Functional and pharmacological properties of human and rat NaV1.8 channels

Neuropharmacology 56 (2009) 905–914

Contents lists available at ScienceDirect

Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm

Functional and pharmacological properties of human and rat NaV1.8 channels Liam E. Browne a, Jeff J. Clare b, Dennis Wray a, * a b

Faculty of Biological Sciences, University of Leeds, Leeds LS2 9JT, United Kingdom Ion Channel Group, Millipore, Cambridge CB5 8PB, United Kingdom

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 September 2008 Received in revised form 21 January 2009 Accepted 21 January 2009

The aim of this work is to characterise the functional properties of human and rat NaV1.8 channels and to investigate the action of anti-nociceptive agents. NaV1.8 a-subunits were expressed in mammalian sensory neuron-derived ND7/23 cells, and sodium currents were recorded using whole-cell patch clamp. The current–voltage curves for activation were similar for human and rat NaV1.8 channels. However, for inactivation, human NaV1.8 showed more hyperpolarised voltage-dependence than for the rat channel, faster development of inactivation, slower recovery from the fast component of inactivation, and faster recovery from the slow component. Thus, this would imply that the human channel is more inactivated at normal resting potentials. Compounds 227c89, A-803467, V102862, ralfinamide and tetracaine all showed greater affinity for the inactivated state than for the resting state. Compounds A-803467 and V102862 were the most potent, and A-803467 showed greater inactivated state affinity for human than for rat channels. Surprisingly, during recovery from inactivation, an increase in current was observed for V102862 and A-803467, probably due to disinhibition of resting block. Rather than the use-dependent inhibition normally seen with inactivated state blockers, for A-803467 this disinhibition led to an increase in current during repetitive stimulation, while V102862 showed less inhibition than otherwise expected at lower frequencies. Thus the data supports the suggestion that, while both V102862 and A-803467 are potent inhibitors of NaV1.8, the compound V102862, rather than A-803467, may be useful as an analgesic where physiological firing frequencies are higher. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Sodium channels Electrophysiology Ion channels Anti-nociceptive drugs

1. Introduction Voltage-gated Naþ channels are responsible for the initiation and propagation of action potentials in neuronal, skeletal and cardiac muscle cells (Hille, 2001). Drugs such as local anaesthetics act on these channels mainly by preferential binding to the inactivated state of the channel (Hille, 1977). This results in a voltageand use-dependent mechanism leading to increased drug block during periods of repetitive firing or sustained depolarisation, as may occur during pain signalling (Blair and Bean, 2002; Renganathan et al., 2001). The tetrodotoxin (TTX)-insensitive Naþ currents in nociceptive neurons of the dorsal root ganglion include a current produced by NaV1.8 (Akopian et al., 1996; Rabert et al., 1998; Sangameswaran et al., 1996; Blair and Bean, 2002; Renganathan et al., 2001). This channel has distinct biophysical properties compared to other NaV subtypes, showing both slower inactivation kinetics and more depolarised voltage-dependent activation and inactivation (Akopian et al., 1996; Rabert et al., 1998; Sangameswaran et al., 1996).

* Corresponding author. Tel.: þ44 113 3434320; fax: þ44 113 3434228. E-mail address: [email protected] (D. Wray). 0028-3908/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropharm.2009.01.018

In pain signalling, the NaV1.8 channel plays a well-established role in noxious mechanical and cold stimuli, and also particularly in inflammation-induced thermal and mechanical hyperalgesia (Abrahamsen et al., 2008; Akopian et al., 1999; Kerr et al., 2001; Laird et al., 2002; Nassar et al., 2005). In neuropathic pain, the central role of this channel is less certain, since neuropathic pain is normal in the absence of NaV1.8-expressing neurons (Abrahamsen et al., 2008), although neuropathic pain can be attenuated by antisense oligonucleotides and siRNA (Joshi et al., 2006; Lai et al., 2002; Porreca et al., 1999; Dong et al., 2007), and by compounds relatively selective against NaV1.8 or TTX-resistant currents (Stummann et al., 2005; Jarvis et al., 2007). Functional properties of rat NaV1.8 channels have been investigated in oocytes (Akopian et al., 1996; Sangameswaran et al., 1996; Vijayaragavan et al., 2004), COS-7 cells (Fitzgerald et al., 1999) and ND7/23 cells (Choi et al., 2004; John et al., 2004; Leffler et al., 2007). Human NaV1.8 channels have been studied in oocytes (Rabert et al., 1998), CHO cells (Akiba et al., 2003) and SH-SY5Y cells (Dekker et al., 2005). However, no studies have been carried out comparing the properties of human and rat NaV1.8 channels (or indeed comparing any other NaV subtype) in mammalian cells under the same experimental conditions. Such comparisons are important since it cannot be assumed that the functional roles of closely

906

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

related orthologues are conserved across species, and differences in properties can have an important impact on drug discovery. Indeed, despite sharing considerable sequence identity and nearly identical S6 regions where pore-blocking drugs appear to act (Catterall, 2002; Nau and Wang, 2004), block of human NaV1.8 channels by A-803467 appears significantly more potent than for native rat TTX-resistant channels (Jarvis et al., 2007). We have carried out a detailed comparison of the electrophysiological properties of human and rat NaV1.8 channels expressed in a dorsal root ganglion-derived mammalian cell line, ND7/23. We have also compared the pharmacological actions on human and rat channels of five different known sodium channel blockers: the local anaesthetic tetracaine, the a-aminoamide derivative ralfinamide (with analgesic properties, Stummann et al., 2005), the lamotrigine derivative 227c89 (also with analgesic properties, Liu et al., 2003), the potent anticonvulsant V102862 (with potential utility as an analgesic, Ilyin et al., 2005), and the potent and highly selective NaV1.8 channel inhibitor A-803467 (also with analgesic properties, Jarvis et al., 2007). These blockers were chosen to further understand their action on the NaV1.8 channel, which is known to be involved in pain signalling. 2. Methods 2.1. Human and rat NaV1.8 channel expression in ND7/23 cells The a-subunit cDNAs used in this study were human NaV1.8 (Accession No. Q9Y5Y9, polymorph 1073V Rabert et al., 1998) and rat NaV1.8 (Accession No. Q62968, polymorphs 1896I and 1901D Akopian et al., 1996; Sangameswaran et al., 1996). Both cDNAs were cloned at GlaxoSmithKline and subcloned in pFastBacMam1 vectors (Condreay et al., 1999). ND7/23 neuroblastoma cells were obtained from ECACC (Salisbury, UK) and cultured in Dulbecco’s modified Eagle medium supplemented with 2 mM L-glutamine, 10% heat-inactivated foetal bovine serum and 1 non-essential amino acids. Cells were seeded in a 35 mm dish at 60% confluence and co-transfected the following day with 3.0 mg NaV1.8 channel cDNA and 0.3 mg EBO-pCD-Leu2 cDNA (for CD8 marker) or pEGFP-N1 cDNA (for fluorescence marker) using Lipofectamine 2000 (Invitrogen). Transfection-positive cells were analysed 2–4 days after transfection and identified with immunobeads (anti-CD8 Dynabeads; Invitrogen) or green fluorescence. 2.2. Electrophysiological recording and data analysis Recordings of NaV1.8 currents were obtained in the whole-cell configuration of the patch clamp technique at room temperature (20–22  C). Patch pipettes were fabricated from thin-walled borosilicate glass capillaries (GC150TF-10; Harvard Apparatus) and coated with Sigmacote (Sigma–Aldrich) to give typical tip resistances of 1.5–2.5 MU. The internal solution was composed of (mM): CsF, 120; HEPES, 10; EGTA, 10; and NaCl, 15 adjusted to pH 7.2 with CsOH. As in many previous studies of NaV1.8 currents in cultured cells and TTX-resistant currents in DRG cells (John et al., 2004; Leffler et al., 2007; Zhao et al., 2007; Elliott and Elliott, 1993; Choi et al., 2006; Dekker et al., 2005; Cardenas et al., 2006), we used fluoride in the internal solution. During experiments, cells were continuously perfused with an external solution containing (mM): NaCl, 140; HEPES, 5; MgCl2, 1.3; CaCl2, 1; Glucose, 11; and KCl, 4.7 adjusted to pH 7.4 with NaOH. Endogenous Naþ channel currents were blocked with TTX (200 nM, Tocris). Tetracaine (4-(butylamino)benzoic acid 2(dimethylamino)ethyl ester) was obtained from Sigma, and compounds A-803467 (5-(4-chlorophenyl)-N-(3,5-dimethoxyphenyl)-2-furancarboxamide), V102862 (4-(4-fluorophenoxy)benzaldehyde semicarbazone), ralfinamide ((S)-(-i-)-2-(4-(2fluorobenzoxy)-benzylamino)-propanamide) and 227c89 (5-(2,6-dichlorophenyl)6-methylpyrimidine-2,4-diamine) were all synthesised at GlaxoSmithKline. All compounds were prepared in dimethyl sulphoxide (DMSO) and diluted to the desired concentration in the external solution giving a final concentration of 1% DMSO. Control recordings were made followed by drug application for 3–4 min before test recordings were made in the presence of drug. After forming the whole-cell configuration, cells were allowed to equilibrate for 10 min, during which time the current amplitude increased, as also reported by John et al. (2004). This increase was appreciable during the first few minutes, but then slowed to about 1% per minute. Kinetics were not measured during this initial period, but did not appear to change when compared by eye in preliminary experiments. After the 10 min equilibration, all protocols were carried out at the same time for each cell. Currents were acquired using an Axopatch-1C or Axopatch 200B amplifier (Molecular Devices), filtered at 5 kHz and sampled at 10 kHz. Series resistance was compensated by 80–85%, and linear leak and capacitance currents

