Cannabinoid control of neurogenic inflammation

Abstract

A significant number of cannabinoids are known to have analgesic and anti‐inflammatory properties in various diseases. Due to their presynaptic/terminal location, cannabinoid receptors can inhibit synaptic transmission and have the potential to regulate neurogenic inflammation. Neurogenic inflammation occurs when a noxious signal is detected in the periphery initiating an antidromic axon reflex in the same sensory neurone leading to depolarization of the afferent terminal. Neuropeptides are subsequently released and contribute to vasodilation, plasma extravasation and modulation of immune cells. Endocannabinoids, synthetic cannabinoids and phytocannabinoids can reduce neuroinflammation by inhibiting afferent firing and inflammatory neuropeptide release. Thus, in addition to a direct effect on vascular smooth muscle and inflammatory cells, cannabinoids can reduce inflammation by silencing small diameter neurones. This review examines the neuropharmacological processes involved in regulating antidromic depolarization of afferent nerve terminals by cannabinoids and the control of neurogenic inflammation in different diseases.

Abbreviations

2‐AG

2‐arachidonoylglycerol

ACEA

arachidonyl‐2′‐chloroethylamide

CBD

cannabidiol

CPZ

capsazepine

ECS

endocannabinoid system

NAPE

N‐arachidonoylphosphatidylethanolamide

NAPE‐PLD

N‐arachidonoylphosphatidylethanolamide PLD

NGF

nerve growth factor

PEA

palmitoylethanolamide

PTX

pertussis toxin

SP

substance P

THC

tetrahydrocannabinol

TRP

transient receptor potential

VIP

vasoactive intestinal protein

1. INTRODUCTION

Cannabis is a fast growing flowering plant originating from Central Asia, although its exact origins are obscure (Clarke & Watson, 2007; Gaoni & Mechoulam, 1964). The earliest record of cannabis use was in 2700 bc where the great Chinese deity Shennong described the hallucinatory effects of cannabis, while also noting the plant stimulated appetite and reduced painful gout (Clarke & Watson, 2007). There are three main strains of cannabis, namely, Cannabis sativa, Cannabis ruderalis, and Cannabis indica. C. sativa in particular has been used as an anaesthetic, analgesic, anxiolytic, anti‐epileptic, and an appetite stimulant in many conditions (Russo, 2014). The discovery of other medical properties of cannabis has expanded to encompass anti‐inflammatory, anti‐cancer, and neuroprotective actions (Russo & Marcu, 2017).

Currently, there are over 545 known compounds in C. sativa (El Sohly & Gul, 2014). Over 100 of these molecules are cannabinoids (C₂₁ terpenophenolic skeleton molecules) (El Sohly & Gul, 2014) and over 200 terpenoids (isoprene units with C₁₀ or C₁₅ skeletons) have been identified (Russo & Marcu, 2017). These chemicals are secreted by glandular trichromes on female plants and have been shown to have beneficial effects for the treatment of inflammation and pain (Andre, Hausman, & Guerriero, 2016; Russo & Marcu, 2017).

In addition to these plant‐derived phytocannabinoids, endogenous cannabinoids have been isolated in mammals and are referred to as endocannabinoids (Lu & Mackie, 2016). The endocannabinoid system (ECS) is composed of three main constituents:‐ cannabinoid receptors, endogenous cannabinoids and metabolic enzymes for cannabinoid synthesis and degradation (Lu & Mackie, 2016). A list of the cannabinoids described in this review can be found in Table 1.

TABLE 1.

List of cannabinoids and their known targets

Cannabinoid	Target/Receptor
Endocannabinoids
N‐arachidonoyldopamine	CB1, TRPV1
anandamide	CB1, CB2, TRPV1, GPR55, PPARα, PPARγ
2‐Arachidonoyl glycerol (2‐AG)	CB1, CB2, GPR55
Noladin ether	CB1, CB2, GPR55, PPARα
Phytocannabinoids
Cannabidiol (CBD)	CB2, GPR55, TRP
Δ9‐tetrahydrocannabinol (THC)	CB1, CB2
Synthetic cannabinoids
Arachidonyl‐2′‐chloroethylamide (ACEA)	CB1
CP55,940	CB2
GW833972	CB1, CB2
GW842166	CB2
HU210	CB1, CB2
JWH‐015	CB2
JWH‐133	CB2
R‐(+)‐methanandamide	CB1
WIN55,212‐2	CB1, CB2
Antagonists/inverse agonists
AM251	CB1
AM630	CB2
SR141716A	CB2

Cannabinoid receptors were discovered in 1990 (Matsuda, Lolait, Brownstein, Young, & Bonner, 1990), the CB₁ receptor is a GPCR with seven‐transmembrane spanning domains (Alexander et al., 2019; Howlett et al., 2002). The CB₁ receptor is predominantly located in the CNS (Herkenham et al., 1990) but has also been detected on the terminals of peripheral nerves (Richardson, Aanonsen, & Hargreaves, 1998), the gastrointestinal system and the reproductive system (Zou & Kumar, 2018). Localized to the terminals of nerve fibres (Howlett, 2002; Nyíri, Cserép, Szabadits, Mackie, & Freund, 2005), the CB₁ receptor has been implicated in inhibiting retrograde signalling resulting in the inhibition of neurotransmitter release and synaptic transmission (Howlett et al., 2002). CB₁ receptors have also been discovered on various immune cells and can promote both pro‐inflammatory and anti‐inflammatory actions (Bouaboula et al., 1993; Howlett et al., 2002). The CB₂ receptor was discovered in 1993 (Munro, Thomas, & Abu‐Shaar, 1993) and was previously known as the peripheral cannabinoid receptor and originally thought to be only located on the cell membrane of immune cells and immune‐mediating organs (Alexander et al., 2019; Howlett et al., 2002; Turcotte, Blanchet, Laviolette, & Flamand, 2016). However, CB₂ receptors have also been found in the CNS (Gong et al., 2006) and on sensory nerve terminals in the periphery (Griffin et al., 1997). Pertussis toxin (PTX) sensitivity of both cannabinoid receptors demonstrates coupling to G_i/o proteins (Howlett, Qualy, & Khachatrian, 1986). Both CB₁ and CB₂ receptor activation can inhibit the release of pro‐inflammatory mediators as a result of adenyl cyclase (AC) inhibition (Matsuda et al., 1990), calcium channel deactivation (Guo, 2004) and stimulation of the MAPK/ERK pathway (Howlett et al., 2002; Kobayashi, Arai, Waku, & Sugiura, 2001; Wartmann, Campbell, Subramanian, & Burstein, 1995; Zou & Kumar, 2018).

