Previously, low-frequency stimulation (LFS) of a nonnociceptive touch-sensitive neuron has been found to elicit endocannabinoid-dependent long-term depression (eCB-LTD) in nociceptive synapses in the leech central nervous system (CNS) that requires activation of a presynaptic transient receptor potential vanilloid (TRPV)-like receptor by postsynaptically synthesized 2-arachidonoyl glycerol (2-AG). This capacity of nonnociceptive afferent activity to reduce nociceptive signaling resembles gate control of pain, albeit longer lasting in these synaptic experiments. Since eCB-LTD has been observed at a single sensory-motor synapse, this study examines the functional relevance of this mechanism, specifically whether this form of synaptic plasticity has similar effects at the behavioral level in which additional, intersegmental neural circuits are engaged. Experiments were carried out using a semi-intact preparation that permitted both synaptic recordings and monitoring of the leech whole body shortening, a defensive withdrawal reflex that was elicited via intracellular stimulation of a single nociceptive neuron (the N cell). The same LFS of a nonnociceptive afferent that induced eCB-LTD in single synapses also produced an attenuation of the shortening reflex. Similar attenuation of behavior was also observed when 2-AG was applied. LFS-induced behavioral and synaptic depression was blocked by tetrahydrolipstatin (THL), a diacylglycerol lipase inhibitor, and by SB366791, a TRPV1 antagonist. The effects of both THL and SB366791 were observed following either bath application of the drug or intracellular injection into the presynaptic (SB366791) or postsynaptic (THL) neuron. These findings demonstrate a novel, endocannabinoid-based mechanism by which nonnociceptive afferent activity may modulate nocifensive behaviors via action on primary afferent synapses.
- long-term depression
nociceptive signaling can be modulated by nonnociceptive afferent activity, a process referred to as gate control (Melzack and Wall 1965). Traditionally, gate control was thought to be a short-term process, lasting as long as the duration of the nonnociceptive afferent activity. Nonnociceptive Aβ-fibers activate GABAergic interneurons, which, in turn, attenuate C-fiber signaling (D'Mello and Dickenson 2008; Melzack and Wall 1965; Price et al. 2009; Zeilhofer et al. 2009). This gate control process is the basis for analgesic therapies, such as transcutaneous electrical nerve stimulators (TENS) and spinal cord stimulation (SCS). However, both TENS and SCS treatment have been shown to induce analgesic affects that outlast the actual duration of afferent stimulation, suggesting that other processes may be involved (DeSantana et al. 2008; Sluka et al. 2006; Sluka and Walsh 2003). One potential neurophysiological process that could contribute to this persistent decrease in nociceptive signaling is endocannabinoid-mediated synaptic modulation.
Endocannabinoids are lipid-based neurotransmitters, such as anandamide and 2-arachidonoyl glycerol (2-AG), involved in short- and long-term forms of synaptic depression (Katona and Freund 2012). In most cases, these transmitters are synthesized postsynaptically and act on receptors on the presynaptic neuron (retrograde signaling). One feature of endocannabinoids that makes them an attractive candidate for gate control modulation is that they are synthesized in an activity-dependent manner. Another attractive feature is that endocannabinoid-dependent long-term depression (eCB-LTD) is not necessarily restricted to the activated synaptic pathway but can elicit heterosynaptic long-term depression in other inputs that converge onto the same postsynaptic target (Chevaleyre et al. 2007; Huang et al. 2008). It is possible that the stimulation of nonnociceptive afferents (e.g., Aβ-fibers) induces endocannabinoid synthesis in the postsynaptic targets in the spinal cord that receive input from both nociceptive and nonnociceptive afferents. The resulting activity-induced endocannabinoid synthesis and/or release could subsequently depress both nonnociceptive and nociceptive afferent (e.g., C-fibers) synapses.
