Small interfering RNA-mediated selective knockdown of NaV1.8 tetrodotoxin-resistant sodium channel reverses mechanical allodynia in neuropathic rats

Neuroscience 146 (2007) 812– 821

SMALL INTERFERING RNA-MEDIATED SELECTIVE KNOCKDOWN OF NaV1.8 TETRODOTOXIN-RESISTANT SODIUM CHANNEL REVERSES MECHANICAL ALLODYNIA IN NEUROPATHIC RATS X.-W. DONG,1* S. GOREGOAKER,1 H. ENGLER, X. ZHOU, L. MARK,2 J. CRONA, R. TERRY, J. HUNTER AND T. PRIESTLEY

Voltage-gated sodium channels have long been recognized as being critically important for the initiation and propagation of action potentials in neurons. Nine sodium channel genes have been identified with sufficient sequence homology to justify their consideration as a single family cluster (NaV1.1–1.9; see Catterall et al., 2003). Six of the nine channel variants are highly sensitive to the prototypic sodium channel toxin, tetrodotoxin (TTX), the remaining three family members, NaV1.5, NaV1.8 and NaV1.9, are unusual in their variable but relative insensitivity to TTX. Sodium channels, in general, represent interesting therapeutic targets and, indeed, a number of successful drugs have been developed that owe their therapeutic efficacy to potent sodium channel block (e.g. local anesthetics, anticonvulsants, antiarrhythmics). The utility of several of these drugs in the clinical management of chronic pain has been repeatedly demonstrated (for review see Priestley and Hunter, 2006). However, none of the currently used therapeutics is capable of distinguishing between sodium channel isoforms and several adverse events, attributed to the resulting broad-spectrum sodium channel block, have severely undermined patient compliance when these medications are used in a pain setting. The challenge facing novel sodium channel drug discovery in the pain arena is to identify the primary transcripts responsible for nociceptive signaling and to selectively block such abnormal pathological activity in sensory afferent nerves. The NaV1.8 transcript is one of several channel isoforms that is particularly interesting with respect to sensory nerve pathophysiology because it is only expressed in a subset of primary afferent nerves (Akopian et al., 1996; Sangameswaran et al., 1996) and it has been linked to various pain states (see reviews by Lai et al., 2004; Priestley, 2004). However, restricted expression alone does not obviate the need for selective pharmacology and this is graphically illustrated in the case of the sensory nervous system where individual afferents express several sodium channel transcripts in addition to NaV1.8. To address the issue of specificity, various groups have adopted reverse-genetics strategies, typically using either antisense oligodeoxynucleosides (ODNs) to disrupt post-transcriptional efficiency or, alternatively, gene-knockout approaches. For example, NaV1.8-specific ODNs have been shown to reduce the hyperalgesia provoked by intraplantar injections of either prostaglandin E2 (PGE2; Khasar et al., 1998; Villarreal et al., 2005) or complete Freund’s adjuvant (CFA; Porreca et al., 1999). The same approach has also been applied to neuropathic

Department of Neurobiology, Schering-Plough Research Institute, K-15-2-2600, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA

Abstract—The biophysical properties of a tetrodotoxin resistant (TTXr) sodium channel, NaV1.8, and its restricted expression to the peripheral sensory neurons suggest that blocking this channel might have therapeutic potential in various pain states and may offer improved tolerability compared with existing sodium channel blockers. However, the role of NaV1.8 in nociception cannot be tested using a traditional pharmacological approach with small molecules because currently available sodium channel blockers do not distinguish between sodium channel subtypes. We sought to determine whether small interfering RNAs (siRNAs) might be capable of achieving the desired selectivity. Using Northern blot analysis and membrane potential measurement, several siRNAs were identified that were capable of a highly-selective attenuation of NaV1.8 message as well as functional expression in clonal ND7/23 cells which were stably transfected with the rat NaV1.8 gene. Functional knockdown of the channel was confirmed using whole-cell voltage-clamp electrophysiology. One of the siRNA probes showing a robust knockdown of NaV1.8 current was evaluated for in vivo efficacy in reversing an established tactile allodynia in the rat chronic constriction nerve-injury (CCI) model. The siRNA, which was delivered to lumbar dorsal root ganglia (DRG) via an indwelling epidural cannula, caused a significant reduction of NaV1.8 mRNA expression in lumbar 4 and 5 (L4 –L5) DRG neurons and consequently reversed mechanical allodynia in CCI rats. We conclude that silencing of NaV1.8 channel using a siRNA approach is capable of producing pain relief in the CCI model and further support a role for NaV1.8 in pathological sensory dysfunction. © 2007 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: dorsal root ganglia, CCI, pain, gene expression, sodium current, ND7/23 cell. 1

These two authors contributed equally to this work. Present address: Wyeth-Ayerst Research, 865 Ridge Road, Monmouth Junction, NJ 08852, USA. *Corresponding author. Tel: ⫹1-908-740-4520; fax: ⫹1-908-7403294. E-mail address: [email protected] (X.-W. Dong). Abbreviations: CCI, chronic constriction nerve injury; CFA, complete Freund’s adjuvant; DRG, dorsal root ganglia; I/V, current–voltage; L6 (4, 5), 6th (4th, 5th) lumbar; mm-siRNA, mismatch small interfering RNA; ns-siRNA, non-silencing small interfering RNA; ODN, oligodeoxynucleoside; PGE2, prostaglandin E2; PWT, paw withdrawal threshold; siRNA, small interfering RNA; TTX, tetrodotoxin; TTXr, tetrodotoxin-resistant; TTXs, tetrodotoxin-sensitive; Vh, membrane holding potential; wt, wild type. 2

0306-4522/07$30.00⫹0.00 © 2007 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2007.01.054

812

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821

pain models and convincing ODN-mediated efficacy has been demonstrated in both the spinal nerve ligation (Porreca et al., 1999) and chronic constriction injury (CCI) models (Liu et al., 2005). In each case, efficacy correlated with a corresponding reduction in NaV1.8 mRNA levels. Gene deletion studies have produced a less convincing picture as regards the role of NaV1.8 in chronic pain. An initial phenotypic characterization of a NaV1.8 knockout mouse revealed a delayed and marginally-blunted hyperalgesic response to intraplantar CFA (Akopian et al., 1999). Subsequent studies, using the same mice, suggested a requisite role for the channel in nerve growth factor–induced hyperalgesia but not in PGE2-induced inflammatory thermal hyperalgesia and only an equivocal role in neuropathic pain resulting from a partial sciatic nerve injury (Kerr et al., 2001). Each of the above genetic approaches has its strengths and weaknesses. Gene knockout is frequently criticized because of potential developmental issues and the NaV1.8-null mouse was, indeed, associated with a compensatory increase in the NaV1.7 sodium channel transcript (Akopian et al., 1999). Antisense ODNs have been used extensively as research tools but they suffer from variable silencing efficacy and off-target effects (Dorsett and Tuschl, 2004; Rohl and Kurreck, 2006) which complicate the therapeutic application of this technology. In the present study, we sought to examine the potential of a small interfering RNA (siRNA) approach that exploits a nucleic-acid-based target-selectivity advantage to achieve specific gene silencing (Milhavet et al., 2003). The siRNA approach has been demonstrated to be more tractable than antisense ODN, both as an investigative tool and as a potential therapeutic approach (see Dorsett and Tuschl, 2004 for a comparative review). For example, the siRNA approach has been reported to confer high levels of sequence-specific mRNA degradation of several therapeutically relevant genes, both in vitro and in vivo, without major off-target liabilities. Furthermore, siRNA sequences can be designed into short hairpin structures that can be packaged in viral vectors and delivered to target tissues by infection strategies (Azkur et al., 2005; Hong et al., 2006) that have the potential to yield longlasting gene silencing. In this study, several putative siRNAs designed against the rat NaV1.8 coding sequence (siRNA-NaV1.8) were evaluated for their ability to selectively attenuate the NaV1.8 message as well as the functional expression in a clonal ND7/23 cell line which was stably transfected with the rat NaV1.8 gene. To determine their in vivo efficacy in regulating NaV1.8 expression in dorsal root ganglia (DRGs) and relieving neuropathic pain, one of the effective siRNAs was delivered via an indwelling epidural cannula to lumbar DRGs, in rats that had previously been subjected to a CCI to the left sciatic nerve. The effect on tactile allodynia in CCI rats was evaluated and subsequently the NaV1.8 expression level in the 4th and 5th lumbar (L4 and 5) DRGs was determined.

