- Email: [email protected]
Lack of involvement of capsaicin-sensitive primary afferents in nerve-ligation injury induced tactile allodynia in rats
Pain 79 (1999) 127–133
Lack of involvement of capsaicin-sensitive primary afferents in nerve-ligation injury induced tactile allodynia in rats Michael H. Ossipov, Di Bian, T. Philip Malan Jr., Josephine Lai, Frank Porreca* Department of Pharmacology, University of Arizona Health Sciences Center, Tucson, AZ 85724, USA Received 6 April 1998; received in revised form 14 July 1998; accepted 13 August 1998
Abstract Tactile allodynia and thermal hyperalgesia, two robust signs of neuropathic pain associated with experimental nerve injury, have been hypothesized to be mechanistically distinguished based on (a) fiber types which may be involved in the afferent input, (b) participation of spinal and supraspinal circuitry in these responses, and (c) sensitivity of these endpoints to pharmacological agents. Here, the possibility that nerve-injury induced tactile allodynia and thermal hyperalgesia may be mediated via different afferent fiber input was tested by evaluating these responses in sham-operated or nerve-injured (L5/L6) rats before or after a single systemic injection of resiniferatoxin (RTX), an ultrapotent analogue of the C-fiber specific neurotoxin, capsaicin. Tactile allodynia, and three measures of thermal nociception, tail-flick, paw-flick and hot-plate responses, were determined before and at various intervals for at least 40 days after RTX injection. Nerveinjured, but not sham-operated, rats showed a long-lasting tactile allodynia and thermal hyperalgesia (paw-flick) within 2–3 days after surgery; responses to other noxious thermal stimuli (i.e., tail-flick and hot-plate tests) did not distinguish the two groups at the stimulus intensities employed. RTX treatment resulted in a significant and long-lasting (i.e. essentially irreversible) decrease in sensitivity to thermal noxious stimuli in both sham-operated and nerve-injured rats; thermal hyperalgesia was abolished and antinociception produced by RTX. In contrast, RTX treatment did not affect the tactile allodynia seen in the same nerve-injured rats. These data support the concept that thermal hyperalgesia seen after nerve ligation, as well as noxious thermal stimuli, are likely to be mediated by capsaicin-sensitive C-fiber afferents. In contrast, nerve-injury related tactile allodynia is insensitive to RTX treatment which clearly desensitizes C-fibers and, therefore such responses are not likely to be mediated through C-fiber afferents. The hypothesis that tactile allodynia may be due to inputs from large (i.e. Ab) afferents offers a mechanistic basis for the observed insensitivity of this endpoint to intrathecal morphine in this nerve-injury model. Further, these data suggest that clinical treatment of neuropathic pains with C-fiber specific agents such as capsaicin are unlikely to offer significant therapeutic benefit against mechanical allodynia. 1999 International Association for the Study of Pain. Published by Elsevier Science B.V. Keywords: Neuropathic pain; Tactile allodynia; Thermal hyperalgesia; Resiniferatoxin; Afferent fibers
1. Introduction Peripheral nerve injury is associated with significant neuroplastic changes in the spinal cord which may underlie the development of tactile allodynia and thermal hyperalgesia. Such neuroplasticity includes changes in the expression of neurotransmitters, upregulation of mRNA for substances such as dynorphin, NPY and CCK in the DRG and spinal cord, central sensitization associated with increased NMDA receptor activity and modifications in synaptic terminations * Corresponding author. Tel.: +1-520-626-7421; fax: +1-520-626-4182; e-mail: [email protected]
of primary afferent neurons (see Ossipov et al., 1997 for review). Non-noxious tactile stimulation is transmitted chiefly through low-threshold, large diameter, myelinated Ab fibers, while noxious thermal stimuli is transmitted to the spinal cord through high-threshold, thin unmyelinated primary afferent fibers (e.g., Yeomans and Proudfit, 1996a). After peripheral nerve axotomy (Woolf et al., 1995) or L5/L6 spinal nerve ligation (Lekan et al., 1996), the large diameter, low threshold fibers which normally terminate in lamina III of the dorsal horn, have been shown to sprout into the superficial laminae and may form novel physiopathic synapses with transmission neurons. These second order neurons,
0304-3959/99/$ - see front matter 1999 International Association for the Study of Pain. Published by Elsevier Science B.V. PII: S03 04-3959(98)001 87-0
128
M.H. Ossipov et al. / Pain 79 (1999) 127–133
which normally code for nociceptive input, now appear to receive non-noxious input from the low threshold fibers and in this way, innocuous tactile input might be interpreted as nociceptive. This is one mechanism believed to be responsible for the development of tactile allodynia (Woolf et al., 1995; Lekan et al., 1996). A corresponding sprouting of Cfibers has not been reported. The vanilloids, capsaicin and its ultrapotent analog resiniferatoxin (RTX), have been used as selective tools to study the action of nociceptive sensory neurons. Binding studies with [3H]RTX and the identification of a competitive antagonist (capsazepine) provide evidence for a specific receptor (vanilloid) for capsaicin, although endogenous ligands have not yet been identified (Szallasi and Blumberg, 1996). Recently, expression cloning strategies have led to the identification and expression of a cation channel, designated as the vanilloid receptor or VR1 (Caterina et al., 1997). Co-operativity shown in binding studies with physiologic tissues and with the cloned receptor lead to speculation that more than one subtype of this receptor may exist. The vanilloid receptor(s) are believed to be expressed almost exclusively by afferent C-fibers involved with nociceptive transmission and neurogenic inflammation (see Szallasi and Blumberg, 1996 for review). These substances and their analogs produce an initial depolarization of primary sensory neurons which gives rise to the pungency and irritation associated with capsaicin-containing plants (e.g.; hot peppers), followed by a long-lasting desensitization of the nerve which may be due to Ca2 + accumulation in the neurons which impairs function and ultimately may produce neuronal death (Wood, 1993; Szallasi and Blumberg, 1996). RTX differs from capsaicin in that its initial nociceptive action is about equipotent to capsaicin but its potency to produce desensitization is several thousand-fold greater than capsaicin (Szallasi et al., 1989). Further, desensitization caused by RTX is of much longer duration than that of capsaicin. Szallasi et al. (1989) reported that RTX was more efficacious than capsaicin, producing a greater degree of desensitization that was more rapid in onset and significantly more persistent against chemically-induced nociception. The absence of sensitivity to chemical challenge after initial capsaicin exposure is attributed to a long-lasting desensitization of small diameter, unmyelinated nociceptive afferent (C-fibers) fibers (Szallasi et al., 1989). Likewise, RTX also produced a long-acting reduction in C-fiber mediated nociception in a model of visceral pain (Craft et al., 1995). In a comprehensive study employing electrophysiologic and behavioral techniques, it was determined that RTX produced a long-lasting (.4 weeks) thermal hypoalgesia but did not alter nociceptive mechanical thresholds in normal rats (Xu et al., 1997). Wind-up induced by conditioning stimuli of C-fibers was also attenuated by RTX (Xu et al., 1997), and reflected RTX-induced desensitization of capsaicin sensitive C-fibers. These studies serve to support the use
of RTX as a means of selectively desensitizing C-fiber primary afferents. The long duration and maximal degree of desensitization of C-fibers to further noxious stimuli make RTX an agent of choice for use in the study of capsaicinsensitive (C-fiber) mediated responses. For these reasons, RTX was selected as a means to optimally and selectively desensitize primary afferent C-fibers in order to explore the relative contributions of these nociceptors to several indices of neuropathic pain in animals with a peripheral nerve injury.
2. Methods Male Sprague–Dawley rats (200–350 g) were used in all experiments. This experimental protocol was approved by the Institutional Animal Care and Use Committee of the University of Arizona. Rats were anesthetized with 0.5% halothane in O2. The L5/L6 nerve ligation described by Kim and Chung (1992) was performed by exposing and isolating the L5 and L6 branches of the sciatic nerve from the surrounding fascia. A ‘finger tight’ ligature was made with 4–0 silk suture around each branch distal to the dorsal root ganglion and proximal to the confluence into the common sciatic nerve. Sham control rats were prepared the same way except that the spinal nerves were not ligated. 2.1. Tactile allodynia Tactile allodynia was determined by measuring the paw withdrawal threshold in response to probing with von Frey filaments (Chaplan et al., 1994). Calibrated filaments were applied to the plantar surface of hindpaws of rats kept in suspended cages with wire-mesh floors. The paw withdrawal threshold was determined by applying Dixon’s nonparametric test (Dixon, 1968). 2.2. Thermal hyperalgesia The method of Hargreaves et al. (1988) was employed to assess thermal hyperalgesia. Rats were allowed to acclimate within Plexiglas enclosures on a clear glass plate maintained at 30°C. A radiant heat source (i.e. high intensity projector lamp) which was activated with a timer and focused onto the plantar surface of the affected paw of nerve-injured or sham-operated rats. Paw withdrawal exposed a photocell that halted both lamp and timer. A maximal cut-off of 40 s was used to prevent tissue damage. 2.3. Tail-flick test Nociceptive testing was performed in both sham-operated and L5/L6 ligated rats by the warm water tail-flick test. This test was performed by determining the latency to withdrawal of the tail from a water bath maintained at
M.H. Ossipov et al. / Pain 79 (1999) 127–133
129
52°C. A cut-off latency of 10 s was employed to prevent tissue injury. 2.4. Hot-plate test The hot-plate test was performed by placing the rats in a Plexiglas enclosure on a thermostatically controlled metal plate maintained at 52°C. The latency to licking of a hindpaw was taken as the nociceptive endpoint. A cut-off time of 40 s was employed to prevent tissue damage. 2.5. Data analysis In all tests, baseline data were obtained for the ligated and sham-operated groups before RTX injection (0.3 mg/kg s.c. or 0.1 mg/kg i.p.). Rats were tested at 5 day intervals for 40 days. Within each of the treatment groups, post-injection means were compared with the baseline values by analysis of variance (ANOVA), followed by post-hoc analysis of least significance difference for multiple comparisons. Comparisons between two means were performed by Student’s t-test. A probability level of 0.05 indicated significance.