Invertebrates are known to possess an active endocannabinoid system that includes the same transmitters and many of the biochemical pathways found in vertebrates (Elphick 2012; Leung et al. 2008; Matias et al. 2001; McPartland et al. 2006). In the medicinal leech, identifiable nociceptive and nonnociceptive sensory afferents, along with their postsynaptic targets, are well characterized (Blackshaw et al. 1982; Nicholls and Baylor 1968). Previously, our laboratory has found that low-frequency stimulation (LFS) of a single nonnociceptive touch (T) cell can depress nociceptive (N cell) synapses in an endocannabinoid-dependent manner (Yuan and Burrell 2010, 2012). Both the N and T cells synapse onto the same postsynaptic target, the L motor neuron that innervates the longitudinal muscles (Muller and Nicholls 1974) that mediate defensive whole body shortening (Kristan et al. 2005; Shaw and Kristan 1995). This eCB-LTD required 2-AG synthesis in the postsynaptic L motor neuron and activation of a putative transient receptor potential vanilloid (TRPV)-like receptor in the presynaptic nociceptive afferent (Yuan and Burrell 2010, 2012). These previous studies were carried out in isolated ganglia and therefore were restricted to only one synapse of the shortening neural circuit. However, the whole body shortening circuit includes intersegmental elements that coordinate activation of motor neurons throughout the CNS (see Fig. 1; Shaw and Kristan 1995, 1999). Although the identity of the neuron(s) that form the intersegmental pathway is unknown, it is possible to ascertain changes at this level using semi-intact preparations in which shortening is elicited via intracellular stimulation of a single N cell. With the use of this preparation, it was found that touch afferent activity reduced shortening in parallel with synaptic depression in the intrasegmental portion of the shortening circuit (the N-to-L synapse). Furthermore, this behavioral depression was mediated by the same endocannabinoid-dependent process observed in N-to-L synaptic depression.
MATERIALS AND METHODS
Leeches (Hirudo verbana; 3 g) were obtained from commercial suppliers (Niagara Medicinal Leeches, Cheyenne, WY; Leeches USA, Westbury, NY), maintained in artificial pond water (0.52 g/l H2O Hirudo salt), and kept on a 12:12-h light-dark cycle at 18°C. The dissections and recordings were carried out in normal saline solution (in mM: 114 NaCl, 4 KCl, 1.8 CaCl2, 1 MgCl2, 5 NaOH, and 10 HEPES; pH = 7.4). Drugs for pharmacology experiments were dissolved in leech saline solution from frozen stock solutions, and final concentrations were made just before each experiment. Dimethyl sulfoxide (DMSO) and Orlistat (tetrahydrolipstatin, THL) were obtained from Sigma-Aldrich (St. Louis, MO). Drugs from Tocris (Ellisville, MO) included 2-arachidonoyl glycerol (2-AG) and SB3667912. All drugs were prepared from frozen stocks (dissolved in saline or DMSO where appropriate), and the final concentrations were prepared on the day of the experiment. Vehicle control experiments were carried out in saline that contained equivalent levels of DMSO (0.01%).