813

EXPERIMENTAL PROCEDURES Animals Male adult Wistar rats (150 –250 g) were used in the study. Animals were housed in groups of three (but individually for 1 week after surgery) in plastic cages with free access to food and water under a 12-h light/dark cycle. Animals were acclimated for 1 week before experiments. All experimental procedures were approved by the Schering-Plough Research Institute Animal Care and Use Committee and were in accordance with the guidelines of the National Institutes of Health and the International Association for the Study of Pain. All efforts were made to minimize the number of animals used and their suffering.

Surgery For all surgical procedures, the animals were anesthetized with sodium pentobarbital (50 mg/kg, intraperitoneally for induction, and 15–20 mg/kg every hour for maintenance). An adequate depth of anesthesia was monitored, at regular intervals, by confirming absence of responses to noxious stimulation. Rectal temperature was maintained near 37 °C by a controlled heating blanket. Surgeries for epidural catheterization and nerve injury were performed using the procedures described below. Upon completion of the surgery, hemostasis was confirmed and the incisions were closed in layers with suture. The rats were kept warm during recovery from anesthesia, after which they returned to their home cage in the animal colony. Surgically operated rats were inspected daily for signs of autotomy and apparent paralysis. Their body weight, food and water intake were also monitored. On rare occasions, early signs of autotomy were seen (gnawing of claw tips and some surrounding tissue on the injured hind paw), and the animal was promptly killed. CCI of the sciatic nerve model. A CCI of the sciatic nerve was induced according to the method of Bennett and Xie (1988). Briefly, following a skin incision made on the lateral surface of the left thigh, the common sciatic nerve was exposed at mid-thigh level by blunt dissection through the biceps femoris muscle. Proximal to the trifurcation, a 10 mm portion of nerve was freed of adhering tissue and four loose ligatures (4.0 chromic gut suture) were tied around the sciatic nerve (1–1.5 mm apart). Epidural catheterization. The catheterization of the lumbar epidural space was carried out in the CCI rats on post-operative days 7– 8 using a modified procedure of a previously reported method by van den Hoogen and Colpaert (1981). Briefly, rats were placed in the prone position and a small midline incision was made in the region of the 1st to 3rd lumbar vertebrae. Following the superficial muscles being carefully dissected and laterally retracted, a small hole was made in the intervertebral space between L1–L2. A polyethylene catheter (PE-10; nominal i.d. 0.28 mm, length 26 cm) was then inserted to the lumbar epidural space and gently advanced about 2–2.5 cm caudally so that the catheter tip reached the region of L4 vertebra. The catheter was secured with a 4 – 0 silk suture tied to a 0.5 mm hole made in the L1 spinous process. Animals were excluded if blood or cerebrospinal fluid was aspirated. The remainder of the catheter was subcutaneously tunneled and exited through a small incision at the back of the neck. The catheter was blocked with a metal plug and secured to the skin with a suture. At the end of the procedure the catheter was flushed with 15 ␮l sterile saline, a volume equivalent to the volume of the catheter. Buprenorphine (0.05– 0.1 mg/ kg, s.c.) was applied to the surgical site on the completion of the surgery to provide post-surgical pain relief. To confirm correct catheter positioning, 50 ␮l of 2% lidocaine was administered through the catheter after complete recovery from anesthesia. A correct epidural catheter placement was judged by paralysis of the hindlimbs without affecting normal forelimb motor function. The

814

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821

cannulated animals were allowed 4 –5 days to recover prior to commencing experiments.

Pain behavioral assessment: measurement of tactile allodynia Rats were placed in transparent plastic boxes on an elevated mesh floor and were habituated to this environment daily, commencing at least 2 days before the study and for a minimum of 30 min prior to mechanical sensitivity test. Tactile sensitivity was evaluated using a series of calibrated von Frey filaments (0.4 – 15 g) applied perpendicularly to the mid-plantar surface of the ipsilateral hind paws in CCI rats, with sufficient force to bend the filament slightly for 3–5 s. Care was taken that the same sites were tested consistently across all the animals on all days. An abrupt withdrawal of the foot or licking and vigorously shaking in response to stimulation were considered pain-like responses. The threshold was determined using the up– down testing paradigm. The 50% paw withdrawal threshold (PWT) was calculated using the nonparametric Dixon test (Dixon, 1980; Chaplan et al., 1994). The values of 15 g and 2 g were used as cutoff for pre- and post-surgical conditions, respectively. For siRNA studies, only those rats that demonstrated a threshold less than 4 g by 13 days after surgery were included in the behavioral studies (more than 90% of the animals). All testing was done by experimenters blinded to the siRNA or vehicle treatments.

siRNAs Putative siRNAs against the rat NaV1.8 coding sequence were designed according to Tuschl’s prediction rules (Tuschl et al., 1999; see also Table 1). Five NaV1.8-siRNAs were selected in order to cover various regions of the 5874 nucleotide NaV1.8 coding sequence. Mismatch small interfering RNAs (mm-siRNAs) were designed by introducing a four-nucleotide mismatch in the original siRNA sequence. All siRNAs, designed with complementary dT-modified 3=-overhangs, including a nonspecific control siRNA (ns-siRNA), were purchased from Qiagen Inc (Valencia, CA, USA).

siRNA preparation and delivery for in vivo studies For in vivo studies, siRNA stock solution (1 mg/ml) was prepared using siRNA suspension buffer (Qiagen). The solution was then heated to 90 °C for 1 min followed by 60 min incubation at 37 °C. Aliquots were stored at ⫺20 °C. On the day of the injection, an aliquot of the siRNA was thawed on ice and then mixed with the transfection reagent, iFect (Neuromics Antibodies, Northfield, MN, USA), to yield a final concentration of 0.2 ␮g/␮l. A total of 100 ␮l of the siRNA solution or vehicle (20 ␮l of suspension buffer in 80 ␮l of iFect; 1:4 v/v) was locally delivered to lumbar DRGs through implanted epidural catheter by slow injection. The catheter was flushed with 15 ␮l sterile saline.

