inactivation defects produced by a myopathic ii-s6 mutation of the muscle sodium channel

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS 246, 792–796 (1998)ARTICLE NO. RC988718

Inactivation Defects Produced by a Myopathic II-S6Mutation of the Muscle Sodium Channel

Oscar Moran,1 Mario Nizzari, and Franco ContiIstituto di Cibernetica e Biofisica, CNR, Via De Marini, 6, I-16149 Genova, Italy

Received April 21, 1998

the transient current, or a shift in voltage dependenceWe have studied the expression in frog oocytes of of inactivation, or some combination of these effects

the a-subunit of the rat skeletal muscle sodium chan- (see [2, 3, 5, 6] for reviews).nel mutation S798F, homologous to the mutation We have studied the expression in frog oocytes ofS804F of the human isoform, that causes potassium the a-subunit of the adult rat skeletal muscle sodiumaggravated myotony (PAM), a muscular hereditary channel (rSkM1, m1), which has more than 90% iden-disease in humans. Wild type channels show a bi-

tity with the human one. The frog oocyte expression ofmodal inactivation, with two gating modes that inac-the a-subunit alone shows an abnormally high percent-tivate with time constants that differ at least by oneage of a slowly inactivating mode [7-11], that is sub-order of magnitude and a steady steady-state voltagestantially reduced by co-expression of the b1-subunitdependence of the slow mode shifted by /27 mV rela-[7, 12, 13]. This exacerbation of the slow mode makestive to that of the fast mode. In the myopathy-linkedthis system more appropriate for studying the intrinsicmutant the propensity of the channel to gate in themodal propensity of the a-subunit and the specific prop-slow mode is significantly increased and there areerties of the slow mode. In this communication we com-alterations in the inactivation properties of bothpare the functional properties of the fast and slowmodes. The half inactivation potential of the fast

mode is shifted negatively, and the inactivation ki- modes of the wild type sodium channel with those of thenetics of both modes are slower, with an apparent PAM mutant S798F, that has not been characterisedshift in their voltage dependence. The changes on before. We find that this mutation changes the inacti-the inactivation properties of the mutant channel vation properties and increases the intrinsic propensitymay be related with the muscle fibre hyperexcitabil- of sodium channel a/subunits for the slowly inactivat-ity observed patients affected by PAM. q 1998 Academic ing mode.Press

MATERIALS AND METHODS

Site-directed mutagenesis of the cDNA coding for the rSkM1 so-Twenty one different point mutations of the genedium channel a-subunit [11] was accomplished using the Quik-SCN4A, localised in the chromosome 17q (23.1 to 25.2)Change kit (Stratagene). The mutation was verified by sequencingand coding for the adult skeletal muscle sodium chan-400 bp in the flanking region near the mutagenic point. Wild-type

nel a-subunit (hSkM1), have been linked with so-called (WT) and mutant cDNA’s were transcribed in vitro using the cappedmuscle sodium-channelopathies [1], a group of autoso- mMessage mMachine kit (Ambion). Oocytes were surgically ex-

tracted from anaesthetised Xenopus laevis, isolated enzymalicallymal dominant hereditary muscle diseases, includingwith 1 mg/ml of collagenase-A (Sigma), injected with 50 nl of 125–paramyotonia congenita (PC), potassium aggravated250 ng/ml solutions of cRNA in DEPC-water and incubated 4 to 7myotonia (PAM) and hyperkalemic periodic paralysis days in Barth’s solution at 187C [14].

