Pain Reviews 2001; 8: 113–119

Anti-in ammatory effects of k -opioids: relevance to rheumatoid arthritis
Judith S Walker
School of Physiology and Pharmacology, University of New South Wales, Sydney, Australia

Therapy with opioids is an exciting new development for arthritis, especially since there is the potential for fewer side effects from molecules that act outside the central nervous system. We have found k-opioid drugs to be powerfully anti-in ammatory, reducing disease severity by as much as 80% and attenuating arthritis in a dose-dependent, stereoselective, antagonist-reversible manner. In contrast, opioids acting at other receptors were therapeutic only at near toxic doses. Currently, however, no pure k-opioids are available for clinical use. The hypothalamic-pituitary-adrenal axis was found to be only partially involved; thus we investigated other neural and immune mechanisms. The results showed that the k-opioid anti-in ammatory actions were exerted via: (1) reduced adhesion molecule expression; (2) inhibition of cell trafcking; (3) reduced tumour necrosis factor release; and (4) alterations in mRNA expression and substance P (SP) and calcitonin gene-related peptide protein in joint tissue. The ability of k-opioids to act at multiple sites in the in ammatory cascade, as suggested by the presence of opioid receptors at various locations throughout the cascade, may explain their powerful actions. k-Opioids are, however, most therapeutic during disease onset; thus it is likely they exert their anti-in ammatory effects predominantly via changes in cellular activation and cytokine expression rather than via the nervous system. The involvement of SP and the ef cacy of neurokinin 1 (NK1) antagonists predicts that combined opioid–NK1 regimens have therapeutic promise. Peripherally acting opioids may prove to be a potent new treatment for rheumatoid arthritis sufferers in the future.

Introduction
Opioid drugs are not currently used in the treatment of rheumatoid arthritis, partly because of their range of side effects and because their antiin ammatory (as opposed to analgesic) actions have been largely unrecognized. The synthesis of peripherally selective k-opioid agonists has allowed the analgesic and anti-in ammatory

Address for correspondence: Judith S Walker, School of Physiology and Pharmacology, University of New South Wales, Sydney NSW 2052, Australia. E-mail: JudithW@medscape.com © Arnold 2001

effects of opioids in arthritis to be studied, while mitigating the problems of tolerance and central side effects. They are powerfully anti-in ammatory in a dose-dependent, time-dependent, stereoselective and antagonist reversible manner.1 This brief report examines the anti-in ammatory effects of k-opioids, both centrally active and peripherally selective k-opioid agonists, with particular relevance to rheumatoid arthritis, and reports data on the mechanisms responsible for the anti-arthritic effects of k-opioids in adjuvant arthritis. Opioids exert their diverse physiological effects through three distinct membrane-bound receptor subtypes mu (µ), delta (d) and kappa (k)
10.1191/0968130201pr181ra

114

JS Walker receptor mediated action.16 This local action has clinical signi cance because intra-articular morphine produced pain relief after knee arthroscopy17 and in patients undergoing dental surgery after submucous injection, without overt systemic effects.15,17,18 A large body of work has also demonstrated that the local administration of low doses of opioid receptor agonists elicit potent analgesic effects in in amed but not nonin amed tissue9,12–21 Clearly, there are functional opioid receptors on the peripheral terminals of afferent nerves that could well be exploited clinically. k-Agonists belong to four chemical classes, namely: the peptides (related to the endogenous ligand dynorphin), the benzomorphans (prototype ethylketocyclazocine), the arylacetamides (prototype PNU50488H), and the benzodiazepine derivative tri uadom. They include the centrally acting and prototype PNU50488H, which has served as the structural starting point for the synthesis of a multitude of compounds such as the centrally acting compounds: PNU62066E, PD117302 and GR89696A. To minimize the problems of tolerance and central side effects, various chemical approaches have been utilized to make opioids less accessible to the brain without reducing k-opioid activity. The structure of PNU50488H has been the basis for the development of many of these compounds.22 Asimadoline (EMD 61753; Merck KGaA), an amphiphilic compound that is orally active, is undergoing Phase II clinical trials in musculoskeletal pain.122,23 Most recently, other arylacetamide-derived peripherally selective agents, ADL 10-0101 and ADL-10-0116 (Adolor Corporation, USA), have been undergoing Phase I safety and ef cacy studies. In our laboratory we induce chronic polyarthritis in the rat by the administration of complete Freund’s adjuvant.12,24 This model of chronic arthritis has proved to be the most common; it is reproducible and shows many similarities to the human disease. We measure disease severity by using three quantitative indicators: paw swelling (oedema), radiological damage, and histological characteristics. We have tested the anti-in ammatory effects of k-opioids in this model. k-Opioids (e.g., PNU50488H and asimadoline) attenuated the progression of

