FINAL ACCEPTED VERSION Neural basis for the ... - CiteSeerX

Articles in PresS. J Neurophysiol (May 19, 2004). 10.1152/jn.00411.2004 JN-00411-2004.R1

FINAL ACCEPTED VERSION

Neural basis for the hyperalgesic action of cholecystokinin in the rostral ventromedial medulla.

Mary M. Heinricher and Miranda J. Neubert, Dept. Neurological Surgery, Oregon Health & Science University, Portland, OR 97239, USA

Running head: Hyperalgesic action of cholecystokinin in medulla

Address for correspondence: Mary M. Heinricher, Ph.D. Dept. Neurological Surgery, L-472 Oregon Health & Science University Portland, OR 97239 503-494-1135, fax: 503-494-7161 email: [email protected]

Copyright © 2004 by the American Physiological Society.

JN-00411-2004.R1, p. 2

Abstract The analgesic actions of opioids can be modified by endogenous “anti-opioid” peptides, among them cholecystokinin (CCK). CCK is also now thought to have a broader, pronociceptive, role, and contributes to hyperalgesia in inflammatory and neuropathic pain states. The aim of the present experiments was to determine whether anti-opioid and pro-nociceptive actions of CCK have a common underlying mechanism. We showed previously that a low dose of CCK microinjected into the rostral ventromedial medulla (RVM) blocked the analgesic effect of systemically administered morphine by preventing activation of “off-cells,” which are the antinociceptive output of this well characterized pain-modulating region. At this “anti-opioid” dose, CCK had no effect on the spontaneous activity of these neurons, or on the activity of “oncells” (hypothesized to facilitate nociception) or “neutral cells” (which have no known role in pain modulation). In the present study, we used microinjection of a higher dose of CCK into the RVM to test whether activation of on-cells could explain the pro-nociceptive action of this peptide. Paw withdrawal latencies to noxious heat and the activity of a characterized RVM neuron were recorded in rats lightly anesthetized with methohexital. CCK (30 ng/200 nl) activated on-cells selectively, and produced behavioral hyperalgesia. Firing of off-cells and neutral cells was unaffected. These data show that direct, selective activation of RVM on-cells by CCK is sufficient to produce thermal hyperalgesia, and indicate that the anti-opioid and pronociceptive effects of this peptide are mediated by actions on different RVM cell classes.

JN-00411-2004.R1, p. 3

Introduction It is now well established that the analgesic actions of opioids can be modified by various endogenous “anti-opioid” peptides, among them cholecystokinin (CCK, see Stanfa and Dickenson (1994) and Wiesenfeld-Hallin et al. (1999) for comprehensive reviews). However, the notion that CCK is recruited to negate or balance the analgesic actions of opioids has now evolved to include a broader, pro-nociceptive, role for this peptide (Ossipov et al. 2003). In support of this, CCK has been shown to contribute to the enhanced nociceptive behavior in models of inflammatory or neuropathic pain, although it is generally agreed that CCK transmission plays little role in nociceptive responses in normal animals (Coudore-Civiale et al. 2000b; Friedrich and Gebhart 2003; Kamei and Zushida 2001; Kovelowski et al. 2000; Nichols et al. 1996; Urban et al. 1996; Xu et al. 1994). From this vantage point, CCK is seen as directly facilitating nociception. This latter perspective raises the question of whether the anti-opioid action of CCK is merely a product of increased nociception (leading to a higher opioid requirement), or whether antiopioid and pro-nociceptive actions of CCK are mediated by distinct neural mechanisms. A logical site at which to address this question is the rostral ventromedial medulla (RVM). The role of this region in opioid analgesia and pain modulation is well documented (Fields and Basbaum 1999; Fields et al. 1991), and functional studies demonstrate that focal application of CCK in the RVM both attenuates the antinociceptive actions of morphine (Heinricher et al. 2001; Mitchell et al. 1998) and facilitates responding on cutaneous and visceral tests of nociception (Friedrich and Gebhart 2003; Kovelowski et al. 2000). In our previous study of CCK in the RVM, we microinjected CCK at a dose (10 ng) chosen specifically to interfere with the antinociceptive effect of a range of doses of systemically administered morphine (Heinricher et al. 2001). RVM neurons can be divided into three classes:

JN-00411-2004.R1, p. 4

“off-cells”, “on-cells” and “neutral cells” (Fields and Heinricher 1985). In our earlier experiments, local administration of CCK had no effect on the ongoing activity of any RVM cell class, but attenuated opioid activation off-cells. Off-cell activation is sufficient to produce antinociception, and known to be critical to opioid analgesia (Heinricher et al. 1999; Neubert et al. in press). The ability of morphine to inhibit firing of on-cells was unaffected by CCK. Because opioid activation of off-cells is indirect (Heinricher et al. 1992; Pan et al. 1990), the selective interference with opioid activation of these neurons by CCK suggested that the peptide acted presynaptically. However, CCK is also known to have post-synaptic, usually excitatory, effects in a number of CNS regions. Since there is now direct evidence that on-cells exert a net facilitatory effect on nociception (Neubert et al. in press), this raises the possibility that a more extensive analysis of the effects of CCK within the RVM might identify direct activation of oncells as underlying the pro-nociceptive (i.e., hyperalgesic) effects of higher doses of CCK in the RVM. In the present study, we used local infusion of a higher (30 ng) dose of CCK within the RVM to test this possibility. We recorded the effect of the peptide on identified RVM neurons and, as a measure of nociceptive responsiveness, the latency of the paw withdrawal reflex to radiant heat. Materials and methods Animals and surgical preparation. All experimental procedures followed the guidelines of the Committee for Research and Ethical Issues of the International Association for the Study of Pain, and were approved by the Institutional Animal Care and Use Committee at Oregon Health & Science University. Male Sprague-Dawley rats (Sasco, 250-300g) were anesthetized with pentobarbital (60 mg/kg, i.p.), and a catheter inserted into an external jugular vein for administration of anesthetic. The rat was