Atypical cannabinoid receptors are also involved in cannabinoid cell signalling (Caterina et al., 1997). Transient receptor potential (TRP) channels are a family of ligand‐gated cation channels mainly found on primary afferent neurones/terminals (Caterina, 2014; Caterina et al., 1997). TRP channels have been implicated in nociceptive signalling through activation by noxious stimuli in the peripheral environment and subsequent depolarization of the sensory nerve terminal (Caterina et al., 1997). Endocannabinoids have been shown to bind to TRP channels where they can modulate these nociceptive responses (Akopian, Ruparel, Jeske, Patwardhan, & Hargreaves, 2008). Other atypical cannabinoid receptors include the GPCR 55 (GPR55) which binds several cannabinoid ligands including HU210, JWH‐015 and R‐(+)‐methanandamide (Mackie & Stella, 2006; Sawzdargo et al., 1999). Unlike classical cannabinoid receptors, GPR55 does not appear to couple with G_i/o proteins (Mackie & Stella, 2006). Opioid receptors located on peripheral nerves are known to be involved in reducing neurogenic inflammation (Barin & McDougall, 2003) and cannabinoids have the ability to bind allosterically to these receptors. Cannabinoids can also target peroxisome proliferator‐activated receptors (PPARs; NRC1) which are known to be involved in mediating inflammation and pain. For example, in a model of multiple sclerosis, the synthetic cannabinoid ligand R‐(+)‐WIN55,212‐2 was found to inhibit the expression of adhesion molecules in brain endothelial cells and attenuate lymphocyte extravasation from the cerebral vasculature. This effect was blocked by the PPARγ (NCRC3) antagonist GW9662 (Mestre et al., 2009), thereby implicating this family of receptors in cannabinoid‐mediated neuroinflammation. Taken together, it is evident that the physiological effects of cannabinoids go far beyond the endocannabinoid system, which makes targeted therapy using these compounds a challenge.

1.1. Neurogenic inflammation

The concept of neurogenic inflammation was first postulated by Bayliss (1901) and is summarized in Figure 1. The study stated that peripherally located nerve fibres caused vasodilation in the hind limbs of dogs when stimulated by mechanical, chemical or thermal stimuli, suggesting that afferent fibres could also fire in an antidromic direction (Bayliss, 1901). These local antidromic currents were termed the “axon reflex” which is responsible for the vascular flare observed following tissue injury. This activation of sensory neurones leads to an inflammatory response called neurogenic inflammation (Bayliss, 1901). Primary afferents have been further implicated in neurogenic inflammation by application of the chemical irritant capsaicin which is known to selectively activate transient receptor potential vanilloid‐1 (TRPV1) ion channels (Caterina, Rosen, Tominaga, Brake, & Julius, 1999; Jancsó, Jancsó‐Gábor, & Szolcsányi, 1967). Local injection of capsaicin into skin can cause pain, tissue oedema and hyperaemia which persists despite sectioning of the peripheral nerve at the site of injection (Jancsó et al., 1967).

FIGURE 1.

Antidromic neurogenic release of neuropeptides from primary afferent nerve terminals and resulting effects on vascular and immune targets. CB₁ receptor (R), CB₂ receptor, CGRP receptor; GPR55, GPCR 55; H₁ receptor; NK₁ receptor and TRPV1

Several pro‐inflammatory neuropeptides have been shown to be released in response to antidromic nerve stimulation (Brain, 1997; Escott & Brain, 1993) and capsaicin injection (Brain, 1997; Hughes & Brain, 1994). These include substance P (SP), neurokinin A (NKA), CGRP and vasoactive intestinal peptide (VIP) (Brain, 1997). Both SP and NKA are members of the tachykinin family and act through NK₁ receptors on endothelial cells in post‐capillary venules (Brain, 1997) leading to an increase in vascular permeability (Brain & Williams, 1989; Foreman, Jordan, Oehme, & Renner, 1983; Lembeck & Holzer, 1979). This increased permeability leads to fluid leakage resulting in oedema (Brain & Williams, 1989; Foreman et al., 1983; Lembeck & Holzer, 1979). SP can also indirectly influence plasma extravasation by activating and promoting degranulation of mast cells resulting in histamine release (Foreman et al., 1983). SP has been implicated in neutrophil accumulation through up‐regulation of adhesion molecules (Zimmerman, Anderson, & Granger, 1992), specifically intercellular adhesion molecule‐1 (Nakagawa, Sano, & Iwamoto, 1995). CGRP is a potent dilator of peripheral arterioles (Brain, Williams, Tippins, Moris, & MacIntyre, 1985) through both NO‐dependent and ‐independent mechanisms (Russell, King, Smillie, Kodji, & Brain, 2014). VIP contributes to neurogenic inflammation by inducing vasodilation (Wilkins, Chung, Tublitz, Wong, & Minson, 2004) and promoting degranulation of mast cells (Theoharides et al., 2012), similar to the effects of SP (Foreman et al., 1983).

Small diameter afferents not only alter vascular tone and permeability, they also communicate with immunocytes to regulate tissue inflammation (Chiu, von Hehn, & Woolf, 2012). Mast cells, leukocytes and lymphocytes all express neuropeptide receptors on their cell surface which when activated leads to cellular degranulation or exocytosis culminating in the local release of inflammatory mediators (Coderre, Basbaum, & Levine, 1989; Dines & Powell, 1997). Furthermore, these neurotransmitters have the potential to recruit immunocytes to the site of inflammation (Smith, Barker, Morris, MacDonald, & Lee, 1993).

The inflammatory neuropeptides released from nociceptors can bind to their respective receptor on the nerve ending leading to neuronal sensitization and pain. SP, CGRP and VIP have all been shown to sensitize nociceptors in various tissues and are key mediators in the generation of pain (Carlton, Zhou, & Coggeshall, 1996; Schuelert & McDougall, 2006; Schwenger, Dux, de Col, Carr, & Messlinger, 2007). The localization of cannabinoid receptors on afferent nerve terminals means that they are ideally situated to control afferent activity and modulate pain in the periphery.