The semi-intact preparations used in these experiments were similar to those used in previous studies (Burrell and Sahley 2005) and consisted of a dissected portion of the CNS where intracellular recordings were made and an intact portion in which behavior was monitored (Fig. 1A). The leech CNS consists of a chain of ganglia linked by a connective nerve (Muller et al. 1981). Similar to the spinal cord, each ganglion has the same compliment of sensory, motor, and interneurons (notable exceptions are the head and tail ganglia, which were not used in this study). Sensory and motor neurons project to the periphery via two bilateral pairs of segmental nerves. Dissections were carried out in frozen Sylgard-filled dishes containing ice-cold leech saline. The leeches were pinned ventral side up, and an incision was made along the midline from segments 1 to 7, with the head ganglion removed to minimize response variability (Shaw and Kristan 1995, 1997). This portion of the CNS was dissected from the skin but remained connected to segments 8 through 10, which were left intact and could still shorten (Fig. 1A). The semi-intact preparation was pinned into a 3-mm Sylgard-lined dish, and Teflon-coated silver wires (0.008 in.; A-M Systems, Carlsborg, WA) were bared at the tips and inserted into the longitudinal muscle layer just beneath the skin in the intact portion of the preparation (Lewis and Kristan 1998). These electromyogram (EMG) wires were connected to a differential amplifier (model 3600; A-M Systems, Everett, WA) to record the magnitude of the shortening reflex. EMG activity was quantified by measuring the area under the EMG signal during the 1.5-s period beginning at the point where the N cell stimulus train started. To validate that the EMG measurements reflected the magnitude of the shortening reflex, a pilot study was carried out in which shortening was measured using both EMG recordings and an isometric tension transducer (model 72–4481; Harvard Apparatus, Holliston, MA). A nylon monofilament (Surgical Specialties, Reading, PA) was tied to the intact portion of the preparation and then connected to the tension transducer (Fig. 1A). EMG and tension transducer recordings of the shortening reflex were made during intracellular stimulation of a single lateral N cell in segment 4 with a train of 10 action potentials at frequencies of 2, 5, 10, and 25 Hz. For the remaining experiments, whole body shortening was elicited by intracellular stimulation of a single lateral N cell in segment 4 (10 action potentials at 10 Hz), which was found to produce a reliable shortening response in the intact portion of the preparation based on EMG and tension transducer recordings as well as visible observation.
In these same semi-intact preparations, current-clamp (bridge balanced) intracellular recordings were made using sharp glass microelectrodes (25–40 MΩ) fabricated from borosilicate capillary tubing (1.0-mm outer diameter, 0.75-mm inner diameter; FHC, Bowdoinham, ME) using a horizontal puller (P-97; Sutter Instruments, Novato, CA) and were filled with 3M K+ acetate. Impalement of individual neurons at ganglion 4 was carried out by a manual micropositioner (model 1480; Siskiyou, Grants Pass, OR). Current pulses were delivered to the microelectrodes through via a programmable stimulator (STG 1004; Multi-Channel Systems; Reutlingen, Germany), and signals were recorded using a bridge amplifier (BA-1S; NPI, Tamm, Germany). The signals were digitally converted (Digidata 1322A analog-to-digital converter) for observation and analysis (Axoscope; Molecular Devices, Sunnyvale, CA).
Individual cells were identified on the basis of their position and action potential shape. The nociceptive (N) and longitudinal (L) motor cells have two bilateral pairs, whereas the touch (T) cells have three bilateral pairs (Nicholls and Baylor 1968). The touch and nociceptive cell soma are located on the ventral side of the ganglion, whereas the motor neurons are found on the dorsal side. For these experiments, the nerve cord (segments 1–7) of the semi-intact preparation was pinned dorsal side up in the Sylgard chamber. This permits recordings from the L motor neuron and the lateral-most touch and nociceptive neurons at ganglion 4 (Fig. 1, A and B). The N-to-L synapse is mediated by a monosynaptic, chemical synapse, whereas the T-to-L synapse consists of both an electrical component and a polysynaptic chemical component (Fig. 1C).
Heterosynaptic long-term depression of the nociceptive synapse (the N-to-L synapse) was elicited by LFS of the T cell (1 Hz for 900 s) as previously reported (Yuan and Burrell 2010). Recordings of the N-to-L excitatory postsynaptic potentials (EPSPs) were made before LFS (pretest) and 60 min after LFS ended (posttest; Fig. 1D). Separate electrode impalements of the same N and L cells were made for the pre- and posttest recordings because chronic recordings of leech neurons result in the progressive rundown of the EPSP, likely due to damage from the chronic impalements (Eliot et al. 1994). Postsynaptic (L motor neuron) input resistance was recorded at the pre- and posttest levels, and only stable recordings were included in the data analysis. For all experiments, the mean pretest input resistance was 24 ± 2.4 MΩ and posttest input resistance was 22 ± 2.8 MΩ. The peak EPSP amplitude was recorded every 10 s and was calculated by averaging ≈10 EPSPs during both the pre- and posttest.
