Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Originally discovered as a gut protein, cholecystokinin (CCK) has been implicated in a wide variety of physiological functions, including the modulation of nociceptive transmission (Stanfa et al., 1994). Anatomical studies have revealed a striking overlap in the distribution of endogenous CCK and opioid peptides and receptors, particularly within areas involved in pain processing, such as the intralaminar nuclei of the thalamus, periaqueductal gray matter, and superficial lamina of the spinal cord (Gall et al., 1987). Such findings provide anatomical evidence for a functional relationship between the two neurotransmitter systems.

Results from behavioral studies suggest CCK has antiopioid actions that are pronociceptive (Faris et al., 1983; O'Neill et al., 1989). CCK attenuates opioid-induced antinociception (Kellstein et al., 1991), whereas CCK receptor antagonists enhance the antinociceptive effect of morphine and endogenous opioid peptides (Watkins et al., 1985a; Singh et al., 1996). Thus, CCK appears to modulate pain transmission by antagonizing opioid antinociception. Under physiological conditions, very little spinal CCK is released; however, previous studies suggest morphine accelerates the release of CCK (Zhou et al., 1993). In support, CCK receptor antagonists administered alone have no antinociceptive effect; CCK receptor antagonists produce antinociception only in the presence of opioids (Watkins et al., 1985b).

Most research on the effects of CCK in nociceptive transmission has focused on modulation of cutaneous pain. Yet visceral pain, particularly chronic visceral pain, is a major clinical concern, and mechanisms involved in pain transmission from cutaneous and deep structures are most likely distinct (Gebhart, 1995). Persistent pain of the viscera, such as arises from inflammatory bowel disease, is a challenging medical problem. It has been estimated that 50 to 100 of every 100,000 people in the United States suffer from an inflammatory bowel disease (de Dombal, 1986). The neural mechanisms influencing the perception of pain in such inflammatory conditions appear to differ from pain arising from healthy tissues. In animals with inflammatory conditions, the antinociceptive effect of morphine is reported to be enhanced by as much as 30-fold (Stanfa et al., 1992). CCK receptor antagonists were shown to have little to no effect on morphine antinociception in a model of cutaneous inflammation, yet CCK still attenuated morphine antinociception (Stanfa and Dickenson, 1993). These findings suggest that following inflammation, a decrease in CCK activity may allow an increase in antinociceptive potency of morphine (Ossipov et al., 1994).

The objective of the present study was to determine whether spinal CCK is involved in modulating nociceptive transmission from the viscera. We addressed this issue by examining the effect of CCK receptor antagonists on the potency of morphine in a model of visceral nociception, colorectal distension (CRD) (Ness and Gebhart, 1988). We repeated the experiments in a model of inflammatory bowel disease to determine the effects of chronic colonic inflammation on CCK receptor antagonist-morphine interactions. Finally, we confirmed which CCK receptor subtype, CCK_A or CCK_B, is involved in modulating nociceptive transmission from the viscera at the level of the spinal cord. Preliminary reports of some of these data have appeared in abstract form (Friedrich and Gebhart, 1997).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and Surgical Preparation. Experiments were conducted in male Sprague-Dawley rats (Harlan, Indianapolis, IN) weighing 400 to 425 g. All protocols were approved by The University of Iowa Animal Care and Use Committee. Rats were deeply anesthetized with an i.p. injection of sodium pentobarbital (50 mg/kg, Nembutal; Abbott Laboratories, Abbott Park, IL). A small incision was made in the atlanto-occipital membrane to allow insertion of an intrathecal (i.th.) catheter (PE-10 tubing) into the subarachnoid space of the spinal cord. The catheter was 8.5 cm in length and terminated at the level of the lumbosacral spinal cord, the termination site of colonic pelvic nerve afferent input. At the end of each experiment, Fast Green dye was injected into the catheter and postmortem examination of the spinal cord confirmed the location of the distal end of the catheter. Electromyographic (EMG) electrodes (Teflon-coated, 40-gauge stainless steel wires; Cooner Wire, Chatsworth, CA) were sutured into the external oblique musculature, just superior to the inguinal ligament. The electrode leads were tunneled s.c. and externalized with the i.th. catheter at the nape of the neck. Rats were allowed 4 to 7 days for recovery.