The aim of this review is to highlight the profusion of evidence that shows that endocannabinoids, synthetic cannabinoids and phytocannabinoids can alleviate neurogenic inflammation. The potential therapeutic value of cannabinoids in modulating neurogenic inflammation in various disease models will also be discussed.

2. CANNABINOID CONTROL OF NEUROGENIC INFLAMMATION

2.1. Endocannabinoids

The discovery of CB receptors throughout the human body prompted the search for innate compounds that serve to activate these receptors (Zou & Kumar, 2018). Endocannabinoids are endogenous lipid compounds that can interact with cannabinoid and cannabinoid‐like receptors (Battista, Di Tommaso, Bari, & Maccarrone, 2012). Unlike other signalling molecules, endocannabinoids are only synthesized on demand by local precursors within the cell and cell membrane (Di Marzo et al., 1994). A large number of endocannabinoids have been identified including n‐3 and n‐6 polyunsaturated fatty acid N‐acylethanolamines and monoacylglycerols (Alharthi et al., 2018; Battista et al., 2012). Of all the endocannabinoids, anandamide (N‐arachidonoylethanolamine) and 2‐arachidonoylglycerol (2‐AG) are the most commonly researched (Battista et al., 2012; Howlett et al., 2002). The biosynthesis of endocannabinoids is complex and involves a vast array of enzymes and substrates (Muccioli, 2010). Fundamentally, anandamide is synthesized by the hydrolysis of N‐arachidonoylphosphatidylethanolamide (NAPE) by a Ca²⁺‐sensitive NAPE‐selective PLD (NAPE‐PLD). While anandamide is only produced on demand, NAPE‐PLD is constitutively expressed in its active form and can therefore rapidly respond to endocannabinoid demands. Anandamide is broken down by fatty acid amide hydrolase into arachidonic acid and ethanolamine. 2‐AG is formed by the hydrolysis of DAG by the enzyme DAG lipase. The biosynthesis of 2‐AG also occurs on demand in response to cellular depolarization leading to profound Ca²⁺ influx. The enzyme responsible for 2‐AG degradation is monoacylglycerol lipase, which converts 2‐AG into arachidonic acid and glycerol. The enzymes and substrates responsible for endocannabinoid synthesis are located in the cell membranes of postsynaptic neurones, while the molecular machinery responsible for endocannabinoid degradation is located presynaptically. With respect to peripheral nerve endings, endocannabinoid synthesis and degradation occurs in the distal aspect of the axons.

The presence of CB₁ and CB₂ receptors on sensory nerve terminals and their inhibitory effect on neuronal activity suggest that the endocannabinoid system has the potential to moderate neurogenic inflammation and pain (Figure 2a). Richardson, Kilo and Hargreaves (1998) showed that local application of anandamide significantly inhibited carrageenan‐induced paw oedema, reduced protein extravasation and decreased locally released CGRP from rat skin. These anti‐inflammatory effects were not observed when anandamide was administered contralaterally and were reversed when experimental cohorts were pretreated with the CB₁ receptor inverse agonist rimonabant (Richardson, Kilo, & Hargreaves, 1998).

FIGURE 2.

Endocannabinoids, phytocannabinoids, and synthetic cannabinoids modulate neurogenic inflammation through CB₁ receptor‐mediated inhibition of nociceptive firing and neuropeptide release from primary afferents (a), through TRPV1 desensitization of afferent terminals (b), and through CB₁ receptor (R), CB₂ receptor and GPR55‐mediated inhibition of mast cell degranulation (c)

In the pithed rat preparation, anandamide decreased vasodilation following electrical stimulation of primary afferent nerves (Marichal‐Cancino, Altamirano‐Espinoza, Manrique‐Maldonado, Maassenvandenbrink, & Villalón, 2014) and these responses were prevented by a CB₁ antagonist, but not CB₂, GPR55 or TRPV1 antagonists (Marichal‐Cancino et al., 2014). Conversely, anandamide was unable to reverse the vasodilatatory effect of exogenously administered CGRP (Marichal‐Cancino et al., 2014). The inhibitory effects of anandamide on electrically induced vasodilatation but not exogenous CGRP‐induced vasodilatation suggest that anandamide was acting locally on sensory nerve terminals and that the response was peripherally mediated since the spinal cord had been severed (Marichal‐Cancino et al., 2014).

The endocannabinoid noladin ether (Sugiura, Kodaka, & Nakane, 1999) has been found to inhibit sensory neurotransmission in rat mesenteric arterial beds (Duncan, Millns, et al., 2004). At concentrations of 0.1–3 μM, direct application of noladin ether inhibited neurogenically mediated hyperaemia in mesenteric arterioles and this effect was blocked by PTX treatment (Duncan, Millns, et al., 2004). Additionally, no inhibitory effects were observed with exogenously applied CGRP nor did noladin ether have any effect on capsaicin‐induced calcium responses in dorsal root ganglion cells (Duncan, Millns, et al., 2004). The authors concluded that the atypical endocannabinoid noladin ether inhibits sensory neurotransmission by a terminal, PTX‐sensitive (G_i/G_o), non‐cannabinoid or TRPV1‐dependent mechanism (Duncan, Millns, et al., 2004).

Other studies found that endocannabinoid‐dependent neuropeptide modulation is concentration dependent. Ahluwalia, Urban, Bevan, and Nagy (2003) reported that anandamide concentrations under 1 μM inhibit both basal and capsaicin‐induced CGRP release through a CB₁ receptor mechanism. In contrast, higher anandamide concentrations evoked higher levels of CGRP release equivalent to that of capsaicin‐induced peptide release, which was reversed by administration of the TRPV1 antagonist capsazepine (CPZ) (Ahluwalia et al., 2003). These findings were replicated by Nemeth et al. (2003) who reported that inflammatory neuropeptide release could be reduced with low concentrations of anandamide (10⁻⁶ M) in a CB₁ receptor‐dependent manner. However, at higher concentrations of anandamide the inhibitory effect of the endocannabinoid on neuropeptide release was deemed to be TRPV1 dependent. Weller, Reeh, and Sauer (2011) later supported these findings using high concentrations of the stable analogue (R)‐(+)‐methanandamide and confirmed the TRPV1 pathway by using TRPV1 knockout mice. These studies demonstrated that these particular endocannabinoids evoked a dual inhibition of neuropeptide release and hence neurogenic inflammation which was either CB₁ receptor dependent (low concentrations) or TRPV1 dependent (higher concentrations) (Ahluwalia et al., 2003; Nemeth et al., 2003; Weller et al., 2011).