3. Results The sham-operated rats did not show a paw withdrawal response to probing with von Frey filaments up to the maximal cut-off of 15 g either before or throughout the entire observation period after receiving RTX (0.3 mg/kg, s.c.) (Fig. 1). Similar data were obtained following i.p. RTX (0.1 mg/kg, data not shown). The baseline paw withdrawal threshold of L5/L6 ligated rats was 1.0 ± 0.3 g, which was significantly less than that of sham-operated rats indicating the presence of tactile allodynia. The withdrawal thresholds
Fig. 1. Paw withdrawal thresholds were observed in sham-operated (X) and L5/L6 ligated (B) rats over 40 days. Tactile allodynia of ligated rats (indicated by low paw withdrawal thresholds) was not altered by a single injection of RTX (0.3 mg/kg, s.c., given immediately after the baseline reading) over the observation period; i.p. RTX (0.1 mg/kg) produced similar results (data not shown) (n = 6 rats/group).
Fig. 2. Paw withdrawal latencies to radiant heat focused onto the plantar surface of the hindpaw were observed in sham-operated (X) and L5/L6 ligated (B) rats over 40 days. The ligated rats initially demonstrated significantly lower (P ≤ 0.05, Student’s t-test) paw withdrawal latencies than the sham-operated rats, indicating thermal hyperalgesia. Thermal thresholds were elevated by a single injection of RTX (0.3 mg/kg, s.c.) in both sham-operated and nerve-ligated rats. The elevated thresholds were significantly higher (*P ≤ 0.05, ANOVA followed by least significant difference) throughout the entire 40-day observation period; i.p. RTX (0.1 mg/ kg) produced similar results (data not shown) (n = 6 rats/group).
remained within the range of 1.1 ± 0.3 to 2.0 ± 0.7 during the 40 day observation period after s.c. RTX (Fig. 1), and did not differ significantly from the pre-RTX baseline mean, indicating consistent tactile allodynia. The i.p. injection of RTX produced similar results (data not shown). The mean baseline paw withdrawal latency to radiant noxious thermal stimuli to the hindpaw of sham-operated rats was 23 ± 2.1 s. The paw withdrawal thresholds were significantly increased to between 29 ± 4.6 and 40 ± 0 s after s.c. RTX injection (Fig. 2) and remained significantly elevated throughout the entire 40 day observation period. Similarly, the i.p. injection of RTX produced significant increases in paw withdrawal latencies in sham-operated rats (data not shown). The mean baseline paw withdrawal latency of the L5/L6 nerve ligated rats was 18 ± 0.9 s, which was significantly less than that of the sham-operated group, indicating the presence of thermal hyperalgesia (Fig. 2). The s.c. injection of RTX produced significant elevations in paw withdrawal latencies to range between 28 ± 5.3 and 40 ± 0 s (Fig. 2). Likewise, i.p. RTX also significantly elevated paw withdrawal latencies in nerve-ligated rats (data not shown). The mean hot-plate latencies of the sham operated and of the L5/L6 ligated groups, 11.2 ± 1.1 and 12.7 ± 0.88 s, respectively, were not significantly different from each other. After the s.c. injection of 0.3 mg/kg of RTX, the hot-plate latencies were significantly elevated, ranging from 30.5 ± 6.6 to 32.5 ± 5.2 s up to 40 ± 0 s for the sham-operated and nerve ligated groups, respectively (Fig. 3). As noted above, i.p. RTX also produced similar results in the hot-plate test (data not shown). The mean tail-flick latencies of the sham operated and of
130
M.H. Ossipov et al. / Pain 79 (1999) 127–133
Fig. 3. Hot-plate latencies to a 52°C hot-plate were observed in shamoperated (X) and L5/L6 ligated (B) rats over 40 days. There was no significant difference between baseline responses in sham-operated or nerveinjured groups. Response thresholds were elevated by a single injection of RTX (0.3 mg/kg, s.c.) in both sham-operated and nerve-ligated rats. The elevated thresholds were significantly higher (*P ≤ 0.05, ANOVA followed by least significant difference) throughout the entire 40 day observation period; i.p. RTX (0.1 mg/kg) produced similar results (data not shown) (n = 6 rats/group).
the L5/L6 ligated groups were 3.4 ± 0.21 and 3.1 ± 0.19 s, which did not differ significantly. After s.c. RTX, the tailflick latencies were all significantly elevated throughout the observation period and ranged from 7.3 ± 1.13 to 5.4 ± 0.74 s up to 10 ± 0 s for the sham-operated and nerve ligated groups, respectively (Fig. 4). Again, the i.p. injection of 0.1 mg/kg of RTX also produced similar results in the tail-flick test (data not shown).
(Zachariou et al., 1997), and replicate the conditions consistently used in this laboratory. The desensitization of C-fibers to noxious thermal stimuli by systemic RTX was confirmed in the present study by the repeated observations of a significant loss of sensitivity to thermal nociceptive stimuli in all thermal tests in both sham-operated and nerve ligated groups. The duration of desensitization to thermal nociception was maintained throughout the entire 40 day observation period and was seen following a single, systemic (i.e., i.p. or s.c.) injection of RTX. This time-course correlates well with the long time course of desensitization induced by RTX in a number of nociceptive endpoints (Craft et al., 1995; Xu et al., 1997). The VR1 receptor, through which capsaicin and RTX may act, has recently been cloned and characterized (Caterina et al., 1997). The similar pharmacology of the VR1 receptor with regard to capsaicin and noxious heat have led to the conclusion that the VR1 receptor acts as a physiologic transducer for heat-activated nociceptors (Caterina et al., 1997). In keeping with these findings, treatment with RTX was able to significantly increase response latencies to heat, supportive of the idea of desensitization of fibers expressing vanilloid receptors (i.e., C-fibers). Systemic injection of RTX has also been shown to reverse cold hyperalgesia, but not tactile allodynia, in rats with injury to central, rather than peripheral nerves, again suggesting the involvement of C-fibers in cold, but not mechanical, hypersensitivity (Hao et al., 1996). A significant finding of this study is that in spite of the significant loss of responsiveness to noxious thermal stimuli, the tactile allodynia present in the L5/L6 ligated rats was not changed throughout the entire course of the experiment. Since RTX selectively desensitizes those neurons
4. Discussion The results of the present investigation demonstrated a thermal hyperalgesia of the hindpaw to radiant noxious heat, but not in the hot-plate or tail-flick tests, in L5/L6 nerve ligated rats. The 52°C hot-plate and the 52°C hotwater tail-flick tests both provide fairly rapid responses to the nociceptive stimuli, and it is generally difficult to differentiate a hyperalgesic from a normal nociceptive response in these circumstances (see Wegert et al., 1997 and Bian et al., 1998 for discussions). The hot-plate test also provides the additional difficulty in that the rat is able to shift weight from the hindlimb ipsilateral to the nerve injury to the contralateral one, further confounding results if used alone. In contrast, the radiant heat paw-flick test presents the advantage of permitting a control over the heat intensity and rats of rise, allowing for a fine-tuning of the normal paw withdrawal latency. The paw-flick latencies to radiant heat employed in the present study correspond to the slow rate of rise, preferentially activating nociceptive C-fibers (Yeomans and Proudfit, 1996a; Yeomans and Proudfit, 1996b) or capsaicin-sensitive fibers
Fig. 4. Tail-flick latencies to a 52°C water bath were observed in shamoperated (X) and L5/L6 ligated (B) rats over 40 days. There was no significant difference between baseline responses in sham-operated or nerveinjured groups. Response thresholds were elevated by a single injection of RTX (0.3 mg/kg, s.c.) in both sham-operated and nerve-ligated rats. The elevated thresholds were significantly higher (*P ≤ 0.05, ANOVA followed by least significant difference) throughout the entire 40 day observation period; i.p. RTX (0.1 mg/kg) produced similar results (data not shown) (n = 6 rats/group).