were subtracted using an online P/4 subtraction procedure where appropriate. Voltage pulses were applied, and data were obtained using pClamp8 or pClamp9 software (Clampex; Molecular Devices). Offline data analysis was performed using Clampfit v9 or 10 (Molecular Devices) and Origin v5.0 (MicroCal Inc.). Conductance–voltage curves were derived from the peak sodium current (INa), according to the equation: G ¼ INa/(V  ENa), where V is the test voltage and ENa is the reversal potential (ENa was obtained separately for each cell from the I–V curve). Conductances were fit with the Boltzmann expression G ¼ Gmax/(1 þ exp((V1/2  V)/ k)), where V1/2 is the midpoint voltage of the curve, k is the slope factor and Gmax is maximum conductance. For inactivation, test currents following a 4-s conditioning prepulse (holding potential 120 mV) were fit with the Boltzmann expression I ¼ (B  A)/(1 þ exp((V  V1/2)/k)) þ (A), where V is the test voltage, V1/2 is the midpoint voltage of the curve, k is the slope factor, B is the maximum current and A is the amplitude of a non-inactivating component. Inactivation rate time constants were obtained by fitting the decay phase of individual currents, and by fitting test pulse currents in twin pulse protocols (Fig. 2A); usually double exponential fits were carried out. Recovery from inactivation was also studied in twin pulse experiments (Fig. 2B) and again test currents were fitted with double exponentials; for some of the drugs used, a third exponential was required. To directly study the binding of test compounds to resting and inactivated channel states, the dissociation constants for resting channels (Kr) and inactivated channels (Ki) were determined according to the model of Kuo and Bean (1994) as described by Liu et al. (2003). The resting state dissociation constants were calculated as Kr ¼ [D]/((1/IDr)  1), where D is the concentration of test compound and IDr is the fraction of sodium current not blocked by the compound in the resting state. The protocol for obtaining inactivated state dissociation constants is shown in Fig. 3C; the dissociation constants were obtained as Ki ¼ [D](1  h)/((1/IDi)  1) where IDi is the fraction of inactivated sodium channels not blocked by test compound and h is the fraction of non-inactivated sodium current following the 4 s depolarisation (Fig. 3C) before application of the compound. The calculated apparent dissociation constants assume a 1:1 binding. Data are presented as mean  standard error, and the Student’s t test was used to test significance.

3. Results 3.1. Voltage-dependence of activation of human and rat NaV1.8 channels In order to compare their functional properties, human and rat NaV1.8 channels were transiently transfected into ND7/23 cells, a hybrid cell line derived from the fusion of rat DRG neurons with mouse N18Tg2 neuroblastoma cells (Wood et al., 1990). The TTXinsensitive NaV1.8 currents (Fig. 1A) were examined in the presence of TTX (200 nM) in order to inhibit endogenous Naþ currents. The I–V curves were very similar for human and rat NaV1.8 channels, with similar reversal potential. Indeed, when the conductance– voltage curves were fitted (Fig. 1C), the V1/2 was not significantly different between the two channels (Table 1), although there was a very small difference in k values (3.0 mV) between the two channels (Fig. 1C, Table 1). Similar findings were also observed from a more physiological holding potential (80 mV) (data not shown). Thus the voltage-dependence of activation was essentially similar for the two channels. Finally, to check for the effects of b subunits, human b1 or b3 subunits were also expressed. The presence of these subunits did not significantly affect the Boltzmann parameters of inactivation for human NaV1.8 (Table 1), indicating lack of effect of these subunits. 3.2. Inactivation properties of human and rat NaV1.8 channels The voltage-dependence of steady-state inactivation of human and rat NaV1.8 channels was compared using a 4 s prepulse to various potentials followed by a test pulse to 0 mV (holding potential of 120 mV, Fig. 1C). The steady-state inactivation curves show that the human channel was inactivated at more negative potentials than for the rat channel; the V1/2 for inactivation for hNaV1.8 was significantly (p < 0.001) more hyperpolarised (by 15.8 mV) than for rNaV1.8 (Table 1). As can be seen from Fig. 1C, at the more physiological resting potential of 80 mV, considerable inactivation (w50%) is already present for the human channel. Thus

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

A

907

0.5

B Human NaV1.8

Rat NaV1.8 V (mV) -50

50

1nA

1nA 2ms

I/Imax

-100

-0.5 2ms

-1.0

C

D 1.0

8

*

0.5

0.5

τi (ms)

6

G/Gmax

Normalised current

1.0

*

4

* 2

0.0

0.0 -120 -100 -80 -60 -40 -20

0

20

40

*

*

*

0 -10

0

10

20

30

40

V (mV)

V (mV)

