Sigma-1 Receptor Agonists Directly Inhibit NaV1.2/1.4 Channels

(+)-SKF 10047 (N-allyl-normetazocine) is a prototypic and specific sigma-1 receptor agonist that has been used extensively to study the function of sigma-1 receptors. (+)-SKF 10047 inhibits K+, Na+ and Ca2+ channels via sigma-1 receptor activation. We found that (+)-SKF 10047 inhibited NaV1.2 and NaV1.4 channels independently of sigma-1 receptor activation. (+)-SKF 10047 equally inhibited NaV1.2/1.4 channel currents in HEK293T cells with abundant sigma-1 receptor expression and in COS-7 cells, which barely express sigma-1 receptors. The sigma-1 receptor antagonists BD 1063,BD 1047 and NE-100 did not block the inhibitory effects of (+)-SKF-10047. Blocking of the PKA, PKC and G-protein pathways did not affect (+)-SKF 10047 inhibition of NaV1.2 channel currents. The sigma-1 receptor agonists Dextromethorphan (DM) and1,3-di-o-tolyl-guanidine (DTG) also inhibited NaV1.2 currents through a sigma-1 receptor-independent pathway. The (+)-SKF 10047 inhibition of NaV1.2 currents was use- and frequency-dependent. Point mutations demonstrated the importance of Phe1764 and Tyr1771 in the IV-segment 6 domain of the NaV1.2 channel and Phe1579 in the NaV1.4 channel for (+)-SKF 10047 inhibition. In conclusion, our results suggest that sigma-1 receptor agonists directly inhibit NaV1.2/1.4 channels and that these interactions should be given special attention for future sigma-1 receptor function studies.


Introduction
The sigma receptor was originally described as a novel opioid receptor subtype, but it is now considered to be a unique receptor [1,2]. Sigma receptors consist of two subtypes: sigma-1 and sigma-2. The sigma-1 receptor was first cloned from guinea pigs in 1996 [3,4,5], but the sigma-2 receptor has not been cloned. The sigma-1 receptor is widely expressed in the brain and peripheral organs, and it may be involved in numerous processes, such as Alzheimer's disease, schizophrenia, pain, drug addiction, stroke, cancer, depression and anxiety [2,6]. The molecular mechanisms of sigma-1 receptor effects in these diseases are not understood. One of the most important molecular actions of sigma-1 receptors is the modulation of various voltage-and ligand-gated ion channels [2,7,8].
Voltage-gated sodium channels initiate and propagate action potentials in excitable cells. Nine voltage-gated sodium channel isoforms have been identified in mammals [9,10]. Na V 1.2 is the most abundant sodium channel isoform in the central nervous system comprising approximately 80% of the total rat brain voltage-gated sodium channels [11][12], and it controls axonal action potential conduction and neurotransmitter release in presynaptic terminals [13]. Na V 1.2 mutations cause inherited febrile seizures and epilepsy [9]. The Na V 1.4 channel is the predominant voltage-gated Na + channel isoform in skeletal muscle [14], and various channel mutations are associated with muscular diseases, including potassium-aggravated myotonia, paramyotonia congenita, hyperkalemic periodic paralysis, hypokalemic periodic paralysis and normokalemic periodic paralysis [15]. The major cardiac voltage-gated Na + channel is Na V 1.5 [16,17], which is involved in many arrhythmic disorders, such as long-QT syndrome type 3, Brugada syndrome, conduction disease, sinus node dysfunction and atrial standstill [18,19].
We found that (+)-SKF 10047 inhibited Na V 1.2 and Na V 1.4 channel currents, but these inhibitory effects were independent of sigma-1 receptor activation. (+)-SKF 10047 inhibited Na V 1.2/1.4 channel currents equally in HEK293T cells (which have abundant sigma-1 receptor expression) and COS-7 cells (which barely express sigma-1 receptors). The present study is the first report of the direct Na V 1.2/1.4 channel current inhibition by sigma-1 receptor agonists, which should be given special attention for investigation of sigma-1 receptor function.