Experimental Protocol. A model of inflammatory bowel disease was established with intracolonic instillation of 2,4,6-trinitrobenzenesulfonic acid (TNBS) (30 mg/ml; Sigma Chemical Co., St. Louis, MO) in halothane-anesthetized rats (Morris et al., 1989; Duchmann et al., 1996). The TNBS was administered in a 50% ethanol vehicle (1.0 ml/rat); ethanol was used to break the mucosal barrier that normally protects the colon. Control rats received an equal volume of 50% ethanol/saline. Inflammation is maximal 3 to 5 days after TNBS instillation in this model (Morris et al., 1989), and all animals were tested at this time.

View larger version (47K): [in this window] [in a new window]   Fig. 1.   Typical VMR to staircase increases in CRD pressure from 0 to 80 mm Hg. The top panel represents the integrated EMG activity recorded from the external oblique musculature. The bottom panel represents intracolonic pressure. Shaded regions denote the first 5 s of each pressure step from which the responses to CRD were quantified.

Drugs. The ED₅₀ of morphine was determined with a cumulative dosing paradigm (Fig. 2): four staircase distensions were administered 0 (immediately after), 5, 10, and 15 min after i.th. morphine (dissolved in 5 µl of saline; followed by a 10-µl saline flush). The next dose of morphine was administered at 19 min followed by another four staircase distensions, and so forth. The ED₅₀ was defined as the dose that reduced the magnitude of the VMR to 50% of the predrug response. The approximate ED₅₀ dose of 1.0 µg of morphine was used in all subsequent experiments.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Dose-response relationship of cumulative doses of i.th. morphine. Data are illustrated as percentages of control (means ± S.E.).

Assessment of Inflammation. To quantify the extent of TNBS-induced inflammation, colonic myeloperoxidase (MPO) activity in TNBS-treated and ethanol/saline-treated rats was determined. MPO is an enzyme found predominantly in azurophilic granules of polymorphonuclear leukocytes (neutrophils). Neutrophil infiltration is a characteristic feature of inflammation, and greater MPO activity in tissue represents increased neutrophil infiltration into inflamed tissue (Krawisz et al., 1984; Mullane et al., 1985). The MPO assay was performed in distension-naive rats 4 days after intracolonic instillation of either TNBS or ethanol/saline. Approximately 15 mm of the distal colon was removed and weighed. The tissue was then suspended in 0.5% hexadecyltrimethylammonium bromide in 50 mM potassium phosphate buffer and homogenized for 10 min. The homogenate was freeze-thawed three times, centrifuged, and the MPO activity of the supernatant was determined. The supernatant was mixed with a dye solution containing o-dianisidine dihydrochloride and 0.0005% hydrogen peroxide. One unit of MPO activity was defined as that converting 1 µmol of hydrogen peroxide to water in 1 min. The change in absorbance at 460 nm was determined spectrophotometrically.

Data Analysis. All experiments were repeated in at least six animals, and all data are presented as means ± S.E. Data were analyzed with unpaired t tests and repeated-measures ANOVA (with GraphPad, Prism, and Excel computer programs), where applicable. P < .05 was considered statistically significant in all cases. The ED₅₀ dose of morphine was defined as the dose required to attenuate the VMR to 50% of the predrug response. The ED₅₀ dose (and 95% confidence interval) of morphine was calculated using a computer curve-fitting program (GraphPad, Prism). Area under the curve (AUC) was calculated with an Excel computer program.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of Intracolonic TNBS on CRD. Three to five days after instillation of intracolonic TNBS, the magnitude of the VMR to CRD was significantly greater in TNBS-treated rats than in vehicle-treated controls (F = 117.3, P < .05) (Fig. 3).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effect of TNBS-induced inflammation on the VMR to colorectal distension. The VMR to distending pressures from 0 to 80 mm Hg were compared in rats 3 to 5 days after instillation of intracolonic TNBS or ethanol/saline vehicle. Data are expressed as means ± S.E.

Effects of Morphine on CRD. In six naive, noninflamed rats, the ED₅₀ dose of morphine given i.th. was determined to be 0.93 µg (0.27-1.53, 95% CI) (Fig. 2). In subsequent experiments, an approximate ED₅₀ dose of 1.0 µg of morphine was used and found to have a significantly greater antinociceptive effect on TNBS-treated animals than on vehicle-treated controls. As shown in Fig. 4, the peak antinociceptive effect of morphine occurred ~40 min after i.th. morphine administration. Because the maximum VMR usually occurred in response to the 60 mm Hg stimulus, this pressure was chosen to represent the effects of morphine and CCK receptor antagonists on a noxious visceral stimulus. Morphine (1.0 µg) reduced the response to 60 mm Hg CRD to 62 ± 7.9% of the pre-ethanol/saline baseline response (Fig. 4A), and to 24 ± 4.8% of the pre-TNBS baseline response (Fig. 4B, P < .05). When expressed as AUC (Fig. 4C), the effect of 1.0 µg i.th. morphine on the VMR to 60 mm Hg distension was 2.5-fold greater in TNBS-treated rats (1609 versus 4060, P < .05).

