A cluster of hydrophobic amino acid residues ... - Semantic Scholar

3 downloads 0 Views 1MB Size Report
Offord, J., Charbonneau, H., Walsh, K., Goldin, A. L. &. Catterall, W. A. (1992) Science 256, 839-842. 19. Schumperli, D., Howard, B. H. & Rosenberg, M. (1982) ...
Proc. Natl. Acad. Sci. USA Vol. 89, pp. 10910-10914, November 1992 Physiology

A cluster of hydrophobic amino acid residues required for fast Na+-channel inactivation JAMES W. WEST*, D. EARL PATTONt, TODD SCHEUER*, YANLING WANG*, ALAN L. GOLDINt, AND WILLIAM A. CATTERALL* *Department of Pharmacology, SJ-30, University of Washington, Seattle, WA 98195; and tDepartment of Microbiology and Molecular Genetics, University of California, Irvine, CA 92717

Contributed by William A. Catterall, August 3, 1992

The inward Na+ current underlying the acABSTRACT tion potential in nerve is terminated by inactivation. The preceding report shows that deletions within the intracellular linker between domains m and IV remove inactivation, but mutation of conserved basic and paired acidic amino acids has little effect. Here we show that substitution of glutamine for three clustered hydrophobic amino acids, Ile-1488, Phe-1489, and Met-1490, completely removes fast inactivation. Substitution of Met-1490 alone slows inactivation sign tly, substitution of Ile-1488 alone both slows inactivation and makes it incomplete, and substitution of Phe-1489 alone removes inactivation nearly completely. These results demonstrate an essential role of Phe-1489 in Na+-channel inactivation. It is proposed that the hydrophobic cluster of Ile-1488, Phe-1489, and Met-1490 serves as a hydrophobic latch that stabilizes the inactivated state in a hinged-lid mechanism of Na+-channel inactivation.

Depolarization of the membrane of excitable cells results in a transient inward Na+ current that is terminated within a few milliseconds by the process of inactivation (1). Perfusion of the intracellular surface of the sodium channel with proteolytic enzymes prevents inactivation, implicating intracellular structures in the inactivation process (2). The a subunit of sodium channels consists of four homologous domains connected by cytoplasmic linker sequences (3, 4). Antibodies directed against the intracellular linker between homologous domains III and IV (LiII/Iv; see figure 1 of preceding paper, ref. 5) completely block fast inactivation of affected single sodium channels (6, 7). Expression of the sodium channel as two polypeptides with a cut between domains III and IV slows inactivation -20-fold (8), and small insertions in this loop also slow inactivation (9). Phosphorylation of a single serine residue in LIII/IV by protein kinase C slows inactivation (10). The preceding paper (5) shows that highly conserved positively charged residues and a highly conserved pair of negatively charged residues in LiII/Iv are not essential for fast sodium channel inactivation, but supports an integral role for LIII/Iv in inactivation of the Na+ channel since deletions of 10-amino acid segments within it completely block fast inactivation. To assess the role of hydrophobic amino acids within LIII/Iv in the inactivation process, site-directed mutants were constructed in which conserved hydrophobic residues were altered, expressed in Xenopus oocytes or transfected Chinese hamster cells, and analyzed by wholecell voltage clamp and single-channel recording. EXPERIMENTAL PROCEDURES Materials. M13mpl8 was obtained from Bethesda Research Laboratories. pVA2580, containing the full-length cDNA encoding rat brain type IIA Na+ channel a subunit (11,