JN-00411-2004.R1, p. 5

placed in a stereotaxic apparatus, a hole drilled in the skull over the cerebellum, and the dura removed to allow placement of an electrode/microinjection cannula assembly in the medulla. Body temperature was maintained at approximately 37 °C by a circulating water pad. Following surgery, the anesthetic level was allowed to lighten until a paw withdrawal reflex could be elicited by application of noxious heat using a feedback-controlled projector lamp focused on the blackened plantar surface of the paw. The animals were then maintained in a lightly anesthetized state using a continuous infusion of methohexital (15-30 mg/kg per h, i.v.) as previously described (Barbaro et al. 1989). Nociceptive testing. Paw withdrawal latency (PW) was used as a measure of nociceptive responsiveness. Each trial consisted of a linear increase in temperature at approximately 1.8 °C/s from a holding temperature of 34 °C until the paw withdrawal occurred or to a maximum of 52 °C at 10.6 s. Trials were carried out at 5 min intervals throughout the experiment. The holding temperature obviates any concern that apparent effects on paw withdrawal latency were due to changes in skin temperature. Recording and drug administration. A gold- and platinum-plated stainless steel recording microelectrode (Frederick Haer Co., Brunswick, ME) was glued parallel to a single-barrel glass infusion pipette with an outer diameter of 75-80 µm in such a way that the tips were separated laterally and slightly rostrocaudally by 100-300 µm. This separation allowed us to maintain well-isolated recordings of single neurons during and after infusions of 200 nl of saline vehicle or drug solution in the RVM in a reasonable number of experiments. The assembly was oriented in the electrode carrier so

JN-00411-2004.R1, p. 6

that the assembly straddled the midline, with both electrode and infusion pipette within the RVM. The infusion pipette was attached to a 1 µl Hamilton syringe with a length of PE-50 tubing for drug infusion. RVM neurons were classified as previously described (Fields et al. 1983). Spike waveforms were monitored and stored for off-line analysis (Datawave Systems, Thornton, CO) to ensure that the unit under study was unambiguously discriminated throughout the experiment, and to verify that the solutions applied did not have local anesthetic effects. Spike times were stored with a temporal resolution of 0.1 ms. Off-cells were characterized by an abrupt pause in ongoing activity beginning just prior to the occurrence of the PW. On-cells were identified by a sudden burst of activity beginning just prior to the occurrence of the PW. Cells of a third class, neutral cells, were identified by no change in activity associated with paw withdrawal, and they did not respond to noxious or innocuous cutaneous stimulation. Protocol and data analysis. We determined the effect of CCK microinjection into the RVM on PW latency and on the ongoing and reflex-related discharges of RVM neurons. Following three baseline PW trials, CCK (sulfated octapeptide, 30 ng, MW 1142.2, Tocris, Ellisville, MO or American Peptide, Sunnyvale, CA) or saline was infused into the RVM. CCK was dissolved in saline, and aliquots stored at –20 °C until the day of use. All RVM injections were made in a volume of 200 nl over a period of approximately 4 min. PW latency and cell activity were then monitored for a period of 45 min. Only one protocol was performed in each animal. In one case, two on-cells were recorded on a single electrode. The average of the PW latencies and cell parameters obtained in the baseline

JN-00411-2004.R1, p. 7

period were compared with the average of the trials from 20 - 45 min following the microinjection. This period includes six PW trials. Three cell parameters were analyzed. 1) Ongoing activity. Because off-cells and on-cells often show irregular alternations between periods of silence and activity, cell activity integrated over the 30 s prior to each PW trial was used as an overall index of ongoing firing. 2) On-cell PW-related burst. Total spike count in the 3 s period beginning 1 s before the PW was recorded for all PW trials. This approach, rather than counting the number of spikes or duration of the reflex-related burst is necessary because a burst cannot be identified unless the neuron is inactive at the time of heat onset. 3) Cycling. The proportion of time that a given cell was considered to be in an “active” or “silent” period was defined as described previously (Barbaro et al. 1989). Briefly, an active period was defined as any epoch lasting at least 2 s with a minimum of 1 spike/s, and a silent period as any epoch of at least 2 s without any cell activity. The proportion of time in which each off-cell was active or silent was then calculated for the baseline and postinfusion intervals in groups in which there was an overall change in ongoing activity (calculated as described above for the 30 s sample prior to each PW trial). Data are presented as mean + SEM. Student’s t-test for correlated means was used to compare baseline and post-injection values for paw-withdrawal latencies and ongoing discharge if two-way repeated measures ANOVA indicated a significant overall effect of drug over time. A t-test for correlated means was also used to compare baseline and post-injection values for the on-cell burst, proportion of time active and duration of the off-cell pause. P < 0.05 was considered statistically significant.