in the central nervous system (CNS)2 and periphery.3 The different receptors have diverse behavioural characteristics2; for example, euphoria, physical dependence and respiratory depression are mainly associated with µ- and d-receptors. In contrast, opioids acting at k-receptors produce dysphoric rather than euphoric effects, which limit their physical dependence liability.4,5 In this regard, k-opioid agonists possess some advantages over µ-agonists: they are devoid of such side effects as dependence liability, constipation and respiratory depression.4,6 Classically, opioids have been used in the treatment of pain rather than in ammation, partly due to their side effects and because their anti-in ammatory actions have been largely unrecognized. A great deal is known about the analgesic effects of opioids7 and the actions of opioids on the hyperalgesic aspects of in ammation have been comprehensively reviewed.8 Apart from our own work there have been relatively few studies of their peripheral anti-in ammatory effects, so a brief overview of these is presented here. There are con icting reports concerning whether µ-opioids have anti- or proin ammatory properties.9 For example, morphine inhibits carrageenan-induced paw swelling10 and near toxic doses of morphine were able to attenuate adjuvant arthritis in rats.11,12 In contrast, low doses of morphine were proin ammatory in adjuvant arthritis.13 High doses would preclude the clinical use of morphine in arthritis, so speci c attention was given to k-opioids, particularly because they have a more favourable side effect pro le. Historically, opioids have been thought to produce their antinociceptive effects via actions in the CNS, but it is now well appreciated that opioid receptors are synthesized in the dorsal root ganglia and transported towards both central and peripheral nerve terminals. Furthermore, the peripheral axonal transport is upregulated during in ammation.14 Experimental and clinical studies have shown potent analgesic effects after peripheral administration of opioids.15 For example, the pioneering work of Robert Schmidt’s group in Germany has shown the local action of opioids in the knee joint of the cat.16 Morphine (µ-agonist) and PNU50488H (kagonist) reduce the action potential frequency in group III (Ad) bres of an articular nerves; reversal of this action by naloxone indicates an opioid
Pain Reviews 2001; 8: 113–119

k-Opioids in rheumatoid arthritis experimental adjuvant arthritis via speci c opioid receptors in the periphery using rigorous criteria such as reversibility by opioid antagonists, dosedependency and stereospeci city.1,9,12,24,25 The anti-arthritic actions of k-opioids were blocked by the opioid antagonist naloxone methiodide, which does not penetrate the CNS, and by speci c k-opioid antagonists; thus the anti-in ammatory effects are most likely exerted in the periphery via k-opioid receptors. This effect was seen with a number of kopioids, including low doses of centrally acting compounds (PNU50488H administered peripherally into a joint,24 PD11730226 and peripherally selective asimadoline,1 so our work has clinical potential because these drugs could be used with the minimal likelihood of central side effects such as addiction and tolerance. In contrast, morphine was anti-in ammatory only at very high doses (median effective dose ~ 58 mg/kg). The temporal details of the treatment regimens were found to be important. Opioid action is most signi cant in the rst few days of treatment (i.e., at disease onset) and is cumulative over 21 days of treatment. These data support current opinion that aggressive drug therapy needs to be started as soon as possible after disease onset to prevent progressive joint destruction.1,24,27 The nding that k-opioids reduce in ammation in adjuvant arthritis raises interesting questions about the mechanisms and in ammatory mediators involved. The actions of k-opioids are mediated via k-receptors, which are found on immune system cells and on neural cells (for review see Figure 1 in Sharp et al., 199828). The close spatial and functional association between nerves and immune cells suggests that modulation of immune or neural proin ammatory pathways could contribute to the disease suppressing activity of these drugs.9 A major point of neuro-immune convergence is in what is known as the hypothalamic–pituitary–adrenal (HPA) axis (Figure 1). Is it possible therefore that the anti-arthritic action of the k-opioids is mediated via the HPA axis? Using the prototype PNU50488H in adrenalectomized rats, arthritis developed sooner and was more severe. However, PNU-50488H substantially reduced the pooled severity index (combined quantitative oedema, histological and radiological assessments) at day 18 in both control and
Pain Reviews 2001; 8: 113–119

115

adrenalectomized rats to an equal extent. Thus, the HPA axis is only partially involved in the anti-in ammatory actions of opioids.29 We therefore continued our investigations into other neural and immune mechanisms.

