Na_v1.3 channels were previously known as type III Na⁺ channels. They are TTX-sensitive products of the SCN3A gene and are found in neurons and cardiac myocytes with the highest level in embryonic and early postnatal animals (117, 126, 127). In DRG neurons, they exhibit rapid recovery from inactivation or “repriming,” thereby enhancing repetitive discharge (128). Their involvement in neuropathic pain is supported by the attenuation of allodynia seen with intra-ganglionic injection of adeno-associated virus expressing small hairpin RNA targeting Na_v1.3 (129). Nerve injury upregulates and promotes re-expression of Na_v1.3 in adult DRG neurons (127, 130, 131) as well as in spinal dorsal horn and thalamus (132, 133). This may reflect removal of suppression of the SCN3A gene by microRNAs such as miR-384-5p, mir-96 and/or miR-30b suggesting that their targeted delivery may be of use in pain management (41–43).

Because Na_v1.3 is mainly present in embryonic and early neonatal animals and because nerve injury promotes selective upregulation of Na_v1.3 in nociceptive pathways of adults, there is considerable interest in developing Na_v1.3 blockers. Structure activity studies starting with a diphenylmethyl amide adduct of an aryl sulphonamide has led to the development of compounds with good selectivity for Na_v1.3 as well as favorable pharmacokinetics (44).

Na_v1.6 is another TTX-sensitive Na⁺ channel. It is the product of the SCN8A gene (117) and was previously known as PN4. Na_v1.6 channels are expressed along the whole length of sensory unmyelinated axons (134) and are clustered at nodes of Ranvier in myelinated fibers where they participate in “saltatory” conduction (135).

The observation that knockout of Na_v1.6 reduces injury-induced pain behaviors and sensory neuron excitability (136–138) implicates it in the etiology of neuropathic pain. It has recently been implicated in a model of vincristine-induced chemotherapy induced peripheral neuropathy (CIPN) and allodynia (139) and is upregulated in the DRG in a model of diabetic neuropathy (140). These findings are corroborated by the description of a gain-of-function mutation in Na_v1.6 in a case of trigeminal neuralgia (141). Since its role in in pain etiology was established relatively recently (114, 142), there have been as yet no attempts to modulate Na_v1.6 channel activity either in animal models or in the clinic.

The TTX-sensitive Na_v1.7 channel is involved in a multiplicity of neuropathic and nociceptive pain states (8, 48, 54, 114, 123, 143–146). It is the product of the SCN9A gene and was previously known as PN1. Na_v1.7 is the dominant voltage-gated Na⁺ channel in peripheral sympathetic neurons and in all types of DRG neuron (117, 147). Its expression extends from peripheral nerve endings in the skin and viscera to primary afferent terminals in the dorsal horn (148) where it is especially concentrated (147). Na_v1.7 is preferentially expressed in small diameter nociceptors including both the CGRP-positive subcategory and the non-peptidergic subcategory that bind the plant lectin IB4 from Griffonia simplicifolia (114). It is also found in olfactory sensory neurons, magnocellular neurosecretory cells of the hypothalamic supraoptic nucleus and in vagal afferents (51, 149–151). Because it is not found to any great extent in vital non-neuronal tissue such as heart or skeletal muscle (114, 147), Na_v1.7 represents a specially attractive target for therapeutic manipulation. Although it is found in pancreatic alpha and beta cells it may be inactivated at their normal resting potential (152).

Immunohistochemical studies first demonstrated Na_v1.7 upregulation in severed axons within human painful neuromas (122, 153) and Na_v1.7 has been shown to be necessary for the release of the pain modulator substance P from primary afferent terminals (124).

Despite this, Na_v1.7 does not appear to be involved in all manifestations of neuropathic pain. For example oxaliplatin-induced pain and cancer-induced bone pain do not require the presence of Na_v1.7 or the Na_v1.8-positive nociceptors in which Na_v1.7 is enriched (123). By contrast, paclitaxel-induced CIPN involves the direction of Na_v1.7 to cell membranes and axons of primary afferent fibers (154). Also, neuropathic pain produced by constriction injury (CCI) is abolished when Na_v1.7 is selectively deleted in murine sensory neurons and although spinal nerve transection or tight ligation (SNL) also produces cold and mechanical allodynia this is not affected by selective knockout of Na_v1.7 in DRG neurons. By contrast, knockout of Na_v1.7 in both sympathetic and sensory fibers attenuates both forms of allodynia (123). This is because SNL involves sprouting of Na_v1.7 expressing perivascular sympathetic fibers (155, 156) and their ectopic interaction with DRG neurons (157–159).

Patients with a rare, chronic pain conditions such as primary erythromelalgia or paroxysmal extreme pain disorder exhibit gain of function mutations in SCN9A (8, 146, 160–163). As of 2019, 30 mutations in SCN9A genes had been described in inherited erythromelalgia and 13 in paroxysmal extreme pain disorder (114). In the case of inherited erythromelalgia, isoleucine 848 is replaced by threonine. This I848T mutation increases the amplitude of current produced by Na_v1.7 in response to slow, small depolarizations as a result of a hyperpolarizing shift in activation and slowed deactivation (161). Recently, protein kinase C has been found to be responsible for the phosphorylation of T848 found in mutant channels and this accounts for the shift in activation (164). Meents et al. (165) have differentiated human induced pluripotent stem cells (hiPSC) from erythromelalgia patients into sensory nociceptors. This will provide an extensive supply of human nociceptors for further study of erythromelalgia. Mutations seen in Na_v1.7 channels of erythromelagia patients also occur in those with paroxysmal extreme pain disorder with an additional suppression of fast inactivation (163). Gain of function mutations of SCN9A also worsen neuropathic pain in a small cohort of patients with painful diabetic neuropathy (166).

Although some patients with small fiber neuropathy display the I228M gain-of-function mutation in Na_v1.7, a pain phenotype does not appear until they reach adulthood (167). Expression of this same mutation in mice promotes increased DRG excitability without the appearance of a measurable pain phenotype. It is suggested that some compensatory mechanism may restrain the development of pain in the mouse model and the possible existence of a similar process in humans may delay the development of a pain phenotype until adulthood (168).