M.H. Ossipov et al. / Pain 79 (1999) 127–133
expressing the vanilloid receptor, and since this receptor has recently been identified as a physiologic thermal nociceptive transducer (Caterina et al., 1997), then tactile stimuli may well be expected to be mediated by other (i.e. nonvanilloid receptor expressing) fibers. Consequently, the results of the present study strongly support the hypothesis that tactile allodynia and thermal hyperalgesia are mediated through different neuronal pathways. There are several possible mechanistic interpretations that may be invoked to support this reasoning. It is strongly suggested, based on these observations, that tactile allodynia is likely to be mediated through the activation of capsaicininsensitive large-diameter myelinated, or Ab afferent fibers, since this population of afferent nerves is unlikely to conduct nociceptive thermal stimuli (Light and Perl, 1979; Sugiura et al., 1986; Maxwell and Rethelyi, 1987) and is not sensitive to desensitization by capsaicin (or RTX) (Szolcsanyi et al., 1990). In the normal state, these fibers terminate in the deeper laminae (III–V) of the dorsal horn and code for nonnoxious stimuli, and they are not present in lamina II (Sugiura et al., 1986; Maxwell and Rethelyi, 1987). After peripheral nerve section or injury to the L5/L6 spinal nerves, Ab fibers form axonal sprouts arising from intact adjacent nerves and from the DRG of the injured nerves, and form abnormal synaptic connections with second order neurons of lamina II (Woolf et al., 1992; Woolf et al., 1995; Lekan et al., 1996; Doubell et al., 1997). This region normally receives an almost exclusively nociceptive input from unmyelinated Cfibers and is associated with nociceptive transmission (Light and Perl, 1979; Sugiura et al., 1986). Although these studies provide strong evidence for a physiopathic rearrangement of synaptic connections leading to neuropathic pain states, detailed time course analyses correlating sprouting with the development of behavioral signs of tactile allodynia, typically evident within 2–3 days after ligation (Chaplan et al., 1994; Bian et al., 1995), are not available. Mechanistic differences between thermal hyperalgesia and tactile allodynia can also be demonstrated pharmacologically. Spinally given morphine, for example, does not alter tactile responses to probing with von Frey filaments in rats with L5/L6 nerve ligation injury, even at doses considered to be supramaximal against other endpoints (i.e. 100 mg) (Bian et al., 1995; Lee et al., 1995). A complete lack of effect of morphine is not indicated, since the supraspinal or systemic injection of morphine does produce a dose-dependent antiallodynic effect. On the other hand, spinal morphine is effective against noxious thermal stimuli, though at somewhat reduced potency, and produces effective antihyperalgesic actions (Wegert et al., 1997). Likewise, earlier studies also reported on the efficacy of systemic (i.v.) morphine in rats with CCI, and implicated the possibility of a peripheral effect of morphine against signs of neuropathic pain (Lee et al., 1995; Kayser et al., 1995). An important difference between the latter studies and the present experiments is that thermal allodynia (determined with a warm, non-noxious water bath), and mechanical hyperalgesia
131
(determined by the Randall–Selitto method) was employed, whereas we measured thermal hyperalgesia (with noxious radiant heat) and tactile allodynia (von Frey filaments). Based on such findings, it is suggested that thermal hyperalgesia is most likely mediated through slow-conducting, unmyelinated C-fibers, sensitive to RTX. These fibers also have opioid receptors on their nerve terminals and this represents an important spinal site of action of opioids for presynaptic regulation of afferent noxious input (Yaksh et al., 1988; Lombard and Besson, 1989). Evidence indicates that it is unlikely that the Ab afferent neurons possess opioid receptors on the nerve terminals (Taddese et al., 1995). There is a differential distribution of expression of opioid receptors and of mRNA for the m-, d- and k- opioid receptors in DRG cell bodies and opioid receptor binding distributions are predominant in the superficial layers of the dorsal horn, corresponding with terminals of C-fibers rather than the terminal region of larger diameter fibers (e.g. Laminae III and IV) (Mansour et al., 1994; Mansour et al., 1995). Furthermore, activation of m-opioid receptors predictably inhibited Ca++ channels of small-diameter nociceptors and not of large-diameter cells indicating that m-receptor activation selectively inhibits the activity of C-fibers (Taddese et al., 1995). Therefore, morphine and other opioids may attenuate thermal nociception and thus thermal hyperalgesia through the actions on opioid receptors on both the primary nerve terminals and on the cell bodies of second order neurons of the dorsal horn. In contrast to thermal stimuli, it appears that tactile allodynia is mediated through the non-opioid receptor containing Ab fibers. Therefore, the only population of spinal opioid receptors that might be considered to be relevant to the attenuation of transmission of allodynia is that existing post-synaptic to the primary afferent neurons, which consists of 30–40% of the total spinal m- and d-opioid receptor population (Besse et al., 1990). Consequently, the ability of spinally injected opioids to block allodynia would be similar to the situation where the receptor reserve has been greatly decreased by a pharmacological manipulation (e.g. irreversible antagonist or development of tolerance). Such a manipulation would be expected to displace the agonist dose–effect curve to the right, and for an agonist of limited efficacy such as morphine, produce a decrease in the maximal response, a pattern typical of partial, irreversible blockade of a fraction of the relevant receptor population. Such a pattern is seen for spinal morphine (Lee et al., 1995; Bian et al., 1995). On the other hand, a high efficacy opioid m agonist such as DAMGO has been shown to elicit a significant antiallodynic action though with somewhat reduced potency, as would be expected (Nichols et al., 1995). A clinical correlate is that morphine generally provides inadequate pain relief in different neuropathic pain states (Arner and Meyerson, 1988; Rowbotham et al., 1991). However, the highly efficacious m-opioid fentanyl given i.t. provides complete pain relief against established post-amputation stump pain (Jacobson et al., 1990).
132
M.H. Ossipov et al. / Pain 79 (1999) 127–133
Finally, a differentiation in spinal and supraspinal neuronal circuitries in the processing of tactile allodynia and thermal hyperalgesia is strongly suggested. Pertovaara et al. (1996) found that the microinjection of lidocaine into the periaqueductal gray (PAG) or rostroventral medulla (RVM) blocked tactile allodynia, but not thermal nociceptive tail-flick reflexes, in rats with L5/L6 ligation injury. Transection of the spinal cord abolished tactile allodynia of the hindpaw, but facilitated the thermal tail-flick reflex, in rats with neurogenic inflammation (Mansikka and Pertovaara, 1997) or with L5/L6 spinal nerve ligation injury (Bian et al., 1998). Similar effects were seen after sacral spinal nerve ligation (Sung et al., 1997). It is unlikely that the loss of allodynia is an artifact of spinalization since spinal nocifensive responses to heat or pinch were present in the hindpaws and the tail. It was concluded that manifestation of tactile allodynia requires the activation of a spinal-supraspinal loop, possibly activating descending facilitatory and inhibitory systems arising from medullary sites, while the spinal thermal responses (still present after spinal transection) depend on localized circuitry, with thermal hyperalgesia being driven by central sensitization of localized spinal circuitry’s, although it may also be under some degree of descending control (Bian et al., 1998). In seeming contradiction to the results presented here, a recent clinical investigation reported that a high (5–10%) dose of capsaicin provided relief against neuropathic pain in the foot or chest wall arising from a variety of causes (Robbins et al., 1998). However, interpretation of the data is complicated in that the type of pain (i.e. hyperalgesic vs. allodynic vs. spontaneous) was not reported. Also, peripheral nerves neighboring an injured nerve have shown increased spontaneous and evoked activity, and this excitation was blocked by the application of lidocaine to the injured nerve (Sotgiu et al., 1996). Finally, the application of the local anesthetic lidocaine prior to capsaicin could confound the results since capsaicin acts initially by opening cation channels, whereas lidocaine exerts its local anesthetic activity by blocking Na + channels, thus opposing the action of capsaicin. The results presented here strongly suggest that different mechanisms may underlie the development of the different signs of neuropathic pain behavior. No single model is adequate to investigate the possible mechanisms that may be exploited in order to develop treatment modalities for this complex pain syndrome. It is clear that different signs of abnormal pain may be due to different and unique pathophysiological mechanisms. The present study, in concert with recent observations, suggests that treatment of hyperalgesia or other C-fiber related manifestations might be amenable to resolution with capsaicin or its analogs, such as RTX, but one must be cautious in extrapolating such conclusions to the entire array of neuropathic pain states especially to mechanical stimuli which appear to be resistant to manipulations resulting in changes in C-fiber activity.