12), was obtained from Rob Dunn (McGill University, Montreal, Quebec). Zem228 was obtained from Eileen Muhlvihill (Zymogenetics, Seattle). Escherichia coli strains HB101, JM103, and CJ236 and cell lines CHO-K1 and R-1610 were obtained from the American Type Culture Collection. Oligonucleotide-Directed Muagenesi. An EcoRV restriction endonuclease recognition site was added to the polylinker of the M13 vector M13mpl8 (Gibco/BRL) by insertion of the self-complementary oligonucleotide GGATATCCCATG (13). A fragment of the rat brain type IIA Na+ channel a subunit (12), encoding nucleotides 4306-6433, was excised from the full-length cDNA sequence (12) by digestion with the restriction endonuclease EcoRV and inserted into the EcoRV site in MP18RV in an orientation such that the resulting bacteriophage single-stranded DNA would contain the coding strand of Na+ channel a subunit cDNA gene (13). The resulting construct, MP18RVNC-1, was used as the template DNA for making mutations in LiuIIv. Uracilcontaining single-stranded DNA templates were prepared from MP18RVNC-1 as described (14). All mutations were made according to the procedure of Ausubel et al. (15) and confirmed by DNA sequencing. EcoRV DNA fragments containing the mutated sequences were inserted into ZemRVSP6-2580 (16) by standard procedures (13). Sodium-Channel Expression in Xenopus laevis Oocytes. Fulllength 6.4-kilobase wild-type or mutant Na+ channel a subunit RNA was synthesized in vitro with SP6 RNA polymerame from ZemSP6-2580 templates (17). RNA encoding the p1 subunit was transcribed from p13l.ClAa (18). Preparation and maintenance of oocytes and RNA injections were carried out as described in the preceding paper (5). Cell Transfection. Chinese hamster cells were maintained in RPMI 1640 (GIBCO/BRL) containing 5-10o (vol/vol) fetal calf serum (HyClone), penicillin (20 jug/ml), and streptomycin (10 ,ug/ml). R-1610 cells (19) were transfected with 3 j.g of DNA per 105 cells by using N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methyl sulfate (DOTAP; Boehringer Mannheim) as described by the manufacturer. After 7-10 days of selection for resistance to G418 (400 ,ug/ml), 24 colonies were isolated and transferred to tissue culture dishes for expansion and analysis. Expression of the cDNA was measured by isolation of mRNA and analysis by Northern blotting as described (16). Electrophysiological Methods. Electrophysiological recordings from Xenopus oocytes were performed as described by Patton et al. (5). Patch clamp experiments on mammalian cells were performed in the whole-cell or cell-attached recording configurations as described (10, 16) with leak and capacitance compensation by the P/4 method (20) for wholecell experiments. The extracellular solution for whole-cell voltage clamp experiments was 130 mM NaCl, 4 mM KCl, 1.5 mM CaCl2, 1 mM MgCl2, 5 mM Hepes, and 5 mM glucose

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Abbreviation: Lill/jv, linker between homologous domains III and IV.

10910

Physiology: West et aL

Proc. Natl. Acad. Sci. USA 89 (1992)

(adjusted to pH 7.4 with NaOH). The intracellular solution was 105 mM CsF, 40 mM CsCl, 10 mM NaCI, 10 mM Cs4EGTA, and 10 Hepes (pH 7.3 with CsOH). For cell-

attached patch recording experiments, the pipet solution contained 175 mM NaCl, 5 KCl, 1 CaCl2, 2 MgCl2, and 5 Hepes (pH 7.4 with NaOH). The extracellular solution contained 10 mM NaCl, 150 mM KC1, 1.5 mM CaC12, 2 mM MgCl2, 1 mM K4EGTA, and 5 mM Hepes (pH 7.4 with

KOH).

RESULTS Effects of Mutations of Hydrophobic Amino Acids on Na+Channel Inactivation in Xenopus Oocytes. The amino acid sequence of LIII/1v is shown in Fig. 1A. Initially two mutations were constructed in which either Phe-1483 (F1483Q) or the contiguous hydrophobic residues Ile-1488, Phe-1489, and Met-1490 (IFMQ3) were replaced with glutamine, as indicated by the vertical lines in Fig. 1A. RNA encoding mutant Na+ channel a subunits was coinjected into Xenopus oocytes with RNA encoding the 1 subunit (18), and the expressed channels were analyzed by two-microelectrode voltage clamp recording. Na+ currents recorded in oocytes injected with RNA encoding Na+ channel mutant F1483Q (Fig. 1B) had macroscopic inactivation kinetics similar to wild-type, whereas those with mutant IFMQ3 showed a dramatic removal offast inactivation (Fig. 1B). Na+ currents resulting from IFMQ3 coexpressed with the 81 subunit decayed with two time constants of 520 + 70 msec and 3890 + 620 msec (n = 4) at - 10 mV in contrast to a time constant ofabout 1 msec for wild type. Only a minor shift (about 6 mV) was observed in the voltage dependence of peak Na+ conductance (Table 1). Thus, the mutation IFMQ3 results in a specific and potent inhibition of the fast inactivation process. We examined the role of each amino acid in the hydrophobic cluster IFM (Ile-1488, Phe-1489, Met-1490) in Na+channel inactivation by replacement of each individually with glutamine (11488Q, F1489Q, and M1490Q). There was little change in the voltage dependence of the peak Na+ conductance for these mutants, but the voltage dependence of inactivation was shifted to more positive membrane potentials for 11488Q (13 mV) and M1490Q (7 mV; Fig. 2 and Table A 1475