Patients with a rare congenital insensitivity to pain (CIP) express a loss of function mutation in Na_v1.7 (169) and global knockout of Na_v1.7 in mice recapitulates this human phenotype (170). Differentiation of hiPSC's from CIP patients into sensory nociceptors, produced cells where Na_v1.7 was appropriately expressed and trafficked to the cell membrane. Since these cells failed to respond to depolarizing stimuli, CIP can be attributed to changes in the function of the channels per se rather than defects in their expression or trafficking (54). These results also provide new evidence for a role of Na_v1.7 in human nociception. As of 2019, 26 mutations in SCN9A have been reported to contribute to CIP.

In addition to its role in controlling neuronal excitability and neurotransmitter release, Na_v1.7 directly or indirectly affects gene expression (51, 52, 171). Na_v1.7 deletion, leads to upregulation of Penk mRNA for the enkephalin precursor proenkephalin in DRG as well as met-enkephalin protein. Since a similar effect is seen with TTX, the upregulation of endogenous opioid function may be contingent on decreased levels of intracellular Na⁺ (52). These authors also showed that blockade of opioid receptors with naloxone reduces the analgesia seen in both male and female Na_v1.7-null mutant mice and in a human patient with Na_v1.7 dependent congenital insensitivity to pain [see also (51)]. The relationship between increased opioid function and decreased Na_v1.7 function is supported by the observation that the analgesic effect of a selective Na_v1.7 blocker, μ-theraphotoxin-Pn3a (from the tarantula Pamphobeteus nigricolor), is augmented by administration with sub-effective doses of opioids or with an enkephalinase inhibitor (172). Further analysis of this effect showed that Na_v1.7 knockout mice have normal peripheral nociceptor activity but synaptic transmission from nociceptor central terminals is greatly reduced in an opioid-dependent fashion. Analgesia was reversed substantially by central but not peripheral application of opioid antagonists (51). These authors thus concluded inhibition of neurotransmitter release is the principal mechanism of analgesia in mouse and human Na_v1.7-null mutants.

Second order sensory neurons in the spinal dorsal horn express few transcripts of Na_v1.7 mRNA. Despite this, immunoreactivity for channel protein is abundant yet is reduced following rhizotomy (173). This suggests that sensory neurons are the source of Na_v1.7 in spinal dorsal horn neurons and that intercellular transport of the protein occurs between these two neuronal populations. This conclusion was supported by the observation that selective deletion of Na_v1.7 in peripheral neurons reduced the intrinsic excitability of dorsal horn neurons.

Although it may not be involved in all types of neuropathic pain (123) it is absence from non-neuronal tissue such as heart or skeletal muscle (114, 147). Na_v1.7 is therefore clearly an attractive target for therapeutic intervention (48, 114, 145, 154, 174). Moreover, the anticonvulsant lacosamide, which is an inactivated state blocker of Na⁺ channels (175, 176) has been found to be safe and effective, in a randomized, placebo-controlled, double-blind, crossover-design study of Na_v1.7 related small fiber neuropathy [(45), Table 1]. Also, the effectiveness of carbamazepine which is used to treat trigeminal neuralgia (56) may in part reflect its affinity for Na_v1.7 (177).

There is also considerable interest in various sulfonamide analogs which display selectivity toward Na_v1.7 and are effective in pain mitigation in animal models [(178–183); see Table 1]. Therapeutic concentrations of the inactivated state blocker PF-05089771 increase the rheobase of control neurons, but not that of Na_v1.7 knock-out neurons. Despite this selectivity for Na_v1.7 and its effectiveness in animal models in vivo (54), a clinical study of PF-05089771 in subjects with painful diabetic peripheral neuropathy failed to meet defined efficacy criteria (46).

Another broad spectrum non-sulfonamide Na_v blocker, vixotrigine, which was previously known as raxatrigine, or CNV1014802, BIIB074, or GSK-1014802 (184), has shown effectiveness in animal models of Na_v1.7-dependent pain. Its safety in human patients has been established (185). A phase III clinical trial for effectiveness in trigeminal neuralgia and phase II trial for small fiber neuropathy are presently ongoing (47).

Another approach to therapeutic modulation of Na_v1.7 activity involves potential use and/or structural modification of natural toxins (48, 186–190). These are typically gating modifiers as opposed to simple pore blockers so some natural toxins increase channel function whereas others attenuate it [(48); Table 2]. Starting points include the cone snail toxin, μ-conotoxin KIIIA, and PnTx1 (Phoneutria nigriventer toxin 1) from a Brazilian spider. Although structure activity studies of small cyclic peptides derived from the structure of these toxins has not as yet revealed Na⁺ channels subtype ligands, the analgesic effect of many of the ligands involves modulation Na_v1.7 channel function. This result was achieved by observing attenuation of pain produced by the Na_v1.7 selective activator α-scorpion toxin OD1 [(191); Table 2]. Further modifications of small cyclic peptides may reveal more subtype selective ligands with appropriate pharmacokinetics in vivo and improved bioavailability (59).

Studies and modification of arachnoid toxins which display natural selectivity toward Na_v1.7 may also lead to development of effective agents (190). As listed in Table 2, there are several examples.

Venom from the tarantula Grammostola porteri contains the 34-residue peptide, GpTx-1, with high and selective affinity for Na_v1.7 (IC₅₀ = 10 nM). Structural modifications of this peptide led to the identification of [Ala5, Phe6, Leu26, Arg28] GpTx-1 (also known as GpTx-1-71) IC₅₀ = 1.6 nM (192). Both peptides exert powerful antinociception in mouse models of acute, visceral, inflammatory and neuropathic pain without impairment of motor co-ordination or development of tolerance (144). Another modified toxin derived from JzTx-V (from venom of the Chinese tarantula Chilobrachys jingzhao) has a 100-fold improved efficacy compared to GP-Tx-1-71 (193).

Studies of the venom from the Peruvian green-velvet tarantula Thrixopelma pruriens revealed a 33 residue peptide termed μ-TRTX-Tp1a (Tp1a or ProTx-III) with high selectivity and affinity for Na_v1.7 (194). Unlike other spider toxins that inhibit the function of Na_v channels, Tp1a inhibited hNaV1.7 without significantly altering the voltage-dependence of activation or inactivation. Like PnTx1, the analgesic effect of Tp1a was demonstrated by its ability to reverse spontaneous pain induced in mice by intraplantar injection of the Na_v1.7 activator OD1 (194).