In the isolated skin preparation, Engel et al. (2011) found that low concentrations of anandamide inhibited CGRP release while higher concentrations of the endocannabinoid increased neuropeptide levels. Heat‐evoked CGRP release, however, was attenuated by high concentrations of anandamide. The reversal of these effects in a TRPV1 knockout model, but not CB₁ receptor knockouts, substantiates a desensitizing role of anandamide through TRPV1 channels, but only at higher concentrations (Figure 2b).

In addition to attenuating neuropeptide release and subsequent mitigation of neurogenic inflammation, endocannabinoids such as anandamide have also been shown to alter mast cell degranulation (Cantarella et al., 2011; Mazzari, Canella, Petrelli, Marcolongo, & Leon, 1996; Nam et al., 2016; Sugawara et al., 2012). Mazzari et al. (1996) demonstrated that oral treatment of mice with the CB₂ agonist LG2110/1 (PEA or palmidrol) reduced the number of mast cells that underwent SP‐induced degranulation. This reduction in the secretagogue action of connective tissue mast cells by LG2110/1 culminated in a reduction in plasma extravasation and oedema. Similarly, SP‐induced human mast cell degranulation could be inhibited by anandamide administration, however the CB₁ antagonist AM251 had no effect on mast cell degranulation, suggesting that the anandamide had to be exogenously applied (Sugawara et al., 2012). Cantarella et al. (2011) proposed that anandamide attenuated the release of pro‐inflammatory mediators such as nerve growth factor (NGF) from human mast cells by activating GPR55. Nam et al. (2016) also reported a reduction of the pro‐inflammatory mediator histamine by anandamide from the rat basophilic leukaemia mast cell line RBL‐2H3 through a CB₁ receptor mechanism. Preventing the degranulation of mast cells and the subsequent release of inflammatory neuropeptides (Figure 2c) is yet another example of how some endocannabinoids have the potential to ameliorate neurogenic inflammation.

Several endocannabinoids are also involved in regulating immune cell function. The glycoprotein CD200R is expressed on the surface of macrophages and B and T lymphocytes where its activation causes the reduction in inflammatory cytokine production (Jenmalm, Cherwinski, Bowman, Phillips, & Sedgwick, 2006). In a multiple sclerosis model, increasing endocannabinoid tone with the anandamide uptake inhibitor UCM‐707 reduced the production of the inflammatory cytokine IL‐1b while increasing levels of the anti‐inflammatory cytokine IL‐10 (Hernangomez et al., 2014). Microglia, which are the resident macrophage in the CNS, also express cannabinoid receptors (Carlisle, Marciano‐Cabral, Staab, Ludwick, & Cabral, 2002). Activation of microglial CB₂ receptor with anandamide inhibited ERK‐1/2 phosphorylation leading to reduced pro‐inflammatory iNOS production in multiple neuroinflammatory conditions (Rom & Persidsky, 2013).

Mediators that promote endocannabinoid synthesis and inhibitors of endocannabinoid hydrolysis are also key modulators of neurogenic inflammation. For example, the endogenous lipid palmitoylethanolamide (PEA) promotes anandamide and 2‐AG production by inhibiting fatty acid amide hydrolase and stimulating DAGL activity, respectively (Di Marzo et al., 2001; Petrosino et al., 2019). Treatment with palmitoylethanolamide attenuated SP‐induced degranulation of RBL‐2H cells (Petrosino et al., 2019), demonstrating that boosting endocannabinoid tone has the potential to reduce neurogenic inflammation. Elsewhere, blockade of 2‐AG hydrolysis with monoacylglycerol lipase inhibitors caused a reduction in pain reactivity in response to noxious thermal and mechanical stimuli (Mulvihill & Nomura, 2013). Monoacylglycerol lipase inhibitors also reduced prostaglandin E2 (PGE₂) production which is a known promoter of neurogenic inflammation (Jang, Kim, & Hwang, 2020; Kress, Guthmann, Averbeck, & Reeh, 1999). It should be noted that in in vivo experiments, the drugs that modulate endocannabinoid synthesis/breakdown are typically administered systemically, however the alteration in endocannabinoid tone occurs locally in the tissues where they can directly alter peripheral neuronal activity.

2.2. Phytocannabinoids

The cannabis plant contains over 100 different cannabinoids whose levels vary depending upon strain and growing conditions. The majority of phytocannabinoids occur in acidic form in the plant and require decarboxylation by heating and light before becoming bioactive in humans. Physiological characterization of the phytocannabinoids is still in its infancy with the vast majority of research focusing on the psychoactive compound Δ⁹‐tetrahydrocannabinol (THC) and the non‐psychotropic cannabinoid, cannabidiol (CBD). Less abundant phytocannabinoids include cannabigerol, cannabinol and cannabichromene. There is significant anecdotal and scientific evidence which shows that medical cannabis has anti‐inflammatory properties in a number of chronic diseases such as arthritis, multiple sclerosis, cancer and inflammatory bowel disease. Plant‐derived CBD, for example, can act on immunocytes to reduce prostaglandin and cytokine production by interfering with COX activity and stimulating the MAPK pathway, respectively. A few of the plant‐based cannabinoids also have the propensity to modulate neurogenic inflammation by acting on the endocannabinoid system (Duncan, Millns, et al., 2004; Qin et al., 2008; Wilkinson, Kendall, & Ralevic, 2007). The psychoactive cannabinoid THC significantly inhibited both vasorelaxation and CGRP release in mesenteric arterial beds following electrical field stimulation (Wilkinson et al., 2007). However, CGRP release was increased at higher concentrations of THC. The non‐selective TRP channel blocker ruthenium red, but not cannabinoid receptor antagonists, prevented the inhibition, suggesting a TRP channel‐mediated modulation of pro‐inflammatory neuropeptide release (Wilkinson et al., 2007). Low concentrations of THC also inhibited capsaicin and heat‐induced CGRP release in rodent skin, which was prevented by co‐administration with AM251 or in CB₁ receptor knockout mice (Engel et al., 2011). The group also showed that THC concentrations exceeding 100 μM increased basal CGRP release on its own but reduced CGRP release when noxious stimuli were applied. The anti‐neurogenic effects of THC were found to be TRPV1 dependent as this action was mitigated in TRPV1 knockout mice (Engel et al., 2011).