Cell line and transfections Gene transfection and cell culture. The mouse neuroblastoma/rat DRG fusion cell line, ND7/23, was obtained from the European Collection of Cell Cultures (ECACC). Experiments were performed on a clonal variant of ND7/23 cells in which the coding sequence for rat NaV1.8 had been permanently incorporated into the host genome using a retroviral vector infection strategy, the details of which have been described previously (Zhou et al., 2006). Clonal cells expressing NaV1.8 channels, henceforth referred to as ND7/23-NaV1.8, were cultured in Dulbecco’s modified Eagle’s medium (Irvine Scientific, Santana Ana, CA, USA) supplemented with 10% v/v fetal bovine serum and 1% sodium pyruvate and grown in incubators set to 5% CO2 and 37 °C. siRNA transfection. Transfections of ND7/23-NaV1.8 cells with siRNAs were performed using Lipofectamine 2000 (LF2000; Invitrogen, Carlsbad, CA, USA). Briefly, 2–3⫻105 cells were seeded into six-well dishes 24 h prior to transfection to yield a confluency of 60% on the day of transfection. Individual siRNAs were diluted in OptiMEM (Invitrogen) and mixed with LF2000 so as to yield the final desired siRNA concentrations. After incubating the transfection mixture for 20 min at room temperature, 0.5 ml was added to each well of cells containing 1.5 ml of normal growth medium. Transfection was allowed to proceed for 4 – 6 h at 37 °C in a humidified incubator equilibrated with 5% CO2, the transfection mixture was then replaced with fresh culture medium, cultures were returned to the incubator and NaV1.8 expression or function assessments were undertaken 24, 48 or 72 h later.

Membrane potential assays ND7/23-NaV1.8 cells were used in membrane potential assays on a FlexStation spectrofluorimeter (Molecular Devices, Sunnyvale, CA, USA). Cells were plated in black clear-bottomed 96-well polylysine plates (BD Biosciences, San Jose, CA, USA) at 50% confluency in 100 ␮l of media and incubated overnight at 37°C and 5% CO2. At 24 h after siRNA transfection, the plates were used for performing membrane potential assays. On an assay day, the transfection medium in 96 well plates was replaced with 80 ␮l membrane potential dye-buffer which was prepared as follows. Lyophilized membrane potential dye (Molecular Devices) was reconstituted in 10 ml Na⫹-free assay buffer (150 mM tetramethylammonium chloride, 3.25 mM KCl, 2 mM CaCl2, 2 mM MgCl2, 10 mM Hepes, 11 mM glucose, and 100 ␮M CdCl2 pH 7.4 adjusted with KOH). Working stock of the membrane potential dye mixture consisted of 250 ␮l reconstituted dye, 2.5 ml 1.2 ␮M TTX (final concentration⫽300 nM), 2.5 ml of 12 ␮M cypermethrin (final concentration⫽3 ␮M) and Na⫹-free assay buffer to a final volume of 10 ml. After incubating cells with membrane potential buffer for an hour at room temperature, plates were placed in the FlexStation and fluorescence assayed, also at room temperature. TTX was used to remove any signal from constitutively-expressed TTX-sensitive sodium channels. The addition of sodium (28 ␮l of

Table 1. siRNA sequence details siRNA no.

Target gene and position of cds

siRNA sequence (sense strand)

1 MM1 2 3 MM3 4 5 ns-siRNA

Rat NaV1.8 492-512 Rat NaV1.8 492-512 Rat NaV1.8 576-596 Rat NaV1.8 1203-1233 Rat NaV1.8 1203-1233 Rat NaV1.8 2790-2810 Rat NaV1.8 4596-4616 Nonspecific control

r(AAGCAGGACCAUUUCCAGA)dTT r(AAGCACCAGCAUUUGCAGA)dTT r(AGUGUCUGUCCAUUCCUGG)dTT r(CUACACCAGCUUUGAUUCC)dTT r(CUACAGGAGGUUUCAUUCC)dTT r(AUCCAUCUGCCUCAUCCUC)dTT r(UAAGUACCAAGGCUUCGUG)dTT r(UUCUCCGAACGUGUCACGU)dTT

cds, Targeted coding sequence, numbering based on GenBank accession # NM_017247; MM, mismatch sequence with base changes in bold text.

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821 600 nM NaCl per well), triggered a collapse in the membrane potential gradient due to ion flux through NaV1.8 sodium channels that were ‘locked’ in an open configuration by the presence of cypermethrin. The resulting increase in fluorescence was monitored for a period of 5 min at 5-s intervals and data were expressed as a % maximal fluorescence signal (arbitrary intensity units). Pilot experiments revealed a small, but consistent, basal signal to be evident in wt ND7/23 cells, this was generally around 20% (mean⫽20.0⫾0.9%, n⫽12 determinations). We attribute this residual signal to be likely due to non-specific and probably unavoidable changes in membrane potential inherent in the assay. In order to calibrate the assay across test days we therefore used a non-selective sodium channel blocker, riluzole, to define the maximal NaV1.8 sodium channel signal window (see Results). Accordingly, replicate riluzole wells were included as a reference in all plate-based FlexStation assays involving siRNAs and the potencies of each siRNA probe was expressed as a fraction of the riluzole-defined NaV1.8-dependent fluorescence. Peak fluorescent signals obtained for cells transfected with different siRNAs were compared using replicate determinations across multiple assay runs (typically five separate plate assays per condition).

Northern blot analysis Cells were harvested from plates and homogenized on Qiashredder columns (Qiagen). Total RNA was purified from homogenized lysates using the RNeasy Mini Kit (Qiagen) according to manufacturer’s instructions and treated with DNAse to remove residual DNA. RNA samples were quantified using an Agilent BioAnalyzer (RNA Agilent Technologies, Palo Alto, CA, USA). The Northern Max Bright Star Psoralen-Biotin labeling and Bright Star BioDetect Kits (Ambion) were used for Northern blot analysis. Ten micrograms of each total RNA sample and Biotinylated RNA Millenium Markers were electrophoresed on a 1% agarose gel containing 1.85% formaldehyde and ethidium bromide. RNA was transferred overnight to a positively charged membrane in a downward transfer and the membrane was rinsed briefly in 1⫻ gel running buffer followed by cross-linking in a Stratolinker. A 750 nt gel-purified NaV1.8 PCR fragment (nt 229 to nt 987) was used as probe. Psoralen– biotin labeling of the probe was performed according to the manufacturer’s suggested protocol. After a brief pre-hybridization step performed for 45 min at 42 °C, 5 ng labeled probe in a volume of 10 ml was hybridized to the membrane in a hybridization bottle. Hybridization was carried out at 42 °C overnight in a hybridization oven. The membrane was washed at room temperature in low stringency wash buffer, twice for 5 min each, then twice for 15 min each in high stringency wash buffer warmed to 42 °C. Detection was performed according to the recommended Bright Star BioDetect Kit protocol. Quantization of the signal was accomplished with a Fluorchem 8000 Imaging System using the AlphaEaseFC software (Alpha Innotech Corporation, San Leandro, CA, USA).