(HyPP), whose symptoms can be attributed to hyperex- Sodium currents were measured from membrane macro-patchescitability of the sarcolema [2, 3]. The key manifesta- in the cell-attached configuration using an Axopatch-200B amplifier

(Axon Instruments). Aluminium-silicate glass micropipettes (Hil-tions of a muscle sodium-channelopathy are stiffnessgemberg) were coated with silicone rubber and fire polished to a tipand weakness [4]. In all studied mutations, the primarydiameter yielding a resistance of 0.6 to 1.2 MV when willed with thedefect is some modification of normal inactivation, working solution. The pipettes were filled with normal frog Ringer

causing either persistent currents, or a slower decay of (in mM: NaCl 115, KCl 2.5, CaCl2 1.8, HEPES 10; pHÅ7.4). Duringelectrophysiological recording, the oocytes were maintained in a highpotassium solution (in mM: KCl 120, TRIS-Cl 20, EGTA 5; pH 7.4),in which the cell membrane potential was 0{2 mV. Therefore, the1 To whom correspondence should be addressed. Fax:/39-10-6475-

500. E-mail: [email protected]. membrane patch potential, V, was estimated to be just opposite to

0006-291X/98 $25.00Copyright q 1998 by Academic PressAll rights of reproduction in any form reserved.

792

AID BBRC 8718 / 6953$$1121 05-12-98 22:34:48 bbrcg AP: BBRC

Vol. 246, No. 3, 1998 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

the pipette potential. The output of the patch-clamp amplifier wasfiltered with the low-pass 4-pole Bessel filter built into the amplifier,set at a cut-off frequency of 5 kHz. The membrane current was sam-pled at 20 kHz. The membrane was kept at a holding potential of0120 mV, and the current response to depolarising voltage pulseswas measured. Voltage stimulation and data acquisition were doneby 16 bit D-A and A-D converters (ITC-16, Instrutech), controlled bya Macintosh microcomputer, using the Pulse software (Heka Elec-tronik). Linear current responses were measured from sub-thresholdstimulation and digitally subtracted. All measurements were doneat a controlled temperature of 15{0.5 7C.

RESULTS

Both WT and mutant cRNA’s expressed sodium cur-rents in oocytes (Fig. 1). In WT channels a maximum

FIG. 1. Sodium currents expressed in Xenopus oocytes microin-peak current of 800{300 pA (mean{sem; nÅ11) in,jected with cRNA coding WT (A) or the mutant S798F (B) rSkM1 a-measured within 3-4 minutes after having obtained thesubunit sodium channels, recorded within 4–5 minutes from patchgiga-seal. Mutant S798F yielded generally smaller cur- isolation. Currents were elicited by 30 ms depolarising pulses from

rent amplitudes (300{160 pA, nÅ9). In all cases the a holding potential of 0120 mV to a test potential between 050 mVcurrent increased during the experiment up to 2- to and 50 mV in 10 mV steps. The inactivating phase of currents is

characterised by two time constants. (C) Comparison between the3-fold after 20 to 35 min from patch isolation (WT:current of WT and S798F channels, a voltage step near the half1700{600 pA, nÅ9; S798F: 750{400 pA, nÅ8). Theactivation potential ({5 mV). The two responses were scaled to aboutsodium current in response to a step depolarisation is the same peak value. Notice the significant increase of the slow inac-

characterised by a rapid increase followed by a biphasic tivating component in the mutant.inactivation [7-9, 12]. The falling phase of the currentcan be well fitted with a double-exponential functionas I(t)Åa1 exp(0t/t1)/a2 exp(0t/t2), where t1 and t2

the ratio between late current and peak current hasdiffer by about one order magnitude, and the ratio ofbeen described [16-18] and a similar finding has beenthe two amplitudes a1 and a2 is fairly independent ofalso reported for human muscle fibres from patientsthe voltage step. It has been proposed that the twowith PAM s [19]. Our results suggest that the abnor-components are due to the separate contribution of amally late currents for pulses of 50 ms duration or lessmixed population of channels, that at the time of theare in fact the result of an increased propensity of thestimulus are either in a fast, which we shall call M1,mutant channels for the slow gating mode.or in a slow mode of inactivation, called M2 [7, 8]. We

