Structure and Function of a Novel Voltage-gated, Tetrodotoxin-resistant Sodium Channel Specific to Sensory Neurons*

Small neurons of the dorsal root ganglia (DRG) are known to play an important role in nociceptive mecha-nisms. These neurons express two types of sodium current, which differ in their inactivation kinetics and sensitivity to tetrodotoxin. Here, we report the cloning of the (cid:97) -subunit of a novel, voltage-gated sodium channel (PN3) from rat DRG. Functional expression in Xenopus oocytes showed that PN3 is a voltage-gated sodium channel with a depolarized activation potential, slow inactivation kinetics, and resistance to high concentrations of tetrodotoxin. In situ hybridization to rat DRG indicated that PN3 is expressed primarily in small sensory neurons of the peripheral nervous system. Voltage-gated sodium channels play a fundamental role in the regulation of neuronal excitability. In addition to differ-ences in primary structure and kinetic properties (1), these channels can be distinguished pharmacologically on the basis of their relative sensitivity to the neurotoxin, tetrodotoxin (TTX) 1 (2). Two types of sodium currents are expressed by sensory neurons within the dorsal root ganglion (DRG), a fast inactivating TTX-sensitive current and a slow inactivating, TTX-resistant current that appears to be expressed by a high proportion of the small afferent neurons (3–8). Of the large and small neurons of the DRG, the latter is of primary importance in the processing of nociceptive information within the somato-sensory system (9, 10).

Small neurons of the dorsal root ganglia (DRG) are known to play an important role in nociceptive mechanisms. These neurons express two types of sodium current, which differ in their inactivation kinetics and sensitivity to tetrodotoxin. Here, we report the cloning of the ␣-subunit of a novel, voltage-gated sodium channel (PN3) from rat DRG. Functional expression in Xenopus oocytes showed that PN3 is a voltage-gated sodium channel with a depolarized activation potential, slow inactivation kinetics, and resistance to high concentrations of tetrodotoxin. In situ hybridization to rat DRG indicated that PN3 is expressed primarily in small sensory neurons of the peripheral nervous system.
Voltage-gated sodium channels play a fundamental role in the regulation of neuronal excitability. In addition to differences in primary structure and kinetic properties (1), these channels can be distinguished pharmacologically on the basis of their relative sensitivity to the neurotoxin, tetrodotoxin (TTX) 1 (2). Two types of sodium currents are expressed by sensory neurons within the dorsal root ganglion (DRG), a fast inactivating TTX-sensitive current and a slow inactivating, TTX-resistant current that appears to be expressed by a high proportion of the small afferent neurons (3)(4)(5)(6)(7)(8). Of the large and small neurons of the DRG, the latter is of primary importance in the processing of nociceptive information within the somatosensory system (9,10). In order to define the molecular basis of sodium channel conductance in sensory neurons, we have attempted to identify and clone novel sodium channel ␣-subunits. As described here, this work has led to the isolation and functional expression of PN3, a novel voltage-gated, TTX-resistant sodium channel expressed predominantly by small sensory neurons within the peripheral nervous system.

EXPERIMENTAL PROCEDURES
cDNA Cloning-EcoRI-adapted cDNA was prepared from normal adult male Sprague-Dawley rat DRG poly(A) ϩ RNA using the Super-Script Choice System (Life Technologies, Inc.). cDNA (Ͼ4 kilobases) was selected by sucrose gradient fractionation (11) ligated into the Zap Express vector (Stratagene) and packaged with the Gigapack II XL lambda packaging extract (Stratagene). Phage (3.5 ϫ 10 5) were screened by filter hybridization with a 32 P-labeled probe (bases 4637-5868 of rBIIa) (12). Filters were hybridized in 50% formamide, 5 ϫ saline/sodium/phosphate/EDTA, 5 ϫ Denhardt's solution, 0.5% SDS, 250 g/ml salmon sperm DNA, and 50 mM sodium phosphate at 42°C and washed in 0.5 ϫ SSC, 0.1% SDS at 50°C. Positive clones were excised in vivo into pBK-CMV using the ExAssist/XLOLR system (Stratagene). Southern blots of EcoRI-digested plasmids were hybridized with a 32 P-labeled DNA probe representing a novel domain IV segment amplified from DRG RNA by PCR with degenerate oligonucleotide primers. Southern filters were hybridized in 50% formamide, 6 ϫ SSC, 5 ϫ Denhardt's solution, 0.5% SDS, and 100 g/ml salmon sperm DNA at 42°C and were washed in 0.1 ϫ SSC, 0.1% SDS at 65°C. A cDNA clone, 7.3, containing a full-length insert, was identified and sequenced on both strands. In addition, several other PN3 clones were partially sequenced. Sequence analyses were done using the Gap, Best-Fit, PileUp, and Distances programs of the Wisconsin Sequence Analysis Package (Genetics Computer Group, Inc.). For oocyte expression analysis, the PN3 cDNA was excised from the vector and, after blunting the ends, subcloned into pBSTA.
