Neuropathic pain develops normally in mice lacking both Nav1.7 and Nav1.8
© Nassar et al; licensee BioMed Central Ltd. 2005
Received: 04 August 2005
Accepted: 22 August 2005
Published: 22 August 2005
Skip to main content
© Nassar et al; licensee BioMed Central Ltd. 2005
Received: 04 August 2005
Accepted: 22 August 2005
Published: 22 August 2005
Two voltage gated sodium channel α-subunits, Nav1.7 and Nav1.8, are expressed at high levels in nociceptor terminals and have been implicated in the development of inflammatory pain. Mis-expression of voltage-gated sodium channels by damaged sensory neurons has also been implicated in the development of neuropathic pain, but the role of Nav1.7 and Nav1.8 is uncertain. Here we show that deleting Nav1.7 has no effect on the development of neuropathic pain. Double knockouts of both Nav1.7 and Nav1.8 also develop normal levels of neuropathic pain, despite a lack of inflammatory pain symptoms and altered mechanical and thermal acute pain thresholds. These studies demonstrate that, in contrast to the highly significant role for Nav1.7 in determining inflammatory pain thresholds, the development of neuropathic pain does not require the presence of either Nav1.7 or Nav1.8 alone or in combination.
Voltage gated sodium channels (VGSC) underlie the electrical excitability of nerve and muscle. VGSCs consist of pore forming α-subunits and auxiliary β-subunits. There are ten cloned α-subunits and 4 β-subunits. The β-subunits modulate the localisation, expression and functional properties of α-subunits . Different α-subunits have distinct electrophysiological and pharmacological properties . The complex pattern of expression of α-subunits may imply special roles for particular subunits in different cell types . Many loss- as well as gain-of-function mutations of α-subunits have been identified in human conditions characterised with epilepsy, seizures, ataxia and increased sensitivity to pain. This suggests that mutations of VGSC in humans are significant factors in aetiology of neuronal diseases .
Nociceptors are a subset of sensory neurons that respond to noxious thermal, mechanical and chemical stimuli. Nociceptors express multiple subtypes of α-subunits . Tissue and nerve damage leads to changes in expression and function of α-subunits that in turn can lead to change in the excitability of sensory neurons. Changes in the excitability of sensory neurons are thought to underlie some chronic pain conditions [5, 6]. Nav1.8 and Nav1.7 are two α-subunits that are abundant in nociceptive sensory neurons [3, 7, 8]. Nav1.8 is expressed exclusively in sensory neurons and is not found in the CNS . Functional characterization of Nav1.8 positive neurons revealed that more than 85% are nociceptors . Nav1.7 is expressed principally in peripheral neurons with very weak expression detected in CNS [10, 11]. As there are no subunit specific blockers, gene knockouts in mice have provided insights into the role of individual α-subunits genes in pain . Deletion of the Nav1.8 gene  and nociceptor-specific knockout of the Nav1.7  gene have identified a role for these two α-subunits in setting mechanical and, to a lesser extent, thermal pain thresholds. In addition, behavioral studies have revealed deficits in inflammatory pain models, most dramatically in the nociceptor-specific Nav1.7 knockout [12, 13]. These findings suggest that these α-subunits could be targets for new anti-inflammatory drugs.
Peripheral nerve injury leads to lowered pain thresholds, enhanced responsiveness and/or ectopic activity in sensory neurons that ultimately leads to hyperalgesia and allodynia [5, 6]. Changes in expression of α-subunits of VGSCs have been documented in models of peripheral nerve injury [5, 6]. This has lead to the hypothesis that modulation of α-subunits in sensory neurons may underlie the increased neuronal excitability of sensory neurons following peripheral nerve injury. Pharmacological blockade of sodium channel activity has been shown to attenuate ectopic activity [14, 15] and reverse hyperalgesia following nerve injury . While the role of Nav1.7 in neuropathic pain remains to be investigated, analysis of a Nav1.8 knockout mouse indicated that it is not involved in alteration of pain threshold following peripheral nerve injury . This is in contrast to the finding of Lai et al who reported that antisense oligonucleotides directed against Nav1.8 administered intrathecally completely reverse neuropathic pain behavior . It is possible that this discrepancy could be due to the up-regulation of the Nav1.7 channel seen in the Nav1.8 knockout mouse  which might mask an otherwise important role for Nav1.8 in neuropathic pain.
In the present study we investigated the role of the Nav1.7 channel in neuropathic pain using nociceptor-specific deletion of Nav1.7 in mouse. In addition, we readdressed the role of Nav1.8 in neuropathic pain by generating a double knockout of Nav1.8 and Nav1.7. We reasoned that the co-deletion of Nav1.7 in Nav1.8-expressing neurons should reveal any potential role for Nav1.8 in neuropathic pain.