Recently another peptide toxin named Tap1a from the Venezuelan tarantula Theraphosa apophysis was shown to reverse colonic mechanical hypersensitivity in a mouse model of irritable bowel syndrome. The toxin's efficacy was shown to reflect selective targeting of Na_v1.7 as well as the T-type Ca²⁺ channel Ca_v3.2 (49).

High-throughput screening has also identified μ-TRTX-Df1a (Df1a) from the venom of the spider Davus fasciatus as an Na_v modulator. This 34-residue peptide inhibits responses mediated by Na_v1.7 that is endogenously expressed in the human neuroblastoma cell line SH-SY5Y. It also inhibits T-type calcium (Ca_v3.1 and Ca_v3.3) currents and other Na_v currents expressed in HEK 293 cells but has no effect on the voltage-gated potassium channel [K_v2.1; (195)]. Df1a is active in vivo and reverses the spontaneous pain behaviors induced by the scorpion venom Na_v activator OD1.

Other investigations have used the venom-peptide ProTX-II (Protoxin II) from the Peruvian green velvet tarantula (Thrixopelma pruriens) as a scaffold, to engineer a library of over 1,500 peptides. This identified JNJ63955918 as a potent, highly selective, closed-state Na_v1.7 blocking peptide which induces insensitivity to pain that closely recapitulates key features of the Na_v1.7-null phenotype seen in mice and humans (196).

More recently attention has been drawn to huwentoxin-IV, from the Chinese bird spider Haplopelma schmidti. Because it has high affinity for sodium channels it is an attractive scaffold for engineering Na_v1.7-selective molecules and several new ligands with high affinity and selectivity have been identified (197).

Other natural products which block Na_v1.7 channels include HNTX I and III from the spider Ornithoctonus hainana (198, 199), bulleyaconitine from aconitum bulleyanum plants (200) and the Japanese traditional medicine goshajinkigan (2, 201).

In general, despite intensive pre-clinical studies with Na_v1.7 blockers, tests of their efficacy in the clinic has yielded rather disappointing results [(48), Table 1] and to the best of our knowledge no studies of tarantula and other toxins in the clinic have appeared. Nevertheless, the continued study of toxins, small molecule blockers and monoclonal antibodies (202) should and will continue (2, 8). In particular, further structural modification of small molecule blockers such as CNV1014802 (vixotrigine) and PF-05089771 as well as chemical modification of natural toxins (48, 50) may provide a route to the development of more efficacious therapeutic entities. The tarantula toxin Tap1a shows particular promise as it appears to selectively target both Na_v1.7 and Ca_v3.2 (49).

Since the consequences of Na_v1.7 blockade are mediated at least in part by endogenous opioids (51, 52), benefit may be obtained by combining small molecule blockers or toxins with low doses of opioids (48, 53).

The development of monoclonal antibodies and the delivery of the inhibitory micoRNA miR-182 (203) or modifiers of Na⁺ channel β subunits (204) may reveal additional therapeutic approaches. This approach may be especially attractive as three different types of β subunits are differentially and selectively expressed in small, medium, and large diameter DRG neurons (205, 206).

An approach that has proved particularly effective for targeting Na_v1.7 uses CRISPR-dCas9 technology (clustered regularly interspaced short palindromic repeats) (55). Epigenome engineering platforms were introduced intrathecally in mice via adeno-associated viruses. A novel approach that prevented expression of Na_v1.7 by editing a regulatory sequence successfully repressed Na_v1.7 expression in lumbar DRG, reduced thermal hyperalgesia in inflammatory pain models and decreased tactile allodynia in the neuropathic pain models without affecting normal motor function. It is anticipated that this “LATER” (long-lasting analgesia via targeted in vivo epigenetic repression technology) might have therapeutic potential in management of persistent pain states. This is important in practical terms as chronic pain patents usually present in the clinic when they have suffered for many months. The technology can of course be easily modified to control expression of any potential or central drug target.

The TTX-resistant Na_v1.8 channel is predominant in small DRG neurons (124, 207–210) but its selective association with nociceptors has been questioned (211). It was originally known as SNS or PN3 and is encoded by the SCN10A gene (117). It is characterized by its high threshold for activation and its slow rate of inactivation at depolarized potentials (210). These properties enable it to generate a slow persistent inward current (212).

Although peripheral nerve injury attenuates Na_v1.8 function in injured DRG neurons (213–215) it is thought to accumulate in uninjured neurons (216) and in neuromas that develop at sites of nerve injury (217). Selective blockade of Na_v1.8 function promotes hypoalgesia (213), gain of function mutations of SCN10A in humans can promote painful neuropathy (218) and its optogenetic silencing in DRG attenuates neuropathic pain (219).

The selective Na_v1.8 blockers A803467 and PF-01247324 are being developed as potential antidysrhythmic agents (220). Although both are reported to attenuate allodynia in a rodent model (221, 222), they have yet to be used in clinical studies (223). Encouraging results have been seen with the pro-drug VX-150 which exhibits analgesic activity in healthy volunteers (57), but preclinical literature in support of these studies are not available online. The μO-conotoxins, MrVIA, MrVIB, and MfVIA block Na_v1.8 and ongoing analysis seeks to increase their affinity by structural modifications (224).

Unlike the situation with Na_v1.7, analgesia produced with blockade of Na_v1.8 is not opioid-dependent (52) and may be attributable to decreased excitability of peripheral afferents and their central terminals (225).

Although the efficacy of the non-psychoactive cannabinoid, cannabidiol in management of neuropathic pain remains to be established (106), it was recently reported to decrease the excitability of DRG neurons by binding to the slow inactivated state of Na_v1.8 channels (102).

Expression of Na_v1.8 in peptidergic DRG neurons is controlled by nerve growth factor (NGF) (215) whereas its expression in non-peptidergic neurons is controlled by glial colony derived neurotrophic factor (GDNF) (226). This may account in part for the effectiveness of the NGF antagonist tanezumab in various pain states (58). In fact, its safety and efficacy in humans identifies tanezumab as one of more the promising new drug candidates for chronic and neuropathic pain (see Clinical Trials Government Identifiers: NCT02528188 and NCT02528188).