We have found in most experiments a negative shiftfind that the slow component of the current is systemat-of 5 to 30 mV in the activation curves within 20-40ically more pronounced in mutant S798F than in WT,minutes from the first recording on a single patch, simi-as shown in Fig. 1. This results in a significantly largerlarly to that has been observed on HEK 293 cells per-late current for pulses of 30 ms duration. To quantifymanently transfected with rSkM1 channels [20] or inthe propensity of the channels to gate in either mode,cardiac muscle fibres [21]. To avoid this problem, wewe define the parameter PM2Åa2/(a1/a2), as a measurehave chosen the half activation potential, V1/2 (the po-of the probability of finding a channel in mode M2 dur-tential at which the peak conductance reaches 50% ofing the test stimulus. PM2 was measured from the dou-their maximum value), measured periodically and sys-ble exponential fit of the decaying phase of the sodiumtematically during each experiment, as a reference po-current evoked by 30 to 150 ms depolarisation to volt-tential. The absolute values of V1/2 could vary in ourages in the range between 020 and /40 mV, where themeasurements between 07 mV and 040 mV in 25 min-value of PM2 was quite voltage independent. Mutantutes. In agreement with previous reports [9, 12], weS798F had significantly larger values of PM2 (0.47{found no significant differences between the activation0.07; nÅ8), compared to WT (0.104{0.06; nÅ15). Thecurves of mode M1 and mode M2, so we do not expectM2 component of sodium channels expressed by a-sub-that V1/2 changes with PM2. Therefore, we expressedunits in oocytes is more pronounced [8, 11, 12] than inthe voltage-dependencies as a function of VrelÅV0V1/2 ,the heterologous expression of muscle sodium channelswhere V1/2 , if needed, was obtained by interpolation ofin mammalian cells [15], where the channels gating inmeasurements at earlier and later times. When re-mode M2 are less than 2%. This allows to interpretferred to V1/2 , voltage-dependent data obtained at dif-our results unambiguously in terms of modal gatingferent times during an experiment were fully reproduc-propensity. For channels expressed in mammalian cellsible and gave much more confidence in the comparisontransfected with rat or human SkM1 channels with

mutations linked with PAM a 2- to 5-fold increase of of the results from different experiments.

793



The activation kinetics of the WT and mutant chan- The peak response to the test pulse vs. The duration ofthe recovery interval was fitted with a double-exponen-nels were characterised by measurements of the time

to reach half of the peak current amplitude, t0.5 . This tial function, yielding the time constants trec1 and trec2 .The inactivation time constant data, pooled in 20 mVparameter was measured for values of Vrel between

020 mV and 70 mV and binned in 10 mV intervals, bins, are presented in Figs. 2B and 2C.The voltage-dependence of the inactivation time con-as shown in Fig. 2A. In WT channels, t0.5 measured

for Vrel in the 010 to 0 mV interval is 0.64{0.05 ms stant for the M1 mode in the S798F mutant is slightlyaltered, compared with the WT. For Vrel°050 mV the(nÅ14), and decreases monotonically, to a value of

0.19{0.01 ms for Vrelú60 mV. We find that the acti- values of t1 for the two phenotypes are very similar.However, t1 tends to be larger for S798F for Vre1¢010vation kinetics of the PAM mutant S798F are sig-

nificantly slower (t0.5 Å1.05{0.10 ms for 010mVõ mV, with a sensible reduction of the slope of its voltagedependence for Vrelú0 mV, and a significantly high pla-Vrel°0 mV, nÅ8). Similar results, taking into account

the differences on the measured V1/2 , have been re- teau for Vrel¢30 mV (Fig. 2B). A similar phenomenon,referred generically to a single inactivation process,ported in other PAM-linked mutants expressed in

mammalian cells [17, 18]. As described above, the has been described in other PAM linked mutations ex-pressed in mammalian cells [17, 18]. In WT channelsdecaying phase of the sodium currents can be well

fitted with a double-exponential function. We have t1 decreases by a factor of 1.8 for Vrel measured from 30mV (t1Å0.58{0.05 ms, nÅ30) to 70 mV (t1Å0.33{0.03,systematically fitted the currents evoked by 30 ms