Fig. 1. Electrophysiological properties of human and rat NaV1.8 channels. (A) Examples of whole-cell patch clamp recordings of human and rat NaV1.8 channel currents from ND7/23 cells in the presence of TTX (200 nM). Currents were elicited by voltage steps (100 mV to þ60 mV in 10 mV increments) from a holding potential of 120 mV. (B) Current–voltage (I–V) relationships are shown for the peak current for human (C, n ¼ 79) and rat (B, n ¼ 57) NaV1.8 channels and normalised to the maximum current (Imax). (C) Conductance– voltage curves for human (C, n ¼ 79) and rat (B, n ¼ 57) were calculated from I–V relationships, fit with the Boltzmann equation and normalised to Gmax (see Methods). Curves for steady-state voltage-dependence of inactivation for human (:, n ¼ 84) and rat (D, n ¼ 51) NaV1.8 channels are shown with normalised currents (see Methods) fit with the Boltzmann equation. Pulse protocols are shown in the insets. (D) The time course of inactivation, si, is shown for human (C, n ¼ 56) and rat (B, n ¼ 50) NaV1.8 channels for the test potentials, V, shown. *, p < 0.001.

all our reported experiments were conducted at 120 mV holding potential. It is noteworthy that, at 80 mV holding potential in our experiments, the value of V1/2 for rNaV1.8 was 58 mV, which is a similar value to that of John et al. (2004) at 90 mV holding potential (54 mV). Our more negative value for V1/2 for rNaV1.8 at 120 mV holding potential (64 mV) probably reflects the true value more closely than at 80 mV holding potential, where some resting inactivation is already present to distort the results. For voltage steps of 15 ms duration from a holding potential of 120 mV, hNaV1.8 currents showed faster inactivation time course as compared to rNaV1.8 (sample traces, Fig. 1A), and indeed the inactivation time course (Fig. 1D) showed that inactivation was significantly faster for hNaV1.8 for all potentials examined. The inactivation time course was examined over relatively short 15 ms pulses, and although they were better fitted with two exponentials, any component with much slower time course was too small in amplitude to be measured. Therefore, the time course of development of inactivation was further studied by examining the current following a 0 mV prepulse of varying length, a 4 ms recovery

Table 1 Voltage-dependent activation and inactivation. The Boltzmann parameters V1/2 and k are shown for activation and inactivation curves. NaV1.8 channel

Human Rat Human þ hb1 Human þ hb3

Activation

Inactivation

V1/2 (mV)

k (mV)

n

V1/2 (mV)

k (mV)

n

2.7  0.8 2.4  0.8

14.7  0.3 11.7  0.3 ND ND

79 57

79.9  1.0 64.1  1.2 79.0  4.5 75.3  5.7

10.5  0.4 8.0  0.4 9.0  1.2 11.1  1.9

84 51 5 4

period, and a test pulse to 0 mV (protocol as in Fig. 2A). Detailed fits to the test current with two exponential time constants showed that both fast and slow time constants were smaller for hNaV1.8 than for rNaV1.8 (Fig. 2A and C), so that inactivation developed significantly faster for hNaV1.8. Also, the human channel showed relatively larger amplitude of the fast component. For both channels, inactivation is complete by 4 s, so that inactivation is indeed steady-state by this time. Recovery from inactivation was investigated using a similar protocol as in Fig. 2A but with varying time between prepulse and test pulse (Fig. 2B). The fast component of recovery occurred with time constant of around 5–10 ms, indicating recovery from the fast inactivation state. Recovery from this fast inactivation state was 1.7fold slower (Fig. 2D) for the human channel. However, by contrast, at later times (presumably corresponding to the slow component of inactivation) recovery from inactivation was faster for the human channel (Fig. 2B and D). The relative amplitudes of fast and slow components of recovery were similar between rat and human channels (Fig. 2D). Use-dependent current inhibition during a train of pulses arises from the accumulation of inactivated channels until the rate of development of inactivation and the recovery from inactivation reaches equilibrium. Upon application of a train of pulses at 10 Hz (10 ms pulses to 0 mV from a holding potential of 120 mV), at steady-state, current inhibition was 12  1% (n ¼ 121) of the initial pulse for human NaV1.8, but 25  1% (n ¼ 86) for the rat channel. Thus, to summarise, in contrast to activation, where similar properties were observed for human and rat NaV1.8, the voltagedependence of steady-state inactivation, its time course and recovery are all different between human and rat channels.

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

B

1.0 2-4000 ms 0 mV

Normalised current

0mV

-120 mV

-120 mV 4 ms

0.5

1.0 0 mV 1.0

0.5

0.0 10 20 30 40 Recovery time (ms)

2000

3000

4000

0

500

0

HumanRat

0

*

HumanRat

500

0

1500

D

*

60

HumanRat

40

80

*

20 0

HumanRat

400

10

5

0

* HumanRat

Ifast (%)

5

50

τslow (ms)

10

Ifast (%)

τfast (ms)

*

1000

τfast (ms)

100

15

Islow (%)

C

1000

2000

Recovery time (ms)

Prepulse length (ms)

40

0

HumanRat

*

200

0

40

Islow (%)

1000

-120 mV

0.5

0.0 0

0 mV

1-2000 ms

0

0.0

50 ms

-120 mV

τslow (ms)

Normalised current

A

Normalised current

908

HumanRat

20

0

HumanRat

Fig. 2. Development and recovery of inactivation for human and rat NaV1.8 channels. (A) Time-course for the development of inactivation for human (C, n ¼ 5) and rat (B, n ¼ 5) NaV1.8 channels. To generate the curves the current amplitude during the test pulse (second pulse) was normalised as a fraction of the current amplitude during the control pulse (first pulse), and plotted against the duration of the prepulse. The curve was fit with a double exponential equation. The pulse protocol is shown inset. (B) Time-course for the recovery from inactivation for human (C, n ¼ 71) and rat (B, n ¼ 55) NaV1.8 channels. The pulse protocol is shown inset. The current amplitude during the test pulse was normalised as a fraction of the current amplitude during the control pulse, and plotted against the duration of the recovery period. The curve was fit with a double exponential equation. (C) Bar diagrams of the double exponential fit parameters for the development of inactivation curves, showing time constants sfast and sslow and the respective amplitudes Ifast and Islow, for human (filled bars) and rat (unfilled bars) NaV1.8 channels. (D) Bar diagrams of the double exponential fit parameters for the recovery from inactivation curves, showing time constants sfast and sslow and the respective amplitudes Ifast and Islow, for human (filled bars) and rat (unfilled bars) NaV1.8 channels. Parameters Ifast and Islow shown in both C and D are the relative % of the fast and slow components.

3.3. Affinities of tetracaine, 227c89, ralfinamide, V102862 and A-803467 for resting and inactivated channels Here we study the binding affinity of a range of drugs for resting and inactivated NaV1.8 channels. The structures for the test compounds 227c89, ralfinamide, tetracaine, V102862 and A-803467 are shown in Fig. 3A. The binding affinity for the resting state was investigated from a holding potential of 120 mV, from which virtually all channels are in the resting state. For this, peak current was measured during single 10 ms test pulses to 0 mV in the absence and presence of test compound, and the ratio of these, IDr, was determined, and hence resting binding affinities obtained. The concentrations chosen for each compound were such as to produce approximately 10–60% block. The values for concentrations of test compound used and IDr (see Methods) are given in Table 2, and the calculated binding affinities are shown in Fig. 3B. Compounds 227c89 and ralfinamide were the least potent with rather similar affinities (Kr approx. 80–130 mM). Compound A-803467 had the highest resting state affinity (Kr approx. 130 nM), while tetracaine and compound V102862 had intermediate dissociation constants (Fig. 3B). The binding affinity for resting channels was similar between hNaV1.8 and rNaV1.8 for 227c89, ralfinamide and A-803467. In contrast, tetracaine and V102862 showed significantly greater affinity for resting hNaV1.8 channels compared to rNaV1.8 channels (Fig. 3B). In order to measure affinity constants for test compounds for the inactivated state, a twin pulse protocol (Fig. 3C) was used comprising a control pulse, a 4 s period of inactivation at a depolarised potential (chosen so as to give approx. 60–80% inactivation) followed by a test pulse. These twin pulses were applied before and after application of test compound. An example trace is shown in Fig. 3D; bold traces show the control and test currents before drug action and the finer traces are control and test currents after drug

