- Email: [email protected]
Comparison of prostaglandin E1- and prostaglandin E2-induced hyperalgesia in the rat
Pergamon
0306-4522(94)EO167-3
Xeurmcience Vol. 62. No. 2, pp. 345-350, 1994 Elsevier Science Ltd Copyright Q 1994 IBRO Printed in Great Britain. All rights reserved 03064522/94 $7.00 + 0.00
COMPARISON OF PROSTAGLANDIN E,- AND PROSTAGLANDIN EJNDUCED HYPERALGESIA IN THE RAT S. G. KHASAR, T. HO, P. G, GREEN and J. D. LEVINE* Departments of Medicine, Anatomy and Oral and Maxillofacial Surgery, University of California at San Francisco, San Francisco, CA 94143-0452A.U.S.A. Abstract-We have studied prostaglandin E,-induced mechanical hyperalgesia in the rat hindpaw, by assessing paw-withdrawal thresholds, before and after injecting prostaglandin E, alone or with other agents, in normal and st~~ozot~in-induct diabetic rats. In normal and diabetic rats, pros~glandin E, (I-IO00 ng) produced a dose-dependent decmase in mechanical nociceptive threshold. In diabetic rats, prostaglandin E, was more potent than in normal rats, in producing hyperalgesia, whereas prostaglandin E, hyperalgesia was not changed in normal and diabetic rats. Prostaglandin E,-induced hyperalgesia was not inhibited by E-type 1 prostaglandin receptor antagonists, SC19220or X51089, either in normal or diabetic rats. In fact, in the presence of SCl9220, prostaglandin E, produced enhanced hyperalgesia. in normal rats. Prostaglandin E, hyperalgesia was not significantly modified by sympathectomy or indomethacin. Unlike prostaglandin E,, prostaglandin E, hyperalgesia was not blocked by the inhibitor of the stimulatory guanine nucI~otide-binding regulatory protein, guanosine S-~-(2-~i~iphosphate). It is suggested that prostagiandin E, decreases primary at&rent nociceprive threshoId directly, by activating a prostaglandin receptor other than the E-type I prostaglandin receptor, and that this receptor is not coupled lo a stimulatory guanine nucleotide-binding regulatory protein.
The release of inflammatory mediators following tissue injury results in the lowering of nociceptive threshold, such that stimuli which were previously not noxious become noxious, and noxious stimuli are perceived as more intense; this phenomenon has been termed hyperalgesia. Hyperalgesic mediators may act directly on primary a&rent nociceptors to produce hy~ralges~a; examples include prostagland~n E, (PGE:), prostacyclin (PGI,), adenosine, serotonin and 8R,l 5S-diHETE.2h,12,‘6~‘8~41 Others such as norepinephrine, bradykinin and leukotriene B, are believed to act indirectly, on neural or non-neural cells, to release prostaglandins which then act on primary alIerent te~inals, since their effects are eliminated by elimination of known indirect mechanisms?5,‘7.‘8.33”Y Among the agents believed to have a
*To whom correspondence should be addressed at: Division of presented Rheumatology, Department of Medicine, S-1334 (Box 0452A), University of California at San Francisco. San Francisco, CA 94143, U.S.A. Abbreviations: DAMGO, [n-Ala’,N-Me-Phe4,Gly5-ol]enkephahn; DHLA, dihomo-y-linolenic acid; EP,., mcep tor, E-type 1-3 prostaglandin receptor; GDP@, guanosine 5’-O-(2-thiodiphosphate); G-protein, guanine nucleotide-binding regulatory protein; PGE,, prostaglandin E,; PG$, prostaglandin E,: PGI,, prostacychn; 8R,l5S-diHET~, 8R,lS~dihydoxyicosatetroenoic acid; SC19220, l-acetyl-2-(8-ch~oro-~O,l1 -dihydrodi~nz lb,~[l,4]ox~epine-IO-carbonyl()hydr~ine; SC51089, 8chlorodibenz [b,fJ[ 1,4]oxazepine-lO(1I,4)-carboxylic acid; STZ, streptozotocin; STZ-D, streptozotocin-induced diabetes.
direct action on primary afferent nociceptors, PGE, has been the most extensively studied due, in part, to the fact that anti-inflammatory analgesic drugs used clinically block the synthetic pathway for prostaglandins. Hyperalgesia induced by prostagiandin E, (PGE,), an in~ammatory mediator of similar chemical structure as PGE,, has been generally assumed to be qualitatively similar to that of PCIE~,‘~.“~‘~but the specific mechanisms involved have not been elucidated. PGE, is derived from dihomo-y-linolenic acid (DHLA), which is widely distributed in mammalian tissues. This acid, like arachidonic acid, derives from hnolenic acid in the diet. In humans, synthesis of PGE, from endogenous DHLA has been reported in the stomach, skin, leukocytes, platelets, and other tissues.‘2~20~2’ Although PGE, itself has been shown to be devoid of algesic effect, small amounts strongly enhance the effects of other algesic substances such as bradykinin, substance P, and serotonin.13J0 It has also been shown to produce hy~ralgesia in manI and monkey.30 B-type prostaglandin receptors (EP receptors) have been classified into EP, , EP,, and EP, subtypes,?-11.22.23.29 Studies with an EP,-type receptor antagonist, SCl9220, suggest that PGE, hyperalgesia is mediated predominantly by the EP, receptor subtype. Thus, PGE2-induced hyperalgesia is significantly attenuated by SCl9220, at a dose which does not block PGI,-induced hyperalgesia.” The downstream mediator of PGE, hyperalgesia appears to be a
345
stimulatory guanine nucleotide-binding protein (Gprotein)x7 and the CAMP second messenger system.?‘,” In spite of this involvemen of the CAMP second messenger system in PGEz hyperalgesia and the suggested altered production of. or increased sensitivity of primary affercnt terminals to CAMP in streptozotocin-induced diabetic (STZ-D) rats, PGEz hyperalgesia is not altered in this model.’ Painful diabetic neuropathy IS a common complication in patients with diabetes mellitus.‘.h The mechanism of this condition is not well understood. Models of diabetes in the rat demonstrate a lower nociceptive threshold to both thermal and mechanical stimuli.‘“.‘4 The mechanism(s) underlying the decrease in nociceptive threshold are still not clear. Ahlgren and Levine’ have shown that alteration in the production of and increased sensitivity to CAMP may account, at least in part, for the lowered nociceptive threshold. It has also been observed that the use of synthetic analogs of PGE, (e.g., misoprostol and enisoprost) in the treatment of impotence is accompanied by pain, as a side-efl‘ect, in diabetx patients but not in normal ones (T. Lue, personal communication). It was therefore decided to study the effect of the naturally occurring PGE, in normal and in diabetic rats, with the view to identify the receptor(s) mediating PGE, hyperalgesia as well as differences and similarities between the hyperalgesia produced by PGE, and PGE!. EXPERIMENTAL
PROCEDURES
