Histamine, histamine receptors, and neuropathic pain relief

Histamine, acting via distinct histamine H1, H2, H3, and H4 receptors, regulates various physiological and pathological processes, including pain. In the last two decades, there has been a particular increase in evidence to support the involvement of H3 receptor and H4 receptor in the modulation of neuropathic pain, which remains challenging in terms of management. However, recent data show contrasting effects on neuropathic pain due to multiple factors that determine the pharmacological responses of histamine receptors and their underlying signal transduction properties (e.g., localization on either the presynaptic or postsynaptic neuronal membranes). This review summarizes the most recent findings on the role of histamine and the effects mediated by the four histamine receptors in response to the various stimuli associated with and promoting neuropathic pain. We particularly focus on mechanisms underlying histamine‐mediated analgesia, as we aim to clarify the analgesic potential of histamine receptor ligands in neuropathic pain. Linked Articles This article is part of a themed section on New Uses for 21st Century. To view the other articles in this section visit http://onlinelibrary.wiley.com/doi/10.1111/bph.v177.3/issuetoc

It is estimated that the worldwide prevalence rate of neuropathic pain in the general population lies between 7% and 10%; however, this figure differs for different countries (van Hecke, Austin, Khan, Smith, & Torrance, 2014). The highest prevalence rates for neuropathic pain were recorded in Canada (17.9%) and in the United States (9.8-12.4%;VanDenKerkhof et al., 2016;Yawn et al., 2009), while a relatively low prevalence rate was noted in Austria and Netherlands (3.3% and 1%, respectively;Bouhassira, Lanteri-Minet, Attal, Laurent, & Touboul, 2008;Gustorff et al., 2008;Harifi et al., 2013). In the United Kingdom, France, and Brazil, it is reported that 7-10% of chronic pain sufferers have been affected by neuropathic pain (Fayaz, Croft, Langford, Donaldson, & Jones, 2016). As a consequence of extended life expectancy, it is predicted that the worldwide prevalence rate of neuropathic pain is likely to increase further, because this type of chronic pain occurs with many common age-related diseases. Neuropathic pain is triggered by a lesion within the somatosensory system, trauma, or to toxic effects of certain medications (Colloca et al., 2017;Yu et al., 2013). Pathological conditions that are responsible for the development of neuropathic pain through injury include metabolic diseases (e.g., diabetic neuropathy), infection (e.g., postherpetic neuralgia), vascular disease (e.g., stroke), trauma (e.g., orofacial neuropathy), and cancer (Campbell & Meyer, 2006). Neuropathic pain is a complex condition that can either be constant or periodic and presents with a range of different symptoms. These symptoms can increase throughout the day with clinically relevant morning-evening differences and can be affected by gender and underlying aetiology (Gilron, Bailey, & Vandenkerkhof, 2013). Sufferers of neuropathic pain have ongoing, spontaneous pain that has a significant negative impact on quality of life and daily functioning, including physical, emotional, and social well-being (Jensen, Chodroff, & Dworkin, 2007).
The mechanisms underlying neuropathic pain are complex and multidimensional. Numerous pathophysiological and biochemical changes cause morphological and functional adaptations in the nervous system, including an increase in excitatory neurotransmitters and neuropeptides, for example, histamine, bradykinin, 5-HT, and glutamate, which leads to hyperexcitability (Baron, Binder, & Wasner, 2010). Likewise, injured peripheral nerve fibres give rise to an intense and prolonged input of ectopic activity to the CNS and may induce secondary changes to the excitability of the spinal cord dorsal horn neurons (Colloca et al., 2017;Ossipov & Porreca, 2005).
These morphological changes and functional adaptations lead to abnormal sensory signs in patients with neuropathic pain presented as, for example, allodynia (pain due to a stimulus that does not normally activate the nociceptive system), hyperalgesia (an increased response to a stimulus that is normally painful), or sensory loss (hypoesthesia). Clinically, neuropathic pain manifests as evoked pain and presents in many ways such as burning, tingling, prickling, shooting, electric shock-like, jabbing, squeezing, spasm, or cold (Rice, Finnerup, Kemp, Currie, & Baron, 2018).
Animal models of neuropathic pain are essential in understanding the plethora of mechanisms that may drive neuropathic pain, allowing the field to identify potential therapeutic targets for the effective management of this condition. Extensive research in the pain field has developed and characterized a wide variety of animal models of neuropathic pain.
The four most commonly used models, also adopted in some studies discussed in this review, are chronic constriction injury (Bennett & Xie, 1988), partial sciatic ligation (Seltzer, Dubner, & Shir, 1990), spinal nerve ligation (Kim & Chung, 1992), and spared nerve injury (Decosterd & Woolf, 2000) models. These models aim to simulate some of the clinical features of neuropathic pain, in the preclinical setting (e.g., allodynia and hyperalgesia), because a large proportion of peripheral neuropathic pain models, which are currently used in research, share alterations in hindlimb cutaneous sensory thresholds following partial injury of a peripheral (usually sciatic) nerve. This is associated with the development of neuropathic pain symptoms, such as hyperalgesia and allodynia (Ma et al., 2003;Wall et al., 1979). There are also limitations associated with the use of animal models of neuropathic pain. These limitations are linked to challenges associated with (a) reliable and objective measures of behavioural responses to noxious stimuli, since animals cannot self-report and the experimenter can be biased, (b) appropriateness of the outcome measures, for example, sleep disturbance to reflect spontaneous pain, and (c) complexity of mechanisms underlying the development of neuropathic pain and their relevance to humans (Colleoni & Sacerdote, 2010).