application. For the calculation (see Methods) of inactivated state affinities (obtained as dissociation constants, Ki) from IDi and h, the value of IDi was calculated from the ratio of test currents before and after application of test compound, and h was calculated from the ratio of control and test currents before drug action (Table 3). Inactivated state affinities (Fig. 3E) for all compounds tested were higher than for the resting states. This can also be seen from the sample trace in Fig. 3D, where the concentrations of tetracaine used gave negligible effect for the control pulse but a large inhibition for the test pulse. The affinities for inactivated states showed a similar rank order as for resting state affinities, where 227c89 and ralfinamide were the least potent (Ki 10–15 mM) and A-803467 showed the highest potency (Ki 15 nM for hNaV1.8 and 43 nM for rNaV1.8) (Fig. 3E). Inactivated state affinities were similar between human and rat forms of the channel, except for compound A-803467, which showed greater affinity for the human channel. Ratios of inactivated state affinities to resting affinities were all in the range 5- to 9-fold for the human channel, whereas for the rat channel this ratio was lower for A-803467 (4-fold), somewhat higher for ralfinamide (12-fold) and higher still for tetracaine and compound V102862 (18-fold) (Fig. 3B and E). These results therefore suggest differences in drug binding properties between the human and the rat forms of the channel. 3.4. Effects of drug block on the recovery from inactivation Recovery from inactivation is expected to be altered by drugs that preferentially bind to inactivated states. In these experiments, this was examined using the same protocol as in Fig. 2B (and shown in Fig. 4F) used before and after the application of test compound. As discussed above, recovery from inactivation in the absence of inhibitor involved an initial fast component (time constant 6–10 ms) followed by a slower component (time constant

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

A

227c89

Ralfinamide

Tetracaine

A-803467

V102862

B

40

200

200

200

3000

*

* 2000

20

100

100

Kr (nM)

Kr (μ μM)

909

100 1000

0

0

227c89

0

Ralfinamide

C

0

Tetracaine

A-803467

Tetracaine (1μM)

D 0 mV

0

V102862

0 mV 4s

-120 mV

60-80% inactivated

0.5 nA

-120 mV

2 ms

E

15

200

2

20

Ki (nM)

Ki (μM)

10 1

10

60

40 100 20

5

0

*

227c89

0

Ralfinamide

0

Tetracaine

0

V102862

0

A-803467

Fig. 3. Dissociation constants for each drug for the resting and inactivated states. (A) The structures of the compounds 227c89, ralfinamide, tetracaine, V102862 and A-803467. (B) Bar diagrams are shown representing resting state dissociation constants (Kr) for each drug for human (filled bars) and rat (unfilled bars) NaV1.8 channels. Values were calculated using concentrations which gave 10–60% resting state block; 227c89 (10 mM, hNaV1.8, n ¼ 5, rNaV1.8, n ¼ 3), ralfinamide (50 mM, hNaV1.8, n ¼ 3, rNaV1.8, n ¼ 5), tetracaine (10 mM, hNaV1.8, n ¼ 6, rNaV1.8, n ¼ 3), V102862 (1 mM, hNaV1.8, n ¼ 6, rNaV1.8, n ¼ 5) and A-803467 (100 nM, hNaV1.8, n ¼ 5, rNaV1.8, n ¼ 5). (C) The pulse protocol is shown; this was used before and after the application of a drug to determine inactivated state affinities (Ki) (the 4 s pulse indicated was taken at membrane potentials of 50 to 80 mV such that 60–80% inactivation occurred). (D) Example current traces are shown using the pulse protocol in C, where currents in bold are before the application of drug and fine traces are after the application of drug; the currents larger in magnitude correspond to the control pulses and currents smaller in magnitude correspond to the test pulse following a 4 s depolarisation. (E) Bar diagrams are shown representing inactivated state dissociation constants (Ki) for each drug for human (filled bars) and rat (unfilled bars) NaV1.8 channels. Values were calculated using concentrations which gave appreciable inactivated state block; 227c89 (10 mM, hNaV1.8, n ¼ 5), ralfinamide (10 mM, hNaV1.8, n ¼ 5, rNaV1.8, n ¼ 4), tetracaine (1 mM, hNaV1.8, n ¼ 7, rNaV1.8, n ¼ 8), V102862 (1 mM, hNaV1.8, n ¼ 6, rNaV1.8, n ¼ 5) and A-803467 (10 nM, hNaV1.8, n ¼ 7, rNaV1.8, n ¼ 3). The Ki value for compound 227c89 for rNaV1.8 channel was not done. *p < 0.05.

910

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

Table 2 Resting block of NaV1.8 channels. The table shows the values of IDr (see Methods) for each drug at the concentration used. NaV1.8 channel

Drug

Concentration (mM)

IDr

n

Human Rat Human Rat Human Rat Human Rat Human Rat

227c89 227c89 Ralfinamide Ralfinamide Tetracaine Tetracaine V102862 V102862 A-803467 A-803467

10 10 50 50 10 10 1 1 0.1 0.1

0.87  0.02 0.91  0.03 0.69  0.08 0.70  0.03 0.44  0.07 0.73  0.03 0.42  0.02 0.67  0.02 0.51  0.09 0.64  0.02

5 3 3 5 6 3 6 5 5 5

120–350 ms). Compound 227c89 (10 mM) did not substantially affect recovery from inactivation as compared with untreated channels (Figs. 4A and 5), even though this concentration produced substantial inactivated state block (as above). Ralfinamide (50 mM) showed small differences from untreated recovery time course (Fig. 4B), mainly due to a slowing of the fast component and an increase in the relative amplitude of the slow component (Fig. 5). Tetracaine (10 mM) gave marked slowing of recovery from inactivation (Fig. 4C), caused by slowing of both fast and slow components, as well as increased amplitude of the slow component (Fig. 5), effects particularly marked for the rat channel. Compound V102862 (1 mM) markedly increased the relative amplitude of the slow component and slowed its time course, but in the case of the human NaV1.8 channel there was also an additional slower component that gave a surprising increase (‘‘disinhibition’’) in current above the initial control value (Fig. 4D). In this case a fit with three-exponential time constants was required (see Methods). The additional third exponential component had an amplitude of 30  7% (of final recovered value) and slow decay time constant of 1070  230 ms (n ¼ 5). In the continual presence of the compound, there is a resting current block, and the effect of the drug during stimulation may be to remove some of this resting block; the third component of the time course may then represent the reinstatement of the resting block. The amplitude of 30% for this component compares with the resting current block of 58% for the human channel (Table 2), suggesting substantial removal of resting block by stimulation in the presence of V102862. The rat form of the channel did not show disinhibition. As for V102862, compound A-803467 (100 nM, Fig. 4E) also showed an additional slow component with marked increase in current for the human channel (though not obvious for the rat channel), again requiring a three-exponential fit of the time course of recovery. This additional component had an amplitude of 22  3% and time constant of 1670  180 ms (n ¼ 5). For A-803467 (100 nM), resting current block was 49  9% for the human channel, so that again the third component is consistent with substantial