The experiments were performed on lightly restrained (250-350 g) male SpragueeDawley rats (Bantin and Kingman, Fremont, CA). The noclceptive flexion reflex was quantified with a Basile Analgesymeter ’ (Stoelting, Chicago), which applies a linearly increasing force to the dorsum of the rat’s hindpaw. During the week preceding the experiments, rats were trained in the paw-withdrawal reflex test at 5-min intervals for 2 h each day. This adaptation procedure produces a stable baseline threshold measurement and enhances the ability to detect the effect of agents which modulate hyperalgesic mechanisms.” On the day of the experiments the rats were exposed to the test stimulus at 5-min intervals for 2h. The baseline threshold is defined as the mean of the last six determinations before injection of the test agent. After measuring the baseline threshold, the test agents were injected intradermally. in volumes of 2.5 ~1, in each hindpaw. Thresholds for paw-withdrawal were then re-determined at 10, 15 and 20 min post injection. The mean of these consecutive readmgs. in each hindpaw, is the threshold at this dose of the test agent. The difference between this mean and the baseline is expressed as the percentage change from baseline threshold of these three consecutive determinations. Increasing doses of hvDeraleesic agents were injected at 2%min intervals. Whenantaionistswere used, thky were co-injected with the agonists. A sin& dose of PGE, (100 ng) was used in the ditermination o?onset of action by taking readings at 1-min intervals for 5 min and at 5.min intervals for 15 min after the injection. In order to ascertain that the PGE, effect was not produced indirectly by stimulating the release of PCiEz from sympathetic postganglionic neuronsI a group of rats that had been previously trained was surgically sympathec-
tomized under pentobarhital ‘mesthesta (50mgikg body wt). The sympathetic chain from L, L, was surgically removed by an extraperitoneal approach. from the left side. This procedure accomplishes virtually complete sympathetic denervation of the hindlimbs Two weeks after surgery. the rats were used in the study. After obtaining baseline thresholds, rats were pretreated with indomethacin (4mglkg body wt) to block prostaglandin synthesis. Dose--response relationship to PGE, was established as described earlier. After one week of training. as described above. diabetes mellitus was induced in some rats by a subcutaneous injection of the pancreatic fi-cell toxm. sh-eptozotocm (STZ; 70mg/kg body wt).” Control rats were Injected with an equal volume of saline vehicle. Onset of diabetes was assessed by the presence of glucosuria (urine glucose >SOO mgidl: Ames Ketostix. Miles Inc., Elkhart, IN). STZ-treated rats usually exceeded the urme glucose of > 500 me,‘dl limit within 24 h after iniection. The rats were, however:not used until 21 days afte; STZ injections. when a decrease in mechanical nociceptive threshold in the hmdpaws was clearly detectable and paw-withdrawal thresholds more stable.’ Baseline thresholds of saline-treated control rats were also determined. Unless otherwise stated, changes in nociceptive threshold m diabetic rats were calculated Hith respect to their baseline threshold at the time of the experiments. Agents used in this study were: PGE, (Biomol. Plymouth Meeting,, PA), PGE,, DAMGO. the uc-opioid recentor agonist-and stimulat& of inhibitory G-proteins, STZ. and euanosine 5’.0-(2-thiodiohosuhatel (GDPBS). the inhIbItor of stimulatory 6-protein (Sigma, St‘Louis, MO), SC19220 and SC51089, generous gift from G. D. Searle (Skokle. IL). PGE, (4 mg/ml) stock solution was made in ethanol. A fine suspension of SC19220 in saline was made using a sonicator, SCSI089 was dissolved in saline. while GDPBS was dissolved in distilled water. Injections of GDPgS were preceded by distilled water, to produce hypo-osmotic shock. This procedure enhances the permeability of cell membranes.’ and facilitates the entry of GDPBS, but does not produce hyperalgesia or Interfere with production of hyperalgesia.” In all cases, drug vehicles were tested for their effect alone on paw-withdrawal threshold.
RESULTS
Injected intradermally into the hindpaw of the rat, the onset of action of PGE, was rapid (Fig. 1A). In the dose range I-1000 ng, PGE, dose-dependently decreased the nociceptive threshold, but its vehicle (O.Ol-10% ethanol in saline) is of the same concentration as that contained in the various doses of PGE, , had no significant effect (Fig. 1B). In contrast to PGE,, which was antagonized by SC19220 (7.5 ng; Fig. lD), PGE, (IO-1000ng) hyperalgesia was significantly potentiuted in the presence of SC19220 (7.5 ng) (Fig. IB). In contrast, another EP, receptor antagonist, SC51089 (100 ng), which also antagonized PGE,-induced hyperdlgesia (data not shown) had no significant effect on PGE, (I-1000 ng) hyperalgesia (Fig. 1C). Sympathectomy with administration of indomethacin did not significantly modify PGE, hyperalgesia (Fig. LB). GDPgS (1 pg), an antagonist of stimulatory Gprotein (G,) did not significantly affect PGE, (100 ng) hyperalgesia in normal rats (PGE, = -41.96 2 3.31%, n = 8; GDPgS/PGE, = -42.67 + 2.63%. n = 8; P > 0.05). In contrast, GDPfiS inhibited PGE,
response by about 90% (PGE2 = - 35.28 k 2+34%, y1= 6; GDP@/PGE2 = -4.0 & 4.03%, n = 6; P < U.OOQ1,unpaired r-test). DAMCXI (1 j@, an
activator of inhibitory G-protein @3iiG,] did not significantly affect PGEl hyperalgesia in normal rats (PGE,= ---37,70&2.89%, n -9; ~~~~~~~~, = -39.60 f 2.45%, n = IO; P > 0.05, unpaired btest). Under similar conditions, DAMGO significantly inhibited PGE, h~~ral~esia (P6E, = -34.22 _I 0.92’S& n = 8; DAMG~~P~E~ = -8.33 & 3.67%, n = 8; P < B.BOI, unpaired t-test).
Baseline
paw-withdrawal
threshold
(77.20;t;
2.00 g, n = 37) of diabetic rats used in this study was
lower than that crf the pre-diabetic (112.31 + 1.63 g, sats Norman si&ite-treated n=42) Of baseline (108.9 1 + 2.45 & n = 42). Pre-diabetic thresholds and values from saline-treated normal rats were each significantly different from baselirre thresholds of diabetic rats (P -=zO.MM~)but not significantly d&rent from each other [P r 0.05). In diabetic rats, PCS& induced a d~s~-~~~nd~~t
hyperalgesia, significantly greater than that of normal
I
6
-10
--%
-5
0
5
IQ
15
20
Time (min) ccc)
PGE, dose (ng) Fig. 1. (A) Latency to ansetaf mechanical hy~ral~~~a afteran ~t~a~~r~l injection of K.iE, (I00 ng). maw-withdrawalthMxGi$ were taken at I-dn intervals for the first 5 min. Thereafter paw-withdrawal thresholds were taken at 5-min intervals for a total of 20 min. Each point is w mean f S.E.M&of at least six observations. (B) Dose-response relationship of intradermally injected PGE, alone @lid squares), in tie For aresence d the EP, antagonist, SC19220(7.5 np;)(solid diamonds), PGE, in ~~p~th~~~orni~ rats {solid inverted triangles), and RX, vchkte (solid circks), normal rais. Repeated measures analysis of variance @NOVA) showti a s&&ant difference V < &O#l~ between the groups. Fisher’s LSDposl h~c test &SW& significant differences between XX, and PGE, with SC19220(P < O,C@S). Unpaired ~-test of each dose teveakd significant difirerxes in the dose range ~~~1~ I@. Each pint is a mean I SJLM, af at least 18 observations. There was no significant difference between PGE, in normal and sympathectomited rats (P > 0.05)+(C) Dose--response relationship of intradermally injected PGE, alone (solid square) and in the presence of another EP, antagonist, SC51089(1 ~8) (open diamonds), in normal rats. Repeated measures ANQVA showed no significant difference {p > 0.05) between the two curves. Each point is a mean 2 S.E.M. ofat feast 12 observations.