Ethics statement
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of the Fudan University (Permit Number: 2007-0002

Molecular biology
Site-directed F1764A mutagenesis was achieved in the Na V 1.2 Na + channel clone (the rat brain type IIA Na + channel clone kindly provided by Alan L. Goldin [22]) using the KOD-Plus mutagenesis system (TOYOBO, Japan). The Na V 1.2 mutant Y1771A was kindly provided by Professor William A. Catterall [23]. The Na V 1.4 Na + channel clone was incorporated into the pEGFP-N1 in our lab using the rat Na V 1.4 gene mRNA sequence (NM013178) on the NCBI website. Site-directed F1579A and Y1586A mutageneses were achieved in the Na V 1.4 Na + channel using the KOD-Plus mutagenesis system (TOYOBO, Japan). The homo sigma-1R gene (NM_005866) with a flag tag was incorporated into a pCDNA3 vector. The siRNA sequence that corresponded to nucleotides 500-519 of the human sigma-1 receptor open-reading frame (NM005866) was inserted into a pGPU6/GFP/Neo plasmid (GenePharm, Shanghai) to generate vector-based siRNA. All of these constructs were confirmed by sequencing.

Electrophysiology
Whole-cell currents in the HEK293T and COS-7 cells were recorded using an Axopatch 200B amplifier (Axon Instruments, Sunnyvale, CA). The bath solution contained (in mM) 145 NaCl, 2.5 KCl, 10 HEPES and 1 MgCl 2 (pH adjusted to 7.4 using NaOH). The internal solution contained (in mM) 140 CsCl, 4 KCl, 10 HEPES, and 5 EGTA (pH adjusted to 7.4 using CsOH). The pipettes were created from capillary tubing(BRAND, Wertheim, Germany) and had resistances of 5 to 7 MV under these solution conditions. All of the recordings were performed at room temperature. A superimposed Na + current was evoked by a 30-ms depolarizing pulse from a holding potential of 2100 to 220 mV. Steady-state Na + channel activation was obtained using the following protocol. The cells were held at 2100 mV and depolarized in 10-mV steps from 270 to +30 mV at 10-s intervals. The normalized conductance was plotted as a function of the command potential. The conductance was calculated as G Na = I Na /(Vm-Vrev). The data points were fitted using the Boltzmann function: Steady-state inactivation of Na + channel was achieved using the following protocol. Five hundred millisecond conditioning prepulses ranging from 2130 to 220 mV in 10-mV increments were applied prior to the 220 mV test pulse. The peak current amplitudes were normalized to the maximum current and plotted against the pre-pulse potential. The normalized current points were fitted using the Boltzmann function: I Na /I Na-max = 1/ {1+exp[(V m -V m1/2 )/k]}+A. The currents were sampled at 10 kHz and filtered at 3 kHz. The currents were corrected online for leak and residual capacitance transients using a P/4 protocol.

Data acquisition and analysis
The data acquisition and analysis were performed with pClamp 8.01 (Axon Instruments) and/or Origin 7.5 software (MicroCal, Northampton, MA). The statistical analysis consisted of unpaired or paired (depending on the circumstances) Student's T tests. Values are given as the means6SEM, and n indicates the number of tested cells. P,0.05 was defined to be a statistically significant difference between groups. Multiple comparisons were analyzed using a one-way analysis of variance (ANOVA) followed by the post-hoc Tukey test. The dose-response curve was fitted by a sigmoidal dose-response equation: I/I max = 1/(1+10ˆ(log IC50 -[SKF]) *p), where IC 50 is the concentration producing half maximal block, [SKF] is the SKF 10047 concentration, p is the Hill coefficient.