RT-PCR Analysis-Tissues were isolated from anesthetized, normal adult male Sprague-Dawley rats and were immediately frozen at Ϫ80°C. RNA was isolated from tissue samples using RNAzol (Tel-Test, Inc.). Random-primed cDNA was reverse-transcribed from 500 ng of RNA from each tissue. PCR primers targeted the 3Ј-untranslated region of PN3 and defined a 410-base pair amplicon. Thermal cycler parameters were: 30 s/94°C, 30 s/57°C, 1 min/72°C (24 cycles); 30 s/94°C, 30 s/57°C, 5 min/72°C (1 cycle). A positive control (1 ng of pBK-CMV/PN3) and a no-template control were also included. cDNA from each tissue was also PCR amplified using primers specific for glyceraldehyde-3phosphate dehydrogenase (13) to demonstrate template viability. PN3 PCR amplicons from nodose ganglia and sciatic nerve were confirmed by nucleotide sequence analysis.
In Situ Hybridization Histochemistry-Oligonucleotide probe sequences were synthesized from the unique 3Ј-untranslated region of PN3 (sense and antisense probes were complementary to each other). Normal rats were perfused with 4% paraformaldehyde; L4-L5 DRG were removed, postfixed in the same solution, and cryoprotected in 20% sucrose. Frozen sections (10 m) were cut and hybridized overnight at 39°C in a solution containing 35 S-ATP-labeled oligonucleotides (specific activity, 5 ϫ 10 7 -1 ϫ 10 8 cpm/g), 50% formamide, 4 ϫ SSC, 0.5 mg/ml salmon sperm DNA, and 1 ϫ Denhardt's solution. Sections were washed over a period of 6 h in 2 ϫ-0.1 ϫ SSC containing 0.1% ␤-mercaptoethanol, dehydrated in a series of ethanols (50 -100%) containing 0.3 M ammonium acetate, and apposed to sheet film (Amersham B max ) or dipped in liquid emulsion (Amersham LM-1) and then developed for 2 and 5 weeks, respectively. The cell area of neurons (hybridized and non-hybridized) with a distinct nucleolus was measured from sections through the lumbar ganglia with the aid of a Macintosh Quadra 840 using the public domain NIH Image Program (W. Rasband, NIH).
Oocyte Recording-Capped cRNA was prepared from the linearized plasmid using a T7 in vitro transcription kit (Ambion, mMessage mMachine) and injected into stage V and VI Xenopus oocytes (14) using a Nanojector (Drummond). After 2.5 days at 20°C, oocytes were impaled with agarose-cushion electrodes (0.3-0.8 megaohm) (15) and voltageclamped with a Geneclamp 500 amplifier (Axon Instruments) in TEV (two-electrode voltage clamp) mode. Stimulation and recording were controlled by a computer running pClamp (Axon Instruments). Oocytes were perfused with a solution containing (in mM): 81 NaCl, 2 KCl, 1 MgCl 2 , 0.3 CaCl 2 , 20 Hepes-NaOH (pH 7.5). The data in Fig. 5A were collected using the Geneclamp hardware leak subtraction, filtered at 5 kHz with a 4-pole Bessel filter, and sampled at 50 kHz. For the experiment in Fig. 5C, the oocytes were depolarized from Ϫ100 mV to ϩ20 mV for approximately 10 ms at 0.1 Hz; P/Ϫ4 leak subtraction was used * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 1 The abbreviations used are: TTX, tetrodotoxin; DRG, dorsal root ganglion; PCR, polymerase chain reaction; RT, reverse transcription; SCN␤1, sodium channel ␤1-subunit; h, human; r, rat. (16). There was a slow "rundown" of the current with time, and a correction was made for the resulting sloping baseline. Varying concentrations of TTX (Sigma) in bath solution were perfused over the oocyte, and the current amplitude was allowed to attain steady state before the effect was recorded.