1525

DNFNQQKKKFGGQDIFMTEEQKKYYNAMKKLGSKKPQKPIPRPANKFQGMVFD

I II

F1 483Q IFMQ3

QQQ

B Wild Type

F1

483Q

IFMQ3

rr" F 5

msec

5

msec

20 msec

FIG. 1. Effects of mutations of a cluster of hydrophobic residues within Lm/Iv on inactivation. (A) The single-letter amino acid code of the primary sequence of LIII/Iv is shown, and the positions of mutations F1483Q and IFMQ3 are indicated by vertical lines. (B) Whole-cell Na+ currents recorded in Xenopus oocytes injected with RNA encoding PI and either wild-type IIA or mutant a subunits as indicated. The currents were elicited by step depolarizations from a holdingpotential of -100 mV to test potentials of -50 mV to +10 mV in 10-mV increments.

10911

1). The macroscopic Na+ currents recorded in oocytes expressing mutant M1490Q activated and decayed rapidly (Fig. 3A). During longer test pulses, inactivation was essentially complete (Fig. 3B); on average, 3% or less of the Na+ current

remained at the end of a 50-msec pulse, which was similar to wild type. Na+ currents recorded in oocytes expressing 11488Q also had a rapid phase of inactivation that was similar to wild type (Fig. 3A). However, inactivation was clearly incomplete during longer test pulses (Fig. 3B); a mean of 11% ± 2% (n = 5) of the Na+ current remained at the end of a 50-msec test pulse. In contrast to the mild effects of mutations 11488Q and M1490Q, mutant F1489Q displayed greatly slowed, biphasic inactivation. For strong depolarizations, a small fraction of the current inactivated quickly, but most of the current failed to inactivate by the end of a 50-msec pulse to -10 mV (Fig. 3A). At -10 mV, the rapid component of the Na+ current decayed too rapidly for precise measurement of a time constant while the slow component decayed with time constants of 55 ± 30 msec and 790 ± 50 msec (n = 5). A mean of 86% ± 2% (n = 5) of the Na+ current remained at the end of a 50-msec pulse (Fig. 3B). The time course of decay of the Na+ current was almost as slow as for mutant IFMQ3 during long test pulses to -10 mV (Fig. 3C). These results identify Phe-1489 as the critical amino acid residue within the hydrophobic cluster IFM. Analysis of Na+-Channel Inactivation in Transfected Mammalian Cells. We also analyzed the effects of single amino acid mutations on inactivation by transfection and stable expression in Chinese hamster cells. Chinese hamster R-1610 cells (19) were transfected with the expression plasmid ZemRVSP6 (16) containing the wild-type or mutant a subunit cDNAs. Clonal cell lines were picked for resistance to the antibiotic G418 and screened for expression of transfected Na+ chahnel a subunit mRNA by high stringency Northern blot analysis. No endogenous Na+ channel a subunit mRNA was detected in untransfected R-1610 cells; however, each transfection yielded multiple cell lines expressing Na+channel a subunit mRNA. Consistent with Northern blot results, R-1610 cells expressed very low levels of endogenous Na+ currents when examined in the whole-cell configuration. Small Na+ currents (30 cells that were analyzed by voltage clamp while the remainder had no detectable Na+ currents. The Na+ currents recorded in transfected R-1610 cells were similar to those recorded in CHO-K1 cells expressing the same type IIA Na+ channel (16). Expression of the a subunit alone resulted in currents having the characteristic rapid Na+-current inactivation observed in brain neurons (e.g., refs. 21 and 22). Na+ currents recorded from wild-type channels activate and inactivate completely within 2 msec at voltages positive to -20 mV (Fig. 4A). The time constant for current decay was 0.26 ± 0.05 msec at +20 mV (n = 4). In contrast, whole-cell Na+ currents recorded in cells expressing the mutants activate rapidly but display altered inactivation kinetics (Fig. 4 B and C). The improved voltage clamp fidelity obtained with R-1610 cells compared to oocytes shows that the fast component of macroscopic inactivation of Na+ currents is significantly slower than wild type in all three cell lines expressing mutant channels. Mutant M1490Q inactivates more slowly (T = 0.94 ± 0.14 msec, n = 5), but inactivation is essentially complete (Fig. 4B). Mutant 11488Q also inactivates more slowly than wild type (T = 0.87 ± 0.18 msec, n = 4), but its inactivation is incomplete (Fig. 4B). As observed in Xenopus oocytes, inactivation of mutant F1489Q is dramatically slowed, and most of the Na+ current remains at the end of a 20-msec test pulse (Fig. 4C). Depolarization to -40 mV or -30 mV produces Na+ currents that do not decay appreciably during the 20-msec test pulse, whereas depolarization to more positive voltages yields Na+ currents with a small component that decays within 5 msec