Results
(+)-SKF 10047 inhibits Na V 1.2 channel currents in a dosedependent manner I Na currents were elicited by a 30-ms depolarizing pulse to 220 mV from a holding potential of 2100 mV at 10 s intervals. The currents were recorded for 1 min to obtain a stable baseline during blank solution perfusion, and the drug solution was perfused until a stable inhibition level was achieved. (+)-SKF 10047 inhibited the Na V 1.2 channel currents in a dose-dependent manner (0.01 mM: 2.862.7%, n = 5; 0.1 mM: 5.762.1%, n = 5; 1 mM: 12.761.5%, n = 6; 50 mM: 3161.5%, n = 6; 100 mM: 39.561.3%, n = 6; 300 mM: 7161.7%, n = 6; 1000 mM : 82.463.1%, n = 6; P,0.05, Fig 1A). The dose-response curve was fitted using equation given in the method, and the IC 50 value and Hill coefficient are 140 mM and 21.0 respectively (Fig1 A). The inhibitory effect of (+)-SKF 10047 on I Na began quickly and reached a maximum effect within 90 s. The inhibition was reversible within 1-2 min (Fig 1B).
The following experiments were performed to investigate the role of G proteins in (+)-SKF 10047 inhibition of Na V 1.2 channel currents. CTX (a G s activator) and PTX (a G i/o inhibitor) were added to the culture medium at different concentrations and for various lengths of time. The cells were incubated with 1 mg/ml CTX for 2 and 22 h and with 200 ng/ml or 500 ng/ml PTX for 4 or 12 h, respectively. The inhibitory effects of 100 mM (+)-SKF 10047 on Na V 1.2 channel currents were not altered in the presence of these toxins. The average I Na inhibitions were 35.262.8% (1 mg/ml CTX for 2 h, n = 5), 36.262.3% (1 mg/ml CTX for 22 h, n = 5), 39.264.8% (200 ng/ml PTX for 4 h, n = 5) and 34.863.4% (500 ng/ml PTX for 12 h, n = 5), which were not significantly different with effect of 100 mM (+)-SKF 10047 alone (P.0.05) (Fig 3B). The inhibitory effect of (+)-SKF 10047 was also not significantly altered by the presence of GTPcS (a G-protein activator), NF 023 (a G i/o antagonist) or NF 449 (a G s antagonist) in the internal solution (Fig 3A).
These results suggested that (+)-SKF 10047 directly inhibited the Na V 1.2 channel currents.
Therefore, we investigated the frequency-and use-dependency of (+)-SKF 10047 Na V 1.2 channel current inhibition.
(+)-SKF 10047 inhibits Na V 1.2 channel currents in a frequency-and use-dependent manner The effect of frequency on (+)-SKF 10047 inhibition of Na V 1.2 channel currents was investigated using different frequencies of 30ms pulses that depolarized the holding potential from 2100 to 220 mV. The cells were stimulated for 2 min in the continued presence of 100 mM (+)-SKF-10047. The average Na V 1.2 channel current inhibitions for stimulations at 0.1, 0.5, 1 and 2 HZ were 39.561.3% (n = 6), 54.664.0% (n = 7), 67.663.2% (n = 9), and 64.163.2% (n = 7),respectively (P,0.05 compared to 0.1 Hz) (Figs 4A and B). These results suggested that (+)-SKF 10047 exerted frequency-dependent effects on the Na V 1.2 channel. The use-dependency of the inhibition was evaluated using 5-Hz trains of five pulses that depolarizing the holding potentials from 2100 to 210 mV (Fig 4C). The trains were repeated every 20 s. (+)-SKF 10047 (100 mM) was applied after 6 control trains and washed out, and the 6 trains were repeated (Fig 4D). The use-dependence (UD) was determined by the peak amplitudes of the first and last evoked currents of the last train in the presence and absence of (+)-SKF 10047 (UD = (A3/A4)/(A1/A2), Fig 4E). The UD of 7 samples ranged from 1.10 to 1.19 (Fig 4E). These results suggested that the inhibitory effect of (+)-SKF 10047 on the Na V 1.2 channel was usedependent and that the drug may preferentially bind to the depolarized (i.e., open or inactivated) channel.
The effect of (+)-SKF 10047 on the Na V 1.2 channel is altered by the site-directed mutation of F1764A and Y1771A Local anesthetic (LA) drugs, such as lidocaine (LD), may directly inhibit rat Na V 1.2 channel currents, and the amino acid residues Phe-1764 and Tyr-1771 in the S6 transmembrane segment of domain IV are critical for this inhibition [23,27]. LD induced less inhibition in the F1764A mutant Na V 1.2 channel than in the wildtype channel (Figs 5A and B).