removal of resting block during stimulation, followed by slow recovery. While at first glance this might seem to contradict the finding that both V102862 and A-803467 have higher affinities for the inactivated state than the resting state, the disinhibitory state may occur just after recovery from the inactivated states, and seemingly not while the channels are in the same higher affinity state. This disinhibition of current seen for A-803467 and V102862 is an unexpected result and has not been previously observed for other inhibitors acting on sodium channels. 3.5. Comparison of use-dependent properties of tetracaine, 227c89, ralfinamide, V102862 and A-803467 Examples of mean currents recorded during trains of 0 mV depolarising pulses at 10 Hz are shown in Fig. 6A for tetracaine and A-803467; mean currents for the last pulse of the trains are also shown as bar diagrams in Fig. 6C. It can be seen that tetracaine showed use-dependent block with rather slow onset. By contrast, compound A-803467 did not produce a use-dependent block in rat, and in the case of the human channel even increased the currents by some 18% (Fig. 6A and C), which may occur by removal of resting block, as discussed above. The lack of use-dependent block in the case of the human channel is consistent with a previous report (Jarvis et al., 2007), and indeed supplementary data in the latter paper suggests a small increase in current. By contrast, compound V102862 showed substantial block at 10 Hz (Fig. 6B and C), with fast onset. A similar extent of block was produced by compound 227c89 for the human channel, but significantly less for the rat channel. Ralfinamide showed little use-dependent block (Fig. 6C), although block was greater (18% for hNaV1.8 and 14% for rNaV1.8) at a holding potential of 80 mV (data not shown), suggesting the action of ralfinamide is increased when a greater fraction of inactivated channels is present in the recovery periods. At 5 Hz train frequency, the compounds generally had less effect; most notably there was a removal of inhibition for V102862, which may reflect a ‘‘disinhibitory’’ component at this lower stimulation rate (Fig. 6B and D). 4. Discussion We have compared the properties of human and rat NaV1.8 channels under the same conditions in a mammalian expression system. We have shown that activation properties are similar for the two forms of the channel, whereas inactivation properties differ. For instance, human NaV1.8 inactivated at 16 mV more hyperpolarised potential than the rat channel. This suggests that, in nociceptive dorsal root ganglion neurons where these channels are predominantly expressed, a larger proportion of NaV1.8 channels will be inactivated at normal cell resting potentials in humans as compared to rats.

Table 3 Block of inactivated NaV1.8 channels. The table shows IDi and h (see Methods) for each drug at the concentration used, together with the calculated inactivated state dissociation constant. NaV1.8 channel

Drug

Concentration (mM)

IDi

h

n

Human Human Rat Human Rat Human Rat Human Rat

227c89 Ralfinamide Ralfinamide Tetracaine Tetracaine V102862 V102862 A-803467 A-803467

10 10 10 1 1 1 1 0.01 0.01

0.61  0.05 0.60  0.08 0.56  0.04 0.62  0.05 0.59  0.09 0.16  0.03 0.15  0.02 0.63  0.06 0.83  0.06

0.32  0.03 0.31  0.04 0.24  0.01 0.28  0.05 0.31  0.05 0.28  0.03 0.36  0.04 0.34  0.03 0.35  0.04

5 5 4 7 8 6 5 7 3

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

911

B

A

Ralfinamide (50μM) Normalised current

Normalised current

227c89 (10μM) 1.0

0.5

1.0

0.5

0.0

0.0 0

500

1000

1500

2000

0

500

1000

1500

2000

Recovery time (ms)

Recovery time (ms)

C

D V102862 (1μM) Normalised current

Normalised current

Tetracaine (10μ μM) 1.0

0.5

0.0

1.0

0.5

0.0 0

500

1000

1500

2000

0

500

Recovery time (ms)

E

A-803467 (100nM) Normalised current

1000

1500

2000

Recovery time (ms)

F

1.0 0 mV

50 ms

-120 mV

0.5

0 mV -120 mV

1-2000 ms

0.0 0

500

1000

1500

2000

Recovery time (ms) Fig. 4. The effects of drugs on the recovery from inactivation. The figures (A–E) show the test pulse amplitude normalised to control pulse during the recovery from inactivation, using the protocol shown in (F). Time courses of recovery from inactivation are shown before (hNaV1.8 C, n ¼ 5–6, rNaV1.8 B, n ¼ 3–6) and after (hNaV1.8 :, rNaV1.8 D) drug application in paired cells. (A) 227c89 (10 mM), (B) ralfinamide (50 mM), (C) tetracaine (10 mM), (D) V102862 (1 mM), and (E) A-803467 (100 nM).

Comparison of our V1/2 values for inactivation with previous work is not easy because of the wide range of experimental conditions used in previous work, particularly the differing holding potential and pulse protocols used. Among the previous publications on cultured cells or dorsal root ganglion cells, all used fluoride for the internal solution, although for a range of concentrations (Elliott and Elliott, 1993; Akiba et al., 2003; John et al., 2004; Dekker et al., 2005; Cardenas et al., 2006; Choi et al., 2006; Leffler et al., 2007; Zhao et al., 2007). Whether for the rat or human channel, our values for V1/2 are generally more negative than previouslyobtained values. However, as already noted, at 80 mV holding potential, for the rat channel our value for V1/2 (58 mV) agrees with that of John et al., 2004, and our more negative value (64 mV) at 120 mV holding potential simply reflects the presence of appreciable inactivation at 80 mV holding potential, indicating distortion of V1/2 values obtained at the less hyperpolarised holding potential in previous work. The strength of the present work is due to the fact that we have compared rat and human channels under identical experimental conditions,

eliminating the effects of pulse protocols and other experimental conditions. Under the identical conditions that we have used, it is clear that inactivation of the human NaV1.8 channel occurs at more negative potentials than for the rat channel. Even for recordings from 2-electrode voltage-clamped oocytes (where values for V1/2 were again less negative than we have found), the value for V1/2 for the human channel (Rabert et al., 1998) was left shifted as compared with the value for the rat channel (Akopian et al., 1996). Both initial and later components of inactivation developed with a faster time course for the human channel as compared with rat (Fig. 2C). For the recovery from inactivation, while the initial time course was slower, the later time course was faster for the human channel (Fig. 2D). Since the extent of inhibition during a train of depolarising pulses is dependent on the rate of development of inactivation and the recovery from inactivation, it would be expected that inhibition would be different between human and rat NaV1.8 channels. Indeed, the human NaV1.8 channel showed less pronounced inhibition following 10 Hz stimulation than the rat channel. In nociceptive C-fibres, firing at around this frequency

912

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

A

1.5

Ifast/Islow

1.0

0.5

0.0

Normalised τfast

B

227c89

Ralfinamide Tetracaine V102862

A-803467

227c89

Ralfinamide Tetracaine

V102862

A-803467

227c89

Ralfinamide Tetracaine

V102862

A-803467

4

2

0

C

Normalised τslow

20

10

0

Fig. 5. The effects of drugs on the recovery from inactivation parameters. The parameters of double exponential fits to the recovery time course are shown for 227c89 (10 mM, hNaV1.8, n ¼ 5, rNaV1.8, n ¼ 3), ralfinamide (50 mM, hNaV1.8, n ¼ 5, rNaV1.8, n ¼ 4), tetracaine (10 mM, hNaV1.8, n ¼ 6, rNaV1.8, n ¼ 3), V102862 (1 mM, hNaV1.8, n ¼ 5, rNaV1.8, n ¼ 6), A-803467 (100 nM, hNaV1.8, n ¼ 5, rNaV1.8, n ¼ 4) for human (filled bars) and rat (unfilled bars) NaV1.8 channels. In the case of V102862 and A-803467 fits were made using three exponentials; here the parameters of the slowest component have been omitted (but given in the text). (A) Ratios of fast to slow components, (B) fast time constants, and (C) slow time constants. All values were normalised in paired untreated cells. The fast component time course could not be obtained for V102862 as the amplitude of this component was too small for analysis.