Non‐psychoactive CBD and cannabinol also demonstrated a TRP‐dependent mechanism in modulating pro‐inflammatory neuropeptide release in peripheral tissues (Qin et al., 2008). These phytocannabinoids heightened CGRP levels significantly higher than baseline measurements, but, similar to THC‐evoked responses, they could only be inhibited by the non‐selective TRP blocker ruthenium red but not cannabinoid receptor antagonists (Qin et al., 2008). Calcium‐dependent, ionotropic glutamate receptor‐dependent (MK‐801, CNQX), and G protein‐dependent (G_i/G_o) signalling were ruled out due to a lack of effect with calcium channel blockers and PTX (Qin et al., 2008).

2.3. Synthetic cannabinoids

Based on the structure–function relationship of natural cannabinoids, a growing number of synthetic ligands have been developed that selectively bind to cannabinoid receptors (Pertwee, 2010). The synthetic cannabinoids are subdivided into four main groups based on their chemical structure. Classical cannabinoid receptor agonists consist of dibenzopyran derivatives such as HU210 and JWH‐133. Non‐classical cannabinoid agonists, such as CP55,940, are bicyclic or tricyclic analogues of THC. The third group is the eicosanoid cannabinoid receptor ligands which includes the endocannabinoid analogues R‐(+)‐methanandamide and arachidonyl‐2′‐chloroethylamide (ACEA). Finally, aminoalkylindole cannabinoids have a unique structure unlike the other synthetic cannabinoid subtypes and include JWH‐015 and R‐(+)‐WIN55,212‐2.

With respect to neurogenic inflammation, synthetic cannabinoids have been found to influence peripheral nerve activity and consequently inflammatory neuropeptide release. Electrical stimulation of isolated rat mesenteric artery beds revealed that the cannabinoid agonists HU210, WIN55,212‐2, and CP55,940 inhibited CGRP‐induced vasorelaxation in a concentration‐dependent manner (Duncan, Kendall, & Ralevic, 2004; Ralevic & Kendall, 2001). Of note, this inhibitory effect was not evident with exogenously applied CGRP confirming a direct effect on peptidergic nerves (Duncan, Kendall, & Ralevic, 2004; Ralevic & Kendall, 2001). Antagonist studies indicate a possible CB₁ receptor‐mediated inhibition of neurogenic inflammation for WIN55,212‐2 and CP55,940 (Duncan, Kendall, & Ralevic, 2004) and a non‐cannabinoid mechanism for HU210 (Ralevic & Kendall, 2001). Contrasting evidence by Tumati et al. (2009) shows that sustained application of WIN55,212‐2 for 24 h increased basal levels of CGRP in rat neonatal dorsal root ganglia. Increased basal CGRP levels were attenuated with treatment of the CB₁ antagonist AM251 and the PKA inhibitor H‐89 (Tumati et al., 2009). This research suggests that, unlike the anti‐neurogenic effects of acute treatment, sustained cannabinoid administration may contribute to pain and inflammation by stimulating inflammatory neuropeptide release (Tumati et al., 2009).

Dvorak, Watkinson, McGlone, and Rukwied (2003) also demonstrated that transdermal application of HU210 prevented histamine‐induced increases in blood flow, flare responses, and itching sensations (Dvorak et al., 2003). In contrast, HU210 seemingly increased plasma extravasation when co‐administered with histamine (Dvorak et al., 2003). The authors postulated that the reductions in blood flow and flare may stem from reduced excitability of the sensory afferents resulting in reduced production and release of neuropeptides, subsequently preventing vasodilation and NO production (Dvorak et al., 2003).

Selective CB₂ receptor agonists have also been used in characterizing neurogenic inflammation (Duncan, Kendall, & Ralevic, 2004). For example, JWH‐015 inhibited vasorelaxation generated by electrical stimulation of small diameter sensory nerve fibres but did not inhibit vasorelaxation induced by exogenously applied CGRP (Duncan, Kendall, & Ralevic, 2004). The anti‐neurogenic effects of JWH‐015 were not reversed by application of CB₁/CB₂ antagonists (Duncan, Kendall, & Ralevic, 2004), suggesting a non‐cannabinoid mechanism for this neurovascular inhibition. Anand et al. (2008) demonstrated a CB₂‐mediated inhibition of capsaicin‐induced neurogenic currents and calcium uptake in both human and guinea pig dorsal root ganglia (DRG) samples. Elsewhere, the CB₂ agonists GW833972 and GW842166 reduced whole‐cell currents in response to capsaicin application, while also providing a significant inhibition of calcium influx (Anand et al., 2008). This calcium inhibition was ultimately reversed by the CB₂ antagonist AM630 and excess cAMP but was unaffected by SR141716A and naloxone (Anand et al., 2008). The authors hypothesized that these agonists may work by depleting cAMP, thus preventing the capsaicin response through TRPV1 (Anand et al., 2008).

Synthetic cannabinoids have also been shown to alter neurogenic inflammation induced by TRP channel activation. Peripheral administration of the CB₁ agonist ACEA inhibited mustard oil‐induced immunoreactive CGRP release in rat hind paw samples highlighting a link between cannabinoid receptors and transient receptor potential ankyrin 1 (TRPA1) ion channels (Ruparel, Patwardhan, Akopian, & Hargreaves, 2011). Interestingly, these effects were not observed in TRPV1 knockout mice or in cells expressing mutated TRPV1 cation channels (Ruparel et al., 2011).