Drugs were bath-applied during LFS or for 15 min without LFS during control experiments. Bath application of drugs was isolated to the CNS, with a Sylgard wall surrounding the semi-intact skin portion of the preparation. Iontophoretic injection was applied for 5 min before LFS and consisted of −1-nA, 500-ms current pulses delivered at 1 Hz. In experiments with iontophoresis, injection of the drug was applied on either the presynaptic N cell or postsynaptic L cell (Fig. 1D).
EPSP amplitude measurements of the pre- and posttest recordings were normalized and presented as means ± SE. Statistical analyses using a one-way analysis of variance (ANOVA) were performed to determine main effects with Newman-Keuls post hoc tests to confirm the ANOVA results. Linear regression analysis was used in experiments examining the relationship between EMG and tension transducer measurements of the shortening response. The analyses were carried out with Statistica analysis software (StatSoft). All significance was determined at an α level of at least P < 0.05.
EMG recordings accurately reflect strength of the shortening response.
To determine whether the EMG recordings provided a reliable measure of the strength of the whole body shortening reflex, experiments were carried out in which shortening was monitored using EMG electrodes and a tension transducer simultaneously. Measurements were made of the shortening in response to increasing frequency of N cell stimulation. Based on the tension transducer recordings, no shortening was observed when the N cell was stimulated at the lower frequencies (2 and 5 Hz), but significant shortening was observed when the N cell was stimulated at 10 and 25 Hz (n = 3). There was a statistically significant correlation between the EMG recording and the amplitude of the response measured with the tension transducer (Fig. 2). These results indicate that the EMG electrodes provide a reliable measure of the magnitude of whole body shortening.
Nonnociceptive stimulation depressed both the shortening behavior and the nociceptive synapse.
Previously, we found that LFS of the nonnociceptive T cell resulted in depression of nociceptive synapses in isolated ganglia (Yuan and Burrell 2010, 2012). To determine whether this depression was behaviorally relevant, experiments were carried out to monitor changes in shortening following T cell LFS. N cell stimulation in ganglia (10 spikes at 10 Hz) was used to induce shortening in the semi-intact preparation. Pretest recordings of the N-to-L synapse and shortening response were carried out before T cell LFS (15 min, 1 Hz). After a 60-min rest period, posttest recordings were carried out to determine any changes in synaptic transmission and behavior. T cell LFS induced significant depression in the nociceptive N-to-L synapse, consistent with previous research (Fig. 3). In the same preparations, a significant decrease in the shortening reflex was also observed (Fig. 3). Control experiments without LFS found no change at the synaptic or behavioral level. These data indicate that a single nociceptive neuron in a single ganglion can effectively induce the shortening reflex but that this reflex can be attenuated through LFS activation of a single nonnociceptive neuron. In isolated ganglia, this process was observed to be endocannabinoid dependent; therefore, experiments were carried out to determine whether these transmitters mediated depression of the shortening behavior.
Depression of the nociceptive synapse and behavior is 2-AG dependent.
Previous research has shown that depression of the nociceptive synapse by nonnociceptive LFS could be mimicked by 2-AG or occluded by prior 2-AG treatment (Yuan and Burrell 2010, 2012). Therefore, the ability of 2-AG to alter the shortening response was also examined. After pretest recordings of N cell-elicited shortening and the N-to-L synapse, 2-AG (60 μM) was bath-applied selectively to the CNS portion of the preparation (ganglia from segments 1–7) for 15 min. After 2-AG treatment and a subsequent 60-min washout period, posttest recordings of the N-to-L synapse and shortening behavior were measured. Similar to the LFS-induced depression, 2-AG produced depression in the nociceptive synapse that was accompanied by a decrease in shortening (Fig. 4A). No change was observed in vehicle controls (0.01% DMSO). This suggests that 2-AG attenuates N cell-elicited shortening in parallel with depression in the N-to-L synapse.