Na_v1.9 is also TTX-resistant (227) and is encoded by the SCN11A gene. It was previously known as NaN. Unlike genes encoding other voltage-gated Na⁺ channels, murine SCN11A is only 75% identical to the human gene (114). Na_v1.9 was previously known as NaN or SNS-2 (117) and because it inactivates extremely slowly, it is capable of producing a persistent inward current (228). This means that gain of function mutation of Na_v1.9 causes decreased excitability because other voltage-gated Na⁺ channels are inactivated by persistent Na_v1.9 mediated depolarization (8). In the peripheral nervous system, NaN/Na_v1.9 was first detected in small DRG neurons of SNS/Na_v1.8—null mice (228) where it is preferentially expressed in non-peptidergic neurons which bind the plant lectin IB4 (229). Channels are found in free nerve endings, along axons, in DRG cell bodies and in primary afferent terminals in spinal lamina II (substantia gelatinosa) (230). Unlike Na_v1.7, sciatic injury reduces expression of mRNA and channel protein for Na_v1.9 (231) and this may be attributable to loss of trophic support by GDNF (226). Since Na_v1.9 knockout mice continue to display allodynia following nerve injury (232), this channel is unlikely to play a role in injury-induced neuropathic pain. This contrasts with the situation for inflammatory pain where a role for Na_v1.9 is well-established (114, 232).

Protein and mRNA for K_v1.1 is reduced in DRG following sciatic nerve injury (238, 245, 255) and this is associated with redistribution of channels away from nodal regions of A-δ fiber axons (245). Although expression of a dominant negative phenotype of K_v1.1 causes allodynia in mice (256), certain glycine derivatives act as Kv1.1 channel openers (63), and substances have been identified that attenuate the time dependent inactivation of K_v1.1 (257), its ubiquitous distribution in brain, heart, retina, skeletal muscle and pancreatic islets (247) may preclude the use of K_v1.1 activators in pain management.

Knockdown of K_v1.2 by siRNA induces mechanical and thermal hypersensitivity in naive rats (258). mRNA for K_v1.2 is also downregulated in several neuropathic pain models (28, 238, 255, 259, 260), and overexpression of Kv1.2 impairs neuropathic pain but does not attenuate acute pain in rats (261). These findings correlate with injury-induced reduction of whole-cell K_v1.2 current (260) and reduced channel protein expression as demonstrated by immunohistochemistry (261, 262) and/or immunoblot (245, 263).

Six different mechanisms have been hitherto suggested to underlie decreased K_v1.2 expression in DRG after peripheral nerve injury.

No small molecule activators of K_v1.2 have been identified (118) and given their documented presence throughout the brain, in spinal cord, mechanoreceptors and proprioceptors, Schwann cells, the heart, vascular smooth muscle and retina (247), direct pharmacological manipulation of these channels is not a viable means of treatment of neuropathic pain. There are some reports of alleviation of pain in animal models by attenuation of HDAC2 action (64, 65) but these may reflect modulation of its actions in the spinal cord as well as upregulation of K_v1.2 in the periphery. The HDAC inhibitor and antineoplastic agent, suberoylanilide hydroxamic acid (vorinostat) has been shown to alleviate pain in a bone cancer model (66) but to the best of our knowledge no trails of its efficacy in any form of neuropathic pain have as yet appeared.

These channels which also exhibit delayed rectification are expressed at relatively low levels compared to K_v1.1 and 1.2 in naïve DRG (238, 245). mRNA for K_v1.3 is decreased but that for K_v1.5 and 1.6 is little affected by nerve injury (238, 255). In view of the relatively limited expression of these channels in DRG, augmentation of their function would not seem to be a desirable therapeutic strategy for pain mitigation.

Channel protein and mRNA are reduced by nerve injury as is K_v2 whole-cell current comprising K_v2.1 and 2.2 (262, 269). These changes may, in part, reflect the influence of the silent subunit K_v9.1 in hetero-tetramers with both K_v2.1 and K_v2.2 (67, 247, 270, 271). Nerve injury downregulates K_v9.1 in DRG neurons and this may alter behavior of K_v9.1~K_v2.1~K_v2.2 hetero-tetramers (270). Selective downregulation of the Kcns gene in DRG in vivo but not in other tissues, reduces K_v9.1 expression and promotes changes in pain behavior consistent with its role in onset of neuropathic pain (67, 270). This suggests that restoring Kcns1 activity in the periphery has therapeutic potential in chronic pain (67).

As seen with K_v2.1, nerve injury downregulates mRNA for K_v2.2 in DRG (255, 269). Since K_v2.2 currents are also affected by the presence of K_v9.1 in hetero-tetramers this give further credibility to potentiation of K_v9.1 as a therapeutic approach (Table 1).

Although immunohistochemical, biophysical and Western immunoblot studies have identified these isoforms in DRG (272), there is little or no evidence for injury-induced changes in their expression or function (255).

M-channels are the K_v7.2 and K_v7.3 products of the KCNQ2/3 genes (273). They are activated by depolarization in a similar fashion to delayed rectifiers but do not inactivate over periods of many minutes. This and the fact that M-channels start to activate at normal resting potential means that they play an important role in determining neuronal excitability and accommodation of firing (274, 275). Whole-cell M-current is reduced in a model of bone cancer pain (276), selective knockdown of K_v7.2 in DRG causes hyperalgesia (277) and peripheral nerve injury induces substantial downregulation of K_v7.2 protein (239). The observation that the M-channel openers such as flupirtine and retigabine alleviate hyperalgesia in several rodent pain models (239, 278, 279) initiated considerable interest in the potential therapeutic use of this type of drug (280–285).

Although a clinical study of retigabine in post herpetic neuralgia failed to meet its efficacy endpoint (68), at least 200 K_v activators are currently under development (285). It has also been observed that the natural products, mallotoxin (MTX) and isovaleric acid (IVA), act synergistically to open neuronal KCNQ channels. This combination has been shown to suppress pentylenetetrazole-induced tonic seizures in mice but has not yet been examined in pain models (71). Similar effects were seen with (E)-2-dodecenal (E-2-D), a natural product derived from cilantro leaves (72). It has been suggested that co-administering MTX, IVA or E-2-D with retigabine may be highly effective in opening of KCNQ2/3 channels (71) (Table 1).