References Arner, S. and Meyerson, B.A., Lack of analgesic effect on neuropathic and idiopathic forms of pain, Pain, 33 (1988) 11–23. Besse, D., Lombard, M.C., Zajac, J.M., Roques, B.P. and Besson, J.M., Pre- and postsynaptic location of mu, delta and kappa opioid receptors in the superficial layers of the dorsal horn of the rat spinal cord, Prog. Clin. Bio. Res., 328 (1990) 183–186. Bian, D., Nichols, M.L., Ossipov, M.H., Lai, J. and Porreca, F., Characterization of the antiallodynic efficacy of morphine in a model of neuropathic pain in rats, NeuroReport, 6 (1995) 1981–1984. Bian, D., Ossipov, M.H., Malan, T.P. and Porreca, F., Tactile allodynia, but not thermal hyperalgesia, of the hindlimbs is blocked by spinal transection in rats with nerve injury, Neurosci. Lett., 241 (1998) 79– 82. Caterina, M.J., Schumacher, M.A., Tominaga, M., Rosen, T.A., Levine, J.D. and Julius, D., The capsaicin receptor: a heat-activated ion channel in the pain pathway, Nature, 389 (1997) 816–824. Chaplan, S.R., Bach, F.W., Pogrel, J.W., Chung, J.M. and Yaksh, T.L., Quantitative assessment of tactile allodynia in the rat paw, J. Neurosci. Meth., 53 (1994) 55–63. Craft, R.M., Cohen, S.M. and Porreca, F., Long-lasting desensitization of bladder afferents following intravesical resiniferatoxin and capsaicin in the rat, Pain, 61 (1995) 317–323. Dixon, W.J., Efficient analysis of experimental observations, Ann. Rev. Pharmacol. Toxicol., 20 (1968) 441–462. Doubell, T.P., Mannion, R.J. and Woolf, C.J., Intact sciatic myelinated primary afferent terminals collaterally sprout in the adult rat dorsal horn following section of a neighbouring peripheral nerve, J. Comp. Neurol., 380 (1997) 95–104. Hao, J.X., Yu, W., Xu, X.J. and Wiesenfeld-Hallin, Z., Capsaicin-sensitive afferents mediate chronic cold, but not mechanical, allodynia-like behavior in spinally injured rats, Brain Res., 722 (1996) 177–180. Hargreaves, K., Dubner, R., Brown, F., Flores, C. and Joris, J., A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia, Pain, 32 (1988) 77–88. Jacobson, L., Chabal, C., Brody, M.C., Mariano, A.J. and Chaney, E.F., A comparison of the effects of intrathecal fentanyl and lidocaine on established postamputation stump pain, Pain, 40 (1990) 137–141. Kayser, V., Lee, S.H. and Guilbaud G., Evidence for a peripheral component in the enhanced antinocicpetive effect of a low dose of systemic morphine in rats with peripheral mononeuropathy, Neuroscience, 64 (1995) 537–545. Kim, S.H. and Chung, J.M., An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat, Pain, 50 (1992) 355–363. Lee, Y.-W., Chaplan, S.R. and Yaksh, T.L., Systemic and supraspinal but not spinal opiates suppress allodynia in a rat neuropathic pain model, Neurosci Lett., 199 (1995) 111–114. Lekan, H.A., Carlton, S.M. and Coggeshall, R.E., Sprouting of A beta fibers into lamina II of the rat dorsal horn in peripheral neuropathy, Neurosci. Lett., 208 (1996) 147–150. Light, A.R. and Perl, E.R., Spinal termination of functionally identified primary afferent neurons with slowly conducting myelinated fibers, J. Comp. Neurol., 186 (1979) 133–150. Lombard, M.C. and Besson, J.M., Attempts to gauge the relative importance of pre- and postsynaptic effects of morphine on the transmission of noxious messages in the dorsal horn of the rat spinal cord, Pain, 37 (1989) 335–345. Mansour, A., Fox, C.A., Thompson, R.C., Akil, H. and Watson, S.J., mOpioid receptor mRNA expression in the rat CNS: comparison to mreceptor binding, Brain Res., 643 (1994) 245–265. Mansikka, H. and Pertovaara, A., Supraspinal influence on hindlimb withdrawal thresholds and mustard oil-induced secondary allodynia in rats, Brain Res. Bull., 42 (1997) 359–365. Mansour, A., Fox, C.A., Akil, H. and Watson, S.J., Opioid-receptor mRNA
M.H. Ossipov et al. / Pain 79 (1999) 127–133 expression in the rat CNS: anatomical and functional implications, Trends Neurosci., 18 (1995) 22–29. Maxwell, D.J. and Rethelyi, M., Ultrastructure and synaptic connections of cutaneous afferent fibres in the spinal cord, Trends neurosci., 10 (1987) 117–123. Nichols, M.L., Bian, D., Ossipov, M.H. and Porreca, F., Evidence for regulation of opioid activity by CCK in a model of neuropathic pain, J. Pharmacol. Exp. Ther., 275 (1995) 1337–1345. Ossipov, M.H., Malan, T.P., Lai, J., Porreca, F., Opioid pharmacology of acute and chronic pain. In: A. Dickenson and J.-M. Beson (Eds.), Handbook of Experimental Pharmacology, Vol. 130, The Pharmacology of Spain, Springer-Verlag, Berlin, 1997b, pp. 305–333. Pertovaara, A., Wei, H. and Hamalainen, M.M., Lidocaine in the rostroventromedial medulla and the periaqueductal gray attenuates allodynia in neuropathic rats, Neurosci. Lett., 218 (1996) 127–130. Rowbotham, M.C., Reisner-Keller, L.A. and Fields, H.L., Both intravenous lidocaine and morphine reduce the pain of post-therapeutic neuralgia, Neurology, 41 (1991) 1024–1028. Robbins, W.R., Staats, P.S., Levine, J., Fields, H.L., Allen, R.W., Campbell, J.N. and Pappagallo, M., Treatment of intractable pain with topical large-dose capsaicin: preliminary report, Anesth. Analg., 86 (1998) 579–583. Sotgiu, M.L., Biella, G. and Lacerenza, M., Injured nerve block alters adjacent nerves spinal interaction in neuropathic rats, NeuroReport, 7 (1996) 1385–1388. Sugiura, Y., Lee, C.L. and Perl, E.R., Central projections of identified, unmyelinated (C) afferent fibers innervating mammalian skin, Science, 234 (1986) 358–361. Sung, B., Na, H.S., Kim, Y.I., Yoon, Y.W., Han, H.C., Nam, S.H. and Hong, S.K., The behavioral sign of mechanical allodynia from peripheral nerve injury is not a spinal reflex, Abstr. Soc. Neurosci., 23 (1997) 1804. Szallasi, A., Joo, F. and Blumberg, P.M., Duration of desensitization and ultrastructural changes in dorsal root ganglia in rats treated with resiniferatoxin, an ultrapotent capsaicin analog, Brain Res., 503 (1989) 68– 72. Szallasi, A. and Blumberg, P.M., Vanilloid receptors: new insights enhance potential as a therapeutic target, Pain, 68 (1996) 195–208.
133
Szolcsanyi, J., Szallasi, A., Szallasi, Z., Joo, F. and Blumberg, P.M., Resiniferatoxin: an ultrapotent selective modulator of capsaicin-sensitive primary afferent neurons, J. Pharmacol. Exp. Ther., 255 (1990) 923– 928. Taddese, A., Nah, S.-Y. and McCleskey, E.W., Selective opioid inhibition of small nociceptive neurons, Science, 270 (1995) 1366–1369. Wegert, S., Ossipov, M.H., Nichols, M.L., Bian, D., Vanderah, T., Malan, T.P. and Porreca, F., Differential activities of intrathecal MK-801 or morphine to alter responses to thermal and mechanical stimuli in normal or nerve-injured rats, Pain, 71 (1997) 57–64. Wood, J.N. (Ed.), Capsaicin in the Study of Pain, Academic Press, New York, 1993. Woolf, C.J., Shortland, P. and Coggeshall, R.E., Peripheral nerve injury triggers central sprouting of myelinated afferents, Nature, 355 (1992) 75–78. Woolf, C.J., Shortland, P., Reynolds, M., Ridings, J., Doubell, T. and Coggeshall, R.E., Reorganization of central terminals of myelinated primary afferents in the rat dorsal horn following peripheral axotomy, J. Comp. Neurol., 360 (1995) 121–134. Xu, X.J., Farkas-Szallasi, T., Lundberg, J.M., Hokfelt, T., WiesenfeldHallin, Z. and Szallasi, A., Effects of the capsaicin analogue resiniferatoxin on spinal nociceptive mechanisms in the rat: behavioral, electrophysiological and in situ hybridization studies, Brain Res., 752 (1997) 52–60. Yaksh, T.L., Al-Rodhan, N.R. and Jensen, T.S., Sites of action of opiates in production of analgesia, Prog. Brain Res., 77 (1988) 371–394. Yeomans, D.C. and Proudfit, H.K., Nociceptive responses to high and low rates of noxious cutaneous heating are mediated by different nociceptors in the rat: behavioral evidence, Pain, 68 (1996a) 133–140. Yeomans, D.C. and Proudfit, H.K., Nociceptive responses to high and low rates of noxious cutaneous heating are mediated by different nociceptors in the rat: electrophysiological evidence, Pain, 68 (1996b) 141– 150. Zachariou, V., Goldstein, B.D. and Yeomans, D.C., Low but not high rate noxious radiant skin heating evokes a capsaicin-sensitive increase in spinal cord dorsal horn release of substance P, Brain Res., 752 (1997) 143–150.