Quantitative PCR DRG tissues taken from the rats were preserved in RNAlater (Ambion) and were homogenized using a Tissuelyser (MM300, Qiagen) with a single 6.5 mm stainless steel ball per sample. Total RNA was isolated from the homogenate using an RNeasy Plus Mini Kit according to the manufacturer’s recommended protocol. TaqMan primers and probes were designed with PrimerExpress software (ABI). The sequences of the rat primers and probe for NaV1.8 (GenBank accession # NM_017247) are as follows: forward primer 5=-CACCGTGTTTTTCACAATGGAG, reverse primer 5=-GGAAGGTACGGAGCACAGACA, PROBE 6-FAM-CTGTGTCATCGTCACCGTGAGCCT. Quantitative PCR was carried out with an ABI Prism 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA, USA). The PCR reactions were prepared using the components from the Platinum qRT-PCR kit

815

and assembled according to the manufacturer’s instructions (Qiagen). The final concentrations of the primers and probe in the PCR reactions were 200 nM and 100 nM respectively. The fluorogenic probes were labeled with 6-carboxyfluorescein (6FAM) as the reporter and 6-carboxy-4,7,2,7=-tetramethylrhodamine (TAMRA) as a quencher. Each 10 ␮l PCR reaction contained 2 ␮l (10 ng) of total RNA. Six replicates of each RT-PCR reaction were performed in a 384-well plate according to the following protocol: 1 cycle for 10 min at 50 °C, followed by 1 5-min cycle at 95 °C, followed by 40 cycles at 95 °C for 15 s and 60 °C for 1 min. A eukaryotic 18S rRNA endogenous control probe/primer set (ABI) was used as an internal control for RNA quality.

Whole-cell voltage-clamp electrophysiology All whole-cell patch-clamp recordings were carried out at room temperature (19 –21 °C). At the time of the experiment, an individual glass coverslip was removed from the culture plate and placed into a perfusion chamber mounted on the stage of an inverted phase-contract microscope equipped with fluorescence optics. The cells were continuously perfused with a salt solution of the following composition: NaCl (129 mM), KCl (3.25 mM), CaCl2 (2 mM), MgCl2 (2 mM), Hepes (10 mM), D-glucose (11 mM), TEA-Cl (20 mM), pH 7.4, 345 mOsm. Conventional whole-cell voltage-clamp electrophysiology was used to record voltage-activated currents, patch electrodes contained a solution of the following composition: CsF (120 mM), NaCl (10 mM), Hepes (10 mM), EGTA (11 mM), TEA-Cl (10 mM), CaCl2 (1 mM), MgCl2 (1 mM), pH 7.3, 325 mOsm. To allow for selection of siRNAtransfected ND7/23-NaV1.8 cells for voltage-clamp recordings, a GFP-encoding plasmid (peGFP, BD Biosciences Clontech) was co-transfected with siRNAs (1 ␮g GFP plasmid and siRNA to a final concentration of 25 nM in six-well tissue culture dishes). Cells were viewed briefly for GFP fluorescence using U.V. illumination (FITC filter set) and arbitrarily selected on the basis of their relative fluorescence intensity. In all experiments, membrane holding potential (Vh) was held at ⫺90 mV between voltage command protocols. Conventional current–voltage (I/V) relationships were established for each cell studied by applying a series of increasing depolarizing command pulses (50 ms duration) from a prepulse potential of ⫺120 mV (150 ms duration). Previous work in this laboratory has established that step commands from ⫺120 mV ensure complete sodium channel availability in ND7/23-NaV1.8 clonal cells (Zhou et al., 2003). After completion of the control I/V data acquisition, the experiment was repeated in the presence of 300 nM TTX. This experimental design therefore yielded total peak sodium current followed by total peak tetrodotoxin-resistant (TTXr) current. Data were acquired online (sampling frequency 10 kHz, filtered at 2 kHz) using an Axopatch 200b amplifier and pClamp 8.0 software (Axon Instruments). Series resistance (80%) and capacitance currents were electronically compensated. Total tetrodotoxin-sensitive (TTXs) current was calculated off-line by subtracting TTXr current from total sodium current.

Data analysis In the in vitro electrophysiology studies, the mean values of peak amplitude of sodium currents were calculated from the ND7/23 cells in each condition. The differences were expressed as percentage of the mean value from mock-transfected cells. Statistical values are reported as mean⫾S.E.M. Differences between means were assessed by Student’s t-test and a value of P⬍0.05 was considered significant. The PCR data were quantified based on a 12-point standard curve generated using fourfold serial dilutions of a cDNA containing the gene of interest. The fourfold dilutions began at 20,000 fg. This procedure provides an absolute quantification of the amount of rat NaV1.8 mRNA in a given sample.

816

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821

In the pain behavioral studies, the tactile thresholds evaluated by von Frey filaments were expressed in grams. Data were compared between pre- and post-drug conditions using a repeatedmeasures ANOVA followed by Wilcoxon test (matched pairs) when appropriate. A P value of less than 0.05 was considered to be statistically significant. All the data are presented as mean⫾S.E.M.

RESULTS Characterization of rNaV1.8 channels stably expressed in ND7/23 cell line A ND7/23 clonal cell line permanently expressing NaV1.8 was used for evaluation and selection of optimal siRNA variants from a number of candidate probes. Macroscopic sodium currents recorded from the ND7/23-NaV1.8 clonal cell line comprised two conductances, one being TTXs, a native component found in the wild-type (wt) ND7/23 cells (Zhou et al., 2006), and a second TTXr component which was attributed to the recombinant rat NaV1.8 sodium channels. The NaV1.8-specific currents could be studied in isolation from the background TTXs currents by the inclusion of TTX (300 nM) in the bathing medium (Fig. 1; see also Zhou et al., 2003). The NaV1.8-currents in the ND7/ 23-NaV1.8 cell line were kinetically slower in both activation and inactivation time courses (Fig. 1A). The I/V relation curve revealed a positively-shifted voltage dependency of channel activation compared with TTXs channels (Zhou et al., 2006), with a peak inward current occurring at

around ⫹20 mV (Fig. 1B). The NaV1.8-currents were characteristically resistant to TTX but sensitive to sodium channel blockers such as Riluzole (Fig. 1C), consistent with our previous observations using NaV1.8 sodium channels transiently transfected in ND7/23 cells (Zhou et al., 2003). The level of NaV1.8 in the ND7/23-NaV1.8 cell line was robust with an averaged peak current above 800 pA, this retained over more than 60 passages without significant decline in amplitude (Fig. 1D). Identification of optimal siRNAs for knockdown of NaV1.8 using a high-throughput membrane potential assay and Northern blot analysis Five siRNAs were designed to target various regions of the rat NaV1.8 coding sequence (Table 1). Following transfection to ND7/23-NaV1.8 cells, the ability of each siRNA to attenuate NaV1.8 activity was evaluated using a FlexStation membrane potential assay. Addition of sodium to the Na⫹free assay buffer caused a robust change in membrane potentials in mock-transfected ND7/23-Nav1.8 cells, as indicated by a large increase in fluorescent signals (Fig. 2A) with an average of 199⫾16⫻103 (n⫽6) fluorescence units. To determine the extent of any non-Nav1.8 components in the FlexStation-reported membrane potential change, experiments were performed in wt ND7/23 (ND7/23-wt) cells under the same assay conditions including the presence of TTX (300 nM). ND7/23-wt cells consistently displayed a detectable but substantially smaller increase in fluorescent

Fig. 1. Electrophysiological characterization of recombinant NaV1.8 sodium channels expressed in ND7/23 cell line. (A) Whole-cell NaV1.8-currents, recorded from a ND7/23-NaV1.8 cell, in response to a series of increasing depolarizing command pulses using the voltage protocol shown in the inset (parameters: Vh⫽⫺90 mV, prepulse⫽⫺120 mV for 150 ms, Vcommand⫽⫺60 to ⫹90 mV in 10 mV increments and 50 ms duration). NaV1.8 currents were isolated by the addition of TTX (300 nM) to the bath solution. (B) I/V relationship. Data were generated from six cells using the voltage protocol shown in (A). (C) Superimposed traces of peak NaV1.8-currents showing their resistance to TTX (300 nM, solid trace; 30 ␮M, dotted trace) and block by a sodium channel blocker, riluzole (100 ␮M, gray trace). Currents were evoked by a voltage step from a holding potential of ⫺90 mV to 20 mV. (D) Preservation of robust expression of functional NaV1.8 channels, in the ND7/23 cell line, over passages. The averaged peak current amplitude of the cells in passage 65 (P-65, n⫽23) was comparable to that in passage 1 (P-1, n⫽43).