(A)
52
2
x
z
PGE, dose (ng)
PGE, dose (ng)
Fig. 2. (A) Dose-response relationship of PGE, jn normal (filled squares), STZ-D (Med triangles), and diabetic rats treated with PGE, and SC19220 (7.5ng) (filled diamonds), Repeated measures ANOVA showed a significant difference between the groups (P < 0.0001). Fisher’s LSD post hnc test showed significant differences between PGE, in normal and diabetic rats (P <0.02). There was no significant difference between FGE, and PGE, with SC19220 (7~5ng) in diabetic rats (P > 0.05). Each point is a mean F S.E.M. of IB observations. (S) Dose-response rela~jonsh~pof PGE, in normaf (filled squares) and STZ-D rats @ilIed triangles). Repeated measures ANOVA showed no significant difference between the two groups (P > O.OSj. Each pint is the mean i S.E.M. of 10 abservations.
rats (Fig. 2A). This contrasts with PGE,-induced hyperalgesia, which is the same in normal and diabetic rats (Fig. 2B). PGE,-induced hyperalgesia was neither blocked nor enhanced by SC19220 in diabetic rats, unlike normal rats, where SC19220 enhances PGE,-induced hyperalgesia (Fig. 2B).
receptor mediating PGE, hyperalgesia may not be coupled to G proteins. In diabetic rats, PGE,-induced hyperalgesia is greater than in normal rats, unlike PGEz hyperalgesia, which does not differ in normal and diabetic rats.’ The inability of SC19220 to enhance PGE, hyperalgesia in diabetic rats may be due to a ceiling effect, DISCUSSiON since the response of diabetic rats to PGE, is already close to the maximal hyperalgesia detectable by The results of this study show that PGEI is a potent the testing method used. Thus, in normal and diabetic hy~~al~~s~~ agent, since when injected int~ad~~ally rats PGEt appears to act on an EP receptor subtype, into the rat hindpaw, it produces a dose-dependent which is not inhibited by the EP, antagonists SC19220 decrease in nociceptive threshold. The fact that and SC.51089. Unfortunately, there are currently no PGE, has a short latency to onset of action and its selective antagonists for either EP, or EP, receptor hyperalgesia is not inhibited by sympathectomy and subtypes, so we cannot yet assess whether PGE, proindomethacin suggest a direct effect on primary noduces hyperalgesia via an action at one of these recepciceptors. In this way PGE, is similar to PGE,, PGIl , tors. However, our evidence that PGE, and PGE, serotanin, 8R, I SS-diHETE and adenosine.?B,~~~3”,~~.~’ (both of which are said to be non-selective EP receptor However, unlike PGE,, PGE, hyperalgesia is not agonists) activate different second messenger systems antagonized by the selective EP, receptor antagonists, supports the hypothesis that PGE, acts on an EP SC192203’ or X51089,” suggesting that PGE, is actreceptor other than EP, sWhether the PGE, receptors ing on a receptor other than EP, . En the presence of in normal and diabetic rats are similar is, as yet, SC19220 there is, in fact, an enhancement of PGEl unknown. hyperalgesia. SC19220, administered alone, does not affect nociceptive threshold (data not shown). Since CONCWSION SC51089, another inhibitor of EP, receptors, does not enhance PGE, hyperalgesia, it seems that SC19220 might have actions other than its ability to antagonize EP, receptor, The nature of this effect(s) is currently unknown. The receptor involved in PGE,-induced hyperalgesia also does not appear to be coupled to G-proteins, since GDP,!% does not significantly affect PGE, hyperaigesia. In additon, whereas DAMGO, the p-opioid receptor agonist and an activator of inhibitory G-protein has been shown to inhibit PGE, hyperatgesia,37 it does not si~j~~ntly affect PGEl hyperalgesia, further supporting the view that the
PGEt and PC& both produce hyperalgesia when injected into the hindpaw of the rat, but the receptors with which they bind may be different. Similarly, the second messengers involved appear not to be the same. The sequence of events leading to production ofhyperalgesia by PGE, is still unclear, and further studies are needed. ~~~~#~~~~g~~Q~s-~e thank Dr Erik Kinnrnan for performing the surgical s~mpath~torn~. Supported by NIH grant NS21647 and G. D. Searle.