Lack of involvement of capsaicin-sensitive primary afferents in nerve-ligation injury induced tactile allodynia in rats Michael H. Ossipov, Di Bian, T. Philip Malan Jr., Josephine Lai, Frank Porreca* Department of Pharmacology, University of Arizona Health Sciences Center, Tucson, AZ 85724, USA Received 6 April 1998; received in revised form 14 July 1998; accepted 13 August 1998
Abstract Tactile allodynia and thermal hyperalgesia, two robust signs of neuropathic pain associated with experimental nerve injury, have been hypothesized to be mechanistically distinguished based on (a) fiber types which may be involved in the afferent input, (b) participation of spinal and supraspinal circuitry in these responses, and (c) sensitivity of these endpoints to pharmacological agents. Here, the possibility that nerve-injury induced tactile allodynia and thermal hyperalgesia may be mediated via different afferent fiber input was tested by evaluating these responses in sham-operated or nerve-injured (L5/L6) rats before or after a single systemic injection of resiniferatoxin (RTX), an ultrapotent analogue of the C-fiber specific neurotoxin, capsaicin. Tactile allodynia, and three measures of thermal nociception, tail-flick, paw-flick and hot-plate responses, were determined before and at various intervals for at least 40 days after RTX injection. Nerveinjured, but not sham-operated, rats showed a long-lasting tactile allodynia and thermal hyperalgesia (paw-flick) within 2–3 days after surgery; responses to other noxious thermal stimuli (i.e., tail-flick and hot-plate tests) did not distinguish the two groups at the stimulus intensities employed. RTX treatment resulted in a significant and long-lasting (i.e. essentially irreversible) decrease in sensitivity to thermal noxious stimuli in both sham-operated and nerve-injured rats; thermal hyperalgesia was abolished and antinociception produced by RTX. In contrast, RTX treatment did not affect the tactile allodynia seen in the same nerve-injured rats. These data support the concept that thermal hyperalgesia seen after nerve ligation, as well as noxious thermal stimuli, are likely to be mediated by capsaicin-sensitive C-fiber afferents. In contrast, nerve-injury related tactile allodynia is insensitive to RTX treatment which clearly desensitizes C-fibers and, therefore such responses are not likely to be mediated through C-fiber afferents. The hypothesis that tactile allodynia may be due to inputs from large (i.e. Ab) afferents offers a mechanistic basis for the observed insensitivity of this endpoint to intrathecal morphine in this nerve-injury model. Further, these data suggest that clinical treatment of neuropathic pains with C-fiber specific agents such as capsaicin are unlikely to offer significant therapeutic benefit against mechanical allodynia. 1999 International Association for the Study of Pain. Published by Elsevier Science B.V. Keywords: Neuropathic pain; Tactile allodynia; Thermal hyperalgesia; Resiniferatoxin; Afferent fibers
1. Introduction Peripheral nerve injury is associated with significant neuroplastic changes in the spinal cord which may underlie the development of tactile allodynia and thermal hyperalgesia. Such neuroplasticity includes changes in the expression of neurotransmitters, upregulation of mRNA for substances such as dynorphin, NPY and CCK in the DRG and spinal cord, central sensitization associated with increased NMDA receptor activity and modifications in synaptic terminations * Corresponding author. Tel.: +1-520-626-7421; fax: +1-520-626-4182; e-mail: [email protected]
of primary afferent neurons (see Ossipov et al., 1997 for review). Non-noxious tactile stimulation is transmitted chiefly through low-threshold, large diameter, myelinated Ab fibers, while noxious thermal stimuli is transmitted to the spinal cord through high-threshold, thin unmyelinated primary afferent fibers (e.g., Yeomans and Proudfit, 1996a). After peripheral nerve axotomy (Woolf et al., 1995) or L5/L6 spinal nerve ligation (Lekan et al., 1996), the large diameter, low threshold fibers which normally terminate in lamina III of the dorsal horn, have been shown to sprout into the superficial laminae and may form novel physiopathic synapses with transmission neurons. These second order neurons,
0304-3959/99/$ - see front matter 1999 International Association for the Study of Pain. Published by Elsevier Science B.V. PII: S03 04-3959(98)001 87-0
128
M.H. Ossipov et al. / Pain 79 (1999) 127–133
which normally code for nociceptive input, now appear to receive non-noxious input from the low threshold fibers and in this way, innocuous tactile input might be interpreted as nociceptive. This is one mechanism believed to be responsible for the development of tactile allodynia (Woolf et al., 1995; Lekan et al., 1996). A corresponding sprouting of Cfibers has not been reported. The vanilloids, capsaicin and its ultrapotent analog resiniferatoxin (RTX), have been used as selective tools to study the action of nociceptive sensory neurons. Binding studies with [3H]RTX and the identification of a competitive antagonist (capsazepine) provide evidence for a specific receptor (vanilloid) for capsaicin, although endogenous ligands have not yet been identified (Szallasi and Blumberg, 1996). Recently, expression cloning strategies have led to the identification and expression of a cation channel, designated as the vanilloid receptor or VR1 (Caterina et al., 1997). Co-operativity shown in binding studies with physiologic tissues and with the cloned receptor lead to speculation that more than one subtype of this receptor may exist. The vanilloid receptor(s) are believed to be expressed almost exclusively by afferent C-fibers involved with nociceptive transmission and neurogenic inflammation (see Szallasi and Blumberg, 1996 for review). These substances and their analogs produce an initial depolarization of primary sensory neurons which gives rise to the pungency and irritation associated with capsaicin-containing plants (e.g.; hot peppers), followed by a long-lasting desensitization of the nerve which may be due to Ca2 + accumulation in the neurons which impairs function and ultimately may produce neuronal death (Wood, 1993; Szallasi and Blumberg, 1996). RTX differs from capsaicin in that its initial nociceptive action is about equipotent to capsaicin but its potency to produce desensitization is several thousand-fold greater than capsaicin (Szallasi et al., 1989). Further, desensitization caused by RTX is of much longer duration than that of capsaicin. Szallasi et al. (1989) reported that RTX was more efficacious than capsaicin, producing a greater degree of desensitization that was more rapid in onset and significantly more persistent against chemically-induced nociception. The absence of sensitivity to chemical challenge after initial capsaicin exposure is attributed to a long-lasting desensitization of small diameter, unmyelinated nociceptive afferent (C-fibers) fibers (Szallasi et al., 1989). Likewise, RTX also produced a long-acting reduction in C-fiber mediated nociception in a model of visceral pain (Craft et al., 1995). In a comprehensive study employing electrophysiologic and behavioral techniques, it was determined that RTX produced a long-lasting (.4 weeks) thermal hypoalgesia but did not alter nociceptive mechanical thresholds in normal rats (Xu et al., 1997). Wind-up induced by conditioning stimuli of C-fibers was also attenuated by RTX (Xu et al., 1997), and reflected RTX-induced desensitization of capsaicin sensitive C-fibers. These studies serve to support the use
of RTX as a means of selectively desensitizing C-fiber primary afferents. The long duration and maximal degree of desensitization of C-fibers to further noxious stimuli make RTX an agent of choice for use in the study of capsaicinsensitive (C-fiber) mediated responses. For these reasons, RTX was selected as a means to optimally and selectively desensitize primary afferent C-fibers in order to explore the relative contributions of these nociceptors to several indices of neuropathic pain in animals with a peripheral nerve injury.
2. Methods Male Sprague–Dawley rats (200–350 g) were used in all experiments. This experimental protocol was approved by the Institutional Animal Care and Use Committee of the University of Arizona. Rats were anesthetized with 0.5% halothane in O2. The L5/L6 nerve ligation described by Kim and Chung (1992) was performed by exposing and isolating the L5 and L6 branches of the sciatic nerve from the surrounding fascia. A ‘finger tight’ ligature was made with 4–0 silk suture around each branch distal to the dorsal root ganglion and proximal to the confluence into the common sciatic nerve. Sham control rats were prepared the same way except that the spinal nerves were not ligated. 2.1. Tactile allodynia Tactile allodynia was determined by measuring the paw withdrawal threshold in response to probing with von Frey filaments (Chaplan et al., 1994). Calibrated filaments were applied to the plantar surface of hindpaws of rats kept in suspended cages with wire-mesh floors. The paw withdrawal threshold was determined by applying Dixon’s nonparametric test (Dixon, 1968). 2.2. Thermal hyperalgesia The method of Hargreaves et al. (1988) was employed to assess thermal hyperalgesia. Rats were allowed to acclimate within Plexiglas enclosures on a clear glass plate maintained at 30°C. A radiant heat source (i.e. high intensity projector lamp) which was activated with a timer and focused onto the plantar surface of the affected paw of nerve-injured or sham-operated rats. Paw withdrawal exposed a photocell that halted both lamp and timer. A maximal cut-off of 40 s was used to prevent tissue damage. 2.3. Tail-flick test Nociceptive testing was performed in both sham-operated and L5/L6 ligated rats by the warm water tail-flick test. This test was performed by determining the latency to withdrawal of the tail from a water bath maintained at
M.H. Ossipov et al. / Pain 79 (1999) 127–133
129
52°C. A cut-off latency of 10 s was employed to prevent tissue injury. 2.4. Hot-plate test The hot-plate test was performed by placing the rats in a Plexiglas enclosure on a thermostatically controlled metal plate maintained at 52°C. The latency to licking of a hindpaw was taken as the nociceptive endpoint. A cut-off time of 40 s was employed to prevent tissue damage. 2.5. Data analysis In all tests, baseline data were obtained for the ligated and sham-operated groups before RTX injection (0.3 mg/kg s.c. or 0.1 mg/kg i.p.). Rats were tested at 5 day intervals for 40 days. Within each of the treatment groups, post-injection means were compared with the baseline values by analysis of variance (ANOVA), followed by post-hoc analysis of least significance difference for multiple comparisons. Comparisons between two means were performed by Student’s t-test. A probability level of 0.05 indicated significance.