Despite advances in the understanding of the underlying causes and mechanisms leading to the development and maintenance of neuropathic pain, 40% of Europeans who suffer from chronic pain did not achieve satisfactory pain control (Breivik, Collett, Ventafridda, Cohen, & Gallacher, 2006), and to date, no medication has shown long-term efficacy and tolerability for neuropathic pain conditions. A significant contributing factor to these limited therapeutic strategies is that neuropathic pain has different aetiology and pathophysiology to any other type of chronic pain, making the management of this type of chronic pain particularly difficult and challenging (Finnerup et al., 2015). Consequently, there is a pressing need for the identification of new therapeutic strategies to improve management of neuropathic pain that will directly improve the outcome for pain sufferers.
The histamine system has been a target for multiple therapeutic interventions. Recently, growing evidence has supported the use of selective ligands of histamine H 3 receptor and H 4 receptor for the treatment of neuropathic pain (Bhowmik, Khanam, & Vohora, 2012;Chaumette et al., 2018;Popiolek-Barczyk et al., 2018;Sanna, Mello, Masini, & Galeotti, 2018). Approval in the European Union for the use of pitolisant (Wakix™), an antagonist/inverse agonist of H 3 receptors for the treatment of narcolepsy (Kollb-Sielecka et al., 2017), presents an opportunity to explore its clinical use for other conditions. Consequently, it seems an appropriate time to reconsider the histamine system as a therapeutic target for the management of neuropathic pain. This review aims to summarize the most recent findings on the role of histamine and its effects, mediated by different subtypes of histamine receptors, on neuropathic pain, with particular regard to the mechanisms underlying histamine-mediated analgesia.

| Histamine and pain
Histamine (2-(4-imidazolyl)-ethylamine), one of the most extensively studied amino acid-derived neurotransmitters in the CNS and PNS, is involved in various physiological and pathological processes, including sleep-waking cycle, homeostasis, synaptic plasticity, and learning (Panula et al., 2015;Parsons & Ganellin, 2006;Pini, Obara, Battell, Chazot, & Rosa, 2016). Histamine is synthesized from the amino acid L-histidine through oxidative decarboxylation via catalysis with the histidine decarboxylase enzyme (Bodmer, Imark, & Kneubuhl, 1999;Haas, Sergeeva, & Selbach, 2008) and is arguably the most pleiotropic molecule in the human and animal body, being present in many cell types (Lindskog, 2017). Histamine is released by neuronal and non-neuronal sources and is responsible for many physiological processes, including the contraction of smooth muscles of the lungs, uterus, and intestine, secretion of gastric acid in the stomach, and vasodilation, and modulation of heart rate and contractility (Parsons & Ganellin, 2006). Histamine also functions as a neurotransmitter, within the nervous system, regulating a variety of body functions, such as temperature control, memory, wakefulness, and pain sensation (Panula & Nuutinen, 2013).
Histamine is a key mediator in the processing of nociceptive information, acting in an antinociceptive manner in the CNS while, conversely, in a nociceptive manner in the PNS (Khalilzadeh, Azarpey, Hazrati, & Vafaei Saiah, 2018). In the PNS, histamine is released in response to tissue injury/damage, and, through the sensitization of polymodal nociceptors resulting in increased firing rates, it contributes to the generation of pain hypersensitivity. In neuropathic pain, histamine released in the periphery by mast cells has been shown to play an important role in the development of hypersensitivity following nerve injury. This pathological process is associated with recruitment of macrophages and neutrophils, and as histamine is a powerful chemoattractant of mast cells, it regulates this recruitment (Smith, Haskelberg, Tracey, & Moalem-Taylor, 2007;Zuo, Perkins, Tracey, & Geczy, 2003). Interestingly, it was also observed that peripherally acting histamine could interact with mechanisms underlying pruritus (itch) and pain. Findings suggest that low concentrations of histamine, acting on sensory neurons, produce pruritus with a high concentration leading to pain (Baron, Schwarz, Kleinert, Schattschneider, & Wasner, 2001;Hough & Rice, 2011;LaMotte, Simone, Baumann, Shain, & Alreja, 1987;Parsons & Ganellin, 2006;Pini et al., 2016;Simone, Alreja, & LaMotte, 1991). There is also evidence to show that histamine-induced itch can convert into pain associated with neuropathic hyperalgesia (Baron et al., 2001). Indeed, multiple itch pathways were identified indicating the presence of distinct itch-generating types of neuron, one responsible for transmitting itch sensation and the other, ultimately, for transmitting pain (Usoskin et al., 2015). In contrast to histamine activity in PNS, multiple behavioural studies have shown that histamine injected directly into the various brain areas (e.g., somatosensory cortex or hippocampus) attenuated pain (Erfanparast, Tamaddonfard, Farshid, & Khalilzadeh, 2010;Tamaddonfard & Hamzeh-Gooshchi, 2014).