JN-00411-2004.R1, p. 8

Histology. At the conclusion of the experiment, recording sites were marked with an electrolytic lesion, and infusion sites by microinjection of Pontamine sky blue dye. Animals were euthanized with an overdose of methohexital, and perfused intracardially with physiological saline followed by 10% formalin. Recording/infusion sites were histologically verified and plotted on standardized sections (Paxinos and Watson 1997). Infusion/recording sites are shown in Figure 1. RVM coordinates were 2.2 caudal to the interaural line, 7 to 9 mm below the brain surface. The RVM was defined as comprising the nucleus raphe magnus as well as the laterally adjacent reticular formation at the level of the facial nucleus, including the nucleus reticularis gigantocellularis pars alpha. Results Microinjection of 30 ng CCK in the RVM produces thermal hyperalgesia in lightly anesthetized rats. Because previous behavioral studies of the effects of high doses of CCK in the RVM had been performed in awake animals (Friedrich and Gebhart 2003; Kovelowski et al. 2000), and because we had previously seen no effect of a lower dose of CCK (10 ng) in the RVM on nociceptive reflexes in anesthetized animals (Heinricher et al. 2001), we first determined whether 30 ng CCK would produce thermal hyperalgesia when microinjected into the RVM in lightly anesthetized animals. We established that application of 30 ng CCK within the RVM in lightly anesthetized rats produced a robust decrease in PW latency (Fig. 2). Hyperalgesia was evident within 10 min of the microinjection, and lasted throughout the 45 min observation period. Infusion of saline vehicle into the RVM had no effect. Importantly, PW latency was unchanged following microinjection of CCK into the reticular formation surrounding the RVM (see Fig. 1

JN-00411-2004.R1, p. 9

for locations), indicating that the site of action of the peptide was within the RVM rather than the adjacent medullary reticular formation. CCK activates on-cells selectively to produce hyperalgesia Seven off-cells, nine neutral cells and eleven on-cells were successfully recorded in experiments in which CCK was microinjected into the RVM during the recording. As shown in the examples in Figs. 3 and 4, the ongoing firing of both off-cells and neutral cells was unchanged following infusion of CCK or vehicle into the RVM. In addition, CCK had no effect on the duration of the off-cell pause or the delay from the final spike prior to the paw withdrawal (pause onset) to the withdrawal (Table 1). The latency from heat onset to the beginning of the pause was shorter after CCK, which is not surprising given that the paw withdrawal occurred at a shorter latency. Note that these values could be measured only on trials on which the off-cell under study was spontaneously active at the time of heat onset. By contrast with off-cells and neutral cells, on-cells displayed a substantial increase in ongoing activity following infusion of CCK (Fig. 5). As can be seen in Table 1, the overall increase in ongoing activity of these neurons reflected an increase in the proportion of time active as well as a substantial increase in firing rate during active periods. Vehicle microinjection had no effect on on-cell firing (n = 7, Fig. 6). CCK injections outside of the RVM (n = 6, see Fig. 1 for locations) also had no effect on RVM on-cell discharge (p > 0.05). Firing rate of the on-cells at the time of the PW reflex was also significantly increased (Table 1). This increase is related to the overall increase in ongoing activity, since there was rarely a discrete on-cell “burst” following CCK (see for example Fig. 5). Indeed, 8 of 11 cells were already active at the time of heat onset on at least five of the six trials falling in the post-infusion analysis period. Overall, the cell under study was already active at heat onset on only 42.4% of

JN-00411-2004.R1, p. 10

baseline trials, but on 84.8% of post-CCK trials. This increase in the probability of being in an active period at the time of the PW trial precluded analysis of the burst onset or duration. Vehicle microinjection had no effect on reflex-related firing rate (13.3 ± 2.4 sp/s at the time of the reflex in baseline, 14.5 ± 2.7 sp/s after saline, p > 0.05). Injections adjacent to the RVM were similarly ineffective (17.3 ± 5.2 sp/s at the time of the reflex in baseline, 21.5 ± 7.1 sp/s after saline, p > 0.05). The effects of 30 ng CCK in the RVM on the ongoing firing of on- and off-cells and neutral cells are summarized in Fig. 7, which shows ongoing firing rate before and after CCK infusion for all three cell classes. Discussion At low doses, microinjection of CCK within the RVM blocks morphine-induced activation of off-cells and attenuates the analgesic effect of the opioid (Heinricher et al. 2001; Mitchell et al. 1998), However, at higher doses, CCK produces behaviorally measurable hyperalgesia in awake behaving animals (Friedrich and Gebhart 2003; Kovelowski et al. 2000). The principle finding of the present study is that focal administration of a higher dose, 30 ng, of CCK in the RVM activates on-cells selectively, producing thermal hyperalgesia. Firing of off-cells and neutral cells is unaffected. These data thus confirm previous behavioral observations in awake animals, and point to activation of on-cells as the mechanism underlying the hyperalgesic action of this peptide. The role of CCK in descending control of nociception. The anti-opioid or effects of CCK are well documented, although whether CCK also modulates “non-opioid” analgesia, having a more general “anti-analgesic” effect is unknown. Administration of CCK agonists can diminish opioid analgesia (Faris et al. 1983; Li and Han

JN-00411-2004.R1, p. 11

1989; O'Neill et al. 1989), whereas CCK antagonists or CCK2 receptor antisense treatment enhances the antinociceptive effects of exogenous and endogenous opioids. CCK blockade can also slow or prevent the development of opioid tolerance in some paradigms (Chapman et al. 1995; Dourish et al. 1990; Hoffmann and Wiesenfeld-Hallin 1994; Maldonado et al. 1993; Mitchell et al. 2000; Watkins et al. 1985a; Watkins et al. 1985b, 1984). In addition, CCK is implicated in environmentally induced changes in nociception (Chen et al. 1998; Hendrie et al. 1989; Lavigne et al. 1992; Wiertelak et al. 1992; Wiertelak et al. 1994). Moreover, there is evidence that alterations in the availability of CCK contribute to the changes in analgesic potency of opioids seen in conditions of stress, inflammation and in some pathological pain states (Coudore-Civiale et al. 2000a; Friedrich and Gebhart 2000; Hawranko and Smith 1999; Nichols et al. 1995; Stanfa and Dickenson 1993). Although the focus of much of the work concerned with CCK-opioid interactions has been at the level of the spinal cord, specific brain regions, including the RVM, have been implicated in this anti-opioid action of CCK (Heinricher et al. 2001; Kovelowski et al. 2000; Mitchell et al. 1998; Pu et al. 1994; Tortorici et al. 2003; Watkins et al. 1985b). We previously showed that infusion of a low dose of CCK (10 ng) into the RVM prevents the analgesic effect of systemically administered morphine by interfering with opioid activation of off-cells (Heinricher et al. 2001). At this dose, CCK alone had no effect on ongoing or reflex-related firing of any cell class, off-cells, on-cells or neutral cells. The more recent evidence that CCK in the RVM contributes to hyperalgesia seen in inflammatory and neuropathic pain states (Friedrich and Gebhart 2003; Kovelowski et al. 2000) raised the question of whether this hyperalgesia requires recruitment of additional neural mechanisms distinct from those underlying the anti-opioid effects of the peptide. The present finding that local application of a higher dose of CCK within the RVM activates on-cells