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Effects of the nonselective CCK receptor antagonist proglumide (pro) on morphine (mor) antinociception. The VMR to 60 mm Hg CRD was determined in (A) ethanol/saline vehicle-treated rats or (B) TNBS-treated rats. The VMR to 60 mm Hg CRD before either TNBS or ethanol/saline vehicle was instilled into the colon was averaged and used as the 100% control (pre). The response to 60 mm Hg CRD was measured following i.th. mor alone or mor after i.th. pretreatment with pro 3 to 5 days after intracolonic TNBS or vehicle. The  represent the responses in either TNBS-treated or vehicle-treated rats following i.th. administration of vehicle (10% DMSO in saline). Data are expressed as percentages of control (means ± S.E.). C, AUC from (A) and (B) were determined for mor and mor plus pro in both ethanol/saline vehicle-treated rats and TNBS-treated rats.

Effects of Proglumide on Morphine Antinociception. Proglumide dose dependently increased the duration and maximum effect of morphine in vehicle-treated rats (Fig. 4A). In the absence of proglumide, the peak antinociceptive effect of morphine occurred at 40 min postinjection and lasted ~80 min. Pretreatment with i.th. proglumide enhanced morphine's antinociceptive effect in vehicle-treated but not TNBS-treated rats. Because time of peak effect, duration of effect, and magnitude of effect of morphine all changed in the presence of proglumide, an AUC was used for analysis of data (Fig. 4C). In vehicle-treated rats, pretreatment with proglumide (0.01, 0.1, or 1.0 µg) dose dependently enhanced the effect of 1.0 µg of morphine from 1609 ± 146.1 to 1784 ± 123.5 (P > .5), 2835 ± 352.3 (P < .05), and 3117 ± 402.6 (P < .05), respectively. No change in response to CRD was observed following administration of proglumide (not illustrated) or DMSO vehicle alone.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Effects of proglumide (pro), a nonselective CCK receptor antagonist, on the antinociceptive effects of the approximate ED30 dose of morphine (mor; 0.6 µg) in TNBS-treated rats. Intrathecal mor was administered alone or following i.th. proglumide (pro), and the VMR to 60 mm Hg CRD was determined. The VMR to 60 mm Hg CRD before TNBS was instilled into the colon was averaged and used as the 100% control. The  represent the responses in TNBS-treated rats following i.th. administration of 10% DMSO vehicle. Data are expressed as percentages of control (means ± S.E.).

Effects of Specific CCK_A and CCK_B Receptor Antagonists on Morphine Analgesia. To elucidate which CCK receptor is involved in modulating spinal opioid antinociception, CRD was performed in animals following administration of morphine after pretreatment with specific CCK_A or CCK_B receptor antagonists (L-364,718 and L-365,260, respectively). Administration of the specific CCK_A receptor antagonist did not enhance morphine antinociception. Figure 6 shows the VMR to 60 mm Hg pressure of distension in ethanol/saline vehicle-treated and TNBS-treated rats following morphine alone or morphine plus the CCK_A receptor antagonist (0.001, 0.01, and 1.0 µg). The antinociceptive effect of morphine (1.0 µg) was again significantly greater in TNBS-treated rats compared with vehicle-treated rats (P < .05). There was no significant effect of the CCK_A receptor antagonist on morphine-produced antinociception in vehicle-treated or TNBS-treated rats (Fig. 6C).

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Effects of the specific CCKA receptor antagonist L-364,718 (L-364) on morphine (mor) antinociception. Responses to 60 mm Hg CRD were determined in (A) ethanol/saline vehicle-treated rats or (B) TNBS-treated rats. The VMR to 60 mm Hg CRD before either TNBS or vehicle was instilled into the colon was averaged and used as the 100% control. The response to 60 mm Hg CRD was measured following i.th. mor alone or mor after i.th. pretreatment with L-364 3 to 5 days after intracolonic TNBS or vehicle. The unfilled circles represent the responses in either TNBS-treated or vehicle-treated rats following i.th. administration of 20% DMSO vehicle. Data are expressed as percentages of control (means ± S.E.). C, AUC from (A) and (B) were determined for mor and mor plus L-364 in both ethanol/saline vehicle-treated rats and TNBS-treated rats.