Comparison
of PGE, and PGE, hyperalgesia
349
REFERENCES
1. Ahlgren S. C., White D. M. and Levine J. D. (1992) Increased responsiveness of sensory neurons in the saphenous nerve of the streptozotocin-diabetic rat. J. Neurophysiol. 68, 2077-2085. rats. Neuroscience 52, 2. Ahlgren S. C. and Levine J. D. (1993) Mechanical hyperalgesia in streptozotocin-diabetic 1040-1055. 3. Archer A. G., Watkins P. J., Thomas P. K., Sharma A. K. and Payan J. (1983) The natural history of acute painful neuropathy in diabetes mellitus. J. Neural. Neurosurg. Psychiat. 46, 91-99. 4. Baron R., Jamg W. and Kollman N. (1988) Sympathetic afferent somata projecting in hindlimb nerves and the anatomical organization of lumbar sympathetic nervous system of the rat. J. camp. Neural. 275, 460468. prostaglandin formation from phosphatidyli5. Burch R. M. and Axelrod J. (1987) Dissociation of bradykinin-induced nositol turnover in Swiss 3T3 fibroblasts: evidence for G-protein regulation of phospholipase A,. Proc. natn. Acad. Sci. U.S.A. 84, 6374-6378. 6, Brown M. I. and Ashbury A. K. (1984) Diabettc neuropathy. Ann. Neural. 15, 2-12. development of a 1 Coleman R. A,, Humphrey P. P. A., Kennedy I. and Lumley P. (1984) Prostanoid receptors-the working classification. Trends phnrmnc. Sci. 5, 303-306. 8 Coleman R. A., Kennedy I. and Sheldrick R. L. G. (1985) AH6809, a prostanoid EP, receptor blocking drug. Br. J. Pharmac. 85, 273P. 9 Coleman R. A., Kennedy I. and Sheldrick R. L. G. (1987) Evidence for the existence of three subtypes of PGE, sensitive (EP) receptors in smooth muscle. Br. J. Pharmac. 91, 323P. 10 Coleman R. A., Kennedy I., Sheldrick R. L. G and Tolowinska I. Y. (1987) Further evidence for the existence of three subtypes of PGE,-sensitive (EP-) receptors. Br. J. Pharmac. 91, 407P. 11 Coleman R. A. (1988) Prostaglandin receptors. Prostaglandin Perspecr. 4, 33-35. 12 Dutta-Roy A. K., Kahn N. N. and Sinha A. K. (1989) Prostaglandin E,: the endogenous physiological regulator of platelet mediated blood coagulation. Prostagl. Leuko/r. Essent. Fatty Acids: Rev. 35, 189-195. 13 Ferreira S. H. (1972) Prostaglandins, aspirin-like drugs and analgesia. Narure 240, 200-203. 14 Ferreira S. H. (1981) Inflammatory pain, prostaglandin hyperalgesia and the development of peripheral analgesics. Trend pharmac. Sci. 2, 183-186. 15 Forman L. I., Estilow S., Lewis M. and Vasilenko P. (1988) Streptozotocin diabetes alters immunoreactive beta-endorphin levels and pain perception after 8 weeks in female rats, Diabetes 35, 1309-1313. 16 Gonzales R., Goldyne M. E., Taiwo Y. 0. and Levine J. D. (1989) Production of hyperalgesic prostaglandins by sympathetic postganglionic neurons. J. Neurochem. 53, 1595-l 598. 17 Green P .G., Luo I., Heller P. H. and Levine J. D. (1994) Effect of E-type prostaglandins on bradykinin-induced plasma extravasation in the knee joint of the rat. Eur. J. Pharmac. 252, 1277132. 18 Hallinan E. A., Hagen T. J.. Husa R. K., Tsymbalov S., Rao S. N., van Hoeck J.-P., Rafferty M. F., Stapelfeld A., Savage M. A. and Reichman M. (193) N-Substituted dibenzoxazepines as analgesic PGE, antagonists. J. med. Chem. 36, 3293-3299. 19. Johnson R. A., Yeung S.-M. H., Stubner D., Bushfield M. and Shoshani 1. (1989) Cation and structural requirements for P site-mediated inhibition of adenylate cyclase. Molec. Phnrmuc. 35, 681-688. 20. Juan H. (1978) Prostaglandins as modulators of pain. Gen. Phurmac. 9, 403409. 21. Kirkland S. J. (1988) Prostaglandin E,: a review. Prostugl. Leukotr. Essent. Fatty Acids: Rev. 32, 165-174. 22. Lawrence R. A., Jones R. L. and Wilson N. H. (1992) Characterization of receptors involved in the direct and indirect actions of prostaglandins E and I on the guinea-pig ileum. Br. J. Pharmac. 105, 271-278. 23. Lawrence R. A. and Jones R. L. (1992) Investigation of the prostaglandin E (EP-) receptor subtype mediating relaxation of the rabbit jugular vein. Br. J. Pharmac. 105, 817-824. 24. Lee J. H., Cox D. J., Mook D. J. and McCarty R. C. (1990) Effect of hyperglycemia on pain threshold in alloxan-diabetic rats. Pain 40, 105-107. 25. Lembeck F., Popper H. and Juan H. (1976) Release of prostaglandins by bradykinin as an intrinsic mechanism of its analgesic effect. Naunyn-Schmiedeberg’s Arch Pharmac. 294,69-73. 26. Levine J. D., Lam D., Taiwo Y. O., Donatoni P. and Goetzl E. J. (1986) Hyperalnesic properties of 15-lipoxvgenase - ._ products of arachidonic acid. Proc. nom Acad. Sci. U.S.A. 83, 5331-5334, ._ _ _ 27. Levine J. D., Taiwo Y. 0.. Collins S. D. and Tam J. K. (1986) Noradrenaline hvoeralaesia is mediated throueh interaction with sympathetic postganglionic neurone terminals. N&e 323, 1588166.’ 28. Levine J. D., Lau W., Kwiat G. and Goetzl E. J. (1984) Leukotriene B, produces hyperalgesia that is dependent on polymorphonuclear leukocytes. Science 225, 743-745. 29. Mantelli L., Amerini S., Rubino A. and Ledda F. (1991) Prejunctional prostanoid receptors on cardiac adrenergic terminals belong to the EP, subtype. Br. J. Pharmac. 102, 573-576. 30. Raja S. N., Mitzel E. L. and Campbell J. N. (1990) Prostaglandin E, sensitizes cutaneous nociceptors in the monkey. Sot. Neurosci. Abstr. 16, 415. 31. Sanner J. H. (1969) Antagonism of prostaglandin E2 by I-acetyl-2-(-8-chloro-IO,1 I-dihydro dihenz[b,fl[I,4]oxazepineIO-carbonyl) hydrazine (SC-19220). Arch. Int. Pharmacodyn. Ther. 180, 46-56. 32. Taiwo Y. O., Goetzl E. J. and Levine J. D. (1987) Hyperalgesia onset latency suggests a hierarchy of action. Brain Res. 423, 333-337. 33. Taiwo Y. 0. and Levine J. D. (1988) Characterization of the arachidonic acid metabolites mediating bradykinin and noradrenaline hyperalgesia. Brain Res. 458, 402-406. 34. Taiwo Y. O., Bierkness L. K.. Goetzl E. J. and Levine J. D. (989) Mediation of primary afferent neriuheral hvneralnesia by the CAMP second messenger system. Neuroscience 32, 577-580. _ . - _ __ 35. Taiwo Y. O., Coderre T. J. and Levine I. D. (1989) The contribution of trainina to sensitivitv in the nociceutive paw-withdrawal test. Brain Res. 487, 14&151. 36. Taiwo Y, 0. and Levine J. D. (1989) Prostaglandin effects after elimination of indirect hyperalgesic mechanisms in the skin of the rat. Brain Res. 492, 397-399. 37. Taiwo Y. 0. and Levine J. D. (1989) Contribution of guanine nucleotide regulatory proteins to prostaglandin hyperalgesia in the rat. Brain Res. 492, 40&403.
3511
S. G. Khasar
to NI.
38. Taiwo Y. 0. and Levme J. D. (1990) Direct cutaneous hyperalgesia induced ba adenosine. Neum.rci~~cr 38, 757 762. 39. Taiwo Y. 0.. Heller P. H. and Levine J. D. i 1990) Characterization of distinct uhosphohoases mediating _ bradvkinin . _ and noradrenaline hyperalgesia. Neuroscience 39, >23-53 I. 40. Taiwo Y. 0. and Levine .I. D. (1991) Further confirmation of the role of adenylyl cyclase and of CAMP-dependent protein kinase in primary afferent hyperalgesia. Neuroscience 44, 131-135. 41. Taiwo Y. 0. and Levine J. D. (1992) Serotinin is a directly-acting hyperalgesic agent in the rat. Neuroscience 48, 485490. 1
(Accepted
I I March 1994)
.