Proc. Natl. Acad. Sci. USA 89 (1992)

Physiology: West et al.

10912

Table 1. Voltage-dependent gating parameters for wild-type and mutant Na+ channels Inactivation Activation n a, mV V1/2, mV n z, eo Mutant V1/2, mV 6 -44 ± 2 6.5 ± 0.6 5 5.2 ± 0.9 -19 ± 3 Wild type 7 8.4 ± 0.3 -48 ± 2 5 4.8 ± 0.7 -19 ± 2 F1483Q ND ND 6 6.4 ± 0.7 -24 ± 1 IFMQ3 5 7.0 ± 1.5 -31 ± 1 5 4.6 ± 0.6 -17 ± 3 11488Q ND ND 5 5.4 ± 0.6 -20 ± 2 F1489Q 6 5.7 ± 0.7 -38 ± 2 6 4.6 ± 0.6 -16 ± 2 M1490Q V1/2, voltage for half-maximal activation or inactivation; z, effective gating charge; a, slope factor; ND, not determined; n, number of experiments; e., elementary charge units. The values given are the means ± S.E.

and a large component that remains at the end of the test pulse. Representative traces of single Na+-channel currents recorded during pulses to -20 mV from F1489Q cells are shown in Fig. 4D. Single wild-type Na+ channels expressed in Chinese hamster cells open once or twice early in a depolarization and then inactivate and remain silent for the remainder of the pulse (22). Single F1489Q channels open early in the pulse but continue to reopen for the duration of the pulse instead of inactivating. The increased probability of reopening of single channels evidently causes the noninactivating component of Na+ current observed at the macroscopic level.

DISCUSSION

identify the cluster of hydrophobic amino acid residues at positions 1488-1490 as a critical structural component for fast inactivation. We propose that these residues are an essential part of the fast inactivation gate of the Na+ channel. Comparison to the Predictions of the Bali-and-Chain Model of Inactivation. Treatment of the cytoplasmic surface of Na+ channels with proteases removes fast inactivation (2, 24-26), and the inactivation process itself has little gating charge movement associated with it (27). Based on these two characteristics, Armstrong and Bezanilla (2, 27) proposed a "ball-and-chain" model of Na+-channel inactivation with three essential structural features: a positively charged inactivation particle (the ball), a polypeptide tether (the chain), A

11 488Q

Wild Type

An Essential Cluster of Hydrophobic Residues in the Inactivation Gate of Na+ Channels. In contrast to the nonessential role of positively charged amino acids in LIuI/Iv (5, 23), our results show that replacement of three contiguous hydrophobic amino acids in the amino-terminal region of LmI/Iv, Ile-1488, Phe-1489, and Met-1490, with glutamine (mutant IFMQ3) completely removes fast inactivation. Replacement of Phe-1489 alone with glutamine (mutant F1489Q) removes fast inactivation nearly completely, and substitution of glutamine for Ile-1488 (11488Q) and Met-1490 (M1490Q) results in slowed Na+-current inactivation with a significant sustained component in the case of 11488Q. These results

(-3 L3

-33 0

(r-

Q) -o

F

L.