View larger version (31K): [in this window] [in a new window]   Fig. 7.   Effects of the CCKB receptor antagonist L-365,260 (L-365) on morphine (mor) antinociception. Responses to 60 mm Hg CRD were determined in (A) ethanol/saline vehicle-treated rats or (B) TNBS-treated rats. The VMR to 60 mm Hg CRD before either TNBS or vehicle was instilled into the colon was averaged and used as the 100% control. The response to 60 mm Hg CRD was measured following i.th. mor alone or mor after i.th. pretreatment with L-365 3 to 5 days after intracolonic TNBS or vehicle. The  represent the responses in either TNBS-treated or vehicle-treated rats following i.th. administration of 20% DMSO vehicle. Data are expressed as percentages of control (means ± S.E.). C, AUC from (A) and (B) were determined for mor and mor plus L-365 in both ethanol/saline vehicle-treated rats and TNBS-treated rats.

Assessment of TNBS-Induced Inflammation. TNBS-treated rats developed diarrhea after TNBS instillation, and colons from TNBS-treated rats appeared thickened and reddened. No differences were observed between ethanol/saline-treated rats and naive rats. To further assess the extent of inflammation caused by TNBS, colonic MPO activity was determined in both TNBS-treated and vehicle-treated rats. MPO activity in vehicle-treated rats was significantly less than MPO activity in TNBS-treated rats (0.01 ± 0.01 versus 25.3 ± 8.2 U MPO/g wet weight of colon, respectively, P < .05).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study demonstrates that spinal CCK is involved in modulating spinal opioid antinociception following a noxious visceral stimulus. Intrathecal CCK receptor antagonists, both nonspecific and the specific CCK_B receptor antagonist, significantly enhanced morphine antinociception in a dose-dependent manner. These results are consistent with studies on the effects of spinal CCK in models of cutaneous nociception (Watkins et al., 1985b; Stanfa and Dickenson, 1993). Proglumide was effective in the CRD model of visceral nociception in dose ranges similar to those reported to be effective in models of cutaneous nociception. Although mechanisms involved in pain modulation of visceral and cutaneous structures are most likely distinct, we have shown that spinal CCK modulates the transmission of noxious visceral input similarly to what has been documented for cutaneous stimulation.

The development of selective CCK receptor antagonists led to the classification of two CCK receptor subtypes, CCK_A and CCK_B. The CCK receptor antagonist L-364,718 (also known as devazepide) is 150 times more selective for the CCK_A receptor than the CCK_B receptor, whereas L-365,260 is ~50 times more selective for the CCK_B receptor (Hughes et al., 1990). CCK_A receptors are predominantly found in the periphery, whereas CCK_B receptors are principally located in the central nervous system (Moran et al., 1986; Wank et al., 1992). Previous studies suggest that interactions between CCK receptor antagonists and opioids are primarily mediated via CCK_B receptors (Valverde et al., 1994). For example, Vanderah et al. (1994) showed that pretreatment with i.c.v. CCK_B receptor antisense oligonucleotide in mice enhanced the antinociceptive potency of i.c.v. morphine. We report herein that the CCK_A receptor antagonist L-364,718 had no effect on the antinociceptive potency of morphine. However, the CCK_B receptor antagonist L-365,260 enhanced the antinociceptive effect of morphine in a manner similar to proglumide, the nonspecific CCK receptor antagonist, confirming that the CCK receptor associated with modulation of spinal visceral nociception is the CCK_B receptor.

The mechanism whereby CCK attenuates morphine antinociception is not known; however, two major theories predominate. Following CCK receptor activation, Ca²⁺ may be mobilized from intracellular stores and facilitate neurotransmitter release (Wang et al., 1992). This would oppose the inhibition of Ca²⁺ channels and concurrent reduction in transmitter release observed after opioid receptor activation (Piros et al., 1995). Alternatively, CCK may attenuate morphine antinociception by inhibiting the release and/or availability of endogenous enkephalins (Ossipov et al., 1994; Vanderah et al., 1996; Skinner et al., 1997), which act principally at the -opioid receptor. Because an additive interaction has been observed between µ- and -opioid receptor agonists (Barrett and Vaught, 1982), a decrease in availability of endogenous enkephalins following morphine administration and subsequent CCK release would lessen the overall opioid antinociceptive effect.