Role of the nervous system
Neuropeptides such as substance P (SP) and calcitonin gene-related peptide (CGRP) are strongly implicated in the pathogenesis and/or spread of in ammatory arthritis.30 By using two differently acting drugs (the k-agonist asimadoline, and the neurokinin 1 (NK1) antagonist GR205171) the roles that SP may play in the pathogenesis and maintenance of experimental arthritis were investigated. Depending upon the timing of their administration, both agents signi cantly attenuated adjuvant arthritis, putatively via a mechanism that involves SP. We hypothesize that asimadoline acts on peripheral terminals to modulate SP release, while GR205171 antagonizes the action of SP, either peripherally or centrally. Time-dependent multiphasic effects were found with asimadoline on SP tissue levels (i.e., protein) in the joints from rats with adjuvant arthritis. Treatment with asimadoline decreased SP levels in the joint during early arthritis (day 3) and increased levels were found during established disease (day 2125). Drug-induced changes in SP content could result from a change in either the release or synthesis from neural cells. We therefore further investigated neuropeptide production in cell bodies innervating the joints as a function of time. Asimadoline reduced the expression of both SP and CGRP mRNA in the dorsal root ganglion on either day 13 or day 21 postadjuvant.31 This reduction in peptide mRNA is likely to be a consequence of a negative feedback mechanism resulting from an increase in SP in the nerve terminals (and therefore enhanced axonal transport) rather than a direct action on peptide synthesis by the k-agonist.31 Our studies con rm literature ndings on the inhibitory actions of opioids on SP release from the peripheral endings of primary afferent bres.32

116

JS Walker

Figure 1: Schematic illustration of neuro-immune interactions in the control of in ammation. Opioid receptors are identi ed by the purple blocks and suggest possible sites of action. IFN-g, interferon-g; ICAM-1, intercellular adhesion molecule-1; PGE2, prostaglandin E2 (source: Wilson, 1998,26 reproduced with permission from the author).

Pain Reviews 2001; 8: 113–119

k-Opioids in rheumatoid arthritis

117

Role of the immune system
Opioids have been found to regulate lymphocyte proliferation, antibody production and natural killer cell activity as well as inhibiting the function of neutrophils, monocytes and macrophages.33 Thus the hypothesis that opioids may exert their anti-in ammatory actions via opioid receptors that exist on immune cells, as shown by Sharp’s group (see Figure 1 in Sharp et al., 199828), together with our own observations that opioids decrease the numbers of both mast cells and macrophages in ltrating the joint, is an attractive one. The pretreatment of rats with a selectin blocker, fucoidin, abolishes peripheral opioid analgesia.34 The question arises of whether blockade adhesion molecules such as Intercellular adhesion molecule-1 (ICAM-1) abolishes in ammation in arthritis. The effect of the k-opioid PD117302 on ICAM-1 expression was thus compared with the effects of a prototypic nonsteroidal anti-in ammatory drug, naproxen, in the same model. In animals treated with PD 117302 and naproxen there was a signi cant attenuation of arthritis, however, only PD117302 was able signi cantly to inhibit the upregulation of ICAM1 (a marker of in ammation) expression in arthritic joints.35 Decreased cytokine release may explain the abrogation of the upregulation of ICAM-1 expression, reduced leucocyte recruitment, and less synovial in ammation observed in the adjuvant arthritic rats. In support of this, k-opioids also directly inhibit the release of both interleukin-1 and tumour necrosis factor (TNF) from macrophages.36 We further investigated this using cultured macrophages in vitro and examined the effects of a range of k-opioids on TNF release from these cells. The production of TNF was signi cantly inhibited in the presence of asimadoline, PD117302 and PNU50488H in a dose-dependent, antagonist reversible, manner over the dose range of 10–11 to 10–3 M with 100% suppression at a concentration of 10–3 M. Thus, k-opioids may act directly to inhibit cytokine release from immune cells. Taken together with our earlier ndings of decreased cell recruitment and reduced expression of adhesion molecules, these data help to
Pain Reviews 2001; 8: 113–119

explain the powerful anti-in ammatory effects observed clinically in the adjuvant arthritis model.