3. Results The sham-operated rats did not show a paw withdrawal response to probing with von Frey filaments up to the maximal cut-off of 15 g either before or throughout the entire observation period after receiving RTX (0.3 mg/kg, s.c.) (Fig. 1). Similar data were obtained following i.p. RTX (0.1 mg/kg, data not shown). The baseline paw withdrawal threshold of L5/L6 ligated rats was 1.0 ± 0.3 g, which was significantly less than that of sham-operated rats indicating the presence of tactile allodynia. The withdrawal thresholds
Fig. 1. Paw withdrawal thresholds were observed in sham-operated (X) and L5/L6 ligated (B) rats over 40 days. Tactile allodynia of ligated rats (indicated by low paw withdrawal thresholds) was not altered by a single injection of RTX (0.3 mg/kg, s.c., given immediately after the baseline reading) over the observation period; i.p. RTX (0.1 mg/kg) produced similar results (data not shown) (n = 6 rats/group).
Fig. 2. Paw withdrawal latencies to radiant heat focused onto the plantar surface of the hindpaw were observed in sham-operated (X) and L5/L6 ligated (B) rats over 40 days. The ligated rats initially demonstrated significantly lower (P ≤ 0.05, Student’s t-test) paw withdrawal latencies than the sham-operated rats, indicating thermal hyperalgesia. Thermal thresholds were elevated by a single injection of RTX (0.3 mg/kg, s.c.) in both sham-operated and nerve-ligated rats. The elevated thresholds were significantly higher (*P ≤ 0.05, ANOVA followed by least significant difference) throughout the entire 40-day observation period; i.p. RTX (0.1 mg/ kg) produced similar results (data not shown) (n = 6 rats/group).
remained within the range of 1.1 ± 0.3 to 2.0 ± 0.7 during the 40 day observation period after s.c. RTX (Fig. 1), and did not differ significantly from the pre-RTX baseline mean, indicating consistent tactile allodynia. The i.p. injection of RTX produced similar results (data not shown). The mean baseline paw withdrawal latency to radiant noxious thermal stimuli to the hindpaw of sham-operated rats was 23 ± 2.1 s. The paw withdrawal thresholds were significantly increased to between 29 ± 4.6 and 40 ± 0 s after s.c. RTX injection (Fig. 2) and remained significantly elevated throughout the entire 40 day observation period. Similarly, the i.p. injection of RTX produced significant increases in paw withdrawal latencies in sham-operated rats (data not shown). The mean baseline paw withdrawal latency of the L5/L6 nerve ligated rats was 18 ± 0.9 s, which was significantly less than that of the sham-operated group, indicating the presence of thermal hyperalgesia (Fig. 2). The s.c. injection of RTX produced significant elevations in paw withdrawal latencies to range between 28 ± 5.3 and 40 ± 0 s (Fig. 2). Likewise, i.p. RTX also significantly elevated paw withdrawal latencies in nerve-ligated rats (data not shown). The mean hot-plate latencies of the sham operated and of the L5/L6 ligated groups, 11.2 ± 1.1 and 12.7 ± 0.88 s, respectively, were not significantly different from each other. After the s.c. injection of 0.3 mg/kg of RTX, the hot-plate latencies were significantly elevated, ranging from 30.5 ± 6.6 to 32.5 ± 5.2 s up to 40 ± 0 s for the sham-operated and nerve ligated groups, respectively (Fig. 3). As noted above, i.p. RTX also produced similar results in the hot-plate test (data not shown). The mean tail-flick latencies of the sham operated and of
130
M.H. Ossipov et al. / Pain 79 (1999) 127–133
Fig. 3. Hot-plate latencies to a 52°C hot-plate were observed in shamoperated (X) and L5/L6 ligated (B) rats over 40 days. There was no significant difference between baseline responses in sham-operated or nerveinjured groups. Response thresholds were elevated by a single injection of RTX (0.3 mg/kg, s.c.) in both sham-operated and nerve-ligated rats. The elevated thresholds were significantly higher (*P ≤ 0.05, ANOVA followed by least significant difference) throughout the entire 40 day observation period; i.p. RTX (0.1 mg/kg) produced similar results (data not shown) (n = 6 rats/group).
the L5/L6 ligated groups were 3.4 ± 0.21 and 3.1 ± 0.19 s, which did not differ significantly. After s.c. RTX, the tailflick latencies were all significantly elevated throughout the observation period and ranged from 7.3 ± 1.13 to 5.4 ± 0.74 s up to 10 ± 0 s for the sham-operated and nerve ligated groups, respectively (Fig. 4). Again, the i.p. injection of 0.1 mg/kg of RTX also produced similar results in the tail-flick test (data not shown).
(Zachariou et al., 1997), and replicate the conditions consistently used in this laboratory. The desensitization of C-fibers to noxious thermal stimuli by systemic RTX was confirmed in the present study by the repeated observations of a significant loss of sensitivity to thermal nociceptive stimuli in all thermal tests in both sham-operated and nerve ligated groups. The duration of desensitization to thermal nociception was maintained throughout the entire 40 day observation period and was seen following a single, systemic (i.e., i.p. or s.c.) injection of RTX. This time-course correlates well with the long time course of desensitization induced by RTX in a number of nociceptive endpoints (Craft et al., 1995; Xu et al., 1997). The VR1 receptor, through which capsaicin and RTX may act, has recently been cloned and characterized (Caterina et al., 1997). The similar pharmacology of the VR1 receptor with regard to capsaicin and noxious heat have led to the conclusion that the VR1 receptor acts as a physiologic transducer for heat-activated nociceptors (Caterina et al., 1997). In keeping with these findings, treatment with RTX was able to significantly increase response latencies to heat, supportive of the idea of desensitization of fibers expressing vanilloid receptors (i.e., C-fibers). Systemic injection of RTX has also been shown to reverse cold hyperalgesia, but not tactile allodynia, in rats with injury to central, rather than peripheral nerves, again suggesting the involvement of C-fibers in cold, but not mechanical, hypersensitivity (Hao et al., 1996). A significant finding of this study is that in spite of the significant loss of responsiveness to noxious thermal stimuli, the tactile allodynia present in the L5/L6 ligated rats was not changed throughout the entire course of the experiment. Since RTX selectively desensitizes those neurons
4. Discussion The results of the present investigation demonstrated a thermal hyperalgesia of the hindpaw to radiant noxious heat, but not in the hot-plate or tail-flick tests, in L5/L6 nerve ligated rats. The 52°C hot-plate and the 52°C hotwater tail-flick tests both provide fairly rapid responses to the nociceptive stimuli, and it is generally difficult to differentiate a hyperalgesic from a normal nociceptive response in these circumstances (see Wegert et al., 1997 and Bian et al., 1998 for discussions). The hot-plate test also provides the additional difficulty in that the rat is able to shift weight from the hindlimb ipsilateral to the nerve injury to the contralateral one, further confounding results if used alone. In contrast, the radiant heat paw-flick test presents the advantage of permitting a control over the heat intensity and rats of rise, allowing for a fine-tuning of the normal paw withdrawal latency. The paw-flick latencies to radiant heat employed in the present study correspond to the slow rate of rise, preferentially activating nociceptive C-fibers (Yeomans and Proudfit, 1996a; Yeomans and Proudfit, 1996b) or capsaicin-sensitive fibers
Fig. 4. Tail-flick latencies to a 52°C water bath were observed in shamoperated (X) and L5/L6 ligated (B) rats over 40 days. There was no significant difference between baseline responses in sham-operated or nerveinjured groups. Response thresholds were elevated by a single injection of RTX (0.3 mg/kg, s.c.) in both sham-operated and nerve-ligated rats. The elevated thresholds were significantly higher (*P ≤ 0.05, ANOVA followed by least significant difference) throughout the entire 40 day observation period; i.p. RTX (0.1 mg/kg) produced similar results (data not shown) (n = 6 rats/group).