Histamine exerts its effects via four distinct GPCR subtypes: H 1 , H 2 , H 3 , and H 4 receptors. These receptors differ in their pharmacology and signal transduction properties (Panula et al., 2015;Parsons & Ganellin, 2006;Simons & Simons, 2011). Thus, histamine has differential effects on neuropathic pain dependent upon the histamine receptor subtype it is bound to. As mentioned previously, this review aims to summarize histamine-mediated effects on neuropathic pain. Therefore, the next sections of this review focus on mechanisms underlying histamine-mediated analgesia.

| Histamine receptors and pain
Excitatory histamine receptor signalling in nociceptive pathways is associated with increased pain symptoms (Gangadharan & Kuner, 2013;Mobarakeh et al., 2000), whereas inhibition of histamine receptor signalling predominantly causes neuroprotective and antinociceptive effects (Bhowmik et al., 2012;Chazot & Care, 2005; Popiolek- Barczyk et al., 2018). Different subtypes of histamine receptors are expressed in both presynaptic and postsynaptic neuronal membranes (Brown, Stevens, & Haas, 2001;Parsons & Ganellin, 2006;Zhang et al., 2013). Presynaptic histamine receptors function as autoreceptors or heteroreceptors providing either positive or negative feedback regulation of neurotransmitter release from the axon terminals into the synaptic cleft (Nieto-Alamilla, Marquez-Gomez, Garcia-Galvez, Morales-Figueroa, & Arias-Montano, 2016). It could be concluded that the resultant excitatory or inhibitory physiological effect of histamine receptors depends on the action of the neurotransmitter and the subsequent downstream cascade.
Specificity of localization of histamine receptors in different parts of the nervous system, on either presynaptic or postsynaptic membranes, is determined by their physiological relevance (Parsons & Ganellin, 2006). Among the four subtypes of histamine receptors, H 1 and H 2 receptors are predominantly identified postsynaptically (Brown et al., 2001;Connelly et al., 2009;Zhang et al., 2013), with the location of the H 4 receptor requiring further investigation (Connelly et al., 2009).
Cross-desensitization and agonist-induced heterodimerization of H 1 and H 2 receptors (Alonso et al., 2013) may suggest a possible partnership between histamine receptors. Expression of H 3 receptors was initially reported as exclusively presynaptic in the rat cerebral cortex (Arrang, Garbarg, & Schwartz, 1983;Clark & Hill, 1996), while postsynaptic expression of H 3 receptors could not be completely excluded (Nieto-Alamilla et al., 2016). H 3 receptors are predominantly expressed in neurons and, together with H 4 receptors, have higher affinity (nM range) for histamine than H 1 and H 2 receptors (μM range; Parsons & Ganellin, 2006). Expression of H 3 and H 4 receptors on the opposite sides of the synaptic cleft may contribute to their effects in neuropathic pain, although the neuronal topology of the H 4 receptor still remains controversial. The use of selective ligands for histamine receptors has led to a better understanding of the physiological and pathophysiological roles of these receptors. The next section summarizes the effects produced by histamine receptor ligands on neuropathic pain.
Besides their presynaptic or postsynaptic localization, the physiological effects of histamine receptors are, to a great extent, determined by the type of guanine nucleotide-binding proteins (G-proteins) to which they are coupled (Leung & Wong, 2017). The difference in underlying signalling pathways may directly determine the effect on pain perception produced by selective ligands, even when they act at the same histamine receptor, as described below.  (Mojtahedin, 2016). G s α subunit stimulates AC with consequent augmented production of cAMP and consequent activation of PKA (Brown et al., 2001). Elevated cAMP concentration up-regulates PKA, which as reported for hippocampal neurons, could activate ligand gated α-AMPA receptors with resultant depolarization and increase of neuronal firing (Park et al., 2016). Also, as it was reported for afterhyperpolarization in enteric neurons, PKA inhibits small conductance

| H 3 receptor
Presynaptic H 3 receptors are coupled with G i (AC inhibitory) proteins (Nieto-Alamilla et al., 2016;Schlicker & Kathmann, 2017 phosphorylation. This suppresses the activity of potassium voltage-gated channels type 7 (K v 7 channels), which depolarizes the neurons, and leads to the augmentation of neuronal excitability, which manifests as increased pain symptoms FIGURE 2 Diagram illustrating histamine receptor signalling-G s pathway. Histamine binds to the H 2 receptor subtype that is coupled with the G s -type protein. G s α subunit activates AC, with subsequent production of cAMP, which then stimulates PKA. PKAdependent phosphorylation activates ligand gated α-AMPA receptors, which open and facilitate influx of Na + and, less commonly, Ca 2+ . PKA also suppresses K + efflux through small conductance Ca 2+ -activated potassium channels (SK channels). Both the activation of AMPA receptors and inhibition of SK channels depolarize the neurons, with consequent augmentation of their excitability and increased pain signalling was reported that H 3 receptor activation stimulated dissociation of G i β and γ subunits from G i α subunit, which then inhibited voltage-gated Ca 2+ influx through N-, P-, and Q-type Ca 2+ channels (Zamponi & Currie, 2013), and stimulated G protein-coupled inwardly rectifying potassium (Kir) channels (Luscher & Slesinger, 2010). Both mechanisms could hyperpolarize presynaptic neurons, reduce neuronal excitability, and produce pain relief ( Figure 3).