0

z

0.8

F1 489Q

B

A

r

ci

20 mbrc

--* Wild Type F :-1 483Q 8 u0 1 4880 ,1: A C

0.6 V/O Fl4890 M1490Q

7

'..0 c L- 0.8

m

23

0.6

j

20 mbec

5 msec B

-

Wild Type, M1490Q

C

ci)

0.4 WO IFMQ3

._ 0.4

0.2

F

1i

0.2

488Q

0

0.0

.0 L8

.__1..

...

..

2

-80---60-40 -20 Voltage

!L

_

-

(mV)

0

F1 489Q

L

-80 -60 --40 ---2

?

Jo tcge (

IFMQ3

FIG. 2. Effects of mutation of hydrophobic amino acid residues on the voltage dependence of peak Na+ conductance and steadystate fast inactivation of Na+ currents in Xenopus oocytes. Wholecell currents were recorded in oocytes injected with RNA encoding PI and either wild-type or the indicated mutant a subunits. (A) Normalized peak conductance values from experiments like those described in Fig. 1 are plotted versus test pulse potential. (B) The voltage dependence of steady-state fast inactivation is plotted as a function of conditioning pulse potential. Oocytes were held at -100 mV and depolarized with a 50-msec conditioning pulse to the indicated voltages followed immediately by a test pulse to -5 mV. The peak currents elicited by the test pulse were normalized to the peak current obtained without a conditioning pulse. Symbols in both A and B represent means, and error bars represent standard deviation. Smooth curves represent fits to a two-state Boltzmann distribution (Table 1).

20 msec

400

mosc

FIG. 3. Effects of mutation of single hydrophobic amino acids on the time course of Na+ currents expressed in Xenopus oocytes. (A) Whole-cell Na+ currents recorded in Xenopus oocytes injected with RNA encoding ,B1 and either wild-type 11A a subunit, I1488Q, F1489Q, or M1489Q. The currents were elicited by step depolarizations from a holding potential of -100 mV to test potentials of -50 mV to +10 mV in 10-mV increments. (B) Comparison of the time course of Na+ currents of wild type, 11488Q, F1489Q, and M1490Q depolarized from a holding potential of -100 mV to a test potential of -10 mV. (C) Time course of Na+ currents of IFMQ3 and F1489Q mutants expressed in Xenopus oocytes depolarized from a holding potential of -100 mV to a test potential of -10 mV for 1 sec.

~ ~-

Physiology: West et aL

A

B WILD TYPE 1488 Q M 1490 0

2 ms

5 ms

C

It,

Proc. Nati. Acad. Sci. USA 89 (1992)

D

iomf-

-_0

==~~~~~ ~

r

| 10 O pA ~~~~~

5 ms

5 pA

10 ms

FIG. 4. Na+ currents recorded in Chinese hamster cells expressing wild-type IIA and mutant Na+ channels. (A) Whole-cell currents from an R-1610 cell expressing wild-type Na+ channels elicited by test depolarizations ranging from -60 mV to 0 mV in 10-mV steps. The holding potential was -100 mV. (B) Comparison of normalized whole-cell currents recorded during test depolarizations to +20 mV for wild type and the indicated mutants expressed in R-1610 cells. (C) Whole-cell currents recorded from mutant F1489Q during test depolarizations ranging from -60 mV to 0 mV in 10-mV steps. (D) Representative cell-attached single-channel current recordings from mutant F1489Q during pulses to -20 mV. Records were filtered at 2.5 kHz.