pulses to membrane potentials between 030 to 40 nÅ16), whereas the decrease for S796F is by a factor of1.4, from 0.84{0.08 ms (nÅ14) to 0.59{0.05 ms (nÅ4).mV. When the fast component of the current was too

small, in particular at low membrane potentials, we Much larger significant differences between WT andS798F channels are observed in the inactivation timeattempted to dissect the two gating modes using an

appropriate conditioning protocol that reduced the constant of mode M2 (Fig. 2C). Currents expressed inoocytes by mutant S798F show an overall increase ofamplitude of the M2 component [9]. In this protocol

the test pulse to V¢020 mV was preceded by a condi- the t2 values, by a factor of 2 to 4, and an apparentshift of Ç20 mV in the t2-Vrel curve. This results to-tioning pre-pulse (to 0 mV for 100 ms) that caused a

complete inactivation of all channels, followed by a gether with the increased intrinsic propensity of the a-subunits for the mode M2, would most likely cause the30 ms repolarisation at 0120 mV. A large recovery

from inactivation for channels gating in mode M1 is increase in the late sodium current described for otherPAM mutations expressed in mammalian cells [16-18].obtained after this relatively short repolarisations,

and most of the channels in mode M1 are again avail- The voltage-dependence of steady-state inactivationwas measured with a double-pulse protocol. A 20 msable for activation, while most of the M2 channels are

still inactivated. The time constant of inactivation depolarising test pulse to Vrel¢0 was preceded by 500ms conditioning pulses to membrane potentials be-for channels gating in mode M1, t1 , where therefore

easily measured from conditioned records and the tween 0140 and 030 mV. A value proportional to thecontribution of mode M2 to the test currents was esti-non-conditioned current records were then fitted fix-

ing the fast time constant, t1 . The estimates of t1 mated from the mean current, between 17 ms and 20ms, where practically all the channels gating in modeobtained from conditioned pulses or from double ex-

ponential fit of non-conditioned records with a large M1 are already inactivated (t1 õ 6 ms; see Fig. 2B). Asimilar estimate of the M1 contribution was obtainedfast component were statistically indistinguishable.

Therefore, we used this strategy routinely, in partic- by subtracting from the peak current An estimate ofthe M2 component extrapolated to the time to peak theular for Vrel° 10 mV.

The time constants for the onset of the inactivation standard values of t2 . The relative half inactivationpotential for each mode, Vh1 and Vh2, was calculatedat low membrane potentials (Vrel between 020 mV and

060 mV), were estimated by fitting with a double expo- by fitting the Vrel-dependence of the normalised cur-rents with a Boltzmann function.nential the decay of the peak current response to a

depolarising pulse as a function of the duration of a WT channels gating in the mode M1 inactivate atrelative membrane potentials about 27 mV more nega-conditioning pre-pulse to a fixed membrane potential,

between 0100 mV and 060 mV. The two time con- tive than in mode M2 (Vh1Å056.3{0.8 mV; nÅ15.Vh2Å029.6{1.4; nÅ12) [9]. In the mutant S798F, thisstants, t1 and t2 , were interpreted as characteristics

of the of inactivation of M1 and M2 at the pre-pulse difference is slightly reduced mainly because the rela-tive half-inactivation potential of mode M1 has a sig-potential. At potentials between 0140 mV and 0100

mV, the inactivation time constants were instead mea- nificantly positive shift (Vh1Å047.7{3.1 mV; nÅ15) rel-ative to WT, whereas that of mode M2 is not signifi-sured by inactivating all channels with a conditioning

pulse of 150 ms to 0 mV, and allowing the recovery from cantly changed (Vh2Å026.3{1.9 mV; nÅ10). Despitethe large scattering of the data reported for the heterol-inactivation a variable time (1-500 ms) at the recovery

potential before applying a test pulse of 20 ms to 10 mV. ogous expression of the other mutant channels in mam-