RESULTS AND DISCUSSION
Sequence Analysis of PN3-To identify novel sodium channel ␣-subunits from the peripheral nervous system, we used degenerate oligonucleotide-primed RT-PCR analysis of RNA from rat DRG and homology cloning from a rat DRG cDNA library. PCR fragments from domain I and interdomain I-II were cloned and sequenced. The sequences matched those of clones 7.3 and 17.2 that were isolated from the rat DRG cDNA library by homology cloning. Clone 7.3 (PN3, peripheral nerve sodium channel 3) was sequenced entirely. Several other full-length and partial clones for PN3 were isolated.
Nucleotide sequence analysis of the PN3 cDNA identified a 5868-base open reading frame, coding for a 1956-amino acid protein. In common with other sodium channels, there is an ATG 5 base pairs upstream of the genuine ATG. The deduced amino acid sequence of PN3 (Fig. 1) exhibited the primary structural features of an ␣-subunit of a voltage-gated sodium channel. PN3 contains four homologous domains (I-IV), each consisting of six putative ␣-helical transmembrane segments (S1-S6). The positively charged residues in the voltage sensor (S4 segments) and the inactivation gate between IIIS6 and IVS1 are highly conserved in PN3; sites for cAMP-dependent phosphorylation and N-linked glycosylation shown to exist in other sodium channels (1) are also present in PN3. In PN3, however, there are two unique consensus sites for cAMP-dependent phosphorylation sites, one in domain II between S3 and S4 and another in the interdomain II-III (Fig. 1). Modulation of rBIIa channel function by cAMP-dependent protein kinase A has been demonstrated (1, 17, 18). The significance of the unique cAMP-dependent protein kinase A consensus sites in PN3 in vitro and in vivo remains to be demonstrated. In addition, there is an insertion of an additional Gln between Pro 583 and Ala 584 in several partial clones. The significance of the glutamine insertion has not been determined. Fig. 2A shows an amino acid homology comparison of PN3 with other cloned rat sodium channels. The higher sequence homology between PN3 and the TTX-insensitive cardiac channel and their slow inactivation kinetics (discussed below) suggest that they belong to a unique subfamily of sodium channels. Indeed, the Distance paradigm of the GCG program classifies PN3 and the cardiac sodium channel as a subfamily of sodium channels ( Fig. 2A).
Other Sodium Channels Isolated from the DRG cDNA Library-In addition to PN3, partial clones for the following known sodium channels were isolated: rBI, rBIII, glial channel (19), and a rat ortholog of the human and mouse atypical sodium channels (20,21). A ␤1-subunit was also cloned from the rat DRG by PCR. The rat DRG ␤1-subunit and rat brain sodium channel ␤1-subunit (rSCN␤1) have identical amino acid sequences. Human sodium channel ␤1-subunit (hSCN␤1) and rSCN␤1 (22) are 96% identical.
Expression of PN3-Northern blot analysis using the RT-PCR fragments and the 3Ј-untranslated region of clone 7.3 as probes showed that PN3 is encoded by a 7.5-kilobase transcript (data not shown). This band was absent in both brain and spinal cord. Analysis of RNA from selected rat tissues by RT-PCR (Fig. 2B) suggests that PN3 mRNA expression was limited to DRG, nodose ganglia, and to a lesser extent, sciatic nerve of the peripheral nervous system (35 cycles of PCR). No signal was present in the sciatic nerve after 25 cycles of amplification. PN3 mRNA was not detected in brain, spinal cord, superior cervical ganglia, heart, or skeletal muscle after 35 cycles of amplification. Additional RT-PCR analyses of DRG mRNA Sensory Neuron-specific TTX-resistant Sodium Channel 5954 have detected other sodium channels including rBI (23), rBIII (24), rH1 (25), peripheral nerve sodium channel type 1 (PN1) (26), SCP6 (27), and other novel sodium channel ␣-subunits (data not shown).
