CB₁ receptor antagonist SR141716A increases capsaicin-evoked release of Substance P from the adult mouse spinal cord

Introduction

The antinociceptive effect of cannabinoids exogenously applied in various models of acute and tonic pain, may be partly explained by the abundance of CB₁ receptors in the superficial laminae of the spinal cord, where C and Aδ fibre nociceptors terminate. Many of these terminals contain the tachykinin peptide transmitter Substance P (SP), which is co-released with glutamate after noxious stimulation of nociceptive fibres. SP is an important component of the pain signal, especially after intense noxious stimulation. When released it acts synergistically with glutamate to excite post synaptic dorsal horn neurones (reviewed Hökfelt et al., 2001). CB₁ is found at two cellular locations in the superficial spinal cord, pre-synaptically on primary afferent terminals (Hohmann & Herkenham, 1999; Ahluwalia et al., 2000) and on a population of excitatory interneurons expressing the gamma sub-unit of protein Kinase C (Farquar-Smith et al., 2000). CB₁ receptors are coupled to G(i)/(o) proteins, whose actions tend to inhibit neuronal excitability by activation of hyperpolarizing potassium currents (Shen et al., 1996), or by inhibition of Ca²⁺ channels reducing intracellular Ca²⁺ levels in order to inhibit transmitter exocytosis (Robbe et al., 2001). CB₁ receptor activation has been shown to inhibit transmitter release from hippocampal circuits (Shen et al., 1996). The CB₁ agonist anandamide has been used to inhibit Ca²⁺ influx (Szóke et al., 2000) and the release of CGRP from rat sensory neurone terminals after capsaicin or electrical stimulation (Richardson et al., 1998, Tognetto et al., 2001). In this study, we aimed to investigate the endogenous activity of spinal cord CB₁ receptors and their influence on excitatory transmission from nociceptors after noxious stimulation.

Methods

Spinal cords from adult CD1 mice (20 – 30 g Charles River U.K.) were extruded and 16 mm sections mounted then sealed in a chamber continuously superfused with oxygenated (5% CO₂) Krebs' solution (1 ml min⁻¹) containing (in mM): (NaCl; 118, KCl; 4, MgSO₄; 1.2, KH₂PO₄; 1.2, NaHCO₃; 25, CaCl₂; 2.5 and glucose; 11). Krebs' solution was supplemented with 0.1% bovine serum albumin, phosphoramidon 1 μM, captopril 100 μM, dithiothreitol 6 μM, bacitracin 20 μg ml⁻¹ (all from Sigma). Superfusate outflow from cord sections was sampled in a series of seven 8 ml fractions, each collected for 8 min at room temperature into glass tubes containing acetic acid (0.1 N). The first three fractions were collected in the absence of stimulation to determine basal SP outflow. Sections were stimulated during collection of the fourth fraction by superfusing capsaicin (0.01 – 100 μM) for 3 min. Three fractions were collected post-stimulation. Samples were concentrated and desalinated using Sep-Pak C₁₈ cartridges (Waters, Watford, U.K.) then dried under nitrogen. SP-like immunoreactivity (SP-LI) of reconstituted samples was measured within the detection limits of a radioimmunoassay (1 – 200 fmol per assay tube, see Malcangio & Bowery, 1993).

Results

Capsaicin evoked release of SP from mouse spinal cord

Capsaicin superfusion significantly increases release of SP-L1 measured in superfusate outflow from mouse spinal cords, above basal non-stimulated levels (Figure 1). To verify that the capsaicin-evoked release was mediated by activation of the receptor for capsaicin (VR1), the competitive VR1 antagonist capsazepine (100 μM) was superfused 3 min before stimulation, co-superfused with capsaicin (10 μM) and superfused 2 min after stimulation. Capsazepine treatment significantly inhibited the release of SP-LI, with this concentration of capsaicin (Figure 1).

Using capsaicin concentrations 0.1 – 10 μM, the evoked release of SP-LI increased dose dependently with an estimated EC₅₀ value of 311 nM (95% C.I., 178 – 542 nM) (Figure 2, solid line). Low doses of 0.01 μM capsaicin were ineffective at releasing SP-LI above basal outflow. Peak evoked release of SP-LI was achieved using 5 μM capsaicin. At higher capsaicin doses (50 – 100 μM), the amount of evoked SP-LI was markedly reduced, producing a bell shaped dose-response curve (Figure 2, dotted line).