JN-00411-2004.R1, p. 12

selectively to produce behavioral hyperalgesia indicates that the pro-nociceptive actions of the peptide are mediated by neural elements distinct from those mediating the anti-opioid effects. Identification of the endogenous CCK inputs influencing the two cell classes will be required to determine whether CCK released endogenously activates the anti-opioid and pro-nociceptive circuits independently, or in parallel. However, parallel recruitment seems likely, since opioids are known to induce CCK release, and since prolonged opioid administration can actually facilitate nociception (de Araujo Lucas et al. 1998; Gardell et al. 2002; Gustafsson et al. 1999; Herman et al. 2002; Larcher et al. 1998; Lucas et al. 1999; Ossipov et al. 2003; Simonnet and Rivat 2003; Vanderah et al. 2001a; Vanderah et al. 2001b). The combined microinjection/recording approach employed here made administration of additional drugs impractical. We were thus unable to use selective antagonists to determine the CCK receptor type involved. However, both anti-opioid and pro-nociceptive actions of CCK are generally attributed to the CCK2 (CCK-B) receptor (Ossipov et al. 2003; Wiesenfeld-Hallin et al. 1999). The dose-dependence of the behavioral and neuronal effects of the peptide is therefore not likely due to an action of the peptide at different receptors. It may be that the receptor mediating the CCK block of opioid activation of off-cells is more easily accessed when exogenous peptide is given by microinjection. The role of on- and off-cells in pain modulation The combined microinjection/single cell recording approach allows us to manipulate the firing of RVM neurons and determine the effects of that manipulation on nociceptive responding. The present observation that direct, selective activation of RVM on-cells by local application of CCK leads to thermal hyperalgesia adds to a growing body of correlative and direct evidence that on-cells facilitate nociception, at least as measured by a withdrawal reflex to noxious heat

JN-00411-2004.R1, p. 13

(Bederson et al. 1990; Foo and Mason 2003; Heinricher et al. 1989; McGaraughty et al. 2003; Neubert et al. in press; Ramirez and Vanegas 1989). The conclusion that on-cells facilitate nociception is also consistent with recent findings of Porreca and colleagues (Burgess et al. 2002; Porreca et al. 2001), who have shown that µ-opioid receptor expressing neurons, likely to be on-cells (Heinricher et al. 1992), are required for allodynia and hyperalgesia following spinal nerve ligation. Notably, the influence of the on-cell output is sufficiently potent to produce this behavioral effect in the absence of concomitant or pre-existing sensitization of dorsal horn circuitry, as presumably occurs in inflammatory or nerve injury models (Mansikka and Pertovaara 1997; Pertovaara 1998; Porreca et al. 2002; Urban and Gebhart 1999). Whether these neurons also modulate responding to noxious mechanical or visceral stimulation will require testing the effect of experimental manipulation of their firing rate or pattern on behavioral responses to mechanical or visceral stimulation. RVM mechanisms for hyperalgesia mediated by on-cells include a reduction in the threshold at which the on-cell burst is triggered, or a modification of the pattern of ongoing activity so that the on-cell population is more likely to be active at the onset of the noxious stimulus (Heinricher et al. 1989; McGaraughty et al. 2003; Neubert et al. in press). The primary change following administration of

CCK was a shift in the firing pattern of on-cells such that this population was more likely to be active at any given time following the microinjection. As a consequence, the noxious heat stimulus, which was delivered to the paw at constant five-min intervals, was more likely to fall at a time when the on-cell population was active and off-cells inactive. The reflex response to the heat therefore occurred at a shorter latency.

Only on-cells showed a change in discharge following microinjection of CCK in the RVM. Activation of these neurons is thus either a direct post-synaptic effect of the peptide on these neurons, or a presynaptic effect on a synaptic input. Iontophoretic experiments will be required

JN-00411-2004.R1, p. 14

to determine whether on-cells are directly sensitive to CCK, which would distinguish between these possibilities. However, in either case, the net effect is to activate on-cells selectively and produce thermal hyperalgesia. Initial concepts of RVM circuitry proposed that on-cells served as inhibitory interneurons within the RVM, and that they mediated the reflex-related off-cell pause (Fields and Heinricher 1985). This hypothesis was based on a number of characteristics of the two cell classes. First, on-cells by definition show a burst activity at the time of the withdrawal reflex, that is, at the time of the off-cell pause. Second, opioid activation of off-cells is mediated by disinhibition, whereas on-cells are known to be directly sensitive to µ-opioid agonists (Heinricher et al. 1994; Pan et al. 1990). Third, spontaneous firing of on- and off-cell populations is out of phase (Barbaro et al. 1989). However, recent observations are inconsistent with the idea that on-cells inhibit off-cells. We recently showed that disfacilitation of on-cell firing does not result in increased off-cell discharge (Heinricher and McGaraughty 1998), and that neurotensinergic activation of on-cells does not suppress off-cell firing (Neubert et al. in press). The present finding that selective activation of on-cells by CCK has no effect on off-cell firing extends these observations, and confirms the idea that the on- and off-cell outputs can be controlled independently. The classification of RVM neurons into on-, off- and neutral cells as used by us has been corroborated by the consistent pharmacology within each class, including responses to µ- and δopioids, α2–adrenergic agonists, neurotensin and cannabinoids, as well as block of endogenous excitatory amino acid inputs by NMDA and non-NMDA receptor antagonists (Harasawa et al. 2000; Heinricher et al. 1988; Heinricher and McGaraughty 1998; Heinricher et al. 1992; Meng et