Given that T cell LFS depresses shortening and that N cell-elicited shortening is sensitive to 2-AG, the question as to whether T cell LFS-induced depression requires 2-AG was addressed. 2-AG synthesis from DAG requires DAG lipase (Bisogno et al. 2003). In previous experiments in the leech, multiple inhibitors of DAG lipase were found to inhibit N-to-L eCB-LTD in isolated ganglia, specifically RHC-80267, OMDM-188, and THL (Yuan and Burrell 2010, 2012). To determine whether inhibiting DAG lipase blocks behavioral depression following T cell LFS, experiments were carried out with THL treatment. Pretest recordings of the N-to-L EPSP and N cell-stimulated shortening were followed by bath application of THL (10 μM) for 15 min onto the isolated CNS portion of the preparation during T cell LFS. THL was washed out, and posttest recordings were carried out 60 min later. THL treatment during LFS blocked the depression normally observed in the nociceptive synapse, as well the attenuation of the shortening reflex (Fig. 4B). Control experiments in which THL was applied without LFS did not reveal any nonspecific effects of drug treatment. These findings support the hypothesis that LFS-induced depression of synaptic transmission and shortening is endocannabinoid dependent, specifically requiring 2-AG.
In previous experiments in which THL was selectively injected into either the pre- (N cell) or postsynaptic (L motor neuron) cell, only postsynaptic injection blocked LFS-induced eCB-LTD, indicating that the L motor neuron was the site of 2-AG synthesis (Yuan and Burrell 2010). Postsynaptic injection of THL (10 μM) was repeated in the semi-intact preparation, and the effects on both the nociceptive synapse and the elicited shortening response were tested. As in previous experiments, postsynaptic THL injection blocked LFS-induced depression in the N-to-L synapse (Fig. 4C). Surprisingly, N cell-elicited shortening was also prevented, indicating that 2-AG synthesis in the L motor neuron affected the intersegmental neural circuit mediating whole body shortening. This result suggests that blocking 2-AG synthesis in the L motor neuron is sufficient to block eCB-LTD of other nociceptive synapses that contribute to the intersegmental shortening neural circuit.
As mentioned earlier, pretreatment with 2-AG occludes subsequent LFS-induced depression at the N cell synapse (Yuan and Burrell 2010). These experiments were repeated in the semi-intact preparations; however, rather than bath-applying 2-AG as was done previously, 2-AG (60 μM) was injected into the N cell prior to pretest recordings of the behavior and synapse. In these pretreated preparations, subsequent LFS failed to elicit depression of either the N cell synapse or the shortening reflex (Fig. 4C). These results are consistent with 2-AG occluding subsequent LFS-induced synaptic and behavioral depression and extend previous occlusion experiments by focusing 2-AG pretreatment onto the N cell.
Depression of the nociceptive synapse and behavior involves a presynaptic TRPV-like receptor.
Previous pharmacological studies in our laboratory have indicated that a TRPV-like receptor was involved in eCB-LTD in the leech. Specifically, eCB-LTD could be mimicked and/or occluded by the TRPV activators capsaicin and resiniferatoxin and could be blocked by the TRPV antagonists capsazepine or SB366791 (Li and Burrell 2011; Yuan and Burrell 2010, 2012). Therefore, the role of this TRPV-like receptor was examined during endocannabinoid-dependent depression of the shortening reflex. The selective TRPV1 antagonist, SB366791 (10 μM), has previously been shown to block eCB-LTD in the leech (Yuan and Burrell 2010, 2012). Bath application of SB366791 to the CNS portion of the preparation prevented 2-AG-induced depression of both the nociceptive synapse and N cell-elicited shortening (Fig. 5A). The ability of SB366791 to block LFS-induced depression of shortening was also examined. Bath application of SB366791 to the CNS portion of the preparation blocked LFS-induced depression at the behavioral and synaptic level, suggesting that this activity-induced depression required the TRPV-like receptor (Fig. 5B).