794



FIG. 2. Activation and inactivation kinetics of the WT (circles) and S798F mutant (squares) sodium currents expressed in oocytes. Theactivation kinetics (A), expressed as the time to reach half of the maximum current, t0.5 , and the inactivation time constants of M1 mode(B), t1 , and of M2 mode (C), t2 , are plotted against Vrel . Data represent mean{sem of 5 to 30 measurements.

malian cells, a tendency to a positive shift a loosely channelopathy. In addition, another feature that mightcontribute to the hyperexcitability of muscle fibres ondefined mean inactivation curve seems to be a common

feature of other PAM linked mutations [17, 18]. patients with the S798F mutation is the kinetics ofinactivation. At the potentials that causes inactivationof the channels activated during an action potential,DISCUSSIONi.e. for Vrel ¢ 0 mV, both, the M1 and M2 modes of thePAM mutant have slower time constants of inactiva-

We have successfully expressed the wild type pheno- tion, that should increase further the excitability of thetype and the PAM linked mutation S798F of the a- cell by reducing the length of the refractory period.subunit of rSkM1 channels by cRNA injection in Xeno- It has been established in WT a-subunits that thepus oocytes. The currents mediated by these proteins half activation potential of mode M2 is more positiveare characterised by a double-exponential inactivation, than that of the naturally predominant node M1 [9,due to a bimodal gating behaviour, comprising a fast 12]. The S798F mutant is characterised by a positivemode, M1, and a slower mode of inactivation, M2, with shift of Vh1 by Ç10 mV, that reduces the difference ininactivation kinetics differing by more than 10-fold [7- the voltage dependence of the two modes and causes9, 12]. Similarly to what has been described for other an increase of the channel ‘‘excitability window’’, thatPAM mutants, and for mutations linked to HyPP or

would make the muscle fibres more excitable, a com-PC, the main effects occur on the inactivation proper-mon characteristic of the channelopathies [2, 3, 25].ties. Both animal [22] and theoretical [9, 23] models

In conclusion, we have compared for the first timehave demonstrated that even small defect of the so-the inactivation properties of the WT with that of adium channel inactivation is sufficient to cause myoto-PAM linked mutation S798F, distinguishing the effectnia and/or paralysis in living muscle tissues. In ouron the two gating modes of the sodium channel. Thisanalysis we have examined separately the effects ofdistinction could be easily made by studying the expres-one particular mutation, as not yet characterised elec-sion of the sole a-subunit in frog oocytes, which en-trophysiologically, on the two gating modes of the chan-hances abnormally the contribution of the slowly inac-nels, in view of the fact that the episodic myotonia ortivating mode M2. We find that, besides a statisticallyparalysis affecting patients with these disorders maysignificant increase of its intrinsic propensity, PM2, alsobe explained not only by the presence of a fraction ofthe inactivation properties of mode M2 are changed inchannels that do not inactivate [23], but also by anthe direction of making larger late currents. If the b-increased the proportion of channels that gate in modesubunit simply reduces the mode M2 by a constantM2 [9, 24]. We have characterised the two gating modesfactor, these results would provide a good quantitativeof the WT and mutant channels, using the heterologousdescription of the behaviour of mutant channels, couldexpression of rSkM1 a-subunits in oocytes, which havegive new insights for understanding the physiopathol-a larger propensity of gating in mode M2 (PM2É10%) [8,ogy of sodium channelopathies.9, 11, 12], and allow a more accurate characterisation

of this mode, that is largely depressed in transfectedmammalian cells (PM2õ0.3%) [15], probably by to the

ACKNOWLEDGMENTSco-assembly of a- and b- subunits [16].We have observed a two- to four-fold increase of the

intrinsic propensity of S798F a-subunits channels to We thanks E. Gaggero for electronic technical assistance and R.gate in the mode M2. This might be the most important Melani for the oocyte preparation. This work is supported by Tele-

thon, project 926.cause of the functional defect of the related sodium

795



13. Patton, D. E., Isom, L. L., Catterall, W. A., and Goldin, A. L.REFERENCES(1994) J. Biol. Chem. 269, 17640–17655.

14. Stuhmer, W. (1992) in Methods in Enzymology: Ion Channels1. Bulman, D. E. (1997) Hum. Mol. Genet. 6, 1679–1685. (Rudy, B., and L. E. Iverson, Eds.), pp. 319–339 Academic Press,2. Cannon, S. C. (1996) TINS 19, 3–10. London.