JN-00411-2004.R1, p. 15

al. 1998; Neubert et al. in press). The present finding that CCK acts uniformly and selectively to activate on-cells further validates the relevance of the physiological classification. Technical considerations. There are a number of technical issues that must be considered in any study using the microinjection technique. Diffusion of drugs to sites distant from the intended target is a primary concern. That seems unlikely to be an issue in the present experiments because missed placements in regions adjacent to the RVM had no effect on paw withdrawal latency. A second consideration is pharmacological specificity. However, the hyperalgesia seen in the present study is entirely consistent with thermal hyperalgesia and mechanical allodynia reported by Kovelowski et al. (2000) in awake animals. The differential effect of CCK on on-cells, the association of behavioral and neuronal changes, and the lack of effect of the vehicle also argue against a non-specific effect of the peptide. Finally, it must be recognized that there may be RVM neurons not accessible to extracellular recording, and such neurons may be responsible for the behavioral effects seen here. However, the alterations of on-cell firing are clearly sufficient to explain the behavioral changes, and activation of this cell class remains the most parsimonious explanation of the behavioral effects of CCK. In conclusion, these data, together with previous work (Heinricher et al. 2001), demonstrate that the anti-opioid and pro-nociceptive effects of CCK within the RVM are mediated by distinct circuits. The anti-opioid action of this peptide can be explained by its ability to block opioid activation of off-cells, the inhibitory output neuron of the RVM. By contrast, the pronociceptive action is mediated by activation of on-cells. These data thus confirm the proposal that on-cells are the facilitating output of the RVM, and raise the possibility that anti-opioid and pro-nociceptive aspects of descending modulation could be recruited independently.

JN-00411-2004.R1, p. 16

Acknowledgements Supported by a grant from NIDA (DA05608).

JN-00411-2004.R1, p. 17

REFERENCES Barbaro NM, Heinricher MM, and Fields HL. Putative nociceptive modulatory neurons in the rostral ventromedial medulla of the rat display highly correlated firing patterns. Somatosens Mot Res 6: 413-425, 1989. Bederson JB, Fields HL, and Barbaro NM. Hyperalgesia during naloxone-precipitated withdrawal from morphine is associated with increased on-cell activity in the rostral ventromedial medulla. Somatosens Mot Res 7: 185-203, 1990. Burgess SE, Gardell LR, Ossipov MH, Malan TP, Jr., Vanderah TW, Lai J, and Porreca F. Timedependent descending facilitation from the rostral ventromedial medulla maintains, but does not initiate, neuropathic pain. J Neurosci 22: 5129-5136., 2002. Chapman V, Honoré P, Buritova J, and Besson JM. Cholecystokinin B receptor antagonism enhances the ability of a low dose of morphine to reduce c-fos expression in the spinal cord of the rat. Neuroscience 67: 731-739, 1995. Chen XH, Geller EB, and Adler MW. CCK(B) receptors in the periaqueductal grey are involved in electroacupuncture antinociception in the rat cold water tail-flick test. Neuropharmacology 37: 751-757, 1998. Coudore-Civiale MA, Courteix C, Boucher M, Meen M, Fialip J, Eschalier A, and Ardid D. Potentiation of morphine and clomipramine analgesia by cholecystokinin -B antagonist CI988 in diabetic rats. Neurosci Lett 286: 37-40, 2000a. Coudore-Civiale MA, Courteix C, Fialip J, Boucher M, and Eschalier A. Spinal effect of the cholecystokinin-B receptor antagonist CI-988 on hyperalgesia, allodynia and morphineinduced analgesia in diabetic and mononeuropathic rats. Pain 88: 15-22, 2000b.

JN-00411-2004.R1, p. 18

de Araujo Lucas G, Alster P, Brodin E, and Wiesenfeld-Hallin Z. Differential release of cholecystokinin by morphine in rat spinal cord. Neurosci Lett 245: 13-16., 1998. Dourish CT, O'Neill M, Coughlan J, Kitchener SJ, Hawley D, and Iversen SD. The selective CCK-B receptor antagonist L-365,260 enhances morphine analgesia and prevents morphine tolerance in the rat. Eur J Pharmacol 176: 35-44, 1990. Faris PL, Komisaruk BR, Watkins LR, and Mayer DJ. Evidence for the neuropeptide cholecystokinin as an antagonist of opiate analgesia. Science 219: 310-312, 1983. Fields HL and Basbaum AI. Central nervous mechanisms of pain modulation. In: Textbook of Pain (4th ed.), edited by Wall PD and Melzack R. Edinburgh: Churchill Livingston, 1999, p. 309-329. Fields HL, Bry J, Hentall I, and Zorman G. The activity of neurons in the rostral medulla of the rat during withdrawal from noxious heat. J Neurosci 3: 2545-2552, 1983. Fields HL and Heinricher MM. Anatomy and physiology of a nociceptive modulatory system. Philos Trans R Soc Lond B Biol Sci 308: 361-374, 1985. Fields HL, Heinricher MM, and Mason P. Neurotransmitters in nociceptive modulatory circuits. Annu Rev Neurosci 14: 219-245, 1991. Foo H and Mason P. Discharge of Raphe Magnus ON and OFF Cells Is Predictive of the Motor Facilitation Evoked by Repeated Laser Stimulation. J Neurosci 23: 1933-1940, 2003. Friedrich AE and Gebhart GF. Effects of spinal cholecystokinin receptor antagonists on morphine antinociception in a model of visceral pain in the rat. J Pharmacol Exp Ther 292: 538-544, 2000. Friedrich AE and Gebhart GF. Modulation of visceral hyperalgesia by morphine and cholecystokinin from the rat rostroventral medial medulla. Pain 104: 93-101, 2003.