A novel K_v7.2 activator known as SCR 2682 was described recently (69). Acute application of SCR 2682 augments M-currents in DRG neurons and alleviates nerve injury induced pain in vivo. Both effects are reversed by M-channel inhibitor XE991. SCR 2682 also increases KCNQ2 mRNA and K_v7.2 protein expression in a rodent model of neuropathic pain (70) but its exact mechanism of action is yet to be determined.

K_v7 thus retains its potential as a drug target for neuropathic pain (Table 1); chemical modification of the retigabine structure may provide new and effective therapeutic agents.

The effects of nerve injury on expression of KCNQ depend on the actions of inflammatory mediators (286) and/or inhibition of transcription by repressor element 1-silencing transcription factor (REST also known as neuron-restrictive silencing factor, NRSF) (239, 287). Overexpression of REST in DRG neurons strongly suppresses M-current density, increases excitability induces mechanical and thermal hyperalgesia (288). Specific knockout of REST in DRG prevents injury-induced downregulation of REST target genes and prevents the development of hyperalgesia in various models of neuropathic pain; an effect that can be restored by REST overexpression (288).

REST inhibits transcription by recruiting the co-repressor complexes SIN3A/B and REST corepressor 1; these complexes modify target gene regions through the action of HDAC1/2, the histone demethylase LSD1 and the histone methyltransferase G9a (289, 290). Inhibition or genetic deletion of G9a in DRG abolishes injury-induced down-regulation of K_v7.2 and reduces neuropathic hyperalgesia. G9a may have an important role in K⁺ channel regulation as it has also been implicated in injury induced suppression of K_v1.4, K_v4.2, and BK channels (K_Ca1.1) (250). Two small molecule inhibitors of G9a are available, namely BIX01294 and UNC0638, both of which attenuate neuropathic pain in rodent models (80, 81). Although there is considerable interest in developing histone methyltranferase inhibitors in cancer treatment (291), to the best of our knowledge neither BIX01294 nor UNC0638 have been examined for treatment of pain in the clinic. Further development of drugs of this type may lead to new approaches to pain management (Table 1).

mRNA for K_v1.4 is downregulated in several models of neuropathic pain, including a model of diabetic neuropathy (238, 250, 254, 255). Knockdown of K_v1.4 with siRNA causes allodynia (295) and miR-17-92 overexpression downregulates A-channels and promotes hyperalgesia (296). The molecular mechanism of altered K_v1.4 expression is similar to that for K_v7.2 described above (250). This means that the effectiveness of G9a inhibitors in inhibiting neuropathic pain (80, 81) may involve preservation of function of both K_v7.2 and K_v1.4 after injury (Table 1).

Kv3.4 are high threshold A-channels that are particularly sensitive to 4-AP block. Nerve injury decreases expression of K_v3.4 immunoreactivity (297) and mRNA is reduced in a model of diabetic neuropathy (254). K_v3.4 antisense produces mechanical hypersensitivity (297). It has also been reported that injury to the spinal cord per se causes K_v3.4 dysfunction in DRG (298). This may reflect the action of excitatory mediators released from the spinal site of injury. This raises the possibility that therapeutic control of DRG function may not only be beneficial for peripheral neuropathy, it may also have benefit for managing pain originating from spinal cord injury.

Immunoreactivity and/or mRNA for all three K_v4 channels is found in DRG neurons (28, 255, 294, 297, 299) with differences in their distribution across different neuronal types (300, 301). Decreased function of all K_v4 channels occurs after peripheral nerve injury (75, 244, 254, 293, 297, 299, 302), and knockdown of K_v4.1 and its modulatory subunits or antisense to K_v4.3 causes mechanical hypersensitivity (297, 299). Taken together these observations strongly suggest malfunction of K_v4 channels in neuropathic pain.

The expression and function of K_v4 channels in DRG is controlled by signaling pathways such as MAPK (293), K_v4 channel interacting proteins (KChIPs) and dipeptidyl-peptidase-like proteins (DPPLs) (303–305). The aforementioned neuron restrictor silencer factor (REST), which controls expression K_v7.2, also effects suppression of transcription of the K_v4.3 gene (KCND3) after nerve injury (302).

Since no activators of K_v4 channels are available, targeting accessory subunits of A-channels may provide an alternative strategy (244). DPPLs and KChIPs not only govern the biophysical properties of K_v channels. They also impact channel assembly, channel trafficking to and from the cellular surface, and targeting of channels to different cellular compartments (304). The compound NS5806 has been reported to potentiate K_v4 currents in a KChip dependent manner (73, 74) and has recently been shown to attenuate cold allodynia in a rodent model of trigeminal neuralgia [(75), Table 1].

BK channels are encoded by the KCNMA1 gene and are present in all DRG neurons (240, 306–308). Their functional expression is reduced by peripheral nerve injury (236, 240, 309). This is associated with decreased expression of KCNMA1 and channel protein (250, 310). Their involvement in generation of pain is suggested by the observation that overexpression of BK increases mechanical threshold in a rodent neuropathic pain model (311). Also, the K_Ca1.1. blocker, iberiotoxin reduces mechanical withdrawal threshold.

Intrathecal injection of the K_Ca1.1 channel opener [1,3-dihydro-1-[2-hydroxy-5-(trifluoromethyl)phenyl]-5-(trifluoromethyl)-2H-benzi midazol-2-one] dose-dependently reverses allodynia and hyperalgesia in nerve-injured rats but had no significant effect on nociception in control rats (310). This substance is one of several BK activators available including the highly effective GoSlo-SR family of anthraquinone analogs (312). Others include NS1619 (313, 314), NS11021 (315, 316), NS13558 (317), and 12,14-dichlorodehydroabietic acid (diCl-DHAA) (318). Because these drugs have profound effects on tissues such as cardiac myocytes and certain smooth muscles, they are unlikely to be of practical use in pain management.

On the other hand, there is considerable discussion in the literature relating to the efficacy of cannabinoids in neuropathic pain (102, 106, 107) and it has been suggested that augmentation of BK function may contribute to their potential therapeutic effect (103).

As was described for K_v7.2 and K_v1.4, injury-induced downregulation of KCNMA1 in DRG is a result of G9a activation (250). This underlines the potential therapeutic application of G9a blockers such as BIX01294 and UNC0638 (Table 1).