Next, experiments were carried out in which SB366791 was injected into the N cell that was stimulated to elicit shortening. In previous experiments, injection of a TRPV1 antagonist into the presynaptic, but not postsynaptic, neuron blocked eCB-LTD (Li and Burrell 2011; Yuan and Burrell 2010). Injection of SB366791 (10 μM) into the N cell not only blocked LFS-induced depression in the nociceptive synapse but also prevented attenuation of the shortening behavior following LFS (Fig. 5C). No changes in synaptic signaling or behavior were observed when the N cell was injected with SB366791 but the LFS was omitted. In control experiments in which SB366791 was injected into the T cell, LFS-induced depression of both the nociceptive synapse and N cell-elicited shortening was still observed, providing further evidence that the effects of SB366791 were restricted to the N cell. These results suggest that a presynaptic TRPV-like receptor mediated nociceptive signaling at both the synaptic and behavioral level.
Previous research in our laboratory has shown that, in isolated ganglia, nonnociceptive (T cell) LFS depresses the nociceptive (N-to-L) synapse. On the basis of pharmacological studies, this heterosynaptic depression was mediated through postsynaptic 2-AG synthesis in the L motor neuron, where 2-AG travels in a retrograde manner to bind to putative TRPV-like receptors on the presynaptic nociceptive neuron (Yuan and Burrell 2010, 2012). This eCB-LTD required presynaptic calcineurin activation and both pre- and postsynaptic increases in intracellular Ca2+, similar to TRPV1-mediated eCB-LTD in the hippocampus (Jensen and Edwards 2012; Yuan and Burrell 2012). Novel gene expression was also necessary, specifically RNA synthesis in the presynaptic neuron and protein synthesis in the pre- and postsynaptic neurons (Yuan and Burrell 2013). In the present study, the potential functional relevance of this heterosynaptic eCB-LTD was examined using the whole body shortening reflex.
Whole body shortening in the leech is a defensive withdrawal reflex in response to noxious mechanosensory stimuli that involves a near-synchronous contraction of all body segments (Shaw and Kristan 1999; Stuart 1970). The reflex is initiated by activation of the P or N cells, and though it appears that activity in multiple pressure-sensitive (P) cells is required to elicit shortening (Shaw and Kristan 1995), the present study shows that activation of a single N cell is sufficient to induce shortening, at least in these semi-intact preparations. These primary afferents have direct input onto motor neurons responsible for shortening that are in the same or neighboring segments, and it is in this intrasegmental component of the shortening circuit that eCB-LTD has been previously observed using single-ganglion experiments. However, these same afferents are also thought to activate an intersegmental pathway that carries the signal to shorten throughout the leech CNS (Shaw and Kristan 1999). The identity of the neuron or neurons that make up this intersegmental pathway is unknown, but its presence can be inferred from the activation of shortening-related motor neurons down the length of the nerve cord (Arisi et al. 2001; Shaw and Kristan 1999). Although it is not possible to directly record from the N-to-shortening interneuron synapse, it is possible to observe changes in this signaling pathway by direct intracellular stimulation of a single nociceptive afferent and monitoring of the subsequent level of shortening.
LFS of the nonnociceptive T cell attenuated the leech shortening reflex, and this was accompanied by eCB-LTD at the N-to-L synapse. Application of the DAG lipase inhibitor THL blocked LFS-induced attenuation of N cell-elicited shortening and depression of the N-to-L synapse. THL was applied only to the CNS of the preparation; therefore, the observed effects were at the level of N cell input to the intersegmental neuron(s) and not where these interneurons synapse onto the motor neurons. Exogenous 2-AG application to the CNS reduced N cell-elicited shortening and induced depression at the N-to-L synapse, mimicking the effects of LFS. Injection of 2-AG into the N cell also occluded subsequent LFS-induced synaptic and behavioral depression. Together, these findings are consistent with the hypothesis that the endocannabinoid 2-AG mediates LFS-induced behavioral and synaptic depression.