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821

817

due to knockdown of NaV1.8, we measured changes in NaV1.8 expression, at RNA level, using Northern blot analysis of ND7/23-Nav1.8 cells transfected with siRNA1, siRNA3, siRNA4 and ns-siRNA, a nonspecific siRNA, as well as in mock transfected cells. In agreement with the results obtained from membrane potential assays, siRNA1 and siRNA3 strongly down-regulated NaV1.8 RNA expression, while siRNA4 showed no detectable effect on NaV1.8 expression (Fig. 3). In addition, transfection with ns-siRNA, a nonspecific siRNA, had no effect on NaV1.8 RNA expression (Fig. 3). Attenuation of NaV1.8 currents by siRNAs

Fig. 2. Effects of siRNAs on NaV1.8-mediated membrane depolarization in ND7/23-NaV1.8 cells determined using the FlexStation membrane potential assay. (A) Fluorescent signals (in arbitrary fluorescence units) elicited by addition of Na⫹ to the Na⫹-free assay buffer (containing TTX 300 nM and cypermethrin 3 ␮M) in control ND7/23Nav1.8 cells (1), ND7/23-Nav1.8 cells treated with 25 nM siRNA1 (2), 30 ␮M riluzole (3) and wt ND7/23 cells (4). Cells were preincubated with a membrane potential sensing dye as described in Experimental Procedures. (B) Relative efficacies of the five siRNA variants in attenuation of functional expression of NaV1.8 channels in ND7/23-Nav1.8 cells. The evoked fluorescent signals were recorded 24 h after siRNA transfection and were expressed as percentage of mock-transfected riluzole-sensitive control fluorescence. Data represent mean⫾S.E.M. of five to seven replicate experiments for each condition.

signal (mean⫽20.0⫾0.9% of NaV1.8-expressing cells, n⫽12 determinations; see also Fig. 2A). To further confirm the primary contribution of Nav1.8 channels to the membrane potential change, the effect of riluzole (Song et al., 1997), a nonspecific sodium channel blocker, was examined. Riluzole (30 ␮M) attenuated Na-evoked membrane potential fluorescence signal in ND7/23-NaV1.8 cells to around ND7/23-wt background levels (23.6⫾0.5% of the peak signal in ND7/23-NaV1.8 cells, n⫽12 determinations; a single-well example is also shown in Fig. 2A). In the siRNA-transfected cells, different degrees (24 – 78%) of attenuation of the evoked fluorescent signals were observed with each of the five siRNAs (Fig. 2B). siRNA1 and siRNA3 were found to be the most effective, achieving over 75% reduction of the fluorescent signals relative to that in mock-transfected cells. In contrast, no significant change was detected in the ND7/23-Nav1.8 cells transfected with a control non-silencing small interfering RNA (ns-siRNA; Fig. 2B). To further confirm that the siRNA-mediated attenuation of membrane potential signal in the FlexStation assay was

Functional knockdown of NaV1.8 was further confirmed by whole-cell voltage-clamp recordings of NaV1.8 currents in the ND7/23-NaV1.8 cells transfected with siRNA1, siRNA3 and siRNA4. The peak NaV1.8 currents were significantly smaller in siRNA1- and siRNA3-transfected cells (18.5⫾ 3.0% and 23.2⫾7.0%, respectively; Fig. 4), compared with the currents in mock-transfected cells which were recorded in parallel experiments (see Table 2 for actual mean current amplitudes for all experiments). However, the constitutive TTXs conductances in ND7/23-NaV1.8 cells were not affected (Fig. 4). The siRNA-mediated reduction of NaV1.8 currents persisted for at least 72 h post-transfection (data not shown). In contrast, the

Fig. 3. Northern blot analysis of siRNA-mediated downregulation of NaV1.8 RNA expression. NaV1.8 expression in mock-transfected cells (lane a) is comparable to that seen after transfection with a nonsilencing control siRNA (lane b) or with the weakly-silencing siRNA4 (lane e). Cells treated in the same manner with either siRNA1 or siRNA3 (lanes c and d, respectively) showed a substantial reduction of NaV1.8 RNA expression. The bottom gel indicates that similar amounts of mRNA (18S and 28S) were loaded into each lane and rules out unequal loading as a potential explanation for the siRNA-mediated effects.

818

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821

Fig. 4. Electrophysiological evaluation of change in functional expression of NaV1.8 channels following siRNA transfection. (A) Superimposed traces of NaV1.8 current recorded from ND7/23-NaV1.8 cells transfected with siRNA1 (gray trace) and mm-siRNA1 (mm-siRNA1; black trace). NaV1.8 currents were evoked by a voltage step from a holding potential of ⫺90 mV to 20 mV and isolated by the addition of TTX (300 nM) to the bath solution. (B) Pooled data showing the relative potencies of the various siRNA probes against TTXs and NaV1.8 currents. Data are expressed as the percentage of mean current amplitudes recorded from mock-transfected cells (see Table 2 for absolute mean current amplitudes and number of cells for each condition). Note that, the potent knockdown of NaV1.8 currents by siRNAs 1 and 3 but the lack of effect of these same probes on TTXs currents.

mean amplitudes of peak NaV1.8 currents recorded from cells transfected with mm-siRNA1 and mm-siRNA3 were comparable to that in mock-transfected cells (Fig. 4 and Table 2). Table 2. Mean current amplitudes recorded by whole-cell voltage clamp siRNA experiment

Mock control siRNA1 Mock control mm-siRNA1 Mock control siRNA3 Mock control mm-siRNA3 Mock control siRNA4

TTXs

NaV1.8

I (pA)

n

I (pA)

n

3878⫾898 3798⫾503

10 11

2080⫾466 1950⫾365 2080⫾466 2115⫾430

10 10 10 10

912⫾161 169⫾27 532⫾106 518⫾91 568⫾121 132⫾38 568⫾121 534⫾110 601⫾79 614⫾90

10 11 10 22 10 10 10 10 10 22

Values of I are means ⫾ SEM. cells used in the studies.

n⫽Number of voltage-clamped

Fig. 5. Effect of siRNA-NaV1.8 on neuropathic pain and NaV1.8 message in lumbar DRGs from CCI rats. (A) Mechanical allodynia in CCI rats was significantly reversed by siRNAs (20 ␮g in 100 ␮l) delivered to lumbar DRGs through an implanted epidural catheter. The timing of siRNA deliveries is indicated by the arrows. The 50% PWT was measured by a series of von Frey filaments using the up– down method. Data are presented as mean⫾S.E.M. Statistical significance is assessed between saline and siRNA groups: * P⬍0.05, ** P⬍0.01. Each group consisted of 5 to 12 rats. (B) Bar graph showing the changes in the expression of NaV1.8 mRNA in L4 and L5 DRGs following saline (n⫽3) and siRNA (n⫽6) treatments in CCI rats. Data are expressed as percentage of control value obtained from naïve rats.

siRNA-NaV1.8 reverses mechanical allodynia in CCI rats Nav1.8 sodium channels are primarily expressed in the peripheral sensory nervous system. The functional consequence of knockdown of Nav1.8 sodium channels to neuropathic pain was examined in CCI rats following local delivery of siRNA-Nav1.8 to lumbar DRGs. After CCI surgery, rats developed mechanical allodynia which was characterized by a decreased PWT to tactile stimulation on the ipsilateral (injured) side. On post-operative day 13, mean 50% PWT was significantly (P⬍0.05) reduced to values below 3 g from a pre-surgical value of 13.4 g (⫾1.06, n⫽23; Fig. 5A). siRNA3 (20 ␮g in 100 ␮l) was then administered to lumbar DRGs, through an indwelling epidural catheter, once daily for four consecutive days, starting at post-operative day 13. In the siRNA-treated group, a significant elevation in PWTs was seen as early as day 3 after commencing the treatment. The PWTs were further increased to 10.57 g (⫾2.51, n⫽5) on the 4th day, and then plateaued. In contrast, rats assigned to either saline- or mm-siRNA-treated groups remained in the allodynic state with 50% PWT values below 3.5 g. No significant difference was observed between mm-siRNA and saline-injected groups (Fig. 5A).