Neural mechanisms involved in nociceptive transmission appear to differ in healthy and inflamed tissue. Inflammation sensitizes primary afferent neurons from the skin, joints, and viscera such that they exhibit increased responses to noxious stimuli and/or are activated by lower intensities of stimulation (Kocher et al., 1987; Schaible et al., 1987; Schmelz et al., 1994; Sengupta and Gebhart, 1994). The present study assessed the effects of CCK receptor antagonists on morphine antinociception in a model of visceral hyperalgesia, produced by intracolonic instillation of TNBS. TNBS is a hapten and can elicit immunologic responses when coupled to a substance of high molecular weight. A 50% ethanol/saline vehicle was used to break the mucosal barrier of the colon, so TNBS could penetrate the colon wall, bind to tissue proteins, and elicit an immunologic response. This immune response produces inflammation, and triggers synthesis and/or release of substances that either directly stimulate (such as histamine and bradykinin) or indirectly sensitize (such as cytokines and prostaglandins) visceral afferent fibers (Bueno et al., 1997). VMRs to all intensities of CRD were significantly greater in TNBS-treated rats. Such visceral hyperalgesia is consistent with inflammation of the gastrointestinal tract (Sengupta et al., 1999). We also documented a significant increase in MPO activity in TNBS-treated colons compared with vehicle-treated controls.

Consistent with other reports (Stanfa et al., 1992; Kontinen et al., 1997), the antinociceptive potency of morphine was significantly enhanced in the presence of chronic colonic inflammation. With a carrageenan-induced model of cutaneous inflammation in the rat, Ossipov et al. (1995) reported a significant reduction in the ED₅₀ of morphine in carrageenan-treated rats (4.4 versus 7.0 mg/kg in untreated rats). The potency of morphine to inhibit evoked C-fiber activity has been reported to increase 30-fold in carrageenan-treated rats (Stanfa et al., 1992). Inflammation of the joint, as studied with an arthritic model in the rat, also leads to enhancement of the antinociceptive effects of morphine (Millan et al., 1987). With the CRD model of visceral pain, Sengupta et al. (1999) reported a significant leftward shift in the dose-response function of the -opioid receptor agonist EMD 61,753 in TNBS-treated rats. Herein, we report an enhanced effect of morphine in rats with chronically inflamed colons compared with noninflamed controls. All parameters assessed (time of maximum effect, duration of action, and magnitude of effect) were enhanced in TNBS-treated rats relative to vehicle-treated rats. Most previous studies of enhancement of opioid potency relate to peripheral sites of opioid action (Stein et al., 1999). In the present study, there was a significant enhancement of morphine antinociceptive potency in TNBS-treated rats following i.th. morphine administration, suggesting a central consequence of peripheral, colonic inflammation.

The increase in morphine potency observed following chronic visceral inflammation was not further enhanced by pretreatment with CCK receptor antagonists. This is consistent with previous studies in models of cutaneous inflammation, in which CCK receptor antagonists also do not further enhance morphine antinociception. CCK, however, still attenuates morphine antinociception in the presence of inflammation. Stanfa and Dickenson (1993) propose that this enhancement of spinal morphine potency during inflammation may be explained by a reduction in spinal CCK release by morphine. The source(s) of the spinal CCK released in response to spinal opioids is not known. However, preliminary results suggest that supraspinal sites are necessary for release of spinal CCK following i.th. morphine administration (Friedrich and Gebhart, 1998).

The physiological basis for release of endogenous CCK in response to opioids is unknown. Such endogenous antianalgesic mechanisms may function to return an animal to a basal state of pain responsivity after endogenous or exogenous opiate-induced analgesia. However, morphine tolerance may result from this compensatory increase in activity of endogenous CCK systems after repetitive opiate administration. Previous studies have shown that administration of CCK receptor antagonists in addition to morphine can enhance the antinociceptive effects of morphine as well as prevent morphine tolerance (Watkins et al., 1984; Idanpaan-Heikkila et al., 1997). A decrease in CCK release or availability following inflammation, however, may work to enhance the antinociceptive effect of opioids during a more chronic pain state.