Two isoforms of voltngcsdcpcndcnt Na chunnclr, cloned from rat skclctvI muulc, wcrc cxprcsscd in Xwopw oocyks. The currcma of AkM t and rSkM2 differ functionally in 4 propcrGcs: (i) tctrodotoxin (TTX) scnrilivity, (ii)ficonotoxin (ira) sensitivity, (iii) amplitude of sin& channel currcnls, and (iv) mtc of inactivnGon. rSkM I is scnsitivc to bolh TTX and /LTX. 1, INTRODUCTIONTwo it&arms of volcagc.dcpcndcnt Na chtrnncls arc found in skclctol mu& of rat; both have been cloned, scqucnccd, and expressed in Xcrlupus oocytcs [l-3]. One of these, rSkM 1 (also known OS ~1). is the predominant Na channel found in adult innervated muscle, and is sensitive to block by tctrodotoxin (TTX) at nanomolar concentrations. The TIYX-resistant isoform (rSkM2 or RHl; IC%>l yM) is prcscnt in both developing and dcncrvatcd skeletal muscle as well PS in rat hsart [l&6]. Adult Na channels from skclctal muscle arc also uniquely sensitive to y-sonotoxin @-CTX), a property nor shared either with the TTX-sensitive Nn channels of brain or with TTX-resistant Na channels of muscle [7,8]. Accordingly the currents of rSkMI, but not of rSkM2, arc blocked by nanomolar concentrations of ,u-CTX when the cRNA is expressed in Xcwpus oocytcs [2,3]. At least two ocher functional diffcrcnscs are observed between the Na currents of the wild-type channels txprcsscd ir: ,Ywroptrs oocytcs, namely the singlcchannel conductencc is larger in rSkM1 than in rSkM2 191, and the rate of inactivation is grcutcr in rSkM2 [2,3]. WC have constructed chimeras and mutations that exchange portions of AkMl and rSkM2 to try to idcn= tify the regions of the protein bat are rcsponsiblc for their sirnctionul differences. Most of the mutations were construct4 in the loop between the putative transmcmbranc segments S5 and S6, either in the first (Dl) or the fourth (D4) of the four homologous domains. Portions of the SS-S6 segment from 011 four domains have been shown to contribute to TI'X sensitivity irnd ainglc than= ncl conductance of rat brain Na channels 1101. A resent report dcmonstratcd that full m sensitivity can bc restored to rSkM2 by cxchangc of a single amino acid bctwccn the two isoforms [ 111. This amino acid, tyrosinc in rSkMl and cystcinc in rSkM2, is located in the soculled SS2 region of the SS-S6 loop of Dl. An abstract of our data hus appeared [12]. MATERIALS AND METHODS Chemicals 253

The effects of Brazilian scorpion Tityus serrulatus toxin gamma (TiTx gamma) were studied on voltage-gated Na+ channels from human heart (hHl) and rat skeletal muscle (rSkM1). The Na+ channels were expressed in Xenopus laevis oocytes, and Na+ currents were recorded using two-microelectrode voltage-clamp techniques. In control experiments, the threshold of activation of hH1 is more negative than that of rSkM1 by approximately 20 mV. The toxin induces a shift of the voltage dependence of activation toward more negative potential values and reduces the amplitude of the current when administered to rSkM1. In contrast, TiTx gamma has little discernible effect on the current-voltage curve for hH1 at 100 nmol/L. Chimeric channels formed from these two isoforms were constructed to localize the binding site of TiTx gamma on rSkM1. TiTx gamma shifts the activation of a chimera (SSHH) in which domains 1 (D1) and 2 (D2) derive from rSkM1 and domain 3(D3) and 4 (D4) derive from hH1. This finding suggests that the toxin acts on the activation of rSkM1 by binding either to D1 and/or D2. TiTx gamma shifted the activation of another chimera with D2-D3-D4 from rSkM1 (HSSS) toward more hyperpolarizing potentials and had no effect on the activation of other chimeras with only D1-D3-D4 from rSkM1 (SHSS) or only D3 from rSkM1 (HHSH). Finally, a chimera in which D2 is from rSkM1 and all others domains are from hH1 (HSHH) provides further compelling support for our hypothesis. TiTx gamma shifts the activation of this chimera toward more negative potential values. Thus, TiTx gamma action on chimeras segregates with the source of D2: when D2 is from rSkM1, the toxin affects activation. We infer that D2 plays an important role in the activation process of voltage-gated Na+ channels.

The tetradomain voltage-gated sodium channels from rat skeletal muscle (rSkM1) and from human heart (hH1) possess different sensitivities to the 22-amino-acid peptide toxin, mu-conotoxin GIIIA (mu-CTX). rSkM1 is sensitive (IC50 = 51.4 nM) whereas hH1 is relatively resistant (IC50 = 5700 nM) to the action of the toxin, a difference in sensitivity of >100-fold. The affinity of the mu-CTX for a chimera formed from domain 1 (D1), D2, and D3 from rSkM1and D4 from hH1 (SSSH; S indicates origin of domain is skeletal muscle and H indicates origin of domain is heart) was paradoxically increased approximately fourfold relative to that of rSkM1. The source of D3 is unimportant regarding the difference in the relative affinity of rSkM1 and hH1 for mu-CTX. Binding of mu-CTX to HSSS was substantially decreased (IC50 = 1145 nM). Another chimera with a major portion of D2 deriving form hH1 showed no detectable binding of mu-CTX (IC50 > 10 microM). These data indicate that D1 and, especially, D2 play crucial roles in forming the mu-CTX receptor. Charge-neutralizing mutations in D1 and D2 (Asp384, Asp762, and Glu765) had no effect on toxin binding. However, mutations at a neutral and an anionic site (residues 728 and 730) in S5-S6/D2 of rSkM1, which are not in the putative pore region, were found to decrease significantly the mu-CTX affinity with little effect on tetrodotoxin binding (

The alpha-subunit cDNAs encoding voltage-sensitive sodium channels of human heart (hH1) and rat skeletal muscle (rSkM1) have been expressed in the tsA201 mammalian cell line, in which inactivation properties appear to be normal in contrast to Xenopus oocytes. A series of rSkM1/hH1 chimeric sodium channels has been evaluated to identify the domains of the alpha-subunits that are responsible for a set of electrophysiological differences between hH1 and rSkM1, namely, midpoints and slope factors of steady-state activation and inactivation, inactivation kinetics and recovery from inactivation kinetics and their voltage-dependence. The phenotype of chimeric channels in which each hH1 domain was successively introduced into a rSkM1 alpha-subunit framework confirmed the following conclusions. (i) The D4 and or/C-ter. are responsible for the slow inactivation of hH1 sodium channels. (ii) Concerning the other differences between rSkM1 and hH1: steady-state activation and inactivation, kinetics of recovery from inactivation, the phenotypes are determined probably by more than one domain of the alpha-subunit.