There is little information about the possible role of K_Ca2 channels in pain but several recent reports have drawn attention to the possible role of K_Ca3.1 (Table 1). Although K_Ca3.1 knockout-mice show increased sensitivity to noxious chemical stimuli they exhibit normal behavioral responses to acute nociceptive, persistent inflammatory, and persistent neuropathic pain (319). Despite this, the K_Ca3.1 channel opener, ASP0819, modulates nociceptive processing and in vivo action potential activity in peripheral nerves in an animal model of fibromyalgia (320) and preliminary investigation of its action in the clinic have provided evidence of efficacy with minimal side effects (321).

K_ATP channels play an indispensable role in pancreatic insulin secretion as a result of their inhibition by intracellular ATP and their activation by ADP (248). Sulphonylurea receptors (SUR or ATP binding cassettes) co-assemble with channel proteins (324). K_ATP channel activation can be achieved by the anti-hypertensive agents, diazoxide and pinacidil and their anti-nociceptive actions have been recognized for many years (325). Nerve injury reduces K_ATP currents and channel activity in DRG neurons (323, 326) and although there are several reports of the efficacy of K_ATP openers in neuropathic pain models (76–79), these findings do not appear to have been exploited in the clinic.

Four transmembrane-domain tandem pore domain (K_2p) channels account for K⁺ leak conductance and set the resting membrane potential of most excitable cells including DRG neurons (28, 249, 327). TRESK (k_2p18) channels seem particularly important in this regard (328). Their potential relevance to neuropathic pain is supported by the observation that sciatic nerve transection reduces TRESK/(k_2p18)/KCNK18 mRNA to a greater extent than other K_2p channels in DRG and in vivo knock down decreases threshold to painful mechanical stimuli (329, 330). Other K_2P channels such TASK3 (K_2p9) and TWIK1 (K_2P1) are also down-regulated by spared nerve injury (SNI) (331).

Although activation of K₂P channels contributes to the therapeutic effectiveness of volatile anesthetics such as isoflurane (327, 332) it is obviously impractical to use these drugs for long term pain management. The novel TREK2/K_2p10.1 activator GI-530159 decreases DRG excitability (333), but its possible effectiveness in pain models has not yet been reported.

Although these L-type VGCC are present in rodent DRG (338), gain of function mutations in humans do not express a pain phenotype (93). On the other hand, following CCI of the sciatic nerve, the “classical” dihydropyridine, nitrendipine reduces the frequency of spontaneous EPSC's in rat lamina II (substantia gelatinosa) neurons. It also, albeit rather weakly, attenuates mechanical allodynia. These effects have been attributed to injury-induced upregulation of α2δ-1 and increased expression of Ca_v1.2 after nerve injury (348). Anti-Ca_v1.2 siRNA or selective knockdown of Ca_v1.2 in the spinal dorsal horn but not in DRG has been shown to reverse the nerve injury associated mechanical hypersensitivity of dorsal horn neurons. This implies that postsynaptic effects such as CREB phosphorylation in the spinal dorsal horn may also contribute to the participation of Ca_v1.2 in neuropathic pain (352, 353). It may relate to the finding that α2δ-1 remodels Ca_v1.2 voltage sensors and allows Ca²⁺ influx at physiological resting potentials (354).

Since we could only find one very old report of clinical effectiveness of classical dihydropyridine, nifedipine in complex regional pain syndrome (355), it is presently assumed that L-type Ca²⁺ channels play a far smaller role in the etiology of neuropathic pain than N- or T-types (see below). This position may however need revision in the light of recent descriptions of prevalent nifedipine sensitive channels in human DRG neurons (356).

Some novel benzodiazepines exhibit selective T-channel block (357) whereas others block both Ca_v1.2 L-type and Ca_v3.2 T-type calcium channels (358). To the best of knowledge there are no reports of the effectiveness of these agents in the clinic.

Ca_v2 channels are the main subtype found in primary afferent terminals (93, 359). Ca_v2.1 (N-type) and Ca_v2.2 (P/Q type) both contain a synaptic protein interaction site (synprint) that interacts with SNARE proteins (syntaxin and SNAP-25) (360, 361). By this mechanism, channels can be closely associated with synaptic vesicles that govern release of neurotransmitter from primary afferent terminals. Although suppression of N-type Ca²⁺ channel current increases the excitability of DRG cell bodies by concomitant decrease of BK function (236, 306), this effect is overridden in vivo by the actions of Ca_v2 blockers to prevent neurotransmitter release from primary afferent terminals.

As already mentioned, the Ca_v2.2 blocker ziconotide which is a synthetic version of ω-conotoxin MVIIA from the cone snail Conus magnus (362) is employed in pain management. The main drawback is that it needs to be delivered directly to the spinal cord via an intrathecal drug delivery system. Zicononotide (Prialt) is usually only effective in patients with severe, intractable forms of chronic pain such as that associated with cancer (82, 363).

There is therefore considerable interest in developing small molecule blockers of Ca_v2 channels that may be effective orally or perhaps by intravenous injection (24, 93, 339, 344). In our previous review (3) we drew attention to the state-dependent Ca_v2 blockers ZC88 (83, 84), A-1264087 (85, 364), and TROX-1 (86, 87, 365). Although all of these drugs display anti-allodynic efficacy in rodent models of neuropathic pain (344), there is as yet no evidence of any clinical efficacy.

Two tetrahydroisoquinoline derivatives have also been shown to display effectiveness in animal models (366, 367) but again clinical efficacy has not yet been demonstrated.

A permanently charged cationic derivative of an N-type calcium channel-blocker was recently synthesized (97). These authors anticipated that this charged compound (known as CNCB-2) would only be effective when applied intracellularly by a mechanism analogous to QX-314 block of Na⁺ channels (60). Surprisingly, extracellular application of CNCB-2 was more effective than intracellular application in inhibiting Ca_v2.2 channels. Inhibition was achieved without channel opening. Moreover, and quite unexpectedly, the compound was also highly effective in inhibiting Na_v1.7 when applied extracellularly. CNCB-2 reduced excitability of mouse DRG neurons and produced long lasting analgesia in several pain models. Given the seminal role of Na_v1.7 in the etiology of many forms of neuropathic pain (8, 114), bifunctional compounds such as CNCB-2, show considerable promise as therapeutic agents (Table 1).