Previous studies have indicated that T cell LFS stimulates 2-AG synthesis in the L motor neuron, which then presumably travels in a retrograde manner to modulate the presynaptic N cell (Yuan and Burrell 2010), similar to most other examples of eCB-LTD (Katona and Freund 2012). When these postsynaptic THL injection experiments were repeated in the current study, not only was synaptic depression prevented, but LFS-induced depression of shortening was blocked as well. This is a surprising finding because it implies that 2-AG synthesis and release from the L motor neuron not only alters synaptic input that it receives but also that it also acts on other synaptic targets of the N cell, e.g., those associated with carrying the intersegmental signal to shorten (see shortening circuit diagram in Fig. 1B). One possible explanation is that 2-AG is being released extensively within the neuropil by the L motor neuron so that all synaptic targets of the N cell (and presumably other synapses sensitive to 2-AG) are affected. Alternatively, it is possible that this change in all the N cell synapse occurs through an intracellular signal that is initiated at the N-to-L synapses and propagates throughout the N cell to other connections made by this afferent. eCB-LTD in this neuron is known to require new protein synthesis in the presynaptic N cell (Yuan and Burrell 2013), which could conceivably contribute to a signal that spreads intracellularly to other presynaptic terminals. Such a process would represent an interesting and novel element to eCB-mediated synaptic plasticity.
Because these studies utilize controlled stimulation of nociceptive and nonnociceptive afferents in terms of spike number and frequency, they pose interesting questions in terms of our understanding of this persistent form of gate control modulation. For example, the observed synaptic depression would appear to mediate a decrease in nocifensive behavior, but can this decrease in synaptic signaling be overcome by increased nociceptive afferent activity (e.g., doubling the firing rate), or is nociceptive signaling effectively blunted across a range of activity levels? Put another way, is the gate effectively closed or simply harder to open? Another issue is the pattern of nonnociceptive (T cell) activity that produces depression of nociceptive synapses and nocifensive signaling. Do different levels of T cell activity produce proportionate levels of depression in nociceptive synapses? To what extent do non-eCB-mediated forms of plasticity contribute to this heterosynaptic depression given that T cell activity can produce both eCB- and NMDA receptor-mediated homosynaptic depression (Li and Burrell 2008)? N cell LFS itself is capable of producing homosynaptic LTD (Yuan and Burrell 2010), and we have not yet investigated whether such depression also affects nocifensive shortening.
Invertebrates possess the endocannabinoid neurotransmitters and potentially the same synthesizing/metabolizing enzymes found in vertebrates, but they lack orthologs to the CB1/CB2 receptors (Burke et al. 2006; Elphick 2012; Leung et al. 2008; Matias et al. 2001; Salzet and Stefano 2002). However, endocannabinoids have been observed to bind to TRPs (De Petrocellis et al. 2001; Di Marzo and De Petrocellis 2010), including the TRPV receptor, to mediate synaptic plasticity (Chavez et al. 2010; Gibson et al. 2008; Grueter et al. 2010; Jensen and Edwards 2012; Maione et al. 2009). In previous studies, we have observed that the TRPV1 antagonists capsazepine and SB366791 could block eCB-LTD in the leech (Li and Burrell 2011; Yuan and Burrell 2010, 2012). In the present study, bath application of SB366791 to the CNS portion of the semi-intact preparation was found to block both 2-AG- and LFS-induced eCB-LTD of the N-to-L synapse. SB366791 bath application also prevented 2-AG- and LFS-induced attenuation of the shortening reflex. In addition, intracellular injection of SB366791 into the presynaptic N cell blocked LFS-induced attenuation of shortening and eCB-LTD of the N-to-L synapse. These results are consistent with the hypothesis that activation of a presynaptic TRPV-like receptor is required for 2-AG-mediated depression of synapses and behavior. These findings are also consistent with evidence that spinal TRPV1 receptors can elicit an analgesic effect, possibly as a result of persistent depression of evoked C-fiber EPSPs (Baccei et al. 2003; Di Marzo and De Petrocellis 2010; Kusudo et al. 2006; Yang et al. 1999).