occurs during pain signalling and so this lower level of inhibition in human channels would allow sustained firing (Akopian et al., 1996; Elliott and Elliott, 1993; Jeftinija, 1994; Kajander and Bennett, 1992; Sangameswaran et al., 1996; Schild and Kunze, 1997). Furthermore, because of these differences between rat and human channels, the use of rat models for pain may not exactly reflect the situation in humans. It is not clear which regions of the channel protein structure are important in determining the differences in inactivation properties between rat and human NaV1.8 channels. The transmembrane segments, P-loops and intracellular linkers have all been shown to have roles in inactivation (Ulbricht, 2005), so perhaps some of the differences between human and rat forms of the channel may reside in these regions. The ND7/23 cell line is derived from rat and mouse cells and thus endogenous modulatory proteins potentially may contribute to differences in properties between human and rat channels by species-specific interactions. For instance, ND7/23 cells endogenously express b1 and b3 subunits (John et al., 2004), which might in principle affect channel gating properties. However, co-transfection of human b1 or b3 subunits did not significantly affect the Boltzmann parameters of inactivation for human NaV1.8 (however it is not possible to exclude the possibility that endogenous expression of b subunits is maximal, so that extrinsic expression would have no effect). As mentioned in the Introduction, previous studies have not been carried out comparing properties of rat and human NaV channels in mammalian cells in the same experimental study. In one study, comparison was carried out in oocytes (Rabert et al., 1998) but, perhaps because the study was not extensive, it was not possible to discriminate clearly between the properties of human and rat channels, although the trend of results in the latter paper was similar to ours for voltage-dependence of activation and inactivation. Moreover, Akiba et al. (2003) also noted that human NaV1.8 channels expressed in CHO cells appear to have more hyperpolarised inactivation than rat NaV1.8 from other studies. Thus, it seems unlikely that the use of ND7/23 cells contributed to the observed differences between human and rat NaV1.8 channels, and indeed results are more likely relevant to physiological cases since we have used a mammalian cell line. We have tested a range of drugs on the NaV1.8 channel. Affinities for the resting state of the compounds tested were in the order ralfinamide < 227c89 < tetracaine < V102862 < A-803467 for both human and rat channels, although tetracaine and V102862 showed greater affinity for resting human NaV1.8 as compared with the rat channel (approx. 3-fold). Affinities for the inactivated state were always higher than for the resting state, although in the same rank order of affinity as for resting state for both human and rat channels. For the human channel, inactivated state affinities were in the range 5- to 9-fold more than resting, whereas there was a wider range (4- to 18-fold more than resting) for the rat channel, suggesting differences in binding properties between the two channels. Of particular note, tetracaine and V102862 showed a much greater differential between resting and inactivated state block for rat NaV1.8 (18-fold) as compared to the human channel (5-fold). The resting state affinity of tetracaine for rat NaV1.8 channels is comparable to the value reported by John et al. (2004) under more depolarised conditions (holding potential of 90 mV). On the other hand, Dekker et al. (2005) suggested that tetracaine is 16-fold more potent for human NaV1.8 channels than for rat channels, apparently in contradiction to our results (2-fold more potent in the resting state and almost equal for the inactivated state). However, in the fluorometric assay used by Dekker et al. (2005) in unclamped cells at normal resting potentials, human NaV1.8 channels will be substantially inactivated as compared with rat channels. Thus human NaV1.8 will bind tetracaine with the higher affinity

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

A

913

Normalised current

Normalised current

B 1.0

0.5

0.0

1.0

0.5

0.0 0

10

20

0

10

Pulse number

C

*

* *

*

*

0.5

0.0

227c89

Ralfinamide Tetracaine

V102862

D * Normalised current

Normalised current

* 1.0

20

Pulse number

A-803467

1.0

*

*

*

0.5

0.0

227c89

Ralfinamide Tetracaine

V102862

A-803467

Fig. 6. Use-dependent properties of each compound. (A) Normalised current amplitudes are shown plotted against pulse number during stimulation at 10 Hz (pulses of 10 ms duration to 0 mV from a holding potential of 120 mV) in the presence throughout of tetracaine (1 mM) (hNaV1.8 C, n ¼ 20, rNaV1.8 B, n ¼ 10) or compound A-803467 (100 nM) (hNaV1.8 :, n ¼ 12, rNaV1.8 D, n ¼ 4). (B) Normalised current is shown plotted against pulse number during stimulation at 5 Hz (using the same voltage protocol as in A) in the presence throughout of V102862 (1 mM) (hNaV1.8 C, n ¼ 5, rNaV1.8 B, n ¼ 5) or at 10 Hz (hNaV1.8 :, n ¼ 11, rNaV1.8 D, n ¼ 12). For A and B, currents were normalised to the first pulse and then to their respective control pulses before the application of drug. (C) Currents, normalised as above, are shown for the 20th pulse at 10 Hz for 227c89 (10 mM, hNaV1.8, n ¼ 4, rNaV1.8, n ¼ 5), ralfinamide (50 mM, hNaV1.8, n ¼ 7, rNaV1.8, n ¼ 8), tetracaine (1 mM, hNaV1.8, n ¼ 20, rNaV1.8, n ¼ 10), V102862 (1 mM, hNaV1.8, n ¼ 11, rNaV1.8, n ¼ 12) and A-803467 (100 nM, hNaV1.8, n ¼ 12, rNaV1.8, n ¼ 4) for human (filled bars) and rat (unfilled bars) NaV1.8 channels. (D) Currents, normalised as above, are shown for the 20th pulse at 5 Hz for 227c89 (10 mM, hNaV1.8, n ¼ 4, rNaV1.8, n ¼ 5), ralfinamide (50 mM, hNaV1.8, n ¼ 3, rNaV1.8, n ¼ 3), tetracaine (1 mM, hNaV1.8, n ¼ 10, rNaV1.8, n ¼ 6), V102862 (1 mM, hNaV1.8, n ¼ 5, rNaV1.8, n ¼ 5) and A-803467 (100 nM, hNaV1.8, n ¼ 6, rNaV1.8, n ¼ 4) for human (filled bars) and rat (unfilled bars) NaV1.8 channels. *p < 0.05, compared to the first pulse.