Nociceptor-specific Nav1.7 knockout mice and their littermate controls were generated as described previously . It is not possible to generate global knockouts of both Nav1.8 and Nav1.7 since global deletion of Nav1.7 is lethal at P0 . Therefore, we generated Nav1.7 and Nav1.8 double-knockout (DKO) mice by exploiting the fact that the Nav1.8Cre allele has Cre sequence inserted in exon 1 followed by transcriptional stop signals . Consequently homozygous Nav1.8Cre mice are Nav1.8 global-knockouts (Nav18 KO) and show no Nav1.8 currents in sensory neurons .
Eight weeks old C57BL6 inbred mice were used as wildtype (WT) control (wildtype Nav1.8 and Nav1.7) because it would be impractical and undesirable to generate them as littermates to the test groups (DKO and Nav1.8 KO). The expected ratio of each of desired strains would be only 6.25% if the parents were heterozygous for both Nav1.8Cre and floxed Nav1.7 alleles compared with 25% as in our breeding strategy. The Nav1.8 KO/Nav1.7 heterozygous group was analysed in behavioural tests and was found to be similar to the Nav1.8 KO group (data not shown).
Pain thresholds to punctate mechanical stimulation was measured using calibrated von Frey hairs according to the up and down method . The 50% withdrawal threshold was not different between the three groups (WT 0.56 ± 0.07 gram, n = 11; Nav1.8 KO 0.45 ± 0.05, n = 13 and DKO 0.49 ± 0.06, n = 17), figure 3c. In contrast pain threshold to noxious mechanical pressure applied to the tail using the Randall-Selitto apparatus was much higher in both the Nav1.8 KO (412.4 ± 33.97 gram, n = 7) and DKO (395.7 ± 21.03, n = 7) groups compared to that of the WT (131.4 ± 16.54, n = 6) group, figure 3d. Usually the cut off point (500 gram) was reached without an observed escape response in the Nav1.8 KO and DKO groups, indicating a high resistance to static blunt mechanical pressure. There was no difference between the Nav1.8 and the DKO mice in this test, figure 3d, and similar results have been obtained with the nociceptor-specific Nav1.7 knockout 
To address the issue of compensatory up-regulation of Nav1.7 in Nav1.8 knockout we induced peripheral nerve injury in WT, Nav1.8 KO and DKO. All mice groups studied developed a robust mechanical allodynia starting form the third day post surgery, figure 5b. There were no statistically meaningful differences in the behaviour of the groups of mice.
This study confirms that the two sodium channel α- subunits Nav1.7 and Nav1.8, expressed selectively in nociceptive sensory neurons, have important roles in nociception and in pain. In the absence of specific pharmacological blockers, the use of genetic approaches, a combination of global and nociceptor-specific knockouts, has enabled us to carry out studies exploring the contribution of these two isoforms of VGSCs in different pain conditions. Nav1.8 knockouts are already known to have deficits in inflammatory and visceral but not neuropathic pain . Nav1.7 nociceptor-specific knockouts are almost totally refractory to changes in peripheral pain thresholds evoked by inflammatory mediator . This mirrors the inflammatory phenotype of Nav1.7 gain-of-function mutations in erythermalgia . This heritable disorder is caused by point mutations that lead to altered thresholds of activation in Nav1.7 
analysis of acute mechanical thresholds in DKO mice confirmed previous findings that the knockout of either Nav1.8 or Nav1.7 renders the mice resistant in the Randel-Selitto test of noxious mechanical pressure while the responses to von Frey hairs remain unchanged compared to controls [12, 13]. Furthermore, it has been shown that both Nav1.8 and nociceptor-specific Nav1.7 knockouts have small increases of about 20–40% in their thermal threshold [12, 13]. Interestingly, we found that the thermal threshold in the Hargreave's test was doubled in the DKO compared to Nav1.8 knockout or WT controls, figure 3b. The increase is more than the sum of the phenotype in the individual knockouts and may indicate that Nav1.8 and Nav1.7 are the major VGSC isoforms present in nociceptive terminals and their deletion dramatically increases pain threshold. This is supported by the fact that Nav1.7 is transported to nerve terminals of sensory neurons  and that Nav1.8 currents are localized in the terminals .
We have not studied pain behavior in inflammatory models where the nociceptive-specific Nav1.7 knockout has shown complete deficits in inflammatory pain behavior , because it would not be possible to observe further changes brought about by the co-deletion of Nav1.8. We only investigated the pain response in the formalin model. While the nociceptive-specific Nav1.7 knockout showed a reduction in the second phase to about 50% of controls  the DKO showed a slightly bigger reduction to 30% of controls, figure 4b. This is in contrast to Nav1.8 mice that showed no reduction in the second phase compared to controls, figure 4b. Therefore, the reduction in the second phase in the DKO is a new phenotype rather than a mere summation of the phenotype in the individual knockouts. This highlights further the predominant role for Nav1.7 in inflammatory pain. Expression of Nav1.7 has been reported to increase in sensory neurons subsequent to induction of inflammation .