In this study, we observed that afferent activity that elicits eCB-LTD of a nociceptive synapse also attenuates nocifensive behavior. There is considerable interest in the development of endocannabinoid-based therapies for pain, and the present findings are relevant to understanding both the cellular mechanisms of endocannabinoid-mediated modulation of nociceptive circuits and the role these neurotransmitters play in mediating analgesia that is induced by somatosensory afferent activity. For example, a recent study has shown that LFS of spinal afferents produces eCB-LTD in nociceptive synapses (Kato et al. 2012), similar to our own observations. We and others (Kato et al. 2012) have proposed that this endocannabinoid-mediated depression of nociceptive synapses may play a role in TENS and SCS therapies in which the analgesic effects are produced by repetitive stimulation of somatosensory afferents. As discussed in the Introduction, it has been noted that the analgesic effects of TENS and SCS therapies persist long after the afferent stimulation has ended (DeSantana et al. 2008; Sluka et al. 2006; Sluka and Walsh 2003). Endocannabinoid-mediated neuroplasticity has a number of properties that would be consistent with a role in these therapies, specifically activity-dependent transmitter synthesis, persistent modulatory effects, and the capacity to alter both active and neighboring inactive synapses. The findings of this study provide the most direct evidence to date that endocannabinoids may contribute to TENS/SCS-induced analgesia. Endocannabinoids have also been shown to contribute to stress-induced analgesia, which, similar to gate control, requires somatosensory stimulus, albeit at noxious levels (Gregg et al. 2012; Hohmann et al. 2005; Nyilas et al. 2009). It would be interesting to test whether the noxious stimuli used to elicit stressed-induced analgesia in mammals also produces endocannabinoid-dependent decreases in nociceptive synapses and nocifensive behavior in the leech.
The physiological and behavioral properties of nociception exhibit a remarkable level of evolutionary conservation; therefore, studies using invertebrate nervous systems can provide valuable insights concerning nociception in mammals (Tobin and Bargmann 2004; Walters and Moroz 2009). Similar to vertebrates, nociceptive afferents in the leech can be divided into mechanonociceptors (the medial N cells) and polymodal nociceptors (the lateral N cell used in this study), the latter responding to mechanical stimuli, nociceptive thermal stimuli, low pH, and capsaicin (Pastor et al. 1996). These features, combined with the ability to record from both identifiable afferents for light touch or sustained pressure and their postsynaptic targets, and similarities in the endocannabinoid-mediated synaptic plasticity make the leech a useful model system for discovering fundamental biological processes related to nociception.
This work was supported by National Science Foundation Grant IOS-1051734 and National Institute of Neurological Disorders and Stroke Grant F31NS074473. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Neurological Disorders and Stroke or the National Institutes of Health.
No conflicts of interest, financial or otherwise, are declared by the authors.
S.Y. and B.D.B. conception and design of research; S.Y. performed experiments; S.Y. analyzed data; S.Y. and B.D.B. interpreted results of experiments; S.Y. and B.D.B. prepared figures; S.Y. and B.D.B. drafted manuscript; S.Y. and B.D.B. edited and revised manuscript; S.Y. and B.D.B. approved final version of manuscript.
We thank the three anonymous reviewers whose helpful comments substantially strengthened this manuscript.
- Copyright © 2013 the American Physiological Society