M.H. Ossipov et al. / Pain 79 (1999) 127–133
expressing the vanilloid receptor, and since this receptor has recently been identified as a physiologic thermal nociceptive transducer (Caterina et al., 1997), then tactile stimuli may well be expected to be mediated by other (i.e. nonvanilloid receptor expressing) fibers. Consequently, the results of the present study strongly support the hypothesis that tactile allodynia and thermal hyperalgesia are mediated through different neuronal pathways. There are several possible mechanistic interpretations that may be invoked to support this reasoning. It is strongly suggested, based on these observations, that tactile allodynia is likely to be mediated through the activation of capsaicininsensitive large-diameter myelinated, or Ab afferent fibers, since this population of afferent nerves is unlikely to conduct nociceptive thermal stimuli (Light and Perl, 1979; Sugiura et al., 1986; Maxwell and Rethelyi, 1987) and is not sensitive to desensitization by capsaicin (or RTX) (Szolcsanyi et al., 1990). In the normal state, these fibers terminate in the deeper laminae (III–V) of the dorsal horn and code for nonnoxious stimuli, and they are not present in lamina II (Sugiura et al., 1986; Maxwell and Rethelyi, 1987). After peripheral nerve section or injury to the L5/L6 spinal nerves, Ab fibers form axonal sprouts arising from intact adjacent nerves and from the DRG of the injured nerves, and form abnormal synaptic connections with second order neurons of lamina II (Woolf et al., 1992; Woolf et al., 1995; Lekan et al., 1996; Doubell et al., 1997). This region normally receives an almost exclusively nociceptive input from unmyelinated Cfibers and is associated with nociceptive transmission (Light and Perl, 1979; Sugiura et al., 1986). Although these studies provide strong evidence for a physiopathic rearrangement of synaptic connections leading to neuropathic pain states, detailed time course analyses correlating sprouting with the development of behavioral signs of tactile allodynia, typically evident within 2–3 days after ligation (Chaplan et al., 1994; Bian et al., 1995), are not available. Mechanistic differences between thermal hyperalgesia and tactile allodynia can also be demonstrated pharmacologically. Spinally given morphine, for example, does not alter tactile responses to probing with von Frey filaments in rats with L5/L6 nerve ligation injury, even at doses considered to be supramaximal against other endpoints (i.e. 100 mg) (Bian et al., 1995; Lee et al., 1995). A complete lack of effect of morphine is not indicated, since the supraspinal or systemic injection of morphine does produce a dose-dependent antiallodynic effect. On the other hand, spinal morphine is effective against noxious thermal stimuli, though at somewhat reduced potency, and produces effective antihyperalgesic actions (Wegert et al., 1997). Likewise, earlier studies also reported on the efficacy of systemic (i.v.) morphine in rats with CCI, and implicated the possibility of a peripheral effect of morphine against signs of neuropathic pain (Lee et al., 1995; Kayser et al., 1995). An important difference between the latter studies and the present experiments is that thermal allodynia (determined with a warm, non-noxious water bath), and mechanical hyperalgesia
131
(determined by the Randall–Selitto method) was employed, whereas we measured thermal hyperalgesia (with noxious radiant heat) and tactile allodynia (von Frey filaments). Based on such findings, it is suggested that thermal hyperalgesia is most likely mediated through slow-conducting, unmyelinated C-fibers, sensitive to RTX. These fibers also have opioid receptors on their nerve terminals and this represents an important spinal site of action of opioids for presynaptic regulation of afferent noxious input (Yaksh et al., 1988; Lombard and Besson, 1989). Evidence indicates that it is unlikely that the Ab afferent neurons possess opioid receptors on the nerve terminals (Taddese et al., 1995). There is a differential distribution of expression of opioid receptors and of mRNA for the m-, d- and k- opioid receptors in DRG cell bodies and opioid receptor binding distributions are predominant in the superficial layers of the dorsal horn, corresponding with terminals of C-fibers rather than the terminal region of larger diameter fibers (e.g. Laminae III and IV) (Mansour et al., 1994; Mansour et al., 1995). Furthermore, activation of m-opioid receptors predictably inhibited Ca++ channels of small-diameter nociceptors and not of large-diameter cells indicating that m-receptor activation selectively inhibits the activity of C-fibers (Taddese et al., 1995). Therefore, morphine and other opioids may attenuate thermal nociception and thus thermal hyperalgesia through the actions on opioid receptors on both the primary nerve terminals and on the cell bodies of second order neurons of the dorsal horn. In contrast to thermal stimuli, it appears that tactile allodynia is mediated through the non-opioid receptor containing Ab fibers. Therefore, the only population of spinal opioid receptors that might be considered to be relevant to the attenuation of transmission of allodynia is that existing post-synaptic to the primary afferent neurons, which consists of 30–40% of the total spinal m- and d-opioid receptor population (Besse et al., 1990). Consequently, the ability of spinally injected opioids to block allodynia would be similar to the situation where the receptor reserve has been greatly decreased by a pharmacological manipulation (e.g. irreversible antagonist or development of tolerance). Such a manipulation would be expected to displace the agonist dose–effect curve to the right, and for an agonist of limited efficacy such as morphine, produce a decrease in the maximal response, a pattern typical of partial, irreversible blockade of a fraction of the relevant receptor population. Such a pattern is seen for spinal morphine (Lee et al., 1995; Bian et al., 1995). On the other hand, a high efficacy opioid m agonist such as DAMGO has been shown to elicit a significant antiallodynic action though with somewhat reduced potency, as would be expected (Nichols et al., 1995). A clinical correlate is that morphine generally provides inadequate pain relief in different neuropathic pain states (Arner and Meyerson, 1988; Rowbotham et al., 1991). However, the highly efficacious m-opioid fentanyl given i.t. provides complete pain relief against established post-amputation stump pain (Jacobson et al., 1990).
132
M.H. Ossipov et al. / Pain 79 (1999) 127–133
Finally, a differentiation in spinal and supraspinal neuronal circuitries in the processing of tactile allodynia and thermal hyperalgesia is strongly suggested. Pertovaara et al. (1996) found that the microinjection of lidocaine into the periaqueductal gray (PAG) or rostroventral medulla (RVM) blocked tactile allodynia, but not thermal nociceptive tail-flick reflexes, in rats with L5/L6 ligation injury. Transection of the spinal cord abolished tactile allodynia of the hindpaw, but facilitated the thermal tail-flick reflex, in rats with neurogenic inflammation (Mansikka and Pertovaara, 1997) or with L5/L6 spinal nerve ligation injury (Bian et al., 1998). Similar effects were seen after sacral spinal nerve ligation (Sung et al., 1997). It is unlikely that the loss of allodynia is an artifact of spinalization since spinal nocifensive responses to heat or pinch were present in the hindpaws and the tail. It was concluded that manifestation of tactile allodynia requires the activation of a spinal-supraspinal loop, possibly activating descending facilitatory and inhibitory systems arising from medullary sites, while the spinal thermal responses (still present after spinal transection) depend on localized circuitry, with thermal hyperalgesia being driven by central sensitization of localized spinal circuitry’s, although it may also be under some degree of descending control (Bian et al., 1998). In seeming contradiction to the results presented here, a recent clinical investigation reported that a high (5–10%) dose of capsaicin provided relief against neuropathic pain in the foot or chest wall arising from a variety of causes (Robbins et al., 1998). However, interpretation of the data is complicated in that the type of pain (i.e. hyperalgesic vs. allodynic vs. spontaneous) was not reported. Also, peripheral nerves neighboring an injured nerve have shown increased spontaneous and evoked activity, and this excitation was blocked by the application of lidocaine to the injured nerve (Sotgiu et al., 1996). Finally, the application of the local anesthetic lidocaine prior to capsaicin could confound the results since capsaicin acts initially by opening cation channels, whereas lidocaine exerts its local anesthetic activity by blocking Na + channels, thus opposing the action of capsaicin. The results presented here strongly suggest that different mechanisms may underlie the development of the different signs of neuropathic pain behavior. No single model is adequate to investigate the possible mechanisms that may be exploited in order to develop treatment modalities for this complex pain syndrome. It is clear that different signs of abnormal pain may be due to different and unique pathophysiological mechanisms. The present study, in concert with recent observations, suggests that treatment of hyperalgesia or other C-fiber related manifestations might be amenable to resolution with capsaicin or its analogs, such as RTX, but one must be cautious in extrapolating such conclusions to the entire array of neuropathic pain states especially to mechanical stimuli which appear to be resistant to manipulations resulting in changes in C-fiber activity.
