol 144:119–128.

Roeper J, Lorra C, Pongs O. 1997. Frequency-dependentinactivation of mammalian A-type K� channel KV1.4regulated by Ca2�/calmodulin-dependent protein kinase.J Neurosci 17:3379–3391.

Rudy B, Chow A, Lau D, Amarillo Y, Ozaita A, SaganichM, Moreno H, Nadal MS, Hernandez-Pineda R, Hernan-dez-Cruz A, et al. 1999. Contributions of Kv3 channels toneuronal excitability. Ann NY Acad Sci 868:304–343.

Sanguinetti MC, Curran ME, Zou A, Shen J, Spector PS,Atkinson DL, Keating MT. 1996. Coassembly ofK(V)LQT1 and minK (IsK) proteins to form cardiacI(Ks) potassium channel. Nature 384:80–83.

Spitzer NC, Ribera AB. 1998. Development of electricalexcitability in embryonic neurons: mechanisms and roles.J Neurobiol 37:190–197.

Spruce AE, Moody WJ. 1992. Developmental sequence ofexpression of voltage-dependent currents in embryonicXenopus laevis myocytes. Dev Biol 154:11–22.

Tseng GN, Tseng-Crank J. 1992. Differential effects ofelevating [K]o on three transient outward potassiumchannels. Dependence on channel inactivation mecha-nisms. Circ Res 71:657–672.

Tseng-Crank J, Yao JA, Berman MF, Tseng GN. 1993.Functional role of the NH2-terminal cytoplasmic domainof a mammalian A-type K channel. J Gen Physiol 102:1057–1083.

Zagotta WN, Hoshi T, Aldrich RW. 1990. Restoration ofinactivation in mutants of Shaker potassium channels bya peptide derived from ShB. Science 250:568–571.

Properties of Kv1.10 K� Channels 235