Ca_v2.2 interacts with collapsin response mediator protein 2 (CRMP2) which directs the channels to presynaptic terminals (368). Interestingly it has been reported that impairment of CRMP2 function using a homopolyarginine (R9)-conjugated CBD3-A6K peptide inhibits Ca_v2.2-CRMP2 interaction, diminishes surface expression of Ca_v2.2 and alleviates tactile allodynia and ongoing pain in a rodent model (369). This observation suggests that CRMP2 may be developed as a novel therapeutic target.

N-type Ca²⁺ channels are modulated by G_i/o coupled agonists (157, 370). The α₂-adrenoceptor agonist, clonidine displays anti-allodynic actions in a rodent model (371) and meta-analysis of clinical trials reveals clinical efficacy (372). Effects of clonidine may be mediated by α₂-adrenergic inhibition of neurotransmitter release leading to modulation of pain processing at the spinal level (5) and/or by attenuation of aberrant interactions between sympathetic and sensory nerves in the periphery (156, 157, 373). Its effectiveness is however limited to subsets of patients within the diabetic neuropathy, complex regional pain syndrome or postherpetic neuralgia cohorts (88–92). In view of the restricted effectiveness of clonidine, it does not meet the criteria for first line treatment of neuropathic pain (1) (Table 1).

Gabapentinoids on the other hand are relatively but not completely effective in a variety of manifestations of neuropathic pain; about 31% of patients see clear benefit (96). Their mechanism is still incompletely understood but clearly involves impediment of transport of Ca_v2 channels to nerve terminals and their uncoupling from the neurotransmitter release process following interaction with their α2δ-1 accessory subunits (3, 94, 374). This occurs in both primary afferents and dorsal horn (95). Apart from the introduction of pregabalin (375) and an enacarbil derivative of gabapentin with improved oral bioavailability (376), there have been no major developments in the pharmacology of α2δ-1 ligands since their introduction in the 1990's. Since gabapentinoids act intracellularly, we have suggested that their effectiveness may be increased by allowing them to enter neurons via the open pore of TRPV1 channels (377).

Since Ca_v2.2 channels are found in pancreatic β-cells and are involved in the secretion of insulin (378) it remains to be established whether Ca_v2.2 blockers have undesirable effects on blood glucose levels. On the other hand, Ca_v2.2 has been implicated in microglial function (379, 380). This raises the possibility that some of the beneficial effects of Ca_v2.2 blockers result from actions on microglia.

T-type, LVA, Ca²⁺ channels (Ca_v3.1, Ca_v3.2, Ca_v3.3) play important roles in setting neuronal excitability (93, 336, 381) and in transmitter release from primary afferent terminals (382, 383). As with Ca_v2 channels, this later function may involve interaction of Ca_v3 channels with the synaptic vesicle release proteins syntaxin 1A and SNAP25 (synprint) (384). DRG neurons express Ca_v3.2 and 3.3 but not 3.1 (385–387). T-type calcium currents are increased in rodent DRG neurons after peripheral nerve injury in a model of diabetic neuropathy and after injury to the spinal cord per se (383, 388–390).

Although there are no reported mutations of Ca_v3.2 that produce a painful phenotype in humans, most of the work relevant to pain mechanisms has involved this channel as opposed to Ca_v3.1 or Ca_v3.3 (93, 334, 335, 339, 391–393). Ca_v3.2 is expressed in low-threshold mechanoreceptors and conditional knockout of the channel in this neuronal subtype has implicated Ca_v3.2 in allodynia linked to neuropathic pain (394). Several mechanisms control the functional expression of Ca_v3.2 channels.

(i) Upregulation of the deubiquitinase, USP5 by the action of the inflammatory mediator interleukin-1β. This impairs Ca_v3.2 ubiquitination thereby protecting it from proteasomal degradation and prolonging its surface expression (383, 395, 396). Knockdown of USP5 in vitro increases Ca_v3.2 ubiquitination and reduces Ca_v3.2 whole-cell currents and since impairment of USP5 function in vivo attenuates mechanical hypersensitivity in both inflammatory and neuropathic mouse models, this enzyme may represent a future therapeutic target (335, 383). Progress in this direction involves the observations that Ca_v3.2/USP5 interactions are interrupted by the anti-parasitic agent, suramin and by a TAT-cUBP1-USP5 peptide and both substances show analgesic activity in neuropathic and inflammatory pain models (99, 100) (Table 1).

(ii) Glycosylation and enhancement of channel trafficking in diabetic pain (397, 398). Deglycosylation of Ca_v3.2 with neuramidase reverses hyperalgesia in a model of diabetic neuropathy (398) (Table 1).

(iii) BDNF stimulation of TrkB coupled to PI3K-p38-PKA signaling in trigeminal neurons (399). Although a range of small molecule TrkB inhibitors are available (400), the multiple biological actions of BDNF in the developing and mature nervous system, preclude the use of these agents in pain management (401).

(iv) Ca_v3.2 channels interact with the scaffold protein Rack-1 [receptor for activated C kinase 1 (402)]. Whole-cell Ca_v3.2 current and channel expression in the plasma membrane is reduced when Ca_v3.2 and Rack-1 are co-expressed in tsA-201 cells. Molecular interaction between the two proteins was demonstrated by co-immunoprecipitation. These findings assume special significance in the light of the suggested role for Rack-1 in neuropathic pain (403).

Although T-type Ca²⁺ channel blockers such as the anticonvulsant ethosuximide increases withdrawal thresholds in nerve-injured rats (404), clinical studies of its effectiveness in pain management have been disappointing (98). A similar picture emerges for other small molecule blockers of Ca_v3.2, most of which showed considerable promise in preclinical studies yet failed to exert significant effects in cohorts of pain patients (334).

For example, ABT-639 showed promise in preclinical studies (405–407) but clinical results have been disappointing (86, 334); it did not treat pain in patients with diabetic neuropathy (408) and has now been discontinued.

Also, because TTA-P2 is a highly selective Ca_v3.2 channel blocker that has minimal effects on other cation channels, it is used extensively in laboratory investigations of T-channel function. Although it is effective in rodent models of chronic inflammatory pain and diabetic neuropathy (409) we could find no reports of its efficacy in the clinic.