JN-00411-2004.R1, p. 19

Gardell LR, Wang R, Burgess SE, Ossipov MH, Vanderah TW, Malan TP, Jr., Lai J, and Porreca F. Sustained morphine exposure induces a spinal dynorphin-dependent enhancement of excitatory transmitter release from primary afferent fibers. J Neurosci 22: 6747-6755., 2002. Gustafsson H, Afrah A, Brodin E, and Stiller CO. Pharmacological characterization of morphineinduced in vivo release of cholecystokinin in rat dorsal horn: effects of ion channel blockers. J Neurochem 73: 1145-1154, 1999. Harasawa I, Fields HL, and Meng ID. Delta opioid receptor mediated actions in the rostral ventromedial medulla on tail flick latency and nociceptive modulatory neurons. Pain 85: 255-262, 2000. Hawranko AA and Smith DJ. Stress reduces morphine's antinociceptive potency: dependence upon spinal cholecystokinin processes. Brain Res 824: 251-257, 1999. Heinricher MM, Barbaro NM, and Fields HL. Putative nociceptive modulating neurons in the rostral ventromedial medulla of the rat: firing of on- and off-cells is related to nociceptive responsiveness. Somatosens Mot Res 6: 427-439, 1989. Heinricher MM, Haws CM, and Fields HL. Opposing actions of norepinephrine and clonidine on single pain-modulating neurons in rostral ventromedial medulla. In: Pain Research and Clinical Management, edited by Dubner R, Gebhart GF and Bond MR. Amsterdam: Elsevier, 1988, p. 590-594. Heinricher MM and McGaraughty S. Analysis of excitatory amino acid transmission within the rostral ventromedial medulla: Implications for circuitry. Pain 75: 247-255, 1998. Heinricher MM, McGaraughty S, and Farr DA. The role of excitatory amino acid transmission within the rostral ventromedial medulla in the antinociceptive actions of systemically administered morphine. Pain 81: 57-65, 1999.

JN-00411-2004.R1, p. 20

Heinricher MM, McGaraughty S, and Tortorici V. Circuitry underlying antiopioid actions of cholecystokinin within the rostral ventromedial medulla. J Neurophysiol 85: 280-286, 2001. Heinricher MM, Morgan MM, and Fields HL. Direct and indirect actions of morphine on medullary neurons that modulate nociception. Neuroscience 48: 533-543, 1992. Heinricher MM, Morgan MM, Tortorici V, and Fields HL. Disinhibition of off-cells and antinociception produced by an opioid action within the rostral ventromedial medulla. Neuroscience 63: 279-288, 1994. Hendrie CA, Shepherd JK, and Rodgers RJ. Differential effects of the CCK antagonist, MK-329, on analgesia induced by morphine, social conflict (opioid) and defeat experience (nonopioid) in male mice. Neuropharmacology 28: 1025-1032, 1989. Herman DS, Stiller CO, Gardell LR, Lai J, Malan TP, Porreca F, and Vanderah T. Morphineinduced release of cholecystokinin in the rostral ventromedial medulla. Soc Neurosci Abstr: 351.351, 2002. Hoffmann O and Wiesenfeld-Hallin Z. The CCK-B receptor antagonist Cl 988 reverses tolerance to morphine in rats. Neuroreport 5: 2565-2568, 1994. Kamei J and Zushida K. The role of spinal cholecystokinin B receptors in thermal allodynia and hyperalgesia in diabetic mice. Brain Res 892: 370-375., 2001. Kovelowski CJ, Ossipov MH, Sun H, Lai J, Malan TP, and Porreca F. Supraspinal cholecystokinin may drive tonic descending facilitation mechanisms to maintain neuropathic pain in the rat. Pain 87: 265-273, 2000. Larcher A, Laulin JP, Celerier E, Le Moal M, and Simonnet G. Acute tolerance associated with a single opiate administration: involvement of N-methyl-D-aspartate-dependent pain facilitatory systems. Neuroscience 84: 583-589, 1998.

JN-00411-2004.R1, p. 21

Lavigne GJ, Millington WR, and Mueller GP. The CCK-A and CCK-B receptor antagonists, devazepide and L-365,260, enhance morphine antinociception only in non-acclimated rats exposed to a novel environment. Neuropeptides 21: 119-129, 1992. Li Y and Han JS. Cholecystokinin-octapeptide antagonizes morphine analgesia in periaqueductal gray of the rat. Brain Res 480: 105-110, 1989. Lucas GA, Hoffmann O, Alster P, and Wiesenfeld-Hallin Z. Extracellular cholecystokinin levels in the rat spinal cord following chronic morphine exposure: an in vivo microdialysis study. Brain Res 821: 79-86, 1999. Maldonado R, Derrien M, Noble F, and Roques BP. Association of the peptidase inhibitor RB 101 and a CCK-B antagonist strongly enhances antinociceptive responses. Neuroreport 4: 947-950, 1993. Mansikka H and Pertovaara A. Supraspinal influence on hindlimb withdrawal thresholds and mustard oil-induced secondary allodynia in rats. Brain Res Bull 42: 359-365, 1997. McGaraughty S, Chu KL, Bitner RS, Martino B, Kouhen RE, Han P, Nikkel AL, Burgard EC, Faltynek CR, and Jarvis MF. Capsaicin infused into the PAG affects rat tail flick responses to noxious heat and alters neuronal firing in the RVM. J Neurophysiol 90: 2702-2710, 2003. Meng ID, Manning BH, Martin WJ, and Fields HL. An analgesia circuit activated by cannabinoids. Nature 395: 381-383., 1998. Mitchell JM, Basbaum AI, and Fields HL. A locus and mechanism of action for associative morphine tolerance. Nat Neurosci 3: 47-53, 2000. Mitchell JM, Lowe D, and Fields HL. The contribution of the rostral ventromedial medulla to the antinociceptive effects of systemic morphine in restrained and unrestrained rats. Neuroscience 87: 123-133, 1998.