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821

To confirm the observed reversal of allodynia in CCI rats being attributed to the in vivo knockdown of Nav1.8 channels by siRNA-Nav1.8, RT-PCR analysis was performed to determine the level of expression of Nav1.8 mRNA in L4 and L5 DRGs taken from the siRNA- and saline-treated CCI rats on the day when the maximum reversal of allodynia was obtained (1 day after the last siRNA administration). The Nav1.8 mRNA level in the DRGs from saline-treated CCI rats was lower (63%⫾8, n⫽3) relative to that from uninjured control rats. siRNA treatment caused a further decrease to 32% (⫾4, n⫽6) in the Nav1.8 mRNA level (Fig. 5B).

DISCUSSION The primary finding of the present study is that specific knockdown of Nav1.8 sodium channels in lumbar DRG neurons, using a siRNA, effectively reversed tactile allodynia in a rat neuropathic pain model. While the role of Nav1.8 sodium channels in neuropathic pain has been well established, the strategy of blocking this sodium channel isoform for pain treatment has been hampered by the lack of selective small molecule blockers. In this study, we have identified four siRNAs capable of selectively downregulating the expression and consequently attenuating the conductance of the recombinant NaV1.8 channels stably expressed in clonal ND7/23 cells. Local delivery of one of these siRNAs to lumbar DRGs successfully downregulated Nav1.8 expression and produced an analgesic effect in CCI rats. Our study demonstrates the potential of siRNA targeting Nav1.8 as a novel therapeutic strategy for pain management. NaV1.8 sodium channels are exclusively expressed in the peripheral sensory nervous system, including DRG, trigeminal and nodose ganglia (Akopian et al., 1996; Sangameswaran et al., 1996). The channel has been implicated in various pathological pain states (see reviews by Lai et al., 2004; Priestley, 2004). Under neuropathic conditions, while both transcript and channel protein of NaV1.8 were significantly reduced or abolished in injured DRG cell bodies (Dib-Hajj et al., 1996, 1999; Decosterd et al., 2002), a marked elevation of NaV1.8 immunoreactivity in axons and/or nerve terminals was found in the tissue samples taken from both animal pain models (Novakovic et al., 1998; Gold et al., 2003) and patients suffering a variety of persistent pain conditions (Coward et al., 2000; Shembalkar et al., 2001; Bucknill et al., 2002; Kretschmer et al., 2002). The increase in NaV1.8 protein was believed to occur in the uninjured axons (Gold et al., 2003), despite the fact that there was no change in NaV1.8 expression observed in their cell bodies (Decosterd et al., 2002). Attenuating NaV1.8 expression with antisense ODNs effectively reversed neuropathic pain (Lai et al., 2002), indicating a critical role for NaV1.8 in these uninjured axons/cell bodies in neuropathic pain. In the present study, we observed a significant reduction (37%⫾8, n⫽3) of NaV1.8 expression in ipsilateral L4 –L5 DRGs on 17 days after nerve injury in CCI rats that displayed apparent tactile allodynia. The reduction is in

819

agreement with previous observations in the same injury model (Dib-Hajj et al., 1999) and could well reflect the loss of NaV1.8 expression in the cell bodies of injured DRG neurons (Dib-Hajj et al., 1996, 1999; Decosterd et al., 2002). The residual NaV1.8 mRNA might be attributed to message in uninjured L4 –L5 DRG neurons, including those whose axons traversed the ligation but survived the injury (Basbaum et al., 1991; Carlton et al., 1991). Treatment of lumbar DRGs with a specific NaV1.8 siRNA effectively reversed mechanical allodynia in CCI rats, along with a downregulation of NaV1.8 in uninjured DRG neurons, as reflected by a further reduction of NaV1.8 expression in L4 –L5 DRGs. A similar effect on tactile allodynia in CCI rats has recently been reported following antisense ODNmediated knockdown of NaV1.8 (Joshi et al., 2006) and it is interesting to note that the timeframe for peak effect was similar to that observed in our siRNA experiments. In CCI rats, there is a near complete loss of large myelinated fibers in the sciatic nerve; unmyelinated Cfibers are believed to be the major component of the surviving axons (Basbaum et al., 1991; Carlton et al., 1991). We suggest that the siRNA-mediated reversal of mechanical allodynia observed in this study is largely attributed to the downregulation of NaV1.8 channels in uninjured, NaV1.8-expressing C-fibers, which remain in contact with their peripheral sensory receptors and report stimulusevoked pain, such as tactile allodynia. This conclusion is consistent with the role of uninjured nerve fibers in neuropathic pain (Li et al., 2000; Liu et al., 2000; Wu et al., 2001; Djouhri et al., 2006). Furthermore, there are several lines of evidence supporting the role of NaV1.8 expressed in uninjured C-fibers in neuropathic pain. First, NaV1.8 is involved in the genesis of action potentials in C-type DRG neurons, to which NaV1.8 contributes a substantial fraction (80 –90%) of the inward current that flows during the rising phase of action potentials. In NaV1.8 null mutant mice, the ability of small DRG neurons to generate action potentials was significantly impaired (Renganathan et al., 2001). Second, the NaV1.8 immunoreactivity was drastically increased in uninjured C-fibers of sciatic nerve following spinal nerve injury which led to an increase in TTXr component of the sciatic compound action potentials (Gold et al., 2003). Third, an increase in spontaneous activity has been observed in uninjured C-fibers after spinal nerve ligation (SNL; Wu et al., 2001). In our study, the siRNAmediated specific downregulation of NaV1.8 in sensory nerves, most likely in these uninjured C-fibers, effectively reversed neuropathic pain in CCI rats. Therefore, we suggest that NaV1.8 contributes to neuropathic pain by a mechanism involving uninjured C-fibers/nociceptors that undergo sensitization to chemical, as well as to mechanical and thermal stimuli, following nerve injury (Ali et al., 1999; Sato and Perl, 1991; Schafers et al., 2003; Shim et al., 2005; see also Campbell and Meyer, 2006). Our findings provide further support to the notion that blocking NaV1.8 function might provide an effective treatment of neuropathic pain (Porreca et al., 1999). The development of small molecules to specifically block NaV1.8 sodium channels has presented a significant