0306-4522(94)EO167-3
Xeurmcience Vol. 62. No. 2, pp. 345-350, 1994 Elsevier Science Ltd Copyright Q 1994 IBRO Printed in Great Britain. All rights reserved 03064522/94 $7.00 + 0.00
COMPARISON OF PROSTAGLANDIN E,- AND PROSTAGLANDIN EJNDUCED HYPERALGESIA IN THE RAT S. G. KHASAR, T. HO, P. G, GREEN and J. D. LEVINE* Departments of Medicine, Anatomy and Oral and Maxillofacial Surgery, University of California at San Francisco, San Francisco, CA 94143-0452A.U.S.A. Abstract-We have studied prostaglandin E,-induced mechanical hyperalgesia in the rat hindpaw, by assessing paw-withdrawal thresholds, before and after injecting prostaglandin E, alone or with other agents, in normal and st~~ozot~in-induct diabetic rats. In normal and diabetic rats, pros~glandin E, (I-IO00 ng) produced a dose-dependent decmase in mechanical nociceptive threshold. In diabetic rats, prostaglandin E, was more potent than in normal rats, in producing hyperalgesia, whereas prostaglandin E, hyperalgesia was not changed in normal and diabetic rats. Prostaglandin E,-induced hyperalgesia was not inhibited by E-type 1 prostaglandin receptor antagonists, SC19220or X51089, either in normal or diabetic rats. In fact, in the presence of SCl9220, prostaglandin E, produced enhanced hyperalgesia. in normal rats. Prostaglandin E, hyperalgesia was not significantly modified by sympathectomy or indomethacin. Unlike prostaglandin E,, prostaglandin E, hyperalgesia was not blocked by the inhibitor of the stimulatory guanine nucI~otide-binding regulatory protein, guanosine S-~-(2-~i~iphosphate). It is suggested that prostagiandin E, decreases primary at&rent nociceprive threshoId directly, by activating a prostaglandin receptor other than the E-type I prostaglandin receptor, and that this receptor is not coupled lo a stimulatory guanine nucleotide-binding regulatory protein.
The release of inflammatory mediators following tissue injury results in the lowering of nociceptive threshold, such that stimuli which were previously not noxious become noxious, and noxious stimuli are perceived as more intense; this phenomenon has been termed hyperalgesia. Hyperalgesic mediators may act directly on primary a&rent nociceptors to produce hy~ralges~a; examples include prostagland~n E, (PGE:), prostacyclin (PGI,), adenosine, serotonin and 8R,l 5S-diHETE.2h,12,‘6~‘8~41 Others such as norepinephrine, bradykinin and leukotriene B, are believed to act indirectly, on neural or non-neural cells, to release prostaglandins which then act on primary alIerent te~inals, since their effects are eliminated by elimination of known indirect mechanisms?5,‘7.‘8.33”Y Among the agents believed to have a
*To whom correspondence should be addressed at: Division of presented Rheumatology, Department of Medicine, S-1334 (Box 0452A), University of California at San Francisco. San Francisco, CA 94143, U.S.A. Abbreviations: DAMGO, [n-Ala’,N-Me-Phe4,Gly5-ol]enkephahn; DHLA, dihomo-y-linolenic acid; EP,., mcep tor, E-type 1-3 prostaglandin receptor; GDP@, guanosine 5’-O-(2-thiodiphosphate); G-protein, guanine nucleotide-binding regulatory protein; PGE,, prostaglandin E,; PG$, prostaglandin E,: PGI,, prostacychn; 8R,l5S-diHET~, 8R,lS~dihydoxyicosatetroenoic acid; SC19220, l-acetyl-2-(8-ch~oro-~O,l1 -dihydrodi~nz lb,~[l,4]ox~epine-IO-carbonyl()hydr~ine; SC51089, 8chlorodibenz [b,fJ[ 1,4]oxazepine-lO(1I,4)-carboxylic acid; STZ, streptozotocin; STZ-D, streptozotocin-induced diabetes.
direct action on primary afferent nociceptors, PGE, has been the most extensively studied due, in part, to the fact that anti-inflammatory analgesic drugs used clinically block the synthetic pathway for prostaglandins. Hyperalgesia induced by prostagiandin E, (PGE,), an in~ammatory mediator of similar chemical structure as PGE,, has been generally assumed to be qualitatively similar to that of PCIE~,‘~.“~‘~but the specific mechanisms involved have not been elucidated. PGE, is derived from dihomo-y-linolenic acid (DHLA), which is widely distributed in mammalian tissues. This acid, like arachidonic acid, derives from hnolenic acid in the diet. In humans, synthesis of PGE, from endogenous DHLA has been reported in the stomach, skin, leukocytes, platelets, and other tissues.‘2~20~2’ Although PGE, itself has been shown to be devoid of algesic effect, small amounts strongly enhance the effects of other algesic substances such as bradykinin, substance P, and serotonin.13J0 It has also been shown to produce hy~ralgesia in manI and monkey.30 B-type prostaglandin receptors (EP receptors) have been classified into EP, , EP,, and EP, subtypes,?-11.22.23.29 Studies with an EP,-type receptor antagonist, SCl9220, suggest that PGE, hyperalgesia is mediated predominantly by the EP, receptor subtype. Thus, PGE2-induced hyperalgesia is significantly attenuated by SCl9220, at a dose which does not block PGI,-induced hyperalgesia.” The downstream mediator of PGE, hyperalgesia appears to be a
345
stimulatory guanine nucleotide-binding protein (Gprotein)x7 and the CAMP second messenger system.?‘,” In spite of this involvemen of the CAMP second messenger system in PGEz hyperalgesia and the suggested altered production of. or increased sensitivity of primary affercnt terminals to CAMP in streptozotocin-induced diabetic (STZ-D) rats, PGEz hyperalgesia is not altered in this model.’ Painful diabetic neuropathy IS a common complication in patients with diabetes mellitus.‘.h The mechanism of this condition is not well understood. Models of diabetes in the rat demonstrate a lower nociceptive threshold to both thermal and mechanical stimuli.‘“.‘4 The mechanism(s) underlying the decrease in nociceptive threshold are still not clear. Ahlgren and Levine’ have shown that alteration in the production of and increased sensitivity to CAMP may account, at least in part, for the lowered nociceptive threshold. It has also been observed that the use of synthetic analogs of PGE, (e.g., misoprostol and enisoprost) in the treatment of impotence is accompanied by pain, as a side-efl‘ect, in diabetx patients but not in normal ones (T. Lue, personal communication). It was therefore decided to study the effect of the naturally occurring PGE, in normal and in diabetic rats, with the view to identify the receptor(s) mediating PGE, hyperalgesia as well as differences and similarities between the hyperalgesia produced by PGE, and PGE!. EXPERIMENTAL
PROCEDURES