F1764A and Y1771A single-point mutant Na V 1.2 channels were investigated in HEK293T cells to assess similarities in the effects of (+)-SKF 10047 and LD on the Na V 1.2 channel. The (+)-SKF 10047 inhibition of the F1764A mutant channel currents was significantly less than that of the wild-type channel currents (WT: 39.561.3%; F1764A mutant: 25.961.6%, n = 9; p,0.001, Fig 5B). Interestingly, the (+)-SKF 10047 inhibitory effect on the Y1771A mutant channel currents was significantly higher than that of the wild-type channel currents (Y1771A mutant: 45.261.3%, n = 6, P,0.05 compared to WT). These data suggested the importance of the F1764 and Y1771 residues for (+)-SKF 10047 inhibition of Na V 1.2 channel currents.
The sigma-1 receptor-selective agonists DM and DTG also reduce rat Na V 1.

currents
The sigma-1 receptor-selective agonists DM, DTG and PRE-084 were used to assess the inhibitory effects of other sigma-1 receptor ligands on rat Na V 1.2 channels. PRE-084 exerted little effect on the Na V 1.2 channel currents (data not shown), but DM and DTG reversibly inhibited the Na V 1.2 channel currents (Fig 6A, D). The average inhibition from 100 mM DM in the Na V 1.2-transfected HEK293T cells was 65.862.0% (n = 6). The sigma-1 receptor antagonists BD 1047 and NE-100 were used to investigate the role of sigma-1 receptors in this inhibitory effect. The average DM inhibitions in the presence of 2 mM BD 1047 and 2.5 mM NE-100 were 59.762.9% (n = 5) and 61.261.6% (n = 5), respectively. These results were not significantly different from those observed with DM alone (Fig 6C). DM inhibition of Na V 1.2 channel currents was also investigated in the COS-7 cells. The inhibitory effect of DM on Na V 1.2 channel currents in the COS-7 cells was similar to the effect in the HEK293T cells (COS-7: 68.763.9%, n = 5; HEK293T: 65.862.0%, n = 6; P.0.05; Figs 6B and C). Previous study has been shown that DTG effect (via activation of the sigma-1 receptor) could be blocked by BD 1063 [28]. Our results showed that The DTG inhibition of Na V 1.2 channel currents was not altered by sigma-1 receptor antagonist BD 1063 (100 mM DTG: 69.062.5%, n = 6; 100 mM DTG+2 mM BD 1063: 71.863.6%, n = 4; P.0.05).These results suggested that DM and DTG inhibited Na V 1.2 currents via a sigma-1 receptor-independent pathway.

Discussion
(+)-SKF 10047 is a prototypic and specific sigma-1 receptor agonist that has been extensively used to study sigma-1 receptor function. By activation of sigma-1 receptors, (+)-SKF 10047 inhibits N-methyl-D-aspartate (NMDA) receptors in rat retinal ganglion cells [30], voltage-gated cardiac Na V 1.5 channels [20,21,25], L-type voltage-gated calcium channels [31], various types of voltage-gated K + channels [32,33,34]. (+)-SKF 10047 also inhibits glutamate release in rat cerebral cortex neurons [35]. We found that (+)-SKF 10047 inhibited Na V 1.2 and Na V 1.4 voltagegated sodium channels, and this inhibitory effect was not eliminated by the sigma-1 selective antagonists BD 1047 and NE-100. These results suggest that (+)-SKF 10047 inhibition of Na V 1.2/1.4 channel currents is independent of sigma-1 receptors. Sigma-1 receptor ligands likely exert effects that are unrelated to sigma-1 receptors. NDHEA sulfate, a sigma-1 receptor agonist, inhibits persistent sodium currents in the rat medial prefrontal cortex via the G i protein and the PKC signaling pathways [36]. The sigma-1 receptor ligand PRE-084 amplifies dopamine D1 receptor signaling in the prelimbic cortex [37]. G-protein blockade or activation using G i/o and G s inhibitors did not alter the (+)-SKF 10047 -induced inhibition of Na V 1.2 channel currents. The inhibitory effect of (+)-SKF 10047 on Na V 1.2 channel currents persisted in the presence of PKA and PKC inhibitors (Fig3). These results further suggest that (+)-SKF 10047 inhibition of Na V 1.2 channel currents is independent of sigma-1 receptors, G-proteincoupled receptor (GPCRs) activity or phosphokinase pathways.