3. EVIDENCE OF CANNABINOID CONTROL OF NEUROGENIC INFLAMMATION IN PRECLINICAL DISEASE MODELS

While the fundamental principles underlying the involvement of the endocannabinoid system in altering sensory neurotransmission and neurogenic inflammation have been described, the role of cannabinoids in various disease states is less complete. A summary of the preclinical evidence in animal models is highlighted in Tables 2, 3, 4.

TABLE 2.

Endocannabinoid, synthetic cannabinoid and phytocannabinoid modulation of neurogenic inflammation in preclinical models of arthritis. ECS, endocannabinoid system.

Model	Cannabinoid/ECS mediator	Dose/application	Receptor/proposed mechanism	Result	Reference
Electric stimulation—saphenous nerve	URB 597	0.3 mg·kg−1 Topical—over exposed joint	GPR55	↓ Rolling leukocytes Reversed by O‐1918	Krustev, Muley, & McDougall, 2017
Sodium monoiodoacetate (MIA) rat model of osteoarthritis	CBD	100–300 μg Topical—over exposed joint or s.c.	TRPV1 CB2	Improved weight bearing and mechanical allodynia ↓ Type III and IV nerve firing ↓ Rolling and adherent leukocytes No effect when given contralaterally (allodynia) Improved allodynia, rolling, and adherent reductions reversed by SB366791; rolling also blocked by AM630	Philpott, O'Brien, & McDougall, 2017
	KML29	700 μg i.artic.	CB1 and CB2	↓ Allodynia and leukocyte trafficking	Philpott & McDougall, 2020
Freund's complete adjuvant (FCA) rat model of inflammatory arthritis	CBD	0.6–62.3 mg·day−1 Transdermal	N/A	↓ Joint oedema (6.2—Improved pain scores) ↓ Heat hypersensitivity ↓ CGRP immune‐reactivity (spinal)	Hammell et al., 2016

TABLE 3.

Endocannabinoid, synthetic cannabinoid, and phytocannabinoid modulation of neurogenic inflammation in preclinical models of asthma. ECS, endocannabinoid system; PEA, palmitoylethanolamide

Model	Cannabinoid/ECS mediator	Dose/application	Receptor/proposed mechanism	Result	Reference
Electric field‐, capsaicin‐, NKA‐induced bronchoconstriction Capsaicin‐induced SP release (guinea pig)	AEA PEA	0.0288–28.8 μM i.v. 0.003–3.3 μM i.v.	CB2 K+ channel	↓ (anandamide/PEA) electric bronchoconstriction No effect on NKA‐induced bronchoconstriction ↓ (PEA only) capsaicin‐induced bronchoconstriction Reversed by SR144528 Reversed by K+ channel blocker (NS1619) ↓ (anandamide/PEA) SP release	Yoshihara, Morimoto, Ohori, Yamada, & Abe, 2005
Capsaicin/neurokinin A‐induced guinea pig bronchoconstriction Rat cigarette smoke induction (guinea pig)	WIN55,212‐2	0.1 ml·kg−1 i.v.	CB2 K+ channel	↓ Capsaicin‐induced bronchoconstriction No effect of NKA‐induced constriction ↓ Plasma extravasation No effect on SP‐induced extravasation Both reversed by SR144528 and K+ channel blocker (NS1619)	Yoshihara et al., 2005
Antigen‐induced plasma extravasation (ovalbumin inhalation)	WIN55,212‐2 FK888	0.001–0.1 mg·kg−1 i.v. 1 mg·kg−1 i.v.	CB2 NK1	↓ Plasma extravasation Reversed by SR144528 FK888 + WIN showed a greater ↓ in plasma extravasation	Fukuda, Abe, & Yoshihara, 2010
Electric‐/capsaicin‐/NKA‐activated afferent fibres (guinea pig) Capsaicin‐induced SP release	WIN55,212‐2 JWH‐133	0.0191–19.1 μM Added directly 0.1–100 μM	CB2 K+ channel CB2	↓ Electrical and capsaicin‐induced contraction No effect on NKA‐induced contraction ↓ Electrical‐induced contraction No effect on NKA‐induced contraction Reversed by SR145528 K+ channel antagonists prevented reductions ↓ (JWH and WIN) SP release	Yoshihara, Morimoto, Yamada, Abe, & Arisaka, 2004

TABLE 4.

Endocannabinoid, synthetic cannabinoid, and phytocannabinoid modulation of neurogenic inflammation in preclinical models of migraine. ECS, endocannabinoid system.

Model	Cannabinoid/ECS mediator	Dose/application	Receptor/proposed mechanism	Result	Reference
Dural vessel dilation (electrical, CGRP/capsaicin/NO induced)	anandamide	1–10 mg·kg−1 i.v.	CB1	↓ Electric‐, capsaicin‐ (1–3 mg·kg−1), CGRP‐, NO‐induced (3–10 mg·kg−1) vasodilation Reversed by AM251	Akerman, 2004
Dural electrical stimulation	anandamide WIN55,212‐2	10 mg·kg−1 i.v. 1 mg·kg−1 i.v.	TRPV‐1 and CB1 CB1	↓ A‐fibre firing in presence of CPZ ↓ C‐fibre firing Reversed by SR141716A ↓ A‐ and C‐fibre firing Reversed by SR141716A	Akerman, Holland, & Goadsby, 2007
Basal dural functioning	anandamide	100 nM to 10 μM Added to cranial window	TRPV‐1 and CB1	↑ % blood flow (100 nM to 1 μM) ↓ % blood flow (10 μM)	Dux, Deák, Tassi, Sántha, & Jancsó, 2016
Capsaicin‐/K+‐induced stimulation of the trigeminal nerve afferent	WIN55,212‐2	25–50 μM Direct application	? Not CB1/CB2 or TRPV1	↑ CGRP (WIN55,212‐2) Reversed by ruthenium red No effect on capsaicin‐induced CGRP No effect on forskolin‐stimulated cAMP levels ↓ K+‐induced CGRP levels	Price, Patwardhan, Akopian, Hargreaves, & Flores, 2004