The experiments were performed on lightly restrained (250-350 g) male SpragueeDawley rats (Bantin and Kingman, Fremont, CA). The noclceptive flexion reflex was quantified with a Basile Analgesymeter ’ (Stoelting, Chicago), which applies a linearly increasing force to the dorsum of the rat’s hindpaw. During the week preceding the experiments, rats were trained in the paw-withdrawal reflex test at 5-min intervals for 2 h each day. This adaptation procedure produces a stable baseline threshold measurement and enhances the ability to detect the effect of agents which modulate hyperalgesic mechanisms.” On the day of the experiments the rats were exposed to the test stimulus at 5-min intervals for 2h. The baseline threshold is defined as the mean of the last six determinations before injection of the test agent. After measuring the baseline threshold, the test agents were injected intradermally. in volumes of 2.5 ~1, in each hindpaw. Thresholds for paw-withdrawal were then re-determined at 10, 15 and 20 min post injection. The mean of these consecutive readmgs. in each hindpaw, is the threshold at this dose of the test agent. The difference between this mean and the baseline is expressed as the percentage change from baseline threshold of these three consecutive determinations. Increasing doses of hvDeraleesic agents were injected at 2%min intervals. Whenantaionistswere used, thky were co-injected with the agonists. A sin& dose of PGE, (100 ng) was used in the ditermination o?onset of action by taking readings at 1-min intervals for 5 min and at 5.min intervals for 15 min after the injection. In order to ascertain that the PGE, effect was not produced indirectly by stimulating the release of PCiEz from sympathetic postganglionic neuronsI a group of rats that had been previously trained was surgically sympathec-
tomized under pentobarhital ‘mesthesta (50mgikg body wt). The sympathetic chain from L, L, was surgically removed by an extraperitoneal approach. from the left side. This procedure accomplishes virtually complete sympathetic denervation of the hindlimbs Two weeks after surgery. the rats were used in the study. After obtaining baseline thresholds, rats were pretreated with indomethacin (4mglkg body wt) to block prostaglandin synthesis. Dose--response relationship to PGE, was established as described earlier. After one week of training. as described above. diabetes mellitus was induced in some rats by a subcutaneous injection of the pancreatic fi-cell toxm. sh-eptozotocm (STZ; 70mg/kg body wt).” Control rats were Injected with an equal volume of saline vehicle. Onset of diabetes was assessed by the presence of glucosuria (urine glucose >SOO mgidl: Ames Ketostix. Miles Inc., Elkhart, IN). STZ-treated rats usually exceeded the urme glucose of > 500 me,‘dl limit within 24 h after iniection. The rats were, however:not used until 21 days afte; STZ injections. when a decrease in mechanical nociceptive threshold in the hmdpaws was clearly detectable and paw-withdrawal thresholds more stable.’ Baseline thresholds of saline-treated control rats were also determined. Unless otherwise stated, changes in nociceptive threshold m diabetic rats were calculated Hith respect to their baseline threshold at the time of the experiments. Agents used in this study were: PGE, (Biomol. Plymouth Meeting,, PA), PGE,, DAMGO. the uc-opioid recentor agonist-and stimulat& of inhibitory G-proteins, STZ. and euanosine 5’.0-(2-thiodiohosuhatel (GDPBS). the inhIbItor of stimulatory 6-protein (Sigma, St‘Louis, MO), SC19220 and SC51089, generous gift from G. D. Searle (Skokle. IL). PGE, (4 mg/ml) stock solution was made in ethanol. A fine suspension of SC19220 in saline was made using a sonicator, SCSI089 was dissolved in saline. while GDPBS was dissolved in distilled water. Injections of GDPgS were preceded by distilled water, to produce hypo-osmotic shock. This procedure enhances the permeability of cell membranes.’ and facilitates the entry of GDPBS, but does not produce hyperalgesia or Interfere with production of hyperalgesia.” In all cases, drug vehicles were tested for their effect alone on paw-withdrawal threshold.
RESULTS
Injected intradermally into the hindpaw of the rat, the onset of action of PGE, was rapid (Fig. 1A). In the dose range I-1000 ng, PGE, dose-dependently decreased the nociceptive threshold, but its vehicle (O.Ol-10% ethanol in saline) is of the same concentration as that contained in the various doses of PGE, , had no significant effect (Fig. 1B). In contrast to PGE,, which was antagonized by SC19220 (7.5 ng; Fig. lD), PGE, (IO-1000ng) hyperalgesia was significantly potentiuted in the presence of SC19220 (7.5 ng) (Fig. IB). In contrast, another EP, receptor antagonist, SC51089 (100 ng), which also antagonized PGE,-induced hyperdlgesia (data not shown) had no significant effect on PGE, (I-1000 ng) hyperalgesia (Fig. 1C). Sympathectomy with administration of indomethacin did not significantly modify PGE, hyperalgesia (Fig. LB). GDPgS (1 pg), an antagonist of stimulatory Gprotein (G,) did not significantly affect PGE, (100 ng) hyperalgesia in normal rats (PGE, = -41.96 2 3.31%, n = 8; GDPgS/PGE, = -42.67 + 2.63%. n = 8; P > 0.05). In contrast, GDPfiS inhibited PGE,
response by about 90% (PGE2 = - 35.28 k 2+34%, y1= 6; GDP@/PGE2 = -4.0 & 4.03%, n = 6; P < U.OOQ1,unpaired r-test). DAMCXI (1 j@, an
activator of inhibitory G-protein @3iiG,] did not significantly affect PGEl hyperalgesia in normal rats (PGE,= ---37,70&2.89%, n -9; ~~~~~~~~, = -39.60 f 2.45%, n = IO; P > 0.05, unpaired btest). Under similar conditions, DAMGO significantly inhibited PGE, h~~ral~esia (P6E, = -34.22 _I 0.92’S& n = 8; DAMG~~P~E~ = -8.33 & 3.67%, n = 8; P < B.BOI, unpaired t-test).
Baseline
paw-withdrawal
threshold
(77.20;t;
2.00 g, n = 37) of diabetic rats used in this study was
lower than that crf the pre-diabetic (112.31 + 1.63 g, sats Norman si&ite-treated n=42) Of baseline (108.9 1 + 2.45 & n = 42). Pre-diabetic thresholds and values from saline-treated normal rats were each significantly different from baselirre thresholds of diabetic rats (P -=zO.MM~)but not significantly d&rent from each other [P r 0.05). In diabetic rats, PCS& induced a d~s~-~~~nd~~t
hyperalgesia, significantly greater than that of normal
I
6
-10
--%
-5
0
5
IQ
15
20
Time (min) ccc)
PGE, dose (ng) Fig. 1. (A) Latency to ansetaf mechanical hy~ral~~~a afteran ~t~a~~r~l injection of K.iE, (I00 ng). maw-withdrawalthMxGi$ were taken at I-dn intervals for the first 5 min. Thereafter paw-withdrawal thresholds were taken at 5-min intervals for a total of 20 min. Each point is w mean f S.E.M&of at least six observations. (B) Dose-response relationship of intradermally injected PGE, alone @lid squares), in tie For aresence d the EP, antagonist, SC19220(7.5 np;)(solid diamonds), PGE, in ~~p~th~~~orni~ rats {solid inverted triangles), and RX, vchkte (solid circks), normal rais. Repeated measures analysis of variance @NOVA) showti a s&&ant difference V < &O#l~ between the groups. Fisher’s LSDposl h~c test &SW& significant differences between XX, and PGE, with SC19220(P < O,C@S). Unpaired ~-test of each dose teveakd significant difirerxes in the dose range ~~~1~ I@. Each pint is a mean I SJLM, af at least 18 observations. There was no significant difference between PGE, in normal and sympathectomited rats (P > 0.05)+(C) Dose--response relationship of intradermally injected PGE, alone (solid square) and in the presence of another EP, antagonist, SC51089(1 ~8) (open diamonds), in normal rats. Repeated measures ANQVA showed no significant difference {p > 0.05) between the two curves. Each point is a mean 2 S.E.M. ofat feast 12 observations.