Most local anesthetics are state-dependent blockers of Na + channels. The mechanism of this blockade may be the high affinity of these drugs for a site on the opening, inactivated and resting channel [38,39]. (+)-SKF 10047 inhibited Na V 1.2 channels in a state-dependent manner, which is consistent with Na + channel blockade. (+)-SKF 10047 preferentially interacted with inactivated Na V 1.2 channels to produce a significant hyperpolarizing shift in the voltage-dependent inactivation, which reduced channel availability and slowed recovery. The marked effects of (+)-SKF 10047 on channel inactivation are consistent with the high affinity of local anesthetics and anticonvulsants, such as LD, for inactivated channels [23,27]. (+)-SKF 10047 also produced a use-dependent blockade of the Na V 1.2 channel following highfrequency stimulation, which suggested an (+)-SKF 10047 affinity for open or inactivated Na + channels. This result suggests that (+)-SKF 10047 may directly inhibit Na V 1.2 channels.
The direct interactions between local anesthetics and Na V 1.2 channels have been thoroughly studied, and Phe-1764 and Tyr-1771 mutations in domain IV of the S6 channel transmembrane segment dramatically reduce Na V 1.2 channel current inhibition by local anesthetics, such as lidocaine [23,27]. We found that an F1764A mutation dramatically slowed the fast inactivation of the Na V 1.2 current, which is consistent with a previous report [27]. This mutation also reduced the (+)-SKF 10047 inhibition of Na V 1.2 channel currents by 35%. However, the Y1771A mutation did not reduce (+)-SKF 10047 inhibition of Na V 1.2 channels (Fig 5). These results suggest that the sites or mechanisms of (+)-SKF 10047 binding to Na V 1.2 channels are not identical to those of lidocaine. Previous studies suggested that F1764 and Y1771 are direct LA (local anesthetic) drugs binding sites of sodium channels [27,40]. Both F1764A and Y1771A mutations reduced the LA affinity for open and fast inactivated channels, but the effect of Y1771A was much less than F1764A [27]. Our data showed that Y1771A increased the inhibition of SKF10047, which may similar to the mutations I1761A, V1776A and N1769A increasing the LA drugs inhibition of Na V 1.2 channels [27]. Y1771A may increase closed-state Na V 1.2 channel sensitivity to SKF 10047 in an indirect way. Since SKF 10047 binds preferentially to inactivatedstate channels, Y1771A may partially change the SKF10047 binding site of close-state Na V 1.2 channels towards the inactivated conformation. The specific structural and mechanistic differences that determine the effects of (+)-SKF 10047 on Na V 1.2 may involve multiple binding sites. However, this hypothesis requires detailed structure-function investigations. The F1579A mutation (corresponding to F1764 in Na V 1.2 channel) significantly reduced Figure 5. The effects of (+)-SKF 10047 or lidocaine on wild-type and mutant Na V 1.2/Na V 1.4 in HEK293T cells. A, Sample current traces for wild-type and mutant Na V 1.2 channels (F1764A or Y1771A) were obtained in the presence and absence of 100 mM (+)-SKF 10047 or 100 mM lidocaine. B, The statistical analyses of I Na inhibition for wildtype and mutant Na V 1.2 channels in the presence of 100 mM (+)-SKF 10047 or 100 mM lidocaine. The (+)-SKF 10047 inhibition of the F1764A and Y1771A mutant channel currents was significantly different from the inhibition of the wild-type Na V 1.2 current (*, P,0.05; ***, P,0.001). The lidocaine-induced I Na inhibition also differed significantly between the F1764A and wild-type Na V 1.2 channels (***, P,0.001, Student's t test). C, Sample current traces for wild-type, F1579A and Y1586A mutant Na V 1.4 channels were obtained in the presence and absence of 100 mM (+)-SKF 10047. The mutation of F1579A significantly reduced the (+)-SKF 10047 inhibition of Na V 1.4 channel currents (***, P,0.001). doi:10.1371/journal.pone.0049384.g005 the (+)-SKF 10047 inhibition of Na V 1.4 channel currents, but the Y1586A mutation showed no effect. This is consistent with the previous report that the affinities of local anesthetic drug binding to Na V 1.4 channel are determined primarily by interaction with F1759 [40]. Since it has more restricted conformation than lidocaine and Y1586 is on the bottom of channel pore comparing to F1579, (+)-SKF 10047 may not direct interact with Y1586 site.