This study and our previous  work have explored the role of Nav1.7 in Nav1.8-expressing neurons, since our strategy exploits the Nav1.8 promoter to drive Cre expression. We have shown previously using non-quantitative RT-PCR that not all the Nav1.7 mRNA in DRGs from the nociceptor-specific Nav1.7 knockout is deleted . This indicates that there is a population of sensory neurons, the size of which has yet to be determined, that express Nav1.7 but not Nav1.8. Therefore, the role of Nav1.7 in sensory neurons where Nav1.8 is not expressed (and which do not express Cre) remains to be studied.
From all the above, our results indicate that Nav1.8 and Nav1.7 subunits contribute to the excitability of peripheral nerve terminals and their modulation is critical for peripheral sensitisation in inflammatory pain
We investigated the role of Nav1.7 expressed in nociceptors in neuropathic pain. Analysis of neuropathic pain in the nociceptor-specific Nav1.7 knockout using the Chung model showed that they developed mechanical allodynia to the same extent as the littermate control mice (homozygous floxed Nav1.7), figure 5. This clearly shows that unlike its critical role in inflammatory pain , Nav1.7 does not contribute to neuropathic pain.
In addition, the Nav1.8 knockout and WT as well as the DKO mice developed a profound mechanical allodynia the level of which was indistinguishable between the three groups. This provides further evidence that Nav1.8 does not contribute to mechanical allodynia as reported by Kerr et al . In addition our results rule out Nav1.7 up-regulation in the Nav1.8 knockout as the reason behind the discrepancy between the finding of Kerr et al  and Lai et al . Surprisingly Lai et al  failed to detect the changes in acute pain thresholds observed in the Nav1.8 KO both in this study and that of Akopian et al , which would be expected to occur if Nav1.8 had been down-regulated.
The lack of changes in mechanical allodynia in any individual or even in double knockouts of Nav1.8 and Nav1.7 subunits is consistent with recent data by Flake et al. They reported that changes in sodium currents shortly after nerve injury do not correlate with an increase in neuronal excitability and that the sum of changes to ionic currents, and not a single class of voltage-gated ion channel, underlie increased neuronal excitability . This does not, however, exclude a potential role for both Nav1.7 and Nav1.8 in spontaneous neuropathic pain as suggested by the finding that neuromas in the Nav1.8 knockout mouse display less ectopic discharges than wildtype littermates . These studies confirm the importance of Nav1.7 and Nav1.8 in the physiological processes that underlie altered peripheral thresholds in inflammatory pain. Neuropathic pain that arises as a consequence of nerve damage and neuronal dysfunction is, however, not dependent on the presence of these two sodium channels. In terms of the contribution of other α-subunits to neuropathic pain, a strong correlation has been found between the expression of Nav1.3 and the appearance of neuropathic pain. GDNF that reverses neuropathic pain behaviour also normalises Nav1.3 expression , and recent antisense studies have supported the view that increased Nav1.3 expression contributes to neuropathic pain development 
In summary, our results indicate a critical role for Nav1.7 and Nav1.8 in setting pain thresholds of nociceptive nerve terminals and in their sensitisation following tissue damage and inflammation. However, the presence of these channels, or changes to their expression and function are not required for the establishment of mechanical allodynia arising from nerve injury.
Nav1.8Cre and floxed Nav1.7 lines were produced as described . Both strains were back-crossed at last 5 times onto a C57/BL6 background.
Genomic DNA was prepared from tail biopsies and genotyped using PCR. Primers for Nav1.7 are (CAGAGATTTCTGCATTAGAATTTGTTC) and (AGTCTTTGTGGCACACGTTACCTC) which give a WT band of 317 bp and a floxed band of 461 bp. Primers for Nav1.8 are (TGTAGATGGACTGCAGAGGATGGA) and (ttacccggtgtgtgctgtagaaag) which give a WT band of 460 bp. Nav1.8Cre was detected by primers (aaatttgcctgcattaccggtcga) and (aaatgttgctggatagtttttactgcc) located inside Cre sequence which give a band of 249 bp.
All tests were approved by the United Kingdom Home Office Animals (Scientific Procedures) Act 1986 and performed in a Home Office designated room at 22 ± 2°C. Experiments were performed on animals of at least 8 weeks of age. Behavioral tests were done as before 
The authors would like to acknowledge Mark D Baker for critical reading of the manuscript and Elizabeth A Matthew for great help in establishing the Chung model. We also thank the Wellcome Trust, the MRC and the London Pain Consortium for funding this work.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.