Melittin-induced Inflammatory Pain And Hyperalgesia

In the past 8 years, we have been working on the bee venom (BV) test, a novel inflammatory pain model produced by the experimentally-produced honeybee's sting, which was firstly introduced as a new tonic pain model by Lariviere and Melzack (1996) and followed by a series of Chen's group research work describing the model with both a feature of tonic nociception and a multifaceted feature of inflammatory pain hypersensitivity, including primary heat and mechanical hyperalgesia, secondary heat hyperalgesia and mirror-image heat hyperalgesia in behaviorally awake rodents following subcutaneous (s.c.) injection of a solution containing the whole bee-venom (Apis mellifera) (Chen et al., 1999b; Chen and Chen, 2000, 2001; Chen, 2003). In the in vivo electrophysiological recordings, the unique expressions of tonic or persistent spontaneous nociception (PSN) and inflammatory pain hypersensitivity have been demonstrated to be mediated by plastic changes in functions of spinal dorsal horn wide-dynamic-range (WDR) neurons, which are believed to play pivotal roles in mediation of spinally-organized nociceptive flexion reflex (Chen et al., 1998, 1999a; You and Chen, 1999; You et al., 2002, 2003; Zheng et al., 2002; Chen,2003). Under pathological pain state of peripheral bee venom insult, spinal dorsal horn WDR neurons became hyperexcitable and hypersensitive with initial long-term ongoing spike discharges lasting for 1-2 h followed by a even much longer time of an enhanced responsiveness to both heat and mechanical stimuli applied to the cutaneous receptive field (cRF) (Chen et al., 1998, 1999a; You and Chen, 1999; You et al., 2002; Zheng et al., 2002; Chen, 2003). The time course of functional changes of spinal dorsal horn neurons shown by c-Fos expression is also comparable to that of behavioral and electrophysiological studies (Luo et al., 1998; Chen, 2003). Taken together, the above results implicate that central neuronal change in the spinal cord dorsal horn plays crucial role in development and maintenance of BV-induced pain and hypersensitivity. The BV test has also been proved to be as valid as other existing models of inflammatory pain and is likely to be more valuable than others in evaluation of endogenous and exogenous substances associated with production (algogenic) or inhibition (analgesic) of tonic spontaneous pain, primary heat and mechanical hyperalgesia, secondary and mirror-image hyperalgesia as well as inflammatory process in an individual animal, because it can largely minimize the difference in animal individuals and models (Chen et al., 2000, 2001, 2003; Chen and Chen, 2000; Li et al., 2000, Li and Chen, 2003; Zheng and Chen, 2000, 2001; You et al., 2002, 2003, Chen, 2003; Sun et al., 2004a, b, 2005 in press; Yu and Chen, 2005 in press). The BV model has also been used to evaluate anti-nociceptive, anti-hyperalgesic and anti-inflammatory effect of naturally occurring substances (e.g., Traditional Chinese Medicinal herbs) (Peng et al., 2003; Shang et al., unpublished data).Although a central change has been primarily believed to play an important role in the maintenance of the BV-induced abnormal pain behaviors, peripheral mechanisms are likely to be more crucial because some results of our previous studies suggest that establishment of the long-term central changes is peripherally dependent and requires a time window for temporary summation of the primary afferent input (Chen et al., 1998, 1999a, 2000, 2001; Chen and Chen, 2001; Youet al., 2002; Chen, 2003). These data imply that some specific biotoxins of the whole BV are potential candidates to be involved in production of the BV-induced abnormal pain behaviors and central changes according to the unique characteristics of the BV test. Regarding this, Lariviere and Melzack (2000) compared the effects of six venoms from hymenoptera (including honeybee, yellow jacket, paper wasp, bumblebee, yellow hornet and white-faced hornet) and found that the honeybee venom was the most effective to produce the highest score of PSN, suggesting a difference in components of the hymenoptera venoms. BV is a combination of many chemical components including of melittin, phospholipase A2 (PLA2), apamin, histamine and other substances; however, over 50% of the component of the BV is melittin, a 26 amino acid amphipathic peptide (Habermann, 1972; Dempsey, 1990; Lariviere and Melzack, 1996). The toxic peptide has been assumed to produce pain by directly acting on peripheral nerve terminals and/or by releasing potassium ions through cell lysis (Habermann, 1972). More recently, some researchers demonstrated that intradermal injection of melittin (5 jig in 50 ul saline) could produce spontaneous pain, mechanical hyperalgesia and neurogenic inflammation in human subjects that are similar to the animal responses to s.c. injection of the BV solution (Chen et al., 1999b; Chen and Chen, 2000; Koyama et al., 2000, 2002; Sumikura et al., 2003). Single fiber recordings also showed activation of both mechanical and mechanoheat nociceptors of goat by s.c. injection