JN-00411-2004.R1, p. 22

Neubert MJ, Kincaid W, and Heinricher MM. Nociceptive facilitating neurons in the rostral ventromedial medulla. Pain, in press. Nichols ML, Bian D, Ossipov MH, Lai J, and Porreca F. Regulation of morphine antiallodynic efficacy by cholecystokinin in a model of neuropathic pain in rats. J Pharmacol Exp Ther 275: 1339-1345, 1995. Nichols ML, Bian D, Ossipov MH, Malan TP, Jr., and Porreca F. Antiallodynic effects of a CCKB antagonist in rats with nerve ligation injury: role of endogenous enkephalins. Neurosci Lett 215: 161-164, 1996. O'Neill M, Dourish CT, and Iversen SD. Morphine-induced analgesia in the rat paw pressure test is blocked by CCK and enhanced by the CCK antagonist MK-329. Neuropharmacology 28: 243-247, 1989. Ossipov MH, Lai J, Vanderah TW, and Porreca F. Induction of pain facilitation by sustained opioid exposure: relationship to opioid antinociceptive tolerance. Life Sci 73: 783-800, 2003. Pan ZZ, Williams JT, and Osborne PB. Opioid actions on single nucleus raphe magnus neurons from rat and guinea-pig in vitro. J Physiol 427: 519-532, 1990. Paxinos G and Watson C. The Rat Brain in Stereotaxic Coordinates. Sydney: Academic Press, 1997. Pertovaara A. A neuronal correlate of secondary hyperalgesia in the rat spinal dorsal horn is submodality selective and facilitated by supraspinal influence. Exp Neurol 149: 193-202, 1998. Porreca F, Burgess SE, Gardell LR, Vanderah TW, Malan TP, Jr., Ossipov MH, Lappi DA, and Lai J. Inhibition of neuropathic pain by selective ablation of brainstem medullary cells expressing the µ-opioid receptor. J Neurosci 21: 5281-5288., 2001.

JN-00411-2004.R1, p. 23

Porreca F, Ossipov MH, and Gebhart GF. Chronic pain and medullary descending facilitation. Trends Neurosci 25: 319-325., 2002. Pu SF, Zhuang HX, and Han JS. Cholecystokinin octapeptide (CCK-8) antagonizes morphine analgesia in nucleus accumbens of the rat via the CCK-B receptor. Brain Res 657: 159-164, 1994. Ramirez F and Vanegas H. Tooth pulp stimulation advances both medullary off-cell pause and tail flick. Neurosci Lett 100: 153-156, 1989. Simonnet G and Rivat C. Opioid-induced hyperalgesia: abnormal or normal pain? Neuroreport 14: 1-7, 2003. Stanfa LC and Dickenson AH. Cholecystokinin as a factor in the enhanced potency of spinal morphine following carrageenin inflammation. Br J Pharmacol 108: 967-973, 1993. Stanfa LC and Dickenson AH. Electrophysiological studies on the spinal roles of endogenous opioids in carrageenan inflammation. Pain 56: 185-191, 1994. Tortorici V, Nogueira L, Salas R, and Vanegas H. Involvement of local cholecystokinin in the tolerance induced by morphine microinjections into the periaqueductal gray of rats. Pain 102: 9-16, 2003. Urban MO and Gebhart GF. Supraspinal contributions to hyperalgesia. Proc Natl Acad Sci U S A 96: 7687-7692, 1999. Urban MO, Jiang MC, and Gebhart GF. Participation of central descending nociceptive facilitatory systems in secondary hyperalgesia produced by mustard oil. Brain Res 737: 8391, 1996.

JN-00411-2004.R1, p. 24

Vanderah TW, Ossipov MH, Lai J, Malan TP, and Porreca F. Mechanisms of opioid-induced pain and antinociceptive tolerance: descending facilitation and spinal dynorphin. Pain 92: 59., 2001a. Vanderah TW, Suenaga NM, Ossipov MH, Malan TP, Jr., Lai J, and Porreca F. Tonic Descending Facilitation from the Rostral Ventromedial Medulla Mediates Opioid-Induced Abnormal Pain and Antinociceptive Tolerance. J Neurosci 21: 279-286, 2001b. Watkins LR, Kinscheck IB, Kaufman EF, Miller J, Frenk H, and Mayer DJ. Cholecystokinin antagonists selectively potentiate analgesia induced by endogenous opiates. Brain Res 327: 181-190, 1985a. Watkins LR, Kinscheck IB, and Mayer DJ. Potentiation of morphine analgesia by the cholecystokinin antagonist proglumide. Brain Res 327: 169-180, 1985b. Watkins LR, Kinscheck IB, and Mayer DJ. Potentiation of opiate analgesia and apparent reversal of morphine tolerance by proglumide. Science 224: 395-396, 1984. Wiertelak EP, Maier SF, and Watkins LR. Cholecystokinin antianalgesia: safety cues abolish morphine analgesia. Science 256: 830-833, 1992. Wiertelak EP, Yang HY, Mooney-Heiberger K, Maier SF, and Watkins LR. The nature of conditioned anti-analgesia: spinal cord opiate and anti-opiate neurochemistry. Brain Res 634: 214-226, 1994. Wiesenfeld-Hallin Z, de Arauja Lucas G, Alster P, Xu XJ, and Hokfelt T. Cholecystokinin/opioid interactions. Brain Res 848: 78-89, 1999. Xu XJ, Hao JX, Seiger A, Hughes J, Hökfelt T, and Wiesenfeld-Hallin Z. Chronic pain-related behaviors in spinally injured rats: evidence for functional alterations of the endogenous cholecystokinin and opioid systems. Pain 56: 271-277, 1994.