and a negatively charged receptor site at the intracellular mouth of the Na+ channel that develops a high affinity for the inactivation ball when the channel is activated. The process of inactivation in the ball-and-chain model then involves binding of the loosely tethered ball to its receptor site at the intracellular mouth of the pore and occlusion of the pore. Our results are consistent with some, but not all, of the features of this original formulation of the ball-and-chain model. As in the ball-and-chain model, we propose that the hydrophobic cluster IFM serves as an essential component of the inactivation particle, which occludes the intracellular mouth of the activated Na+ channel. However, in contrast to the original formulation of the ball-and-chain model, positively charged residues are not required, and hydrophobic forces rather than electrostatic interactions seem to mediate the interaction of the inactivation particle with its receptor. In addition, the primary structure of LIII/Iv, which positions the IFM hydrophobic cluster only 14-amino acid residues from transmembrane segment IIIS6 in a loop that is predicted to have an ordered, partially a-helical structure (28) and is tethered at both ends, is most consistent with an ordered conformational transition that folds LiII/Iv into the intracellular mouth of the transmembrane pore. A conformational change of this kind is consistent with previous results (6, 7) showing that antibodies against LIII/Iv that block inactivation can only do so by binding to the noninactivated state of the channel, as if their binding site is made inaccessible by the process of inactivation. Relationship to Inactivation of Shaker K+ Channels. The primary structures and functional properties of mutants that prevent fast inactivation of Shaker K+ channels conform more closely to the predictions of the ball-and-chain model. Both positively charged and hydrophobic amino acid residues are essential components of the inactivation gate (29, 30), although mutations of a single hydrophobic residue, Leu-7, have the most dramatic effects on inactivation as we have found here for mutation of Phe-1489 on Na+ channels. The inactivation particle is located at the amino terminus of the Shaker polypeptide, at the end of a chain of over 200 amino acid residues from the first transmembrane segment of the molecule. More-

10913

over, reducing the length of this chain region accelerates inactivation while increasing it slows inactivation (29), and adding the inactivation particle to noninactivating K+ channels as a soluble peptide restores inactivation (30). These results demonstrate that the inactivation gate peptide has affinity for a receptor located elsewhere in the K+-channel structure and suggest that it may reach its receptor site by restricted diffusion of a tethered-ball structure. While the inactivation gate of the Na+ channel is not at its amino terminus and there is no significant similarity in amino acid sequence, there are some analogies between the inactivation gating structures ofthe two channel types. LiI/Iv links domains III and IV of the sodium channel a subunit and therefore is the amino terminus of domain IV. Its position between domains III and IV may be similar to the position of the much larger Shaker amino terminus in a homotetrameric Shaker K+ channel. Functional analogies between the inactivation particles of the two channels are suggested by recent experiments in which the addition of a Na+-channel inactivation gate sequence to the amino terminus of a noninactivating Shaker K+ channel by construction of a cDNA encoding a channel chimera results in partial restoration of fast inactivation (31). Evidently, the mechanism of Na+-channel inactivation is similar, but not strictly analogous, to the ball-and-chain mechanism proposed for K+ channels. A Hinged-Lid Model of Na+-Channel Inactivation. The loop structure of the Na+-channel inactivation gate differs from that of the K+ channel but closely resembles the hinged-lid structures of allosteric enzymes (32). Hinged lids have been defined structurally by x-ray crystallography and molecular modeling and therefore provide a valuable model for the unknown structure of the Na+-channel inactivation gate. They consist of structured loops of 10-20 residues between two hinge points and serve as rigid lids that fold over the active sites of allosteric enzymes to control substrate access. Binding of allosteric ligands causes a conformational change of the lid to open or close the active site. By analogy, LII,/IV may function as a rigid lid to control Na+ entry to and exit from the intracellular mouth of the pore of the Na+ channel (Fig. 5). This hinged lid may be held in the closed position during inactivation by a hydrophobic latch formed by the hydrophobic cluster IFM. Glycine and proline residues on either side of the IFM domain (Fig. 1A), which are conserved in all five cloned rat Na+ channels that have been functionally expressed, may function as hinge points (33) allowing the inactivation gate region of LIII/Iv to move in and out of the channel pore. Mutation of one of these glycine residues to valine in the skeletal muscle Na+ channel causes paramyotonia congenita (34). Mechanism of the Biphasic Decay of the Na+ Current in Mutant Na+ Channels. Comparison of the Na+ currents recorded in oocytes or cultured cells expressing the mutant Na+ channels reveals interesting kinetic differences in inactivation. Na+ currents recorded from mutant IFMQ3 have no IV

IV

+H3NN *

CO2-

FIG. 5. A hinged-lid model for Na+-channel inactivation. LIjI/IV is depicted as a hinged lid that occludes the transmembrane pore of the Na+ channel during inactivation. Phe-1489 is illustrated in a pore-blocking position in the inactivated state.

10914

Proc. Nati. Acad. Sci. USA 89 (1992)

Physiology: West et al.

fast inactivation at any potential. In contrast, Na+ currents recorded from mutants 11488Q, F1489Q, and M1490Q have a component of fast inactivation with a slowed rate followed by a sustained Na+ current that ranges from