of melittin (Cooper and Bomalaski, 1994). Other BV-contained substances such as apamin, PLA2, histamine and other inflammatory mediators are not likely the major chemical components responsible for the long-lasting effects on the peripheral nociceptors and spinal dorsal horn neurons, since the results of previous studies are not supportive (Keele and Armstrong, 1964; Bleehan and Keele, 1977; Wheeler-Aceto et al., 1990; Tjolsen et al., 1992; Hong and Abbott, 1994; Koyama et al., 2000, 2002). Based on the results of the above human and animal studies, we propose that melittin is one of the major algogenic components of the BV-produced long-term changes inperipheral and central neural plasticity as well as abnormal pain behaviors.In the present Ph.D. dissertation, behavioral surveys, in vivo electrophysiological single unit recording, in vitro whole-cell patchclamp recording and Ca2+-imaging were used and the biological effects of melittin were well studied. Followed are our main results:Part 1, Characteristics of melittin-induced inflammatory ongoing pain: involvement of peripheral thermal nociceptor (TRPV1) in hyperexcitability of spinal dorsal horn nociceptive neurons;Part 2, Characteristics of melittin-induced inflammatory heat and mechanical hypersensitivity: involvement of peripheral phospholipase A2-lipoxygenase pathway in activation of thermal nociceptor TRPV1 and heat hypersensitivity of spinal dorsal horn nociceptive neurons.Results:Part 1: Characteristics of melittin-induced inflammatory ongoing pain: involvement of peripheral thermal nociceptor (TRPV1) in hyperexcitability of spinal dorsal horn nociceptive neurons1. Melittin-induced inflammatory responsesS.c. injection of three doses of both melittin (5, 25 and 50 \xg) and BV (10, 50 and 100 ug) could produce a similar dose-dependent increase in paw volume (edema) and the O.D. value of Evans blue (plasma extravasation) measured in the injected hind paw, while s.c. injection of saline did not produce such effect. This result was further demonstrated by the H-E histological staining of the skin showing a dramatic plasma infiltration with leucocytes and red blood cells from the peripheral blood vessels to the subcutaneous layers at the injection site, while vehicle did not produce observable plasma extravasation at the locus. This result indicated that melittin could induce inflammatory responses similar to BV test.2. Melittin-induced ongoing pain-related responses and long-lasting firing of spinal dorsal horn nociceptive neurons(1) In the behavioral surveys, s.c. injection of three doses of both melittin (5, 25and 50 jj.g) and BV (10, 50 and 100 ug) into the posterior surface of one hind paw of rats produced an immediate tonic nociceptive response displaying as persistent spontaneous paw flinching reflex. Similar to the BV test, the melittin response was also monophasic and dose-dependent in terms of both intensity and time course.(2) On the other hand, WDR neurons are believed to be the major inter-connecting neurons of spinal dorsal horn involved in mediation of the spinally-organized nociceptive flexion reflex (Chen, 2003; You et al., 2003), so we use the extracellular electrophysiological recordings to observe the changes of WDR neurons after melittin. It was found that s.c. injection of the same three doses of melittin into the cRF produced an immediate, dose-dependent increase in spontaneous spike discharges of WDR neurons. The melittin-induced ongoing spike responses are similar to the behavioral flinching reflex in terms of both duration and frequency. The pattern and time course in response to melittin (50 fig) and BV (100 ug) were comparable and showed no significant difference in the total mean number of spikes.These results indicated that melittin could induce PSN and long-term spinal neuronal changes similar to the BV test.3. Involvement of peripheral thermal nociceptor (TRPV1) in hyperexcitability of spinal dorsal horn nociceptive neuronsWe have proved that BV-induced pain is mediated by capsaicin-sensitive primary afferents (Chen et al., 1998, 1999a; Chen and Chen, 2001; You et al., 2002), so next we want to know the roles of the peripheral TRPV1 inmelittin-induced PSN.(1) In the behavioral surveys, peripheral pre-treatment with a TRPVl inhibitor, capsazepine (0.3 mg/ 50 ul), resulted in a suppressive effect on the development of paw flinching reflex compared with the control groups. Post-treatment with capsazepine (0.3 mg/ 50 ul) could partially suppress the paw flinching reflex during the whole time course.