(A)
52
2
x
z
PGE, dose (ng)
PGE, dose (ng)
Fig. 2. (A) Dose-response relationship of PGE, jn normal (filled squares), STZ-D (Med triangles), and diabetic rats treated with PGE, and SC19220 (7.5ng) (filled diamonds), Repeated measures ANOVA showed a significant difference between the groups (P < 0.0001). Fisher’s LSD post hnc test showed significant differences between PGE, in normal and diabetic rats (P <0.02). There was no significant difference between FGE, and PGE, with SC19220 (7~5ng) in diabetic rats (P > 0.05). Each point is a mean F S.E.M. of IB observations. (S) Dose-response rela~jonsh~pof PGE, in normaf (filled squares) and STZ-D rats @ilIed triangles). Repeated measures ANOVA showed no significant difference between the two groups (P > O.OSj. Each pint is the mean i S.E.M. of 10 abservations.
rats (Fig. 2A). This contrasts with PGE,-induced hyperalgesia, which is the same in normal and diabetic rats (Fig. 2B). PGE,-induced hyperalgesia was neither blocked nor enhanced by SC19220 in diabetic rats, unlike normal rats, where SC19220 enhances PGE,-induced hyperalgesia (Fig. 2B).
receptor mediating PGE, hyperalgesia may not be coupled to G proteins. In diabetic rats, PGE,-induced hyperalgesia is greater than in normal rats, unlike PGEz hyperalgesia, which does not differ in normal and diabetic rats.’ The inability of SC19220 to enhance PGE, hyperalgesia in diabetic rats may be due to a ceiling effect, DISCUSSiON since the response of diabetic rats to PGE, is already close to the maximal hyperalgesia detectable by The results of this study show that PGEI is a potent the testing method used. Thus, in normal and diabetic hy~~al~~s~~ agent, since when injected int~ad~~ally rats PGEt appears to act on an EP receptor subtype, into the rat hindpaw, it produces a dose-dependent which is not inhibited by the EP, antagonists SC19220 decrease in nociceptive threshold. The fact that and SC.51089. Unfortunately, there are currently no PGE, has a short latency to onset of action and its selective antagonists for either EP, or EP, receptor hyperalgesia is not inhibited by sympathectomy and subtypes, so we cannot yet assess whether PGE, proindomethacin suggest a direct effect on primary noduces hyperalgesia via an action at one of these recepciceptors. In this way PGE, is similar to PGE,, PGIl , tors. However, our evidence that PGE, and PGE, serotanin, 8R, I SS-diHETE and adenosine.?B,~~~3”,~~.~’ (both of which are said to be non-selective EP receptor However, unlike PGE,, PGE, hyperalgesia is not agonists) activate different second messenger systems antagonized by the selective EP, receptor antagonists, supports the hypothesis that PGE, acts on an EP SC192203’ or X51089,” suggesting that PGE, is actreceptor other than EP, sWhether the PGE, receptors ing on a receptor other than EP, . En the presence of in normal and diabetic rats are similar is, as yet, SC19220 there is, in fact, an enhancement of PGEl unknown. hyperalgesia. SC19220, administered alone, does not affect nociceptive threshold (data not shown). Since CONCWSION SC51089, another inhibitor of EP, receptors, does not enhance PGE, hyperalgesia, it seems that SC19220 might have actions other than its ability to antagonize EP, receptor, The nature of this effect(s) is currently unknown. The receptor involved in PGE,-induced hyperalgesia also does not appear to be coupled to G-proteins, since GDP,!% does not significantly affect PGE, hyperaigesia. In additon, whereas DAMGO, the p-opioid receptor agonist and an activator of inhibitory G-protein has been shown to inhibit PGE, hyperatgesia,37 it does not si~j~~ntly affect PGEl hyperalgesia, further supporting the view that the
PGEt and PC& both produce hyperalgesia when injected into the hindpaw of the rat, but the receptors with which they bind may be different. Similarly, the second messengers involved appear not to be the same. The sequence of events leading to production ofhyperalgesia by PGE, is still unclear, and further studies are needed. ~~~~#~~~~g~~Q~s-~e thank Dr Erik Kinnrnan for performing the surgical s~mpath~torn~. Supported by NIH grant NS21647 and G. D. Searle.