3.1. Arthritis

Various arthritis models exhibit an antidromic neurogenic component which can drive joint disease resulting in joint damage, neutrophil accumulation, vasodilatation and pain (Ahmed, Srinivasan, Theodorsson, Schultzberg, & Kreicbergs, 1995; Levine et al., 1984). In animal knee joints, low frequency stimulation of joint nerves caused synovial hyperaemia indicative of a neurogenic inflammatory response (Karimian, McDougall, & Ferrell, 1995; McDougall, Karimian, & Ferrell, 1994). These vasodilatory effects were mediated by peripheral release of NO, SP, CGRP and VIP (McDougall & Barin, 2005; McDougall & Ferrell, 1996; McDougall, Ferrell, & Bray, 1999; McDougall, Karimian, & Ferrell, 1995). Synovial levels of SP and CGRP are elevated in models of arthritis (Ahmed et al., 1995; Levine et al., 1984), while the number of SP and CGRP containing fibres have also been found to increase in rat, cat and human knees (Hanesch, Heppelmann, & Schmidt, 1997; Larsson, Ekblom, Henriksson, Lundeberg, & Theodorsson, 1991; McDougall, Bray, & Sharkey, 1997). Clearly, there is a strong neurogenic component to joint inflammation.

Cannabinoids have been shown to reduce neurogenic inflammation in joints (Table 2). The first indication that cannabinoids can act locally in the joint was demonstrated electrophysiologically where the synthetic CB₁ agonist ACEA reduced afferent firing in response to noxious mechanical stimulation (Schuelert & McDougall, 2008). The antinociceptive effect of ACEA was augmented in osteoarthritic joints, suggesting that the endocannabinoid system could be a promising target for the alleviation of arthritis pain and inflammation. Preventing the breakdown of anandamide using fatty acid amide hydrolase inhibitors is another strategy that has been used to reduce neurogenic inflammation in arthritis models (Krustev et al., 2017). Treatment of inflamed knees with the fatty acid amide hydrolase inhibitor URB 597 reduced rolling leukocytes following electrical stimulation of the saphenous nerve (Krustev et al., 2017). The endocannabinoid‐induced reduction in leukocyte trafficking was abolished with the co‐administration of the GPR55 receptor antagonist O‐1918 but persisted in the presence of CB₁ and CB₂ antagonists (Krustev et al., 2017). This suggests that a non‐classical cannabinoid receptor was responsible for the endocannabinoid amelioration of neurogenic inflammation in this acute inflammatory model (Krustev et al., 2017).

Treatment of osteoarthritic rats with the monoacylglycerol lipase inhibitor KML29 reduced joint pain and inflammation in a CB receptor‐dependent manner (Philpott & McDougall, 2020). Since 2‐AG can also be broken down by COX‐2, the study tested a combination of KML29 and celecoxib and found that the amalgamation of the two drugs was more efficacious than the individual components.

Local application of phytocannabinoids, such as CBD (Costa, Trovato, Comelli, Giagnoni, & Colleoni, 2007; Hammell et al., 2016; Philpott et al., 2017), has also been shown to reduce joint afferent firing and antidromic neuropeptide release in various arthritis models (Table 2). In a rat model of osteoarthritis, CBD dose‐dependently reduced joint nociceptor firing and reduced leukocyte trafficking in the synovial vasculature (Philpott et al., 2017). This response was attenuated by the TRPV1 antagonist SB366791 (Philpott et al., 2017). Thus, CBD is able to reduce neurogenic inflammation by desensitizing TRPV1 ion channels leading to a reduction in afferent firing rate and subsequent inflammatory neuropeptide release (Philpott et al., 2017). Hammell et al. (2016) demonstrated similar effects in an inflammatory model of arthritis. Transdermal application of CBD reduced knee joint oedema, improved heat hyperalgesia, reduced immune cell filtration and lowered spinal levels of CGRP (Hammell et al., 2016). Since no changes in CGRP levels were detectable in dorsal root ganglia, the alteration of peripheral neurogenic inflammation could not be confirmed (Hammell et al., 2016). Another study using a rat model of inflammatory arthritis showed that CBD reduced endothelial NOS production in a TRPV1‐dependent manner (Costa et al., 2007). The proposed TRPV1 mechanism of analgesia, and reduction of a primary CGRP‐induced vasoactive mediator, suggests an inhibitory effect of CBD on neurogenic inflammation in joints (Costa et al., 2007).

3.2. Asthma

Asthma is considered a chronic inflammatory disease of the airways typically instigated by allergens and chemical irritants (Maggi, Giachetti, Dey, & Said, 1995). Neurogenic mechanisms contributing to asthmatic inflammation have been described in detail (Joos, Germonpre, & Pauwels, 2000; Maggi et al., 1995). Neuropeptide‐containing nerve fibres have been identified throughout the respiratory tract (Maggi et al., 1995) and SP levels are elevated in asthmatic bronchoalveolar lavage fluid (Joos et al., 2000). Furthermore, chemical irritants such as cigarette smoke induce the release of neuropeptides through a TRPA1 mechanism, leading to bronchoconstriction and alveolar plasma extravasation (Andrè et al., 2008).

Asthmatic neurogenic inflammation has most commonly been modelled in guinea pigs (Yoshihara et al., 2004; Yoshihara, Morimoto, Ohori, Yamada, & Abe, 2005; Yoshihara, Morimoto, Ohori, Yamada, Abe, & Arisaka, 2005; Table 3). The endocannabinoid anandamide was able to reduce electrically stimulated guinea pig bronchoconstriction and capsaicin‐induced SP release (Yoshihara, Morimoto, Ohori, Yamada, Abe, & Arisaka, 2005). The CB₂ antagonist SR144528 and Ca²⁺ activated K⁺ channel blockers, iberiotoxin and charybdotoxin, were used to prevent these inhibitory effects (Yoshihara, Morimoto, Ohori, Yamada, Abe, & Arisaka, 2005). The maxi‐K⁺ channel opener, NS1619, showed similar inhibitions of bronchoconstriction and plasma extravasation (Yoshihara, Morimoto, Ohori, Yamada, Abe, & Arisaka, 2005). Similarly, WIN55,212‐2 and JWH‐133 showed a dose‐dependent inhibition of capsaicin‐induced bronchoconstriction, which was reversed by a CB₂ antagonist but not by a CB₁ antagonist (Yoshihara, Morimoto, Ohori, Yamada, & Abe, 2005). The inhibition of cigarette smoke‐induced plasma extravasation was also prevented solely by the CB₂ antagonist (Yoshihara, Morimoto, Ohori, Yamada, & Abe, 2005). Neither NKA‐induced bronchoconstriction or SP‐induced plasma extravasation was affected by the endocannabinoid system (Yoshihara, Morimoto, Ohori, Yamada, & Abe, 2005). These data suggest that improvements in asthma may be due in part to the CB₂‐mediated silencing of sensory nerve terminals, ultimately preventing the release of pro‐inflammatory neuropeptides (Yoshihara, Morimoto, Ohori, Yamada, Abe, & Arisaka, 2005). Further studies showed that combined administration of WIN55,212‐2 with an NK₁ receptor antagonist (FK888; L‐alaninamide) ) produced a marked improvement in pulmonary plasma extravasation compared to the NK₁ antagonist alone (Fukuda et al., 2010). This observation suggests that WIN55,212‐2 is able to reduce SP release from C‐fibre terminals and inhibit neurogenic inflammation in the lung (Fukuda et al., 2010).