(2) In the extracellular electrophysiological recordings, a total of 22 WDR neurons were recorded from the spinal dorsal horn of the L4-5 segments in 22 anesthetized rats. 11 were used for test the effects of pre-treatment with drug or vehicle while the other 11 units were used for post-treatment. Peripheral pre-treatment with capsazepine (0.3 mg/ 50 ul) (n=5) resulted in a suppressive effect on the development of increase in spontaneous spike discharges compared with the control groups (n=6). Post-treatment with capsazepine (0.3 mg/ 50 ul) could partially suppress the increase in spontaneous spike discharges during the whole time course.These results indicated that TRPVl had been involved in melittin-induced PSN and long-term spinal neuronal changes.Part 2, Characteristics of melittin-induced inflammatory heat and mechanical hypersensitivity: involvement of peripheral phospholipase A^lipoxygenase pathway in activation of thermal nociceptor TRPVl and heat hypersensitivity of spinal dorsal horn nociceptive neurons1. Characteristics of melittin-induced inflammatory heat and mechanical hypersensitivity(1) In the behavioral surveys, after s.c. injection of three doses of both melittin (5, 25 and 50 ug) and BV (10, 50 and 100 fig), the two higher doses of melittinproduced a distinct reduction in both paw withdrawal thermal latency (PWTL) and paw withdrawal mechanical threshold (PWMT) in the injection site although the lowest dose had no such effect, suggesting induction of both primary heat and mechanical hypersensitivity at the injection site by local melittin treatment. The melittin-induced heat and mechanical hypersensitivity were comparable to the BV induced ones.(2) In the extracellular electrophysiological recordings, a period of 10 s heat stimulus (42 °C, 45 °C, 47 °C and 49 °C) was applied to the cRF of 12 units prior to and 1 h after melittin injection. Of 12 units tested, 8 units were responsive to heat stimuli. The responsiveness of a WDR unit was gradedly increased with the increase in skin temperature prior to local melittin injection, however, under this normal state, only the local skin temperature beyond 47°C could induce robust spike discharges. In contrast, the responsiveness of the same WDR unit was significantly enhanced by 50 ug melittin injection into the cRF with a distinct leftward shift of the stimulus-response functional curve, and under such abnormal condition, heat stimulus with even 42°C could evoke almost the same amounts of spikes as 47°C did under normal state. Moreover, the melittin-induced hyper-responsiveness of WDR neurons to heat stimuli was similar to what BV produced. A period of 10 s mechanical stimuli were applied to the cRF of 12 units prior to and 1 h after melittin injection. The responsiveness of a WDR unit was gradedly increased with the increase in mechanical intensity (brush, pressure and pinch) prior to local melittin injection, however, in contrast, the mechanical stimuli-evoked spike discharges were also significantly increased by both local melittin (50 ug) and BV (100 jig) injection and there was also a leftward shift of the stimulus-response functional curves for both the melittin- and the BV-induced inflammatory states compared with the normal state.These results indicated that melittin could induce thermal and mechanical hypersensitivity and long-term spinal neuronal changes.2. Involvement of peripheral thermal nociceptor TRPV1 in melittin-induced heat hypersensitivity, but not mechanical hypersensitivity(1) In the behavioral surveys, peripheral pre-treatment with a TRPV1 inhibitor,capsazepine (0.3 mg/ 50 ul) produced a blocking effect on the development of the primary heat hypersensitivity. Peripheral post-treatment with capsazepine (0.3 mg/ 50 ul) could significantly reverse the established primary heat hypersensitivity. In contrast, peripheral pre- or post-treatment with capsazepine (0.3 mg/ 50 ul) had no blocking effect on the development or maintenance of primary mechanical hypersensitivity.(2) In the extracellular electrophysiological recordings, to compare the effects of capsazepine with vehicle on the heat and mechanical responsiveness of spinal WDR neurons, a period of 10 s heat stimulus (42 °C, 45 °C, 47 °C and 49 °C) and mechanical stimulus (brush, pressure and pinch) was applied to the centre of the cRF of 15 units prior to and 1 h after melittin injection. Of 15 units tested, 13 units were responsive to heat stimuli and 15 units were responsive to mechanical stimuli. Peripheral post-treatment with capsazepine (0.3 mg/ 50 ul) could significantly reverse the increased responsiveness of the same WDR unit to the same each heat stimulus; but not significantly reverse the increased responsiveness of the same WDR unit to the same each mechanical stimulus.These results indicated that TRPV1 has been involved in melittin-induced heat hypersensitivity and long-term spinal neuronal changes in response to heat stimuli. For melittin could activate phospholipase A2, so we want to know whether endogenous fatty acids have been involved in melittin-induced heat hypersensitivity. If so, melittin would activate TRPV1 via an indirect pathway.So in the behavioral study, we used a phospholipase A2 inhibitor (antiflammin-1), a lipoxygenase enzymes inhibitor (nordihydroguaiaretic acid) and a cyclo-oxygenase enzymes inhibitor (indomethacin) to test the effects of endogenous fatty acid in melittin-induced heat hypersensitivity. We found that antiflammin-1 (2 (ag/50 fj,l, n=8), nordihydroguaiaretic acid (NDGA; 1 mg/50 (il, n=6) could significantly suppress melittin-induced heat hypersensitivity, but indomethacin (1 mg/50 ul, n=9) had no effect on the melittin-induced heat hypersensitivity. This result indicated that phospholipase A2-lipoxygenase pathway has been involved in melittin-induced heat hypersensitivity, which is mediated by TRPV1.3. Involvement of peripheral phospholipase A2-Iipoxygenase pathway in activation of thermal nociceptor TRPV1 which contributes to melittin-induced heat hypersensitivityAs well known, thermal nociceptor TRPV1 is highly expressed in dorsal root ganglion (DRG) cells (Caterina et al., 1997; Gunthorpe et al., 2002). So we used the whole-cell recording in DRG cells to elucidate the relationship of melittin and TRPV1.