CB₁ antagonist SR141716A modulates capsaicin-evoked SP-LI release

Superfusion of the CB₁ antagonist SR141716A 5 μM (8 min pre-stimulation, 3 min co-stimulation with capsaicin and 13 min post-stimulation), significantly increases the evoked release of SP-LI, obtained with capsaicin doses 0.1 – 100 μM (Figure 3). Following SR1 superfusion 8 min before stimulation, mean SP-LI release was (34.0±5.84 fmol 8 ml⁻¹). SR1 failed to alter basal levels of SP-LI release significantly when compared to samples collected before SR1 superfusion (29.76±5.44 fmol 8 ml⁻¹), (n=21 cord sections), nor did it significantly increase release from cords treated with the non-effective 0.01 μM concentration of capsaicin. After SR1 treatment, SP-LI release was maximized using the 100 μM capsaicin dose, instead of 5 μM. This suggests that the reduction in SP-LI release seen at high capsaicin doses, may be mediated by CB₁ activity.

CB₁ agonist Anandamide inhibits capsaicin-evoked SP-LI release

CB₁ receptors in the spinal cord were activated by superfusion of a receptor agonist anandamide for 3 min before and during superfusion of capsaicin. Anandamide had no effect on basal SP-LI measured in 8 and 16 min samples, but significantly inhibited the release of SP-LI compared to control cords superfused with the same concentration of capsaicin (Figure 4).

Discussion

In this study we show that the CB₁ receptor antagonist SR141716A modulates capsaicin stimulated release of the excitatory transmitter SP from the mouse spinal cord. Release levels evoked by capsaicin are comparable with stimulated SP release from rat dorsal horn preparations (Malcangio & Bowery, 1993). A significant proportion of evoked SP-LI release in mouse superfusates is attributable to a VR1 receptor mediated response to capsaicin, as SP-LI release is significantly reduced by the VR1 antagonist capsazepine. The effective capsaicin dose range used to release SP is also close to the reported 1 μM affinity for capsaicin at VR1 (Caterina et al., 1997). Electrophysiological studies indicate that most capsaicin sensitive neurones are polymodal nociceptors (Holzer, 1991) and most SP and CGRP containing cells of this population express mRNA for VR1 (Michael & Priestley, 1999). It is likely that most capsaicin-evoked SP is released from nociceptive neurones expressing VR1 as these neurones can release peptides after capsaicin stimulation 1 – 100 μM, even when isolated from other cell types in culture (Ulrich-Lai et al., 2001). Increases in capsaicin-stimulated release of SP in the presence of the CB₁ antagonist SR1, implies that CB₁ receptors are tonically active in the spinal cord and function to reduce SP release after stimulation of nociceptive fibres. Activating these receptors by exogenously applying the CB₁ agonist anandamide to spinal cord sections, reduces the capsaicin-evoked release of SP, as was previously shown for the release of CGRP peptide from the rat dorsal horn (Tognetto et al., 2001; Richardson et al., 1998). In addition, some behavioural studies have shown that intrathecal delivery of SR1 to the spinal cord lowers the nociceptive threshold in both normal and hyperalgesic animals (see Walker et al., 2000), suggesting tonic modulation of nociceptive responses by CB₁. In this study, SR1 does not increase either basal SP release or release with a non-effective capsaicin concentration (0.01 μM), suggesting that CB₁ activity is only effective in regulating SP release after sufficient nociceptor stimulation. The SR1 mediated increase in SP-LI, was not related to the effectiveness of the dose of capsaicin used to stimulate its release. This may be indicative of inverse agonist activity reported for SR1 (Walker et al., 2000). Alternatively, this could be explained by a non-specific action of capsaicin, such as activating depolarizing Potassium channels or desensitizing VR1 receptors (Bevan & Szolcsanyi, 1990). The reduction in SP release using high dose capsaicin appears to be CB₁ mediated, as this effect is reversed using SR1, perhaps reflecting that the endogenous activity of CB₁ receptors is increased after intense ‘noxious’ stimulation with high dose capsaicin. This could also be evidence that production of endogenous cannabinoids might be induced by noxious stimulation, in order to account for the increased activation of these receptors. Formation of the putative endogenous CB₁ ligand anandamide seems to occur after Ca²⁺ dependent depolarization of neurones (di marzo et al., 1993) and its noxious stimulus-evoked release has been demonstrated in periaquaductal grey tissue after formalin injections (Walker et al., 1999).

Anandamide has been recently detected in spinal cord tissue, but its release after stimulation has yet to be demonstrated (di marzo et al., 2000), however in vitro studies showed cells over-expressing VR1, released anandamide in response to capsaicin stimulation (di marzo et al., 2001). This study provides further evidence that the release of excitatory transmitters from stimulated nociceptor terminals in the spinal cord is tonically controlled by endogenous cannabinoids. The mechanism by which CB₁ executes this negative control depends on where cannabinoids are acting: by reducing Ca²⁺ dependent neurosecretion directly from their pre-synaptic receptors located on primary afferent terminals, where VR1 is often co-expressed (Ahluwalia et al., 2000), or indirectly at interneuronal CB₁ sites where they might act as retrograde messengers, reducing post-synaptic excitation levels in the cord, in order to prevent potentiation of pre-synaptic transmitter release. This study supports a role for the endogenous cannabinoid signalling system in modulating nociceptive transmission in the spinal cord.