Similarly, TTA-A2 is used extensively in laboratory investigations (395) as it has higher affinity for Ca_v3.2 than Ca_v31.1 (410). Although it is effective in rodent models of irritable bowel syndrome (410), no clinical studies appear to have been done.

Z944 is another high-affinity T-type channel blocker that is effective against Ca_v3.1, Ca_v3.2, and Ca_v3.3 with little affinity for other Ca²⁺ channel types (411). Its effectiveness in murine pain models may reflect it actions on spinal and thalamic neurons (412, 413). So far, the results of phase 1 and phase 2 trials appear promising (101) (Table 1). Although there does not seem to be any preclinical information regarding the effectiveness of the N-(1-benzyl-1H-pyrazol-3-yl)-2-phenylacetamide derivative ACT-709478 in animal models of neuropathic pain (414), it appears to be showing promise in phase 2 trials (101) (Table 1).

As already mentioned the has been shown to also selectively block Ca_v3.2 (49).

Cannabinoids, which are effective in some neuropathic pain cases (108, 109), inhibit recombinant human T-type (Ca_v 3.1, 3.2) Ca²⁺ channels (104) and as mentioned above, augment BK (K_Ca1.1) currents. Intrathecal injection of the CB1/CB2 receptor agonist NMP-7 inhibits injury-induced neuropathic pain in a rodent model. This effect involves CB2 receptors and Ca_v3.2 channels (415). To the best of our knowledge, NMP-7 has not yet progressed to clinical trials but its preclinical effectiveness led to the development of the derivative [N-((1-(2-(tertbutylamino)-2-oxoethyl)piperidin-4-yl)methyl)-9-pentyl-9Hcarbazole-3-carboxamide] (Compound 9) which displays remarkable effectiveness in murine models of inflammatory and neuropathic pain (105).

Several Na⁺ channel blockers show promise as therapeutic agents or as lead compounds for structural refinements (Table 1).

The first is the Na_v1.7 blocker, vixotrigine (CNV1014802, BIIB074, or GSK-1014802) (184, 185). The outcomes of a phase III clinical trial for effectiveness in trigeminal neuralgia (NCT03637387) and phase II trial for small fiber neuropathy are eagerly awaited (47).

The Na_v1.7 blocker PF-05089771 failed to meet defined efficacy criteria in patients with painful diabetic peripheral neuropathy (46). Since its use in clinical trials would have been contingent on establishment of safety for use in humans, it may serve as a safe lead compound for the development of more effective agents.

Certain natural toxins, notably those from various types of tarantula venom show selectivity and high affinity for Na_v1.7 as well as analgesic effects in various pain models. One of the most promising agents is Tap1a as this interacts with both Na_v1.7 and the T-type Ca²⁺ channel, Ca_v3.2 (49, 50). Recent studies of Tap1a have shown that it interacts with voltage-sensor domain II of Na_v channels with nanomolar affinity. Structural modification of Tap1a has produced two peptides Tap1a-OPT1 and Tap1a-OPT2 that exhibit increased affinity for Na_v1.1, Na_v1.2, Na_v1.3, Na_v1.6, and Na_v1.7. Intraplantar injection of Tap1a-OPT1 reduces Na_v1.7/OD1-induced spontaneous pain behaviors in a murine model. Moreover the anti-nociceptive effect of Tap1a-OPT1 is significantly greater than the native peptide (50).

Although the selective Na_v1.8 blockers A803467 and PF-01247324 attenuate allodynia in a rodent model (221, 222), they have not yet been examined in the clinic (223). The pro-drug VX-150 is metabolized into a highly selective Na_v1.8 blocker which exhibits analgesic activity in healthy volunteers (57). Expression of Na_v1.8 in peptidergic DRG neurons is controlled by NGF (215) and the NGF binding antibody tanezumab is effective in various pain states (58). In fact, its safety and efficacy in humans identifies tanezumab as one of more the promising new drug candidates for chronic and neuropathic pain (see Clinical Trials Government Identifiers: NCT02528188 and NCT02528188). Small molecule peripherally acting TrkA inhibitors have recently been described (432, 433).

We have also described the idea of combining cationic local anesthetics with TRPV 1 activators (60, 61, 233), although this seems to work well in animal models, this approach has not yet been demonstrated in a clinical situation.

Although a clinical study of retigabine in post herpetic neuralgia failed to meet its efficacy endpoint (68), there is considerable interest in its structure as template for ligand-based drug design of K_v7.2/3 activators (434); at least 200 K_v activators are currently under development (285). Certain natural products augment K_v currents and it has been suggested that these might augment retigabine effectiveness (71).

The sulphonylurea compound NS5806 which augments K_v4.3 type A-currents in animal models (75) is yet to be examined in the clinic.

A phase 2a clinical trial of the K_Ca3.1 channel opener, ASP0819 for fibromyalgia (NCT03056690), has provided evidence of efficacy with minimal side effects (321). As mentioned above, little, or no success has been realized with other direct activators of K_v1, 2, 3 or 4 or K_Ca1 or 2.

N-type Ca_v2 channels have been recognized as targets for anti-allodynic drugs for many years. The limitations to the use of the channel blocker ziconotide and the α2-adrenoceptor ligand clonidine have already been alluded to (363). Although gabapentinoids interact indirectly with Ca_v2 via their α2δ-1 subunits they are neither universally effective or without undesirable adverse effects (96). As mentioned above, a few small molecule Ca_v2 blockers are in development but none have as yet been tested in a clinical situation. The compound CNCB-2 is of special interest as it blocks both Ca_v2.2 and Na_v1.7 channels (97).

The potential role of Ca_v3 in neuropathic pain was established about 12 years ago (389, 413, 435) but the classical Ca_v3 blocker ethosuximide displays only limited effectiveness in the clinic (98). In the interim, several small molecule blockers have appeared such as TTA-P2 and TTA-A2 which are highly selective for Ca_v3.2. Clinical studies are yet to be initiated or reported. By contrast, phase 1 and 2 clinical studies with two compounds Z944 and ACT-70948 have yielded promising results (101). Interest in cannabinoid modulation of Cav3.1 and 3.2 has led to development of a series of small molecule channel blockers such as “compound 9,” although it is remarkably effective in preclinical studies clinical studies are yet to be initiated (105).