Interestingly, (+)-SKF 10047 directly inhibited both Na V 1.2 and Na V 1.4 channels. These inhibitory effects occurred and were washed out within 1-2 minutes. The IC 50 value of Na V 1.2 for (+)-SKF 10047 was 140 mM. These results are different from previous reports on (+)-SKF 10047 inhibition of cardiac Na V 1.5 channels via sigma-1 receptor activation, which required more than 10 min to occur and wash out and had a much lower IC 50 Value (70 mM) [21]. The inhibitory effects of (+)-SKF 10047 on Na V 1.5 channels also differed between HEK293 cells with abundant sigma-1 receptor expression and COS-7 cells with little sigma-1 receptor expression [21,25]. This study compared SKF-1047 inhibition of Na V 1.2 and Na V 1.4 channels in HEK293T cells and COS-7 cells, and no differences between these two cell types were observed (Fig 2). SKF-10047 directly inhibited the COS-7 cells, which express Na V 1.2 and Na V 1.4 channels but few sigma-1 receptors. The Na V 1.2, Na V 1.4, and Na V 1.5 a-subunit isoforms have greater than 60% amino acid sequence identity. However, these channels exhibit gating, permeability, and conductance functional differences, which result in tissue-specific physiological functions and subtle difference in their pharmacological properties [10]. The markedly different (+)-SKF 10047 inhibition of Na V 1.2/Na V 1.4 and Na V 1.5 currents may be explained by chemical structure differences because the Na V 1.5 protein harbors multiple evolutionary conserved amino acid motifs for N-glycosylation in its extracellular domain. The N-glycosylation affects Na V 1.5 channel gating [41] , and maybe also affect SKF 10047 binding to the channel.
(+)-SKF 10047 may inhibit various ion channels via sigma-1 receptor activation, but a direct interaction has seldom been noted. Lamy et al. have recently reported that the sigma-1 agonist DTG directly inhibits small-conductance Ca 2+ -activated K channels in dopaminergic neurons and HEK-293 cells [42]. In addition, DM directly inhibits brain Na + channels in Xenopus oocytes [43]. The sigma-1 agonists DTG, (+)-SKF 10047 and DM directly inhibited the Na V 1.2 and/or Na V 1.4 channel currents in the present study. DM and (+)-SKF 10047 belong to the homologous family of benzomorphan compounds. Therefore, further investigation of the effects of other benzomorphan compounds is required.
The (+)-SKF 10047 inhibitory effect was also tested in primary cultured rat cerebellar granule neurons. The sigma-1 receptor antagonist BD 1063 and NE-100 failed to block the (+)-SKF 10047 inhibition of sodium currents in the granule neurons, which suggested that the inhibition was independent of the sigma-1 receptor activation. However, rat cerebellar granule neurons express both Na V 1.2 and Na V 1.6 channels [29], the (+)-SKF 10047 effect on Na V 1.6 channel currents need further investigation.
In conclusion, our study found that the sigma-1 receptor agonists DTG, (+)-SKF 10047 and DM directly inhibited Na V 1.2 or Na V 1.4 channels in transfected HEK293T and COS-7 cells. The sigma-1 receptor is involved in many diseases, and the final action of sigma-1 receptor activation is most likely to modulate various ion channels. Therefore, the direct effects of sigma-1 receptor ligands on ion channels should receive special attention.