Summary
In summary, therapy with opioids is an exciting new development for arthritis, especially since there is the potential for fewer side effects from molecules that act outside the CNS. We found kopioid drugs to be powerfully anti-in ammatory, reducing disease severity by as much as 80%, attenuating arthritis in a dose-dependent, stereoselective, antagonist reversible manner. In contrast, opioids acting at other receptors were therapeutic only at near toxic doses. Currently, however no pure k-opioids are available for clinical use. The HPA axis was found to be only partially involved; thus we investigated other neural and immune mechanisms. The results showed that the k-opioid anti-in ammatory actions were exerted via: (1) reduced adhesion molecule expression; (2) inhibition of cell traf cking; (3) reduced TNF release; and (4) alterations in mRNA expression and protein levels of SP and CGRP in joint tissue. The mechanisms involved are summarized in Figure 1. The ability of kopioids to act at multiple sites in the in ammatory cascade, as suggested by the presence of opioid receptors at various locations throughout the cascade, may explain their powerful actions. k-Opioids are, however, most therapeutic during disease onset; thus it is likely that they exert their anti-in ammatory effects predominantly via changes in cellular activation and cytokine expression, rather than via the nervous system. The involvement of SP and the ef cacy of NK1 antagonists predict that combined opioid-NK1 regimens have therapeutic promise. Thus our work supports the ndings of Stein’s group in humans, that opioids do indeed have powerful actions in the periphery via speci c receptors.37 This study examined intra-articular morphine in chronic arthritis. Synovial leucocyte counts were lower after morphine than after saline. Intra-articular morphine produced analgesia of a similar magnitude to dexamethasone and it may therefore have anti-in ammatory actions in human arthritis. Furthermore, unpublished data from our laboratory demonstrates

118

JS Walker
chemiluminescence of leucocytes stimulated by zymosan. J Pharm Pharmacol 1985; 37: 100–104. Levine JD, Moskowitz MA, Basbaum AI. The contribution of neurogenic in ammation in experimental arthritis. J Immunol 1985; 155 : 843s–847s. Walker JS, Chandler AK, Wilson JL, et al. Effect of µ-opioids morphine and buprenorphine on the development of adjuvant arthritis in rats. In amm Res 1996; 45: 557–63. Earl JR, Claxson AW, Blake DR, et al. Proin ammatory effects of morphine in the rat adjuvant arthritis model. Int J Tissue React 1994; 16: 163–70. Machelska H, Binder W, Stein C. Opioid receptors in the periphery. In: Kalso E, McQuay HJ, Wiesen eld-Hallin Z editors. Opioid sensitivity of chronic noncancer pain (Progress in Pain Research and Management, vol. 14) . IASP Seattle: Press, 1999: 47. Likar R, Sittl R, Gragger K, et al. Peripheral morphine analgesia in dental surgery. Pain 1998; 76: 145–50. Russell NSW, Schaible HG, Schmidt RF. Opiates inhibit the discharges of the afferent units from in amed knee joint of the cat. Neurosci Lett 1987; 76: 107–12. Stein C, Comisel K, Haimerl E, et al. Analgesic effect of intraarticular morphine after arthroscopic knee surgery. N Engl J Med 1991; 325: 1123–26. Kalso E, Tramer MR, Carroll D, et al. Pain relief from intra-articular morphine after knee surgery: a qualitative systematic review. Pain 1997; 71: 127–34. Stein C, Hassan AHS, Lehrberger K, Grie ng J, Yassouridis A. Local analgesic effect of endogenous opioid peptides. Lancet 1993; 342 : 321–24. Lehrberger K, Stein C, Hassan A, Yassouridis A. Knee surgery. Sports Traumatol Arthrosc 1994; 2: 174–75. Lee SH, Kayser V, Guilbaud G. Antinociceptive effect of systemic kelatorphan, in mononeuropathic rats, involves different opioid receptor types. Eur J Pharmacol 1994; 264 : 61–67. Barber A, Gottschlich R. Central and peripheral nervous systems: novel developments with selective, non-peptidic kappa-opioid receptor agonists. Exp Opin Invest Drugs 1997; 6: 1351–68. Barber A, Bartosyk GD, Bender HM, et al. A pharmacological pro le of the novel, peripherallyselective kappa-opioid receptor agonist, EMD 61753. Br J Pharmacol 1994; 113 : 1317–27. Wilson JL, Nayanar V, Walker JS. The site of antiarthritic action of the kappa agonist, U50488H: importance of local administration. Br J Pharmacol 1996; 118 : 1754–60. Binder W, Scott C, Walker JS. Involvement of substance P in the anti-in ammatory effects of the peripherally selective k-opioid, asimadoline and the

that the clinically available, oxycodone, which has k-opioid activity, is also anti-arthritic in the Freunds adjuvant model of arthritis. Peripherally acting opioids may prove to be a potent new treatment for rheumatoid arthritis sufferers in the future. Acknowledgements I am very grateful to the dedicated students, postdoctoral fellows and entire research staff of my laboratory over the past decade. In particular, I would like to thank the following persons for their contributions to this research: Professor Richard Day, Associate Professor John Carmody, Dr Bruce Kirkham, Drs Jodie Wilson, Dr Waltraud Binder, Dr Caroline Scott, and Katherine Bush.