representative of inactivated state binding, leading to the apparent 16-fold selectivity for human over rat NaV1.8 observed in that study. Similar arguments would explain IC50 values obtained for ralfinamide in the fluorimetric study, which again are closer to inactivated binding affinities than to resting affinities. In agreement with our data, Stummann et al. (2005) reported an IC50 value of 10 mM for inactivated rat TTX-resistant channels for ralfinamide. Comparing our results for NaV1.8 channels with data for other NaV channel subtypes, tetracaine has a greater inactivated state binding affinity for NaV1.8 channels as compared to rNaV1.3 channels (Li et al., 1999), V102862 has greater affinity for NaV1.8 than for rNaV1.2 channels (Ilyin et al., 2005), and compound A-803467 showed approximately 100-fold greater affinity for human NaV1.8 than for human NaV1.2, NaV1.3, NaV1.5 and NaV1.7 channels (Jarvis et al., 2007). Ralfinamide has greater affinity for NaV1.8 channels as compared with TTX-sensitive channels (Stummann et al., 2005), while compound 227c89 showed similar inactivated state binding affinity to rNaV1.2 channels (Liu et al., 2003). Although the binding sites for all these inhibitors are expected to involve the alreadyidentified amino acid residues in the S6 segment (Catterall, 2002; Nau and Wang, 2004), the greater affinities of tetracaine, V102862, A-803467 and ralfinamide for NaV1.8 suggest differences in relative strengths of binding to these residues for the different NaV channels, despite the near identity of primary structure in the S6 region. The reasons for these differences require further investigation, as do the differences between effects of drugs on rat and human channels. Interestingly, the most potent NaV1.8 channel blockers studied here (V102862 and A-803467) share a striking effect on the

recovery from human NaV1.8 channel inactivation; recovery involved not simply removal of inactivation, but also a period where currents were disinhibited, i.e. larger than initial values. For rat NaV1.8 channels, this was not observed for V102862, and only a small effect for A-803467. Furthermore, such a phenomenon has never been reported for any blocking drugs for other NaV channel subtypes; for example disinhibition did not occur for V102862 at rat NaV1.2 channels (Ilyin et al., 2005). In cases where disinhibition occurs, it seems likely that the increased current on recovery is due to an appreciable disinhibition of the resting current during depolarising pulses in the presence of drug, since the initial control values include a resting drug block. Consistent with this mechanism, the extent of disinhibition did not exceed the extent of resting block. The mechanism for this effect is unclear, but perhaps recovery from inactivation in the presence of certain drugs involves the passage through states not previously observed, perhaps involving dissociation of drug from the resting state and/or shift in the resting state equilibrium. An alternative mechanism, though perhaps less likely, could be the induction of new states by the drug during recovery with greater channel conductance or open time. Further experimental work measuring single channel conductance would settle this point, and enable the construction of a statemodel to fully explain this unusual effect. This disinhibition by A-803467 during recovery from inactivation would be expected to lead to an increase in current during a high frequency train of stimuli, and this ‘‘reverse use-dependence’’ was indeed observed as an increase in current during 10 Hz stimulation. This suggests that compound A-803467 may be less useful than anticipated as an analgesic in humans where

914

L.E. Browne et al. / Neuropharmacology 56 (2009) 905–914

physiological frequencies are of this order (Kajander and Bennett, 1992). For compound V102862, although disinhibition of recovery was observed for human NaV1.8 channels, for high frequency trains no increase in current was observed. The likely reason for this is that, for V102862, the disinhibition is prevalent at later times of recovery than for A-803467. However, at lower frequency stimulation (5 Hz), there was much less inhibition for V102862, with the possibility of a component of disinhibition that partially compensated the background inhibition. One might hypothesise that, at lower frequencies, say around 2 Hz, disinhibition and increased currents would be seen for V102862, and so the compound would be selective for higher frequency firing nerve fibres, just as would be desired for an effective drug for use in pain. In summary, while the activation properties between human and rat NaV1.8 channels were similar, there were differences in inactivation properties between these channels. The compounds investigated here all showed greater affinity for the inactivated state than for the resting state of NaV1.8. Compounds A-803467 and V102862 were the most potent, and showed an interesting property of disinhibition during recovery from inactivation for the human, though not rat, form of the channel. Acknowledgments We thank Andrew Powell and Tim Dale for technical assistance and advice. This work was supported by the Biotechnology and Biological Sciences Research Council and GlaxoSmithKline, Stevenage. References Abrahamsen, B., Zhao, J., Asante, C.O., Cendan, C.M., Marsh, S., Martinez-Barbera, J.P., Nassar, M.A., Dickenson, A.H., Wood, J.N., 2008. The cell and molecular basis of mechanical, cold, and inflammatory pain. Science 321, 702–705. Akiba, I., Seki, T., Mori, M., Iizuka, M., Nishimura, S., Sasaki, S., Imoto, K., Barsoumian, E.L., 2003. Stable expression and characterization of human PN1 and PN3 sodium channels. Receptors Channels 9, 291–299. Akopian, A.N., Sivilotti, L., Wood, J.N., 1996. A tetrodotoxin-resistant voltage-gated sodium channel expressed by sensory neurons. Nature 379, 257–262. Akopian, A.N., Souslova, V., England, S., Okuse, K., Ogata, N., Ure, J., Smith, A., Kerr, B.J., McMahon, S.B., Boyce, S., Hill, R., Stanfa, L.C., Dickenson, A.H., Wood, J.N., 1999. The tetrodotoxin-resistant sodium channel SNS has a specialized function in pain pathways. Nat. Neurosci. 2, 541–548. Blair, N.T., Bean, B.P., 2002. Roles of tetrodotoxin (TTX)-sensitive Naþ current, TTXresistant Naþ current, and Ca2þ current in the action potentials of nociceptive sensory neurons. J. Neurosci. 22, 10277–10290. Cardenas, C.A., Cardenas, C.G., de Armendi, A.J., Scroggs, R.S., 2006. Carbamazepine interacts with a slow inactivation state of NaV1.8-like sodium channels. Neurosci. Lett. 408, 129–134. Catterall, W.A., 2002. Molecular mechanisms of gating and drug block of sodium channels. Novartis Found. Symp. 241, 206–218. discussion 218–232. Choi, J.S., Hudmon, A., Waxman, S.G., Dib-Hajj, S.D., 2006. Calmodulin regulates current density and frequency-dependent inhibition of sodium channel NaV1.8 in DRG neurons. J. Neurophysiol. 96, 97–108. Choi, J.S., Tyrrell, L., Waxman, S.G., Dib-Hajj, S.D., 2004. Functional role of the C-terminus of voltage-gated sodium channel NaV1.8. FEBS Lett. 572, 256–260. Condreay, J.P., Witherspoon, S.M., Clay, W.C., Kost, T.A., 1999. Transient and stable gene expression in mammalian cells transduced with a recombinant baculovirus vector. Proc. Natl. Acad. Sci. U.S.A. 96, 127–132. Dekker, L.V., Daniels, Z., Hick, C., Elsegood, K., Bowden, S., Szestak, T., Burley, J.R., Southan, A., Cronk, D., James, I.F., 2005. Analysis of human NaV1.8 expressed in SH-SY5Y neuroblastoma cells. Eur. J. Pharmacol. 528, 52–58. Dong, X.W., Goregoaker, S., Engler, H., Zhou, X., Mark, L., Crona, J., Terry, R., Hunter, J., Priestley, T., 2007. Small interfering RNA-mediated selective knockdown of NaV1.8 tetrodotoxin-resistant sodium channel reverses mechanical allodynia in neuropathic rats. Neuroscience 146, 812–821. Elliott, A.A., Elliott, J.R., 1993. Characterization of TTX-sensitive and TTX-resistant sodium currents in small cells from adult rat dorsal root ganglia. J. Physiol. 463, 39–56. Fitzgerald, E.M., Okuse, K., Wood, J.N., Dolphin, A.C., Moss, S.J., 1999. cAMPdependent phosphorylation of the tetrodotoxin-resistant voltage-dependent sodium channel SNS. J. Physiol. 516 (Pt 2), 433–446. Hille, B., 1977. Local anesthetics: hydrophilic and hydrophobic pathways for the drug-receptor reaction. J. Gen. Physiol. 69, 497–515. Hille, B., 2001. Ionic Channels of Excitable Membranes. Sinauer, Sunderland, MA.