Besides the G i pathway, postsynaptic activation of H 3 receptors was reported to stimulate PLC in a subpopulation of striatal neurons, with subsequent activation of the IP 3 pathway followed by increased intracellular concentrations of Ca 2+ (Rivera-Ramírez et al., 2016). Thus, it is analogous to the mechanisms described for H 1 /H 2 receptors coupled to G q proteins ( Figure 1). Similarly, the H 1 and H 2 receptor G q cascade PLC signalling pathways modulate neuronal excitability with resultant potential facilitation of pain sensitivity. Furthermore, H 3 receptor activation was established to inhibit glycogen synthase kinase 3β (GSK3β) and MAPK/ERK pathways (Bongers et al., 2007;Morita, Aida, & Miyamoto, 1983). These effects are translated via G i β and γ subunits (Lai et al., 2016), which up-regulate the PI3K pathway with subsequent production of phosphatidylinositol 3,4,5triphosphate (PIP 3 ) from PIP 2 , which results in the recruitment of PKB (Akt). PKB is initially activated by phosphoinositide-dependent kinase 1 (PDK1) and mammalian target of rapamycin complex 2-dependent phosphorylation (Dzamko, Zhou, Huang, & Halliday, 2014). PKB, via phosphorylation, inactivates GSK3β, which decreases neuronal excitability (Paul et al., 2016), inhibits neuronal inflammation (Maixner & Weng, 2013), and, subsequently, relieves pain. PKBdependent phosphorylation that activates the MAPK/ERK cascade was also reported to be a highly efficient neuroprotective mechanism for chronic inflammatory and neuropathic pain (Cruz & Cruz, 2007; Figure 3).

| H 4 receptor
The role of H 4 receptors in the nervous system is poorly understood (Schneider & Seifert, 2016 that are coupled with G i -type protein. The G i α subunit inhibits AC with subsequent suppression of cAMP production and inhibition of PKA activity. Also, G i β and γ subunits can inhibit Ca 2+ influx through voltage-gated N-, P-, and Q-type Ca 2+ channels and stimulate G protein-coupled inwardly rectifying potassium (GIRK) channels, with resultant K + efflux. Both effects on Ntype Ca 2+ and GIRK channels result in the development of hyperpolarization, attenuation of neuronal excitability, and resultant pain relief. Besides G i α subunit effects, H 3 receptor activation could produce analgesic effects through G i β and γ subunits, which up-regulate the PI3K pathway, with the subsequent production of phosphatidylinositol 2,4,5-trisphosphate (PIP 3 ) from phosphatidylinositol 4,5bisphosphate (PIP 2 ). PIP 3 recruits PKB (Akt), which phosphorylates and inactivates glycogen synthase kinase 3β (GSK3β). In parallel, PKB (Akt)-dependent phosphorylation additionally activates the MAPK/ERK cascade. The action on both GSK3β and MAPK/ERK decreases neuronal excitability, inhibits mechanisms of neuronal inflammation, and, therefore, produces pain relief in turn results in the excitation of histamine-sensitive afferents and, therefore, may result in a modulation of pain sensitivity (Rossbach et al., 2011).

| H 1 and H 2 receptor ligands and neuropathic pain
The best-known roles for the H 1 receptor are regulation of vasodilation and bronchoconstriction on multiple cell types, including endothelial and smooth muscle cells, while the H 2 receptor is primarily involved in the modulation of gastric secretion on parietal cells (Barocelli & Ballabeni, 2003;Simons, 2003). There is also evidence for their expression in the nervous system, where they regulate some neuronal functions (Haas et al., 2008;Kashiba, Fukui, & Senba, 2001;Murakami et al., 1999).
Both H 1 and H 2 receptors have been implicated in the role of histamine in nociception and chronic pain (Table 1) (Kashiba et al., 2001). The earliest electrophysiological study to probe the histaminergic system in neuropathic pain transmission reported that daily injections of astemizole, a CNS-sparing H 1 receptor antagonist, acting via peripheral histaminesensitive C-fibres, blocked ectopic spontaneous discharges from the neuroma and suppressed autotomy following sciatic and saphenous neurectomy (Seltzer et al., 1991;Yu et al., 2013). More recently, i.c.
v. injection of histamine blocked mechanical and thermal hypersensitivity associated with neuropathic pain (Sanna et al., 2015;Wei et al., 2016). However, these pain-modulatory effects of histamine can vary depending on the dose of histamine administered i.c.v., due to an action on H 2 receptors and the involvement of adrenoceptors (Wei et al., 2014;Wei et al., 2016).  was prevented by spinal pretreatment with zolantidine, a brain penetrating H 2 receptor antagonist, as well as localized peripheral administration of cimetidine (H 2 receptor antagonist) and chlorpheniramine (H 1 receptor antagonist) into the plantar side of the hindpaw (Zuo et al., 2003).
Taken together, these results indicate that the brain histamine,  pain-control pathway, that are critical for the processing of nociceptive information. Within the CNS, this receptor has been found in various brain areas, such as thalamus, hypothalamus, prefrontal cortex, and periaqueductal grey area (Drutel et al., 2001), and in the spinal cord (Cannon et al., 2007;Heron, Rouleau, Cochois, Pillot, & Schwartz, 2001;Medhurst et al., 2008). In the periphery, the expression of H 3 receptors has been identified in DRG, superior cervical ganglia, and dermal tissues (Cannon et al., 2007;Medhurst et al., 2008).