JN-00411-2004.R1, p. 25

JN-00411-2004.R1, p. 26

Figure Legends Fig. 1.

Histologically verified locations of recording/infusion sites in which CCK (filled

circles) or vehicle (open circles) was infused in to the RVM. CCK microinjection sites outside of the RVM are indicated by open triangles. VII: facial nucleus, SO: superior olive. Distance from the interaural line is indicated. Fig. 2.

Microinjection of CCK (30 ng) within the RVM produces thermal hyperalgesia. CCK

(30 ng in 200 nl) produced a significant decrease in PW latency when microinjected into the RVM. Vehicle (200 nl) had no effect, nor did injections outside of the RVM (primarily dorsal and rostral, CCK, out). There was no difference among the groups in baseline latencies (ANOVA). (** p < 0.01 compared to baseline, t-test for correlated means). Neurons were successfully recorded in 26 of the 39 animals in the CCK-RVM group, 16 of the 17 animals in the vehicle group, and 12 of 18 animals in the CCK placement control group. Fig. 3

Ratemeter records illustrate the lack of effect of CCK on the ongoing discharge of two

off-cells in the RVM. In both cases, firing pattern and rate were completely unaffected by CCK. PW latencies decreased to 71% (top) and 77% (bottom) of these animals’ respective baselines. Triangles indicate PW trials. 1 s bins. Fig. 4.

Ratemeter records illustrate the lack of effect of CCK on the ongoing discharge of two

neutral cells in the RVM. As was the case with off-cells, firing pattern and rate were unaffected by CCK. PW latencies decreased to 39% (top) and 71% (bottom) of these animals’ respective baselines. Triangles indicate PW trials. 1 s bins. Fig. 5

Ratemeter records illustrate effects of 30 ng CCK on ongoing discharge of identified

on-cells in the RVM.

JN-00411-2004.R1, p. 27

Top: Dramatic activation of an on-cell following microinjection of CCK in the RVM This neuron had little spontaneous activity in baseline (silent 81.4% of the baseline period), and firing rate during active periods was 4.8 sp/s. The cell became continuously active after CCK, and mean firing rate over the 20-45 min post-CCK analysis period was 21.7 sp/s. PW latency decreased to 72% of baseline following CCK in this animal. Bottom: This neuron had substantial ongoing activity throughout baseline (silent only 5.8% of the baseline period), but firing rate was increased from 10.6 sp/s in baseline, to 14.9 sp/s following CCK. PW latency decreased to 56% of baseline following CCK in this animal Triangles indicate PW trials. 1 s bins. Fig. 6

Vehicle infusion had no effect on on-cell discharge.

Two on-cells recorded during infusion of saline vehicle in the RVM. Firing pattern and rate were unaltered, as was PW latency (103 and 109% of baseline latencies, in top and bottom record, respectively). Triangles indicate PW trials. 1 s bins. Fig. 7

Mean ongoing firing of on-cells, off-cells and neutral cells in baseline compared to the

post-injection period in animals microinjected with 30 ng CCK (left) or saline vehicle in the RVM. Only on-cells show a significant effect of CCK. Off-cell and neutral cell discharge is unchanged. Baseline activity was comparable between CCK and vehicle groups for each cell class. (* p < 0.05, discharge following CCK compared to that in baseline)

JN-00411-2004.R1, p. 28

Table legend Table 1. Pre- and post-CCK values for parameters describing ongoing and reflex-related firing of on-cells and the reflex-related off-cell pause. *p < 0.05, **p < 0.01. Baseline

Post-CCK

Proportion of sampling epoch that the cell was in an active period:

45.9 ± 9.9 %

81.1 ± 7.5 %**

Spontaneous firing rate during active periods:

7.4 ± 1.4 sp/s

16.8 ± 3.4 sp/s*

Firing rate at time of the withdrawal reflex:

13.3 ± 2.6 sp/s

24.8 ± 3.9 sp/s*

Baseline

Post-CCK

49.3 ± 22.7 s

57.4 ± 25.9 s

Latency from beginning of pause to paw withdrawal:

0.925 ± 0.162 s

0.798 ± 0.101 s

Latency from heat onset to beginning of pause:

2.199 ± 0.217 s

1.622 ± 0.220 s**

ON-CELL Parameters

OFF-CELL Pause Parameters Pause duration:

2.00 1.04

SO

2.30

VII

1.32

VII VII

2.60

1.52 VII VII

CCK in RVM, n = 39 Saline in RVM, n = 17 CCK placement control, n = 18

Baseline

PW Latency (s)

5

Post-Injection

4 3 2

**

1 0

CCK (RVM)

CCK (out)

Saline

Off-Cells

20

Firing Rate (sp/s)

10 0

CCK

3 min

30 20 10 0

CCK

3 min

Neutral Cells

20

Firing Rate (sp/s)

10 0 20

CCK

3 min

CCK

3 min

10

0

On-Cells

50

Firing Rate (sp/s)

25 0 50

CCK

3 min

25

0

CCK

3 min

50

On-Cells

Firing Rate (sp/s)

25 0 40

Sal

3 min

Sal

3 min

20

0

30 20

*

On-Cells

10

Firing Rate (sp/s)

0 20

Off-Cells

10 0 30

Neutral Cells

20 10 0

CCK

Saline

Baseline Post-Injection