Ilyin, V.I., Hodges, D.D., Whittemore, E.R., Carter, R.B., Cai, S.X., Woodward, R.M., 2005. V102862 (Co 102862): a potent, broad-spectrum state-dependent blocker of mammalian voltage-gated sodium channels. Br. J. Pharmacol. 144, 801–812. Jarvis, M.F., Honore, P., Shieh, C.C., Chapman, M., Joshi, S., Zhang, X.F., Kort, M., Carroll, W., Marron, B., Atkinson, R., Thomas, J., Liu, D., Krambis, M., Liu, Y., McGaraughty, S., Chu, K., Roeloffs, R., Zhong, C., Mikusa, J.P., Hernandez, G., Gauvin, D., Wade, C., Zhu, C., Pai, M., Scanio, M., Shi, L., Drizin, I., Gregg, R., Matulenko, M., Hakeem, A., Gross, M., Johnson, M., Marsh, K., Wagoner, P.K., Sullivan, J.P., Faltynek, C.R., Krafte, D.S., 2007. A-803467, a potent and selective NaV1.8 sodium channel blocker, attenuates neuropathic and inflammatory pain in the rat. Proc. Natl. Acad. Sci. U.S.A. 104, 8520–8525. Jeftinija, S., 1994. The role of tetrodotoxin-resistant sodium channels of small primary afferent fibers. Brain Res. 639, 125–134. John, V.H., Main, M.J., Powell, A.J., Gladwell, Z.M., Hick, C., Sidhu, H.S., Clare, J.J., Tate, S., Trezise, D.J., 2004. Heterologous expression and functional analysis of rat NaV1.8 (SNS) voltage-gated sodium channels in the dorsal root ganglion neuroblastoma cell line ND7-23. Neuropharmacology 46, 425–438. Joshi, S.K., Mikusa, J.P., Hernandez, G., Baker, S., Shieh, C.C., Neelands, T., Zhang, X.F., Niforatos, W., Kage, K., Han, P., Krafte, D., Faltynek, C., Sullivan, J.P., Jarvis, M.F., Honore, P., 2006. Involvement of the TTX-resistant sodium channel NaV1.8 in inflammatory and neuropathic, but not post-operative, pain states. Pain 123, 75–82. Kajander, K.C., Bennett, G.J., 1992. Onset of a painful peripheral neuropathy in rat: a partial and differential deafferentation and spontaneous discharge in Ab and Ad primary afferent neurons. J. Neurophysiol. 68, 734–744. Kerr, B.J., Souslova, V., McMahon, S.B., Wood, J.N., 2001. A role for the TTX-resistant sodium channel NaV1.8 in NGF-induced hyperalgesia, but not neuropathic pain. Neuroreport 12, 3077–3080. Kuo, C.C., Bean, B.P., 1994. Slow binding of phenytoin to inactivated sodium channels in rat hippocampal neurons. Mol. Pharmacol. 46, 716–725. Lai, J., Gold, M.S., Kim, C.S., Bian, D., Ossipov, M.H., Hunter, J.C., Porreca, F., 2002. Inhibition of neuropathic pain by decreased expression of the tetrodotoxinresistant sodium channel, NaV1.8. Pain 95, 143–152. Laird, J.M., Souslova, V., Wood, J.N., Cervero, F., 2002. Deficits in visceral pain and referred hyperalgesia in NaV1.8 (SNS/PN3)-null mice. J. Neurosci. 22, 8352– 8356. Leffler, A., Reiprich, A., Mohapatra, D.P., Nau, C., 2007. Use-dependent block by lidocaine but not amitriptyline is more pronounced in tetrodotoxin (TTX)resistant NaV1.8 than in TTX-sensitive Naþ channels. J. Pharmacol. Exp. Ther. 320, 354–364. Li, H.L., Galue, A., Meadows, L., Ragsdale, D.S., 1999. A molecular basis for the different local anesthetic affinities of resting versus open and inactivated states of the sodium channel. Mol. Pharmacol. 55, 134–141. Liu, G., Yarov-Yarovoy, V., Nobbs, M., Clare, J.J., Scheuer, T., Catterall, W.A., 2003. Differential interactions of lamotrigine and related drugs with transmembrane segment IVS6 of voltage-gated sodium channels. Neuropharmacology 44, 413– 422. Nassar, M.A., Levato, A., Stirling, L.C., Wood, J.N., 2005. Neuropathic pain develops normally in mice lacking both NaV1.7 and NaV1.8. Mol. Pain 1, 24. Nau, C., Wang, G.K., 2004. Interactions of local anesthetics with voltage-gated Naþ channels. J. Membr. Biol. 201, 1–8. Porreca, F., Lai, J., Bian, D., Wegert, S., Ossipov, M.H., Eglen, R.M., Kassotakis, L., Novakovic, S., Rabert, D.K., Sangameswaran, L., Hunter, J.C., 1999. A comparison of the potential role of the tetrodotoxin-insensitive sodium channels, PN3/SNS and NaN/SNS2, in rat models of chronic pain. Proc. Natl. Acad. Sci. U.S.A. 96, 7640–7644. Rabert, D.K., Koch, B.D., Ilnicka, M., Obernolte, R.A., Naylor, S.L., Herman, R.C., Eglen, R.M., Hunter, J.C., Sangameswaran, L., 1998. A tetrodotoxin-resistant voltage-gated sodium channel from human dorsal root ganglia, hPN3/SCN10A. Pain 78, 107–114. Renganathan, M., Cummins, T.R., Waxman, S.G., 2001. Contribution of NaV1.8 sodium channels to action potential electrogenesis in DRG neurons. J. Neurophysiol. 86, 629–640. Sangameswaran, L., Delgado, S.G., Fish, L.M., Koch, B.D., Jakeman, L.B., Stewart, G.R., Sze, P., Hunter, J.C., Eglen, R.M., Herman, R.C., 1996. Structure and function of a novel voltage-gated, tetrodotoxin-resistant sodium channel specific to sensory neurons. J. Biol. Chem. 271, 5953–5956. Schild, J.H., Kunze, D.L., 1997. Experimental and modeling study of Naþ current heterogeneity in rat nodose neurons and its impact on neuronal discharge. J. Neurophysiol. 78, 3198–3209. Stummann, T.C., Salvati, P., Fariello, R.G., Faravelli, L., 2005. The anti-nociceptive agent ralfinamide inhibits tetrodotoxin-resistant and tetrodotoxin-sensitive Naþ currents in dorsal root ganglion neurons. Eur. J. Pharmacol. 510, 197–208. Ulbricht, W., 2005. Sodium channel inactivation: molecular determinants and modulation. Physiol. Rev. 85, 1271–1301. Vijayaragavan, K., Powell, A.J., Kinghorn, I.J., Chahine, M., 2004. Role of auxiliary b1-, b2-, and b3-subunits and their interaction with NaV1.8 voltage-gated sodium channel. Biochem. Biophys. Res. Commun. 319, 531–540. Wood, J.N., Bevan, S.J., Coote, P.R., Dunn, P.M., Harmar, A., Hogan, P., Latchman, D.S., Morrison, C., Rougon, G., Theveniau, M., 1990. Novel cell lines display properties of nociceptive sensory neurons. Proc. Biol. Sci. 1302, 187–194. Zhao, J., Ziane, R., Chatelier, A., O’leary, M.E., Chahine, M., 2007. Lidocaine promotes the trafficking and functional expression of Na(v)1.8 sodium channels in mammalian cells. J. Neurophysiol. 98, 467–477.