| H 3 receptor ligands and neuropathic pain
However, while the localization of H 3 receptors strongly suggests its functional involvement in the regulation of nociceptive transmission, pharmacological studies using agonists and antagonists of H 3 receptors are confusing as these drugs have different effects on the nociceptive threshold depending on the pain model used, the nociceptive stimulus selected, together with the affinity and selectivity for the histamine receptors, and the dose and routes of administration (Huang et al., 2007;Smith et al., 2007). Several studies have reported inhibitory effects on pain following activation of H 3 receptors using agonists (Cannon et al., 2003;Hasanein, 2011). The involvement of H 3 receptors in neuropathic pain has been implicated using a range of H 3 receptor antagonists/inverse agonists (Table 1) agonists restores LC and decreases α 2 adrenoceptor activity, respectively, potentially leading to relief in neuropathic pain hypersensitivity (Chaumette et al., 2018;Wei et al., 2010). To further support this proposed mechanism, it was shown that bilateral lesion of the LC, transection of the spinal cord, or direct injection of a α 2 agonist (fadolmidine) into the LC reversed the antihyperalgesic effect produced by H 3 receptor antagonists, A-960656 or GSK189254 (Chaumette et al., 2018;McGaraughty, Chu, Cowart, & Brioni, 2012;Wei et al., 2014 Gemkow et al., 2009;Giannoni et al., 2010).
Interestingly, the majority of the behavioural observations published indicate that H 3 receptor antagonists/inverse agonists do not produce any antinociceptive effects in naïve rodents, suggesting a possibility that H 3 receptors are not involved or tonically activated in nociception (at least in relation to acute mechanical nociception), but are critical for pathological pain states, particularly for mechanical hypersensitivity (Chaumette et al., 2018;McGaraughty et al., 2012).
Also, H 3 receptor KO mice showed unaltered response to mechanical pinch (Cannon et al., 2003), and multiple studies suggest modality (mechanical vs. heat) and intensity (preferential responses to lowintensity tail pinch stimulation; Cannon et al., 2003) with specific antinociceptive effects mediated by H 3 receptors. H 3 receptor antagonists/inverse agonists at a dose that produced a significant reduction of mechanical hypersensitivity in neuropathic pain did not attenuate heat hypersensitivity indicating that the antihyperalgesic effect was due to selective depression of spinal sensory rather than motor neurons (Wei et al., 2014;Wei et al., 2016). To support this, in situ hybridization studies revealed H 3 receptor mRNA transcripts in the sensory neurons of the dorsal horn and DRG (Heron et al., 2001 (Cannon et al., 2007). In addition, Medhurst et al. (2007) showed that GSK207040 and GSK334429, selective H 3 receptor antagonists, blocked the secondary mechanical allodynia in the capsaicin-induced model of pain. Secondary mechanical hypersensitivity is known to be exclusively signalled by A-fibres and amplified by sensitized dorsal horn neurons (Magerl, Fuchs, Meyer, & Treede, 2001;Treede & Magerl, 2000). Thus, presumably reducing the sensitivity of H 3 receptor-positive A-fibres with selective H 3 receptor antagonists resulted in a diminished input to the dorsal horn and the subsequent amplification of the A-fibres response, confirming the potential role for H 3 receptors in the modulation of central sensitization processes. In contrast, thermal (heat) hypersensitivity is generally regarded as a sign of the peripheral sensitization of C-fibres, which do not express H 3 receptors (Cannon et al., 2007;Gold & Gebhart, 2010).
The only study that reported a significant increase in thermal (radiant heat in Hargreaves test), but not mechanical, threshold in the spared nerve injury model of neuropathic pain used the selective H 3 receptor inverse agonist pitolisant (Wakix™). The drug produced this unexpected effect at a dose 5× higher than its clinically relevant dose, and pharmacological analysis of this effect suggested at least partial involvement of transient receptor potential cation channel subfamily V member 1 (TRPV1), without any contribution of H 3 receptors (Zhang et al., 2012). Interestingly, the H 3 receptor antagonist/inverse agonist E-162, at a dose that produced a significant reduction in mechanical hypersensitivity, also attenuated the response to cold in neuropathic pain (Popiolek-Barczyk et al., 2018). The signalling of cool temperatures that become aversive in neuropathic pain is known to be mediated via the TRPM8 receptor, a member of the TRP channel family (Knowlton et al., 2013). It was reported that the number of TRPM8positive Aδ-fibres (but not C-fibres) increases after nerve injury (Ji, Zhou, Kochukov, Westlund, & Carlton, 2007) dose of thioperamide administration. In addition, the involvement of other histaminergic mechanisms of action in the behavioural effects produced by thioperamide is suggested by the observation that thioperamide increases the density of intact mast cells in the injured nerve (Smith et al., 2007). While nerve injury causes a decrease in mast cell numbers as a consequence of degranulation (Zuo et al., 2003), thioperamide's action leads to an opposite effect that may represent a prevention of mast cell degranulation and stabilization or redistribution of mast cells in the injured nerve that theoretically would result in the inhibition of hyperalgesia, rather than its enhancement. This effect could be linked to the observation that histamine (acting through H 1 and H 3 receptors) inhibits the release of the proinflammatory cytokine TNF-α from alveolar macrophages (Sirois, Ménard, Moses, & Bissonnette, 2000), and antagonism of H 3 receptors on macrophages resulted in an increase in TNF-α and, subsequent, enhancement of mechanical hyperalgesia (Smith et al., 2007).