3.3. Migraine

Migraines are a prevalent neurovascular disorder characterized by unilateral head pain often accompanied by nausea and hypersensitivity to light or sound (Burstein, Noseda, & Borsook, 2015). Pathophysiological research of migraines suggests that these symptoms are a result of the activation of trigeminovascular neurons and peripheral dural afferents (Burstein et al., 2015). Initial studies found that stimulation of trigeminal afferents resulted in increased release of neuropeptides such as CGRP within the trigeminal circulation (Goadsby & Edvinsson, 1993). In humans, infusion of CGRP can cause migraine‐like headaches in healthy volunteers suggesting a neurogenic component to the disorder (Lassen et al., 2002).

Localization of CB₁ receptor along the trigeminal tract and trigeminal afferents have sparked interest in determining whether the endocannabinoid system can intervene in neurogenic‐induced migraine (Richardson, Aanonsen, & Hargreaves, 1998). In vivo testing of endocannabinoids has shown promising reductions in neurogenic inflammation in animal migraine models (Akerman, 2004; Akerman et al., 2007; Price et al., 2004; Table 4). For example, intravenous administration of anandamide (1–3 mg·kg⁻¹) reduced the diameter of both capsaicin‐induced and electrically stimulated dilated vessels, while only higher concentrations of anandamide (3–10 mg·kg⁻¹) were able to significantly prevent CGRP‐ and NO‐induced vasodilatation within the dura (Akerman, 2004). Anandamide‐induced amelioration of vasorelaxation was inhibited by a CB₁ antagonist (Akerman, 2004). This study suggests that the endocannabinoid anandamide can have CB₁ receptor‐mediated inhibitory effects on neurogenic vasodilatation of trigeminal blood vessels (Akerman, 2004). A second study demonstrated that anandamide could significantly inhibit C‐fibre firing in the dura mater of the trigeminocervical complex while a combination of anandamide and the TRPV1 antagonist capsazepine was required to inhibit A‐δ fibre activity (Akerman et al., 2007). This study further showed that the non‐selective cannabinoid agonist WIN55,212‐2 imparted a significant inhibition of A‐δ and C‐fibre firing that was subsequently inhibited by the CB₁ antagonist SR141716A; CB₂ receptors were not involved in this process (Akerman et al., 2007).

A more recent study showed that topical anandamide had only a minor effect on meningeal blood flow whereas another endocannabinoid N‐arachidonoyldopamine caused significant hyperaemia (Dux et al., 2016). The vaso‐relaxing effect of N‐arachidonoyldopamine occurred via capsaicin‐sensitive nerves and could be blocked by a TRPV1 ion channel antagonist as well as a CGRP antagonist. Thus, in the meninges, N‐arachidonoyldopamine has the ability to cause neurogenic inflammation. In contrast, the synthetic cannabinoid WIN55,212‐2 produced conflicting results when applied to trigeminal sensory afferent neurones (Price et al., 2004). The synthetic cannabinoid evoked CGRP release from trigeminal afferents when calcium was also present in the extracellular space (Price et al., 2004). The cannabinoid‐induced CGRP release was only prevented by ruthenium red and not capsazepine, suggesting that these actions were mediated by a non‐specific TRP channel (Price et al., 2004). In the same study, WIN55,212‐2 had no effect on capsaicin‐induced CGRP release or forskolin‐induced cAMP accumulation but did reduce potassium‐induced CGRP release from trigeminal afferents (Price et al., 2004). These reductions were not affected by either cannabinoid receptor antagonists and were not PTX sensitive, suggesting a non‐classical receptor/channel mechanism (Price et al., 2004).

4. SUMMARY

Endocannabinoids, synthetic cannabinoids and phytocannabinoids can reduce neurogenic inflammation by inhibiting neuropeptide release from primary afferent nerve endings, although some of these effects are highly dose dependent. Cannabinoids can also inhibit the crosstalk between nerves, immunocytes and mast cells leading to a reduction in cellular degranulation and the release of pro‐inflammatory mediators. The anti‐neurogenic effects of cannabinoids occur within a variety of cell types through interactions with both typical cannabinoid receptors and atypical receptors linked to the endocannabinoid system. Since many chronic inflammatory diseases have a neurogenic component, it will be important for future studies to consider the interaction of the endocannabinoid system and peripheral nerve terminals under pathological conditions. The widespread use of medical cannabis to manage inflammatory disorders also requires greater scrutiny with a particular focus on the interplay between individual cannabis constituents. This entourage effect of phytocannabinoids, non‐cannabinoid molecules, and their multitudinous metabolites on peripheral nerve terminals as they pertain to neurogenic inflammation and pain needs further investigation. Finally, more research is needed to reveal potential synergistic effects of the endocannabinoid system to other non‐cannabinoid‐based therapies such as opioids to improve the safety and efficacy of treatments aimed at reducing neurogenic inflammatory disorders.

4.1. 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 PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2019/20 (Alexander et al., 2019).

CONFLICT OF INTEREST

The authors declare no conflicts of interest.

McKenna M, McDougall JJ. Cannabinoid control of neurogenic inflammation. Br J Pharmacol. 2020;177:4386–4399. 10.1111/bph.15208