820

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821

challenge for conventional medicinal chemistry. The sodium channel family consists of at least nine subtypes, most of which are likely to play a role in the initiation and/or propagation of action potentials in excitable cells. Thus, any significant disruption of their activity is likely to have profound consequences for cell signaling. A broad-spectrum action against multiple ion channel targets is generally not tolerated, especially in a clinical setting (Priestley and Hunter, 2006). Pharmacological selectivity is, therefore, considered an essential requirement for future therapeutics. In the present study, we demonstrated a remarkably selective and potent attenuation of NaV1.8 function by using siRNAs to disrupt production of nascent NaV1.8 protein by taking advantage of the specificity afforded in the coding sequence for the channel. To facilitate the selection of optimal siRNA candidates from a number of potential probes, we used a previously disclosed ND7/23 clonal cell line (Zhou et al., 2003) that was transformed to permanently express the rat NaV1.8 channel. This clonal cell line has proven to be a suitable host for recombinant NaV1.8 for a number of reasons. Firstly, ND7/23 cells do not constitutively express a TTXr (NaV1.8 or NaV1.9) conductance that is detectable at the whole-cell level. Secondly, pilot experiments revealed the ND7/23 cells were amenable to transient transfection and also supported the robust expression of recombinant NaV1.8. Finally, the background sodium conductance in ND7/23 cells, which we attribute to multiple TTX-s transcripts based on RT-PCR detection of mRNA, was easily blocked by TTX at concentrations below those known to exert any effect on NaV1.8. The stable expression ND7/ 23-NaV1.8 system developed for these studies retains all of the features of the previously reported transiently expressed rat recombinant NaV1.8 (Zhou et al., 2003). In our study, we found that our probes, siRNA1 and siRNA3 for example, were highly specific against NaV1.8 and did not affect expression of the closely related NaV1.5 and NaV1.9 sodium channels, consistent with other published work using siRNA technology. The functional reduction of NaV1.8 conductance mirrored the downregulation of this channel at the RNA level. The specificity of both siRNA1 and siRNA3 for NaV1.8 was further confirmed by the lack of their effects on the constitutive TTX-sensitive conductance in ND7/23 cells. We consider it unlikely that NaV1.8 is particularly sensitive to the disruption by siRNA or that our observations reflect a general cellular ‘malaise’ brought about by siRNA in the cytoplasm because neither mm-siRNA sequences nor sequences that were found to be ineffective at the RNA level had any effect on NaV1.8 sodium currents. In our cell culture studies, the downregulation of NaV1.8 function in ND7/23-NaV1.8 cells by siRNA1 and 3 was long lasting (up to 72 h). This is in contrast with a previous report that lipid-mediated transfection of siRNA oligos was effective in silencing genes, but generally shortlived (Elbashir et al., 2002). The short duration likely resulted from a combination of loss of siRNA, degradation over time and a ‘dilution’ effect due to cell replication in culture. In addition, our siRNAs showed greater effective-

ness in gene silencing. Thus, a concentration of 25 nM of siRNA1 and siRNA3 resulted in a greater than 75% knockdown in both NaV1.8 mRNA expression and channel function in our experiments.

CONCLUSION In summary, we have demonstrated that silencing the NaV1.8 gene by siRNA resulted in a functional knockdown of channel activity in the host ND7/23 cell line. This downregulation was highly specific and long lasting. When one of the siRNAs was delivered to the lumbar DRGs via an indwelling epidural cannula, it caused a significant reduction of NaV1.8 mRNA expression in uninjured L4 –L5 DRG neurons and consequently reversed mechanical allodynia in CCI rats. We conclude that functional NaV1.8 channel in uninjured axons of sciatic nerve in CCI rats plays an important role in neuropathic pain and blocking this channel would likely provide an effective treatment for neuropathic pain. Our study also indicates that the siRNA approach presents a promising new therapeutic strategy with a target-selectivity advantage. Acknowledgments—The authors dedicate this manuscript to the memory of Dr. Heidrun Engler, an outstanding scientist and an engaging collaborator.

REFERENCES Akopian AN, Sivilotti L, Wood JN (1996) A tetrodotoxin-resistant voltage-gated sodium channel expressed by sensory neurons. Nature 379:257–262. Akopian AN, Souslova V, England S, Okuse K, Ogata N, Ure J, Smith A, Kerr BJ, McMahon SB, Boyce S, Hill R, Stanfa LC, Dickenson AH, Wood JN (1999) The tetrodotoxin-resistant sodium channel SNS has a specialized function in pain pathways. Nat Neurosci 2:541–548. Ali Z, Ringkamp M, Hartke TV, Chien HF, Flavahan NA, Campbell JN, Meyer RA (1999) Uninjured C-fiber nociceptors develop spontaneous activity and alpha-adrenergic sensitivity following L6 spinal nerve ligation in monkey. J Neurophysiol 81:455– 466. Azkur AK, Kim B, Suvas S, Lee Y, Kumaraguru U, Rouse BT (2005) Blocking mouse MMP-9 production in tumor cells and mouse cornea by short hairpin (sh) RNA encoding plasmids. Oligonucleotides 15:72– 84. Basbaum AI, Gautron M, Jazat F, Mayes M, Guilbaud G (1991) The spectrum of fiber loss in a model of neuropathic pain in the rat: an electron microscopic study. Pain 47:359 –367. Bennett GJ, Xie YK (1988) A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain 33:87–107. Bucknill AT, Coward K, Plumpton C, Tate S, Bountra C, Birch R, Sandison A, Hughes SP, Anand P (2002) Nerve fibers in lumbar spine structures and injured spinal roots express the sensory neuron-specific sodium channels SNS/PN3 and NaN/SNS2. Spine 27:135–140. Campbell JN, Meyer RA (2006) Mechanisms of neuropathic pain. Neuron 52:77–92. Carlton SM, Dougherty PM, Pover CM, Coggeshall RE (1991) Neuroma formation and numbers of axons in a rat model of experimental peripheral neuropathy. Neurosci Lett 131:88 –92. Catterall WA, Goldin AL, Waxman SG (2003) International Union of Pharmacology. XXXIX. Compendium of voltage-gated ion channels: sodium channels. Pharmacol Rev 55:575–578.