(1) Activation of DRG cells by melittin: we performed a series of patch-clamp recording and calcium imaging on DRG cells. We recorded 181 small cells (<25 jam in diameter), 85 middle cells (25-35 urn in diameter) and 36 large cells (>35 Jim in diameter). Among these cells, melittin induced inward current on 150/181 small cells, 35/85 middle cells and 6/36 large cells. This result indicated that melittin mostly activates small and middle DRG cells. We also found that melittin induced inward current in a dose-dependent manner. On DRG cells, melittin (1 ^M) induced an inward current ranging from 200 pA to 1.2 nA. Successive applications of 1 uM melittin produced a progressive increase orsensitization in responses. During the whole-cell configuration, the cells showed robust inward current responses (at a holding potential of -70 mV) that developed with a long latency (40-60 s) after first application of melittin. In the calcium imaging, we observed that capsaicin and melittin both increase the intracellular [Ca2+] level, but the application latency of melittin is longer than capsaicin.(2) Contribution of TRPV1 to melittin-induced activation of DRG cells: we performed a series of patch-clamp recording and calcium imaging on capsaicin-sensitive DRG cells to test the relationship of melittin and TRPV1. Melittin (1 fiM) induced an inward current on capsaicin-sensitive DRG cells. Current-voltage curve showed that reversal potential was near to 0 mV (Erev = 3 ± 0.8 mV, n = 8), which indicated that melittin maybe open a non-selective cation channel. In the calcium imaging, a TRPV1 non-selective inhibitor, Ruthenium red (1 uM) could reduce the melittin-induced increase of intracellular [Ca2+]. In the whole-cell recording, a TRPV1 selective inhibitor, capsazepine (1 uM) could significantly inhibit melittin-induced inward current in capsaicin-sensitive DRG cells. So these results indicated the non-selective cation channel opened by melittin is TRPV1.In brief, melittin could activate TRPV1 on DRG cells, which could mediate behavioral persistent pain and heat hypersensitivity. Now a problem was produced: how does melittin activate TRPV1, directly or indirectly? If indirectly, what is the pathway of melittin activating TRPV1?(3) Involvement of phospholipase A2-lipoxygenase pathway in activation of TRPV1: we proved that melittin activates TRPV1 on DRG cells, but the application latency of melittin is longer than application latency of capsaicin. Compared melittin with capsaicin, their molecular structures are different. So we speculated that melittin may indirectly activate TRPV1. For melittin couldactivate phospholipase A2, then we further test the effects of endogenous fatty acid in melittin-induced inward current.In the whole-cell recording on DRG cells, to further understand the pathway of melittin activating TRPV1, we used some inhibitors of lipid metabolism to test the effects of endogenous fatty acid in melittin-induced inward current. In the whole-cell configuration, melittin-induced inward current could be completely and reversibly blocked by a potent phospholipase A2 inhibitor, antiflammin-1 (500 nM, n=8). While the inhibitor of phospholipids C, U73122 (2 uM, n=9) could not reverse melittin-induced inward current. This result indicated that melittin activates TRPV1 via phospholipase A2, but not phospholipids C. As well known, the product of phospholipase A2 is arachidonic acid, which could be catalyzed by lipoxygenase enzymes and cyclo-oxygenase enzymes respectively to produce different products. Then we tested the roles of lipoxygenase enzymes and cyclo-oxygenase enzymes in melittin-induced inward current. We found that a lipoxygenase enzymes inhibitor, nordihydroguaiaretic acid (NDGA; 10 uM, n=7) could significantly block melittin-induced inward current while a cyclo-oxygenase enzymes inhibitor, indomethacin (10 uM, n=9) would not reduce melittin-induced inward current. This result indicated that lipoxygenase enzymes, but not cyclo-oxygenase enzymes have been involved in melittin activating TRPV1.We also tested the roles of second messengers in melittin-induced inward current. We found that melittin-induced inward current could be completely and reversibly blocked by a potent PKC inhibitor, Bisindolylmaleimide I (BIM 1 uM, n=12). A PKA inhibitor, H-89 (1 ^M, n=9) could partly block melittin-induced inward current while a PKG inhibitor and a mitogen activated protein kinase kinase (MEK) inhibitor, KT5823 (1 uM, n=9) or U0126 (20 uM, n=ll) would not...