Taken together, the interpretation of the thioperamide data is complicated further since the drug has high affinity, not only for H 3 and H 4 receptors but also for 5-HT3 receptors (Leurs et al., 1995).
Studies with more selective H 3 receptor antagonists/inverse agonists suggest that these ligands may be beneficial for the improvement of mechanical and cold hypersensitivity associated with neuropathic pain, particularly given their ability to modulate histamine levels, as well as several neurotransmitters, including ACh, histamine, noradrenaline, dopamine, and SP. However, due to the wide presynaptic and postsynaptic distribution of H 3 receptors throughout the CNS and PNS, more research is certainly needed to clarify the involvement of peripheral, spinal, and brain H 3 receptors in various neuropathic pain states, thus determining their full potential in neuropathic pain.

| H 4 receptor ligands and neuropathic pain
The H 4 receptor, which has low homology with other histamine receptors, can be primarily found in bone marrow, intestinal tissue, spleen, thymus, and also in various immune cells, such as T cells, mast cells, neutrophils, and eosinophils, showing modulatory effects on these cells, including activation, migration, and production of cytokines and chemokines, suggesting its principal role in the regulation of immune/inflammatory mechanisms (Takeshita, Sakai, Bacon, & Gantner, 2003;Zhu et al., 2001). Interestingly, recent reports also indicate the presence of H 4 receptors on peripheral sensory nerves, in the DRG, with more intense staining of small-and medium-diameter cells, and in the spinal cord, especially laminae I and II (Sanna, Lucarini, et al., 2017;Strakhova et al., 2009). This neuronal localization supports H 4 receptors involvement in the regulation of neuronal function related to the modulation of nociceptive transmission (Sanna, Ghelardini, Thurmond, Masini, & Galeotti, 2017;Sanna, Lucarini, et al., 2017).
The involvement of H 4 receptors in both acute (Galeotti, Sanna, & Ghelardini, 2013) and persistent inflammatory pain (Hsieh et al., 2010) is relatively well documented, and recently, the role of H 4 receptors in the modulation of neuropathic pain was identified in H 4 receptor-KO mice through the observation that these animals, when subjected to neuropathic pain, induced by spared nerve injury of sciatic nerve, showed enhanced hypersensitivity to mechanical and thermal stimuli compared to wild-type controls .
Interestingly, H 4 receptor deficiency does not support a role for H 4 receptors in the physiological maintenance of pain threshold, as H 4 receptor-KO mice did not show any change in thermal or mechanical nociceptive thresholds, suggesting that the H 4 receptor is specifically involved in the regulation of hypersensitivity associated with pathological chronic pain induced by nerve injury . This observation in H 4 receptor-KO neuropathic mice is particularly important as H 4 receptor mRNA expression in humans and rodents supports their involvement in the regulation of neuronal function, including regulation of neuropathic pain. The controversy around the generation of consistently specific H 4 receptor antibodies highlights the need for cautious interpretation of some of the immunohistochemical outcomes (Beermann, Seifert, & Neumann, 2012;Gutzmer et al., 2012;Schneider & Seifert, 2016). In line with the observation from H 4 receptor KO mice, blockade of H 4 receptors by the specific H 4 receptor antagonist JNJ7777120, injected s.c. directly into the operated hindpaw, resulted in a significant increase in mechanical hyperalgesia compared to controls (Smith et al., 2007). Subsequently, activation of H 4 receptors by localized administration of potent and selective agonists, ST-1006 (i.c.v.) and VUF8430 (i.c.v., intrathecally, and s.c. directly into the operated hindpaw), resulted in a significantly reduced mechanical and thermal withdrawal threshold in mice subjected to neuropathic pain induced by spared nerve injury or partial nerve ligation of the sciatic nerve (Sanna et al., 2015;Sanna, Lucarini, et al., 2017;Smith et al., 2007). The analgesia produced by VUF8430 has been shown to be associated with a reduction in neuroinflammation and oxidative stress mediated by neuronal H 4 receptors in the spinal cord and sciatic nerve (Sanna, Lucarini, et al., 2017), and the involvement of H 4 receptors in the behavioural effects produced by ST-1006 and VUF8430 was confirmed with JNJ10191584, H 4 receptor antagonist also known as VUF6002, which fully prevented the analgesic effects produced by these H 4 receptor agonists (Sanna et al., 2015).
Interestingly, similar to the H 3 receptor, pharmacological studies using agonists and antagonists of H 4 receptors demonstrate that these drugs can have different effects on the nociceptive threshold depending on the routes of administration and target cells (Popiolek-Barczyk et al., 2018;Sanna et al., 2015). In contrast to the studies above that  (Takeshita et al., 2003;Zhu et al., 2001), it is possible that the antinociceptive action of H 4 receptor antagonists, particularly after systemic administration, BJP may result from a reduction in ongoing inflammatory processes at the site of nerve injury, since the analgesic effect produced by JNJ7777120 was weaker (secondary) than its anti-inflammatory effect (Hsieh et al., 2010). An underlying mechanism may be associated with stabilization of mast cells that are known to regulate the recruitment of neutrophils and macrophages and, subsequently, to modulate the development of hyperalgesia in neuropathic pain (Smith et al., 2007;Zuo et al., 2003). It should be further noted that similar observations have been described for the closely related H 3 receptor.