References Arner, S. and Meyerson, B.A., Lack of analgesic effect on neuropathic and idiopathic forms of pain, Pain, 33 (1988) 11–23. Besse, D., Lombard, M.C., Zajac, J.M., Roques, B.P. and Besson, J.M., Pre- and postsynaptic location of mu, delta and kappa opioid receptors in the superficial layers of the dorsal horn of the rat spinal cord, Prog. Clin. Bio. Res., 328 (1990) 183–186. Bian, D., Nichols, M.L., Ossipov, M.H., Lai, J. and Porreca, F., Characterization of the antiallodynic efficacy of morphine in a model of neuropathic pain in rats, NeuroReport, 6 (1995) 1981–1984. Bian, D., Ossipov, M.H., Malan, T.P. and Porreca, F., Tactile allodynia, but not thermal hyperalgesia, of the hindlimbs is blocked by spinal transection in rats with nerve injury, Neurosci. Lett., 241 (1998) 79– 82. Caterina, M.J., Schumacher, M.A., Tominaga, M., Rosen, T.A., Levine, J.D. and Julius, D., The capsaicin receptor: a heat-activated ion channel in the pain pathway, Nature, 389 (1997) 816–824. Chaplan, S.R., Bach, F.W., Pogrel, J.W., Chung, J.M. and Yaksh, T.L., Quantitative assessment of tactile allodynia in the rat paw, J. Neurosci. Meth., 53 (1994) 55–63. Craft, R.M., Cohen, S.M. and Porreca, F., Long-lasting desensitization of bladder afferents following intravesical resiniferatoxin and capsaicin in the rat, Pain, 61 (1995) 317–323. Dixon, W.J., Efficient analysis of experimental observations, Ann. Rev. Pharmacol. Toxicol., 20 (1968) 441–462. Doubell, T.P., Mannion, R.J. and Woolf, C.J., Intact sciatic myelinated primary afferent terminals collaterally sprout in the adult rat dorsal horn following section of a neighbouring peripheral nerve, J. Comp. Neurol., 380 (1997) 95–104. Hao, J.X., Yu, W., Xu, X.J. and Wiesenfeld-Hallin, Z., Capsaicin-sensitive afferents mediate chronic cold, but not mechanical, allodynia-like behavior in spinally injured rats, Brain Res., 722 (1996) 177–180. Hargreaves, K., Dubner, R., Brown, F., Flores, C. and Joris, J., A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia, Pain, 32 (1988) 77–88. Jacobson, L., Chabal, C., Brody, M.C., Mariano, A.J. and Chaney, E.F., A comparison of the effects of intrathecal fentanyl and lidocaine on established postamputation stump pain, Pain, 40 (1990) 137–141. Kayser, V., Lee, S.H. and Guilbaud G., Evidence for a peripheral component in the enhanced antinocicpetive effect of a low dose of systemic morphine in rats with peripheral mononeuropathy, Neuroscience, 64 (1995) 537–545. Kim, S.H. and Chung, J.M., An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat, Pain, 50 (1992) 355–363. Lee, Y.-W., Chaplan, S.R. and Yaksh, T.L., Systemic and supraspinal but not spinal opiates suppress allodynia in a rat neuropathic pain model, Neurosci Lett., 199 (1995) 111–114. Lekan, H.A., Carlton, S.M. and Coggeshall, R.E., Sprouting of A beta fibers into lamina II of the rat dorsal horn in peripheral neuropathy, Neurosci. Lett., 208 (1996) 147–150. Light, A.R. and Perl, E.R., Spinal termination of functionally identified primary afferent neurons with slowly conducting myelinated fibers, J. Comp. Neurol., 186 (1979) 133–150. Lombard, M.C. and Besson, J.M., Attempts to gauge the relative importance of pre- and postsynaptic effects of morphine on the transmission of noxious messages in the dorsal horn of the rat spinal cord, Pain, 37 (1989) 335–345. Mansour, A., Fox, C.A., Thompson, R.C., Akil, H. and Watson, S.J., mOpioid receptor mRNA expression in the rat CNS: comparison to mreceptor binding, Brain Res., 643 (1994) 245–265. Mansikka, H. and Pertovaara, A., Supraspinal influence on hindlimb withdrawal thresholds and mustard oil-induced secondary allodynia in rats, Brain Res. Bull., 42 (1997) 359–365. Mansour, A., Fox, C.A., Akil, H. and Watson, S.J., Opioid-receptor mRNA
M.H. Ossipov et al. / Pain 79 (1999) 127–133 expression in the rat CNS: anatomical and functional implications, Trends Neurosci., 18 (1995) 22–29. Maxwell, D.J. and Rethelyi, M., Ultrastructure and synaptic connections of cutaneous afferent fibres in the spinal cord, Trends neurosci., 10 (1987) 117–123. Nichols, M.L., Bian, D., Ossipov, M.H. and Porreca, F., Evidence for regulation of opioid activity by CCK in a model of neuropathic pain, J. Pharmacol. Exp. Ther., 275 (1995) 1337–1345. Ossipov, M.H., Malan, T.P., Lai, J., Porreca, F., Opioid pharmacology of acute and chronic pain. In: A. Dickenson and J.-M. Beson (Eds.), Handbook of Experimental Pharmacology, Vol. 130, The Pharmacology of Spain, Springer-Verlag, Berlin, 1997b, pp. 305–333. Pertovaara, A., Wei, H. and Hamalainen, M.M., Lidocaine in the rostroventromedial medulla and the periaqueductal gray attenuates allodynia in neuropathic rats, Neurosci. Lett., 218 (1996) 127–130. Rowbotham, M.C., Reisner-Keller, L.A. and Fields, H.L., Both intravenous lidocaine and morphine reduce the pain of post-therapeutic neuralgia, Neurology, 41 (1991) 1024–1028. Robbins, W.R., Staats, P.S., Levine, J., Fields, H.L., Allen, R.W., Campbell, J.N. and Pappagallo, M., Treatment of intractable pain with topical large-dose capsaicin: preliminary report, Anesth. Analg., 86 (1998) 579–583. Sotgiu, M.L., Biella, G. and Lacerenza, M., Injured nerve block alters adjacent nerves spinal interaction in neuropathic rats, NeuroReport, 7 (1996) 1385–1388. Sugiura, Y., Lee, C.L. and Perl, E.R., Central projections of identified, unmyelinated (C) afferent fibers innervating mammalian skin, Science, 234 (1986) 358–361. Sung, B., Na, H.S., Kim, Y.I., Yoon, Y.W., Han, H.C., Nam, S.H. and Hong, S.K., The behavioral sign of mechanical allodynia from peripheral nerve injury is not a spinal reflex, Abstr. Soc. Neurosci., 23 (1997) 1804. Szallasi, A., Joo, F. and Blumberg, P.M., Duration of desensitization and ultrastructural changes in dorsal root ganglia in rats treated with resiniferatoxin, an ultrapotent capsaicin analog, Brain Res., 503 (1989) 68– 72. Szallasi, A. and Blumberg, P.M., Vanilloid receptors: new insights enhance potential as a therapeutic target, Pain, 68 (1996) 195–208.
133
Szolcsanyi, J., Szallasi, A., Szallasi, Z., Joo, F. and Blumberg, P.M., Resiniferatoxin: an ultrapotent selective modulator of capsaicin-sensitive primary afferent neurons, J. Pharmacol. Exp. Ther., 255 (1990) 923– 928. Taddese, A., Nah, S.-Y. and McCleskey, E.W., Selective opioid inhibition of small nociceptive neurons, Science, 270 (1995) 1366–1369. Wegert, S., Ossipov, M.H., Nichols, M.L., Bian, D., Vanderah, T., Malan, T.P. and Porreca, F., Differential activities of intrathecal MK-801 or morphine to alter responses to thermal and mechanical stimuli in normal or nerve-injured rats, Pain, 71 (1997) 57–64. Wood, J.N. (Ed.), Capsaicin in the Study of Pain, Academic Press, New York, 1993. Woolf, C.J., Shortland, P. and Coggeshall, R.E., Peripheral nerve injury triggers central sprouting of myelinated afferents, Nature, 355 (1992) 75–78. Woolf, C.J., Shortland, P., Reynolds, M., Ridings, J., Doubell, T. and Coggeshall, R.E., Reorganization of central terminals of myelinated primary afferents in the rat dorsal horn following peripheral axotomy, J. Comp. Neurol., 360 (1995) 121–134. Xu, X.J., Farkas-Szallasi, T., Lundberg, J.M., Hokfelt, T., WiesenfeldHallin, Z. and Szallasi, A., Effects of the capsaicin analogue resiniferatoxin on spinal nociceptive mechanisms in the rat: behavioral, electrophysiological and in situ hybridization studies, Brain Res., 752 (1997) 52–60. Yaksh, T.L., Al-Rodhan, N.R. and Jensen, T.S., Sites of action of opiates in production of analgesia, Prog. Brain Res., 77 (1988) 371–394. Yeomans, D.C. and Proudfit, H.K., Nociceptive responses to high and low rates of noxious cutaneous heating are mediated by different nociceptors in the rat: behavioral evidence, Pain, 68 (1996a) 133–140. Yeomans, D.C. and Proudfit, H.K., Nociceptive responses to high and low rates of noxious cutaneous heating are mediated by different nociceptors in the rat: electrophysiological evidence, Pain, 68 (1996b) 141– 150. Zachariou, V., Goldstein, B.D. and Yeomans, D.C., Low but not high rate noxious radiant skin heating evokes a capsaicin-sensitive increase in spinal cord dorsal horn release of substance P, Brain Res., 752 (1997) 143–150.