11

12

13

14

References
1 2 3 4 5 6 Binder W, Walker JS. Effect of the peripherally selective µ-opioid agonist, asimadoline, on adjuvant arthritis. Br J Pharmacol 1998; 124 : 647–54. Law PY, Loh HH. Regulation of opioid receptor activities. J Pharmacol Exp Ther 1999; 289: 607–24. Minami M, Statoh M. Molecular biology of the opioid receptors: structures, functions and distributions. Neurosci Res 1995; 23: 121–45. Horwell DC. Kappa opioid analgesics. Drugs Future 1988; 13: 1061–70. Wollemann M, Benyhe S, Simon J. The kappa opioid receptor: evidence for the different subtypes. Life Sci 1996; 52: 599–611. Porreca F, Mosberg HT, Hurst R, et al. Roles of mu, delta and kappa opioid receptors in spinal and supraspinal mediation of gastrointestinal transit effects and hotplate analgesia in the mouse. J Pharmacol Exp Ther 1984; 230 : 341–48. Reisine T, Pasternak G. Opioid analgesics and antagonists. In: Hardman JG, Limbird L E, Molinoff PB, Ruddon RW, Goodman LS, Gilman A, editors. The pharmacological basis of therapeutics , ninth edition. New York: McGrawHill, 1996: 521–56. Barber A, Gottschlich R. Opioid agonists and antagonists: an evaluation of their peripheral actions in in ammation. Med Res Rev 1992; 12: 525–62. Walker JS, Wilson J, Binder W, et al. The antiin ammatory effects of opioids: their possible relevance to the pathophysiology and treatment of rheumatoid arthritis. Res Alerts Rheum Arthritis 1997; 1: 291–99. Gyires K, Budavari I, Furst S, et al. Morphine inhibits the carrageenan-induced edema and the

15 16

17 18 19 20 21

7

22

8

23

9

24

10

25

Pain Reviews 2001; 8: 113–119

k-Opioids in rheumatoid arthritis
NK1 antagonist GR205171. Eur J Neurosci 1999; 11: 2065–72. Wilson JL. The mechanism for the antiin ammatory and anti-arthritic effects of k-opioid agonist: role of the immune system [Thesis]. Sydney: University of New South Wales, 1998. Pincus T. Rheumatoid arthritis: a medical emergency? Scand J Rheumatol 1994; 23: 21–30. Sharp BM, Roy S, Bidlack JM. Evidence for opioid receptors on cells involved in host defence and the immune system. J Neuroimmunol 1998; 83: 45–46. Wilson JL, Carmody JJ, Walker JS. The importance of the hypothalamo–hypophyseal–adrenal axis to the anti-in ammatory actions of the k-opioid agonist PNU-50,488H in rats with adjuvant arthritis. J Pharmacol Exp Ther 2000; 294 : 1131–36. Kidd BI, Mapp PI, Gibson SJ, et al. A neurogenic mechanism for symmetrical arthritis. Lancet 1989; ii: 1128–30. Walker JS, Scott C, Bush KA, et al. Effects of the peripherally selective k-opioid asimadoline, on substance P and CGRP mRNA expression in chronic arthritis of the rat. Neuropeptides 2000; 34:

119

26

32 33 34 35

27 28 29

36

30 31

37

193–202. Schaible HG, Grubb BD. Afferent and spinal mechanisms of joint pain. Pain 1993; 55: 5–54. Carr DJJ. The role of endogenous opioids and their receptors in the immune system. Proc Soc Exp Biol Med 1991; 198 : 710–20. Machelska H, Cabot PJ, Mousa SA, Zhang Qm Stein C. Pain control in in ammation governed by selectins. Nat Med 1998; 4: 1425–58. Wilson JL, Walker JS, Antoon JS, Perry MA. Intercellular adhesion molecule-1 expression in adjuvant arthritis in rats: inhibition by kappa-opioid agonist but not by NSAID. J Rheumatol 1998; 25: 499–505. Belkowski SM, Alicea C, Eisenstein TK, et al. Inhibition of IL-1 and TNF-a synthesis following treatment of macrophages with the kappa opioid agonist U-50,488H. J Pharmacol Exp Ther 1995; 273 : 1491–