| HISTAMINE AND NON-NEURONAL CELLS IN NEUROPATHIC PAIN
Following peripheral nerve injury, the immune system seems to play a vital role in the development of persistent inflammation and chronic neuropathic pain (Marchand, Perretti, & McMahon, 2005). Nonneuronal astrocytes, satellite glia cells, microglia, and mast cells play key roles in communication between the immune system and the CNS via the production of neuroinflammatory mediators, including histamine, 5-HT, chemokines, and growth factors (Zhuang, Gerner, Woolf, & Ji, 2005). The neuroimmune interactions between these two systems may reflect distinct roles in the development of chronic neuropathic pain (Zhao et al., 2017). Stimulation of H 1 receptors via a PKC/MAPK/MEK1 signalling pathway has recently been shown to elicit release of the key pro-inflammatory cytokines IL-1β and IL-6 with, subsequent, regulation of nerve growth factor release from astrocytes (reviewed recently in Jurič, Kržan, & Lipnik-Stangelj, 2016). Satellite glial cells, prominent in the PNS, including the DRGs, with active roles in persistent neuropathic pain are also known to secrete the cytokine IL-6 in the chronic constriction injury neuropathic pain model, but the identity profile of the histamine receptor in these cells has yet to be established (Dubový, Klusáková, Svíženská, & Brázda, 2010). Activated microglia also release a myriad of pro-inflammatory cytokines, including notably, IL-6, IL-1β, and TNF-α (Kempuraj et al., 2016;Mika, Zychowska, Popiolek-Barczyk, Rojewska, & Przewlocka, 2013). H 3 and H 4 receptor activation of primary and clonal microglia has been shown to inhibit these cytokines (Ferreira et al., 2012).
Mast cells are professional cellular suppliers of histamine and contribute to the histamine-based effects in neuropathic pain. For example, mast cell depletion prevented mechanical allodynia in a mouse model of postoperative pain .
Recently, it was shown that the administration of azelastine hydrochloride, a second-generation H 1 receptor antagonist and mast cell stabilizer, blocked the development of mechanical allodynia and inhibited mast cell degranulation in mice with oxaliplatin-induced mechanical allodynia pain. The H 1 and H 4 receptors are likely molecular players in this process (Sakamoto, Andoh, & Kuraishi, 2016). For example, genetically silencing the H 4 receptor inhibited the production of IL-1β for human mast cells (Ebenezer, Prasad, Rajan, Thangam, & Transduction, 2018), and H 4 receptor activation was shown to stimulate a number of cytokines, including IL-6 (Jemima, Prema, & Thangam, 2014). In addition, while H 1 and H 2 receptor antagonism reduced hypersensitivity following nerve injury, it is possible that histamine released by mast cells contributes to the recruitment of neutrophils and macrophages in neuropathic pain and, acting via these histamine receptors, contributes to the regulation of hypersensitivity in this type of chronic pain Smith et al., 2007;Zuo et al., 2003).
There is an important aspect associated with H 2 receptor antagonism, which should be considered for its therapeutic potential in neuropathic pain control. In vitro studies using CHO and HEK-293 cells identified time-and dose-dependent up-regulation of H 2 receptors upon long-term exposure to H 2 receptor antagonists (e.g., ranitidine), which may underlie the development of tolerance after prolonged clinical use of these ligands and result in the rebound hypersecretion of gastric acid and anaphylaxis that can occur after withdrawal of treatment (Allen, Chazot, & Dixon, 2018;Smit et al., 1996). Thus, side effects linked to pharmacological tolerance may potentially compromise long-term efficacy and tolerability of H 2 receptor antagonists in neuropathic pain. Little is known about the role of the H 3 receptors in non-neuronal cells in neuropathic pain states.
Overall, non-neuronal cells play a key, but poorly, defined role in the mechanisms underlying histamine-mediated neuropathic pain.
We propose that neuronal H 1 and H 4 receptors (Ferreira et al., 2012) may orchestrate these mechanisms, with IL-6 and IL-1β cytokines as common denominator mediators. Further complications arise from the recent observation that activated mast cells trigger microglial activation (Zhang, Wang, Dong, Xu, & Zhang, 2016). These cell types and their interactions may potentially go some way to explain the paradoxical effects of histamine ligands, particularly for the H 4 receptor, seen in animal pain models.

| HISTAMINE AND ITS INTERACTION WITH OPIOID SYSTEM IN NEUROPATHIC PAIN
Interestingly, in neuropathic pain, high doses of opioids are required to achieve pain relief, and pharmacological tolerance to analgesic effect of opioids develops rapidly (Osikowicz, Mika, Makuch, & Przewlocka, 2008). This phenomenon significantly restricts the clinical usefulness of opioids. In addition, the misuse of and addiction to opioids, including prescription pain relievers, morphine and heroin, as well as synthetic opioids, such as fentanyl, is a serious international crisis that affects public health as well as social and economic welfare (Lipman & Webster, 2015).