15. Ukomadu, C., Zhou, J., Sigworth, F. J., and W. S., A. (1992) Neu-3. Hoffman, E. P., Lehmann-Horn, F., and Rudel, R. (1995) Cell 80,ron 8, 663–676.681–686.

16. Mitrovic, N., George, A. L. J., Heine, R., Wagner, S., Pika, S.,4. DeSilva, S. M., Kuncl, R. W., J. W., G., Cornblath, D. R., and S.,Hartlaub, U., Zhou, M., and Lerche, H. (1994) J. Physiol. 478,C. (1990) Nerve Muscle 13, 21–26.395–402.

5. Barchi, R. L. (1995) Ann. Rev. Physiol. 57, 355–385. 17. Mitrovic, N., George, A. L., Lerche, H., Wager, S., Fhalke, C.,6. Cannon, S. C. (1996) Ann. Rev. Neurosci. 19, 141–164. and Lehmann-Horn, F. (1995) J. Physiol. 487, 107–114.

18. Hayward, L. J., Brown, Jr, R. H., and Cannon, S. C. (1996) J.7. Zhou, J., Potts, J. F., Trimmer, J. S., W. S., A., and F. J., S. (1991)Gen. Physiol. 107, 559–576.Neuron 7, 775–785.

19. Lerche, H., Heine, R., Pika, U., George, A., Mitrovic, N., Browat-8. Moorman, J. R., Kirsch, G. E., VanDongen, A. M. J., Joho, R. H.,zki, M., Weiss, T., Rivet-Bastide, M., Franke, C., Lomonaco, M.,and Brown, A. M. (1990) Neuron 4, 243–52.Ricker, K., and Lehmann-Horn, F. (1993) J. Physiol. 470, 13–

9. Moran, O., Nizzari, M., and Conti, F. (1998a) in Neuronal Cir- 22.cuits and Networks (Torre, V., and J. Nicolls, Eds.), pp. Plennum 20. Cummins, T. R., and Sigworth, F. J. (1996) Biophys. J. 71, 227–press, New York. 236.

10. Auld, V. J., Goldin, A. L., Krafte, D. S., Marshall, J., Dunn, J. M., 21. Wend, D. J., Starmer, C. F., and Grant, A. O. (1992) Am. J. Phys-Catteral, W. A., Lester, H. A., Davidson, N., and Dunn, R. J. iol. 263, C1234–C1240.(1988) Neuron 1, 449–61. 22. Cannon, S. C., and Corey, D. P. (1993) J. Physiol. 466,

11. Trimmer, J. S., Cooperman, S. S., Tomiko, S. A., Zhou, J., Crean, 23. Cannon, S. C., Brown, R. H., and Corey, D. P. (1993) Biophys. J.S. M., Boyle, M. B., Kallen, R. G., Sheng, Z., Barchi, R. L., Sig- 65, 270–288.worth, F. J., Goodman, R. H., Agnew, W. S., and Mandel, G. 24. Moran, O., Nizzari, M., and Conti, F. (1998b) Pflugers Arch. 435,(1989) Neuron 3, 33–49. R16.

25. Lehmann-Horn, F., and Rudel, R. (1996) Rev. Physiol. Biochem.12. Ji, S., Sun, W., George, Jr., A. L., Horn, R., and Barchi, R. L.(1994) J. Gen. Physiol. 104, 625–643. Pharmacol. 128, 159–268.

796


inactivation defects produced by a myopathic ii-s6 mutation of the muscle sodium channel

Documents