In situ hybridization using PN3-specific oligonucleotide probes showed that PN3 mRNA was expressed predominantly by small neurons in rat DRG (Fig. 3). Approximately 76% of the small cell population (400 -1000 m 2 ) and 33% of the large cell population (1400 -2000 m 2 ) were hybridized with probes for PN3 (Fig. 4). Recently, Akopian and Wood (28) isolated a partial cDNA clone, G7, from a rat DRG cDNA library with homology to known sodium channels and specific expression in subsets of sensory neurons. Whether PN3 and G7 are the same gene is unknown at present.
Functional Analysis of PN3-Two electrode voltage clamp recordings from Xenopus oocytes injected with PN3 cRNA indicated that expression of PN3 produced an inward current with slow inactivation kinetics (Fig. 5A). This current was voltage-dependent (Fig. 5B) and is carried by sodium ions; reduction of extracellular sodium ion concentration (by substituting N-methyl-D-glucamine) from 91 to 50 and 21 mM resulted in hyperpolarizing shifts in the reversal potential from ϩ43 mV to ϩ12 mV and Ϫ22 mV, respectively. Examination of the current-voltage relationship for PN3 (Fig. 5B) reveals a strikingly depolarized activation potential. In this expression system, PN3 exhibits little or no activation at Ϫ10 mV, whereas most cloned sodium channels begin to activate between Ϫ60 and Ϫ30 mV (29 -31).
The currents produced by injection of PN3 cRNA had slow inactivation kinetics (Fig. 5A). rBIIa, rBIII, and rSkM1 sodium channels also produce currents with slow inactivation kinetics when injected into Xenopus oocytes; coexpression of the ␤1subunit greatly accelerates the inactivation kinetics of these channels (22,(32)(33)(34)(35). However, coinjection of 1.3 ng of human sodium channel ␤1-subunit (hSCN␤1) (36) cRNA, which is homologous to the rat brain and DRG SCN␤1, with PN3 cRNA did not accelerate the inactivation kinetics (data not shown). In contrast, coexpression of this quantity of hSCN␤1 cRNA with rSkM1 cRNA was sufficient to accelerate the inactivation kinetics of rSkM1 maximally. Therefore, PN3 may possess inherently slow kinetics.
When expressed in Xenopus oocytes, the PN3 sodium current is highly resistant to TTX (IC 50 Ն 100 M) (Fig. 5C). The TTX-sensitive brain and skeletal muscle sodium channels are blocked by nanomolar TTX concentrations, whereas the TTXinsensitive cardiac sodium channels are blocked by micromolar TTX concentrations (2). In rat heart sodium channel 1 (rH1), Cys 374 is a critical determinant of TTX insensitivity (37)(38)(39); in the TTX-sensitive rBI, rBII, rBIII, and rSkM1, the corresponding residue is an aromatic amino acid, either Phe or Tyr, the aromatic ring of which facilitates the binding of TTX to the protein. In PN3, this position is occupied by a Ser residue (Ser 356 ), which may explain the unique response to TTX. Sitedirected mutagenesis of this residue to Phe/Tyr or Cys will determine whether this amino acid residue is solely responsible for TTX resistance.
TTX-resistant sodium currents have been implicated in peripheral and central neuronal sensitization mediated by the Sensory Neuron-specific TTX-resistant Sodium Channel 5955 C-fibers of small neurons following peripheral tissue damage and nerve injury. The biophysical and pharmacological properties of PN3 suggest that it contributes to the TTX-resistant sodium currents in small neurons of DRG. PN3 may, therefore, play a role in the the sensory function and dysfunction that is characteristic of pathophysiological pain processing. In addition, we suggest that PN3 may conduct TTX-resistant sodium currents in other sensory ganglia of the peripheral nervous system such as nodose ganglia (40).

FIG. 5. Expression of PN3 in Xenopus oocytes.
A, currents produced by step depolarizations of an oocyte injected with 18 ng of PN3 cRNA from a holding potential of Ϫ100 to Ϫ30 mV through ϩ50 mV in 10-mV increments. The Geneclamp hardware leak subtraction was used. No inward current was observed in oocytes injected with water. B, current-voltage relationship of the data in A. C, concentration dependence for TTX block of PN3 sodium current. Each oocyte was exposed to the full range of TTX concentrations shown, beginning with the lowest concentration and proceeding to the highest. For each concentration, the effect was allowed to attain steady state. Each point represents the mean Ϯ range; n ϭ 2.