An interaction between histaminergic and opioidergic systems within the CNS was suggested nearly 30 years ago, through an observation that morphine administration resulted in the release of histamine and its increased turnover in the periaqueductal grey (Nishibori, Oishi, Itoh, & Saeki, 1985), suggesting that analgesia produced by opioids may be associated with the stimulation of histamine receptors at the supraspinal level. There are also data suggesting that ligands of histamine receptors may modulate the analgesic action of opioids; however, the site and mode of this interaction differ between the spinal or supraspinal level, and depend on the subtype of histamine receptor involved (Mobarakeh et al., 2002;Mobarakeh et al., 2006;Mobarakeh, Takahashi, & Yanai, 2009 To the best of our knowledge, the literature does not provide evidence for the mechanisms underlying histamine and opioid system interactions, in relation to the modulation of morphine analgesic effects. Given that the analgesic effects produced by modulation of the activity of both the histamine and opioid systems could be associated with blocking SP release from peripheral nerve terminals (Barnes et al., 1986;Przewłocki & Przewłocka, 2001), it is possible that an interaction that would result in potentiation of analgesic efficacy of morphine may involve, together with other possible mechanisms, the inhibition of peripheral SP accumulation. Such an outcome may be useful for the management of neuropathic pain, particularly when peripheral administration of drugs is possible, thus affording reduction of the undesired secondary effects associated with opioid administration and peripheral mechanisms of action (e.g., constipation). However, centrally acting drugs administered by peripheral routes should be taken into consideration due to the potential serious interactions related to their pharmacodynamics and central mechanisms of action.
For example, chlorpheniramine (a first-generation H 1 receptor antagonist) was reported to potentiate fentanyl-induced sedation and respiratory depression after surgery (Anwari & Iqbal, 2003).

| CONCLUSIONS AND FUTURE DIRECTIONS
Findings from the last two decades indicate that selective pharmacological antagonism of neurons expressing H 3 receptors could provide important and promising therapeutic approaches for the control of mechanical and cold hypersensitivity in peripheral neuropathies (Table 1). The analgesic effectiveness of H 3 receptor antagonists/ inverse agonists was comparable to gabapentin and pregabalin, firstline treatments for neuropathic pain. Importantly, multiple examples of behavioural, electrophysiological, and molecular evidence strongly support the rationale for this neuropathic pain strategy, particularly given their ability to modulate histamine levels as well as several neurotransmitters critical for chronic pain processing. Moreover, the recent registered approval of pitolisant (Wakix™), an antagonist/ inverse agonist of H 3 receptors, for the treatment of narcolepsy in patients, has opened the door for the potential use of H 3 receptor ligands for other conditions, including chronic neuropathic pain. However, due to the wide presynaptic and postsynaptic distribution of H 3 receptors throughout the CNS and PNS, more research is certainly needed to clarify the involvement of peripheral, spinal, and brain H 3 receptors in various pain states, before determining their full potential in neuropathic pain.
Recent findings also suggest the use of centrally permeable H 2 receptor antagonists as promising new drug candidates for the treatment of neuropathic pain, in view of their analgesic effects and metabolic stability. Interestingly, however, despite the discovery of the most recently discovered histamine receptor, the role of the H 4 receptor in neuropathic pain transmission is still controversial after nearly 20 years, with apparent confounding effects of both agonists and antagonists on hypersensitivity associated with neuropathic pain. This may be due to biased signalling of histamine and H 4 receptor agonist ligands and differential effects on multiple signalling pathways in central and peripheral parts of the sensory nervous system. Furthermore, the paucity of detailed mechanistic definitions of histamine-mediated analgesia, and the additive effects with the opioid system, requires attention to provide a rationale to the field of histamine and development of neuropathic pain control therapeutics.
A better understanding of the interaction between histaminergic signalling pathway molecules (Figures 1-3) and histamine receptors may result in the identification of further novel pharmacological targets to improve neuropathic pain control. The literature available provides some evidence for potential pharmacological target molecules. One potential strategy exploits the role of Ca 2+ channels in the regulation of cellular excitability associated with nociception (e.g., N-type Ca 2+ channels). Evidence has shown that Ca 2+ channel blockers (e.g., ω-conotoxin-MVIIA/Prialt) offer interesting analgesic potential in treating neuropathic pain (Vanegas & Schaible, 2000).
However, there is no evidence for the effect produced by a combination of Ca 2+ channel blockers and histamine receptor ligands, and we propose that their interaction should be taken into consideration.
Another potential target involves the contribution of the MAPK/ERK signalling pathway to the regulation of pain hypersensitivity. Recently, Sanna et al. (2015) showed that H 4 receptor stimulation, which led to analgesic activity in neuropathic pain, was modulated by MAPK/ERK signalling in the neurons of the DRG, spinal cord, and sciatic nerve. While the MAPK/ERK signalling pathway regulates pain sensitivity and, for a while, has been considered as a target for the treatment of neuropathic pain (Ma & Quirion, 2005), further studies on the interaction between this pathway and H 4 receptors may lead to the identification of more efficient therapeutic strategies to control neuropathic pain.

| Nomenclature of targets and ligands
Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org/, the common portal for data from the IUPHAR/BPS Guide to PHARMA-COLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 (Alexander, Christopoulos et al., 2017;Alexander, Fabbro et al., 2017;Alexander, Striessnig et al., 2017).

Supported by the funding from the Medical Research Council
Confidence in Concept scheme to Drs P. L. Chazot and I. Obara (MC/PC/17157). We acknowledge Drs Janet Chazot and Natalie Young for their editorial assistance on the manuscript.