X.-W. Dong et al. / Neuroscience 146 (2007) 812– 821 Chaplan SR, Bach FW, Pogrel JW, Chung JM, Yaksh TL (1994) Quantitative assessment of tactile allodynia in the rat paw. J Neurosci Methods 53:55– 63. Coward K, Plumpton C, Facer P, Birch R, Carlstedt T, Tate S, Bountra C, Anand P (2000) Immunolocalization of SNS/PN3 and NaN/ SNS2 sodium channels in human pain states. Pain 85:41–50. Decosterd I, Ji RR, Abdi S, Tate S, Woolf CJ (2002) The pattern of expression of the voltage-gated sodium channels Na(v) 1.8 and Na(v) 1.9 does not change in uninjured primary sensory neurons in experimental neuropathic pain models. Pain 96:269 –277. Dib-Hajj S, Black JA, Felts P, Waxman SG (1996) Down-regulation of transcripts for Na channel alpha-SNS in spinal sensory neurons following axotomy. Proc Natl Acad Sci U S A 93:14950 –14954. Dib-Hajj SD, Fjell J, Cummins TR, Zheng Z, Fried K, LaMotte R, Black JA, Waxman SG (1999) Plasticity of sodium channel expression in DRG neurons in the chronic constriction injury model of neuropathic pain. Pain 83:591– 600. Dixon WJ (1980) Efficient analysis of experimental observations. Annu Rev Pharmacol Toxicol 20:441– 462. Djouhri L, Koutsikou S, Fang X, McMullan S, Lawson SN (2006) Spontaneous pain, both neuropathic and inflammatory, is related to frequency of spontaneous firing in intact C-fiber nociceptors. J Neurosci 26:1281–1292. Dorsett Y, Tuschl T (2004) siRNAS: Applications in functional genomics and potential as therapeutics. Nat Rev Drug Discov 3:318 –329. Elbashir SM, Harborth J, Weber K, Tuschl T (2002) Analysis of gene function in somatic mammalian cells using small interfering RNAs. Methods 26:199 –213. Gold MS, Weinreich D, Kim CS, Wang R, Treanor J, Porreca F, Lai J (2003) Redistribution of NaV1.8 in uninjured axons enables neuropathic pain. J Neurosci 23:158 –166. Hong CS, Goins WF, Goss JR, Burton EA, Glorioso JC (2006) Herpes simplex virus RNAi and neprilysin gene transfer vectors reduce accumulation of Alzheimer’s disease-related amyloid-beta peptide in vivo. Gene Ther 13:1068 –1079. Joshi SK, Mikusa JP, Hernandez G, Baker S, Shieh CC, Neelands T, Zhang XF, Niforatos W, Kage K, Han P, Krafte D, Faltynek C, Sullivan JP, Jarvis MF, Honore P (2006) Involvement of the TTXresistant sodium channel Nav1.8 in inflammatory and neuropathic, but not post-operative, pain states. Pain 123:75– 82. Kerr BJ, Souslova V, McMahon SB, Wood JN (2001) A role for the TTX-resistant sodium channel Nav1.8 in NGF-induced hyperalgesia, but not neuropathic pain. Neuroreport 12:3077–3080. Khasar SG, Gold MS, Levine JD (1998) A tetrodotoxin-resistant sodium current mediates inflammatory pain in the rat. Neurosci Lett 256:17–20. Kretschmer T, Happel LT, England JD, Nguyen DH, Tiel RL, Beuerman RW, Kline DG (2002) Accumulation of PN1 and PN3 sodium channels in painful human neuroma-evidence from immunocytochemistry. Acta Neurochir (Wien) 144:803– 810. Lai J, Gold MS, Kim CS, Bian D, Ossipov MH, Hunter JC, Porreca F (2002) Inhibition of neuropathic pain by decreased expression of the tetrodotoxin-resistant sodium channel, NaV1.8. Pain 95:143–152. Lai J, Porreca F, Hunter JC, Gold MS (2004) Voltage-gated sodium channels and hyperalgesia. Annu Rev Pharmacol Toxicol 44: 371–397. Li Y, Dorsi MJ, Meyer RA, Belzberg AJ (2000) Mechanical hyperalgesia after an L5 spinal nerve lesion in the rat is not dependent on input from injured nerve fibers. Pain 85:493–502. Liu X, Eschenfelder S, Blenk KH, Janig W, Habler H (2000) Spontaneous activity of axotomized afferent neurons after L5 spinal nerve injury in rats. Pain 84:309 –318. Liu Y, Yao S, Song W, Wang Y, Liu D, Zen L (2005) Effects of intrathecally administered NaV1.8 antisense oligonucleotide on the

821

expression of sodium channel mRNA in dorsal root ganglion. J Huazhong Univ Sci Technolog Med Sci 25:696 – 699. Milhavet O, Gary DS, Mattson MP (2003) RNA interference in biology and medicine. Pharmacol Rev 55:629 – 648. Novakovic SD, Tzoumaka E, McGivern JG, Haraguchi M, Sangameswaran L, Gogas KR, Eglen RM, Hunter JC (1998) Distribution of the tetrodotoxin-resistant sodium channel PN3 in rat sensory neurons in normal and neuropathic conditions. J Neurosci 18:2174–2187. Porreca F, Lai J, Bian D, Wegert S, Ossipov MH, Eglen RM, Kassotakis L, Novakovic S, Rabert DK, Sangameswaran L, Hunter JC (1999) A comparison of the potential role of the tetrodotoxininsensitive sodium channels, PN3/SNS and NaN/SNS2, in rat models of chronic pain. Proc Natl Acad Sci U S A 96:7640 –7644. Priestley T (2004) Voltage-gated sodium channels and pain. Curr Drug Targets CNS Neurol Disord 3:441– 456. Priestley T, Hunter JC (2006) Voltage-gated sodium channels as molecular targets for neuropathic pain. Drug Dev Res 67:360 –375. Renganathan M, Cummins TR, Waxman SG (2001) Contribution of Nav1.8 sodium channels to action potential electrogenesis in DRG neurons. J Neurophysiol 86:629 – 640. Rohl T, Kurreck J (2006) RNA interference in pain research. J Neurochem 99:371–380. Sangameswaran L, Delgado SG, Fish LM, Koch BD, Jakeman LB, Stewart GR, Sze P, Hunter JC, Eglen RM, Herman RC (1996) Structure and function of a novel voltage-gated, tetrodotoxin-resistant sodium channel specific to sensory neurons. J Biol Chem 271:5953–5956. Sato J, Perl ER (1991) Adrenergic excitation of cutaneous pain receptors induced by peripheral nerve injury. Science 251:1608 –1610. Schafers M, Lee DH, Brors D, Yaksh TL, Sorkin LS (2003) Increased sensitivity of injured and adjacent uninjured rat primary sensory neurons to exogenous tumor necrosis factor-alpha after spinal nerve ligation. J Neurosci 23:3028 –3038. Shembalkar PK, Till S, Boettger MK, Terenghi G, Tate S, Bountra C, Anand P (2001) Increased sodium channel SNS/PN3 immunoreactivity in a causalgic finger. Eur J Pain 5:319 –323. Shim B, Kim DW, Kim BH, Nam TS, Leem JW, Chung JM (2005) Mechanical and heat sensitization of cutaneous nociceptors in rats with experimental peripheral neuropathy. Neuroscience 132:193– 201. Song JH, Huang CS, Nagata K, Yeh JZ, Narahashi T (1997) Differential action of riluzole on tetrodotoxin-sensitive and tetrodotoxinresistant sodium channels. J Pharmacol Exp Ther 282:707–714. Tuschl T, Zamore PD, Lehmann R, Bartel DP, Sharp PA (1999) Targeted mRNA degradation by double-stranded RNA in vitro. Genes Dev 13:3191–3197. van den Hoogen RH, Colpaert FC (1981) Long term catheterization of the lumbar epidural space in rats. Pharmacol Biochem Behav 15:515–516. Villarreal CF, Sachs D, de Queiroz Cunha F, Parada CA, Ferreira SH (2005) The role of NaV1.8 sodium channel in the maintenance of chronic inflammatory hypernociception. Neurosci Lett 386:72–77. Wu G, Ringkamp M, Hartke TV, Murinson BB, Campbell JN, Griffin JW, Meyer RA (2001) Early onset of spontaneous activity in uninjured C-fiber nociceptors after injury to neighboring nerve fibers. J Neurosci 21:RC140. Zhou X, Dong XW, Crona J, Maguire M, Priestley T (2003) Vinpocetine is a potent blocker of rat NaV1.8 tetrodotoxin-resistant sodium channels. J Pharmacol Exp Ther 306:498 –504. Zhou X, Dong XW, Priestley T (2006) The neuroleptic drug, fluphenazine, blocks neuronal voltage-gated sodium channels. Brain Res 1106:72– 81.

(Accepted 26 January 2007) (Available online 23 March 2007)