Comparison
of PGE, and PGE, hyperalgesia
349
REFERENCES
1. Ahlgren S. C., White D. M. and Levine J. D. (1992) Increased responsiveness of sensory neurons in the saphenous nerve of the streptozotocin-diabetic rat. J. Neurophysiol. 68, 2077-2085. rats. Neuroscience 52, 2. Ahlgren S. C. and Levine J. D. (1993) Mechanical hyperalgesia in streptozotocin-diabetic 1040-1055. 3. Archer A. G., Watkins P. J., Thomas P. K., Sharma A. K. and Payan J. (1983) The natural history of acute painful neuropathy in diabetes mellitus. J. Neural. Neurosurg. Psychiat. 46, 91-99. 4. Baron R., Jamg W. and Kollman N. (1988) Sympathetic afferent somata projecting in hindlimb nerves and the anatomical organization of lumbar sympathetic nervous system of the rat. J. camp. Neural. 275, 460468. prostaglandin formation from phosphatidyli5. Burch R. M. and Axelrod J. (1987) Dissociation of bradykinin-induced nositol turnover in Swiss 3T3 fibroblasts: evidence for G-protein regulation of phospholipase A,. Proc. natn. Acad. Sci. U.S.A. 84, 6374-6378. 6, Brown M. I. and Ashbury A. K. (1984) Diabettc neuropathy. Ann. Neural. 15, 2-12. development of a 1 Coleman R. A,, Humphrey P. P. A., Kennedy I. and Lumley P. (1984) Prostanoid receptors-the working classification. Trends phnrmnc. Sci. 5, 303-306. 8 Coleman R. A., Kennedy I. and Sheldrick R. L. G. (1985) AH6809, a prostanoid EP, receptor blocking drug. Br. J. Pharmac. 85, 273P. 9 Coleman R. A., Kennedy I. and Sheldrick R. L. G. (1987) Evidence for the existence of three subtypes of PGE, sensitive (EP) receptors in smooth muscle. Br. J. Pharmac. 91, 323P. 10 Coleman R. A., Kennedy I., Sheldrick R. L. G and Tolowinska I. Y. (1987) Further evidence for the existence of three subtypes of PGE,-sensitive (EP-) receptors. Br. J. Pharmac. 91, 407P. 11 Coleman R. A. (1988) Prostaglandin receptors. Prostaglandin Perspecr. 4, 33-35. 12 Dutta-Roy A. K., Kahn N. N. and Sinha A. K. (1989) Prostaglandin E,: the endogenous physiological regulator of platelet mediated blood coagulation. Prostagl. Leuko/r. Essent. Fatty Acids: Rev. 35, 189-195. 13 Ferreira S. H. (1972) Prostaglandins, aspirin-like drugs and analgesia. Narure 240, 200-203. 14 Ferreira S. H. (1981) Inflammatory pain, prostaglandin hyperalgesia and the development of peripheral analgesics. Trend pharmac. Sci. 2, 183-186. 15 Forman L. I., Estilow S., Lewis M. and Vasilenko P. (1988) Streptozotocin diabetes alters immunoreactive beta-endorphin levels and pain perception after 8 weeks in female rats, Diabetes 35, 1309-1313. 16 Gonzales R., Goldyne M. E., Taiwo Y. 0. and Levine J. D. (1989) Production of hyperalgesic prostaglandins by sympathetic postganglionic neurons. J. Neurochem. 53, 1595-l 598. 17 Green P .G., Luo I., Heller P. H. and Levine J. D. (1994) Effect of E-type prostaglandins on bradykinin-induced plasma extravasation in the knee joint of the rat. Eur. J. Pharmac. 252, 1277132. 18 Hallinan E. A., Hagen T. J.. Husa R. K., Tsymbalov S., Rao S. N., van Hoeck J.-P., Rafferty M. F., Stapelfeld A., Savage M. A. and Reichman M. (193) N-Substituted dibenzoxazepines as analgesic PGE, antagonists. J. med. Chem. 36, 3293-3299. 19. Johnson R. A., Yeung S.-M. H., Stubner D., Bushfield M. and Shoshani 1. (1989) Cation and structural requirements for P site-mediated inhibition of adenylate cyclase. Molec. Phnrmuc. 35, 681-688. 20. Juan H. (1978) Prostaglandins as modulators of pain. Gen. Phurmac. 9, 403409. 21. Kirkland S. J. (1988) Prostaglandin E,: a review. Prostugl. Leukotr. Essent. Fatty Acids: Rev. 32, 165-174. 22. Lawrence R. A., Jones R. L. and Wilson N. H. (1992) Characterization of receptors involved in the direct and indirect actions of prostaglandins E and I on the guinea-pig ileum. Br. J. Pharmac. 105, 271-278. 23. Lawrence R. A. and Jones R. L. (1992) Investigation of the prostaglandin E (EP-) receptor subtype mediating relaxation of the rabbit jugular vein. Br. J. Pharmac. 105, 817-824. 24. Lee J. H., Cox D. J., Mook D. J. and McCarty R. C. (1990) Effect of hyperglycemia on pain threshold in alloxan-diabetic rats. Pain 40, 105-107. 25. Lembeck F., Popper H. and Juan H. (1976) Release of prostaglandins by bradykinin as an intrinsic mechanism of its analgesic effect. Naunyn-Schmiedeberg’s Arch Pharmac. 294,69-73. 26. Levine J. D., Lam D., Taiwo Y. O., Donatoni P. and Goetzl E. J. (1986) Hyperalnesic properties of 15-lipoxvgenase - ._ products of arachidonic acid. Proc. nom Acad. Sci. U.S.A. 83, 5331-5334, ._ _ _ 27. Levine J. D., Taiwo Y. 0.. Collins S. D. and Tam J. K. (1986) Noradrenaline hvoeralaesia is mediated throueh interaction with sympathetic postganglionic neurone terminals. N&e 323, 1588166.’ 28. Levine J. D., Lau W., Kwiat G. and Goetzl E. J. (1984) Leukotriene B, produces hyperalgesia that is dependent on polymorphonuclear leukocytes. Science 225, 743-745. 29. Mantelli L., Amerini S., Rubino A. and Ledda F. (1991) Prejunctional prostanoid receptors on cardiac adrenergic terminals belong to the EP, subtype. Br. J. Pharmac. 102, 573-576. 30. Raja S. N., Mitzel E. L. and Campbell J. N. (1990) Prostaglandin E, sensitizes cutaneous nociceptors in the monkey. Sot. Neurosci. Abstr. 16, 415. 31. Sanner J. H. (1969) Antagonism of prostaglandin E2 by I-acetyl-2-(-8-chloro-IO,1 I-dihydro dihenz[b,fl[I,4]oxazepineIO-carbonyl) hydrazine (SC-19220). Arch. Int. Pharmacodyn. Ther. 180, 46-56. 32. Taiwo Y. O., Goetzl E. J. and Levine J. D. (1987) Hyperalgesia onset latency suggests a hierarchy of action. Brain Res. 423, 333-337. 33. Taiwo Y. 0. and Levine J. D. (1988) Characterization of the arachidonic acid metabolites mediating bradykinin and noradrenaline hyperalgesia. Brain Res. 458, 402-406. 34. Taiwo Y. O., Bierkness L. K.. Goetzl E. J. and Levine J. D. (989) Mediation of primary afferent neriuheral hvneralnesia by the CAMP second messenger system. Neuroscience 32, 577-580. _ . - _ __ 35. Taiwo Y. O., Coderre T. J. and Levine I. D. (1989) The contribution of trainina to sensitivitv in the nociceutive paw-withdrawal test. Brain Res. 487, 14&151. 36. Taiwo Y, 0. and Levine J. D. (1989) Prostaglandin effects after elimination of indirect hyperalgesic mechanisms in the skin of the rat. Brain Res. 492, 397-399. 37. Taiwo Y. 0. and Levine J. D. (1989) Contribution of guanine nucleotide regulatory proteins to prostaglandin hyperalgesia in the rat. Brain Res. 492, 40&403.
3511
S. G. Khasar
to NI.
38. Taiwo Y. 0. and Levme J. D. (1990) Direct cutaneous hyperalgesia induced ba adenosine. Neum.rci~~cr 38, 757 762. 39. Taiwo Y. 0.. Heller P. H. and Levine J. D. i 1990) Characterization of distinct uhosphohoases mediating _ bradvkinin . _ and noradrenaline hyperalgesia. Neuroscience 39, >23-53 I. 40. Taiwo Y. 0. and Levine .I. D. (1991) Further confirmation of the role of adenylyl cyclase and of CAMP-dependent protein kinase in primary afferent hyperalgesia. Neuroscience 44, 131-135. 41. Taiwo Y. 0. and Levine J. D. (1992) Serotinin is a directly-acting hyperalgesic agent in the rat. Neuroscience 48, 485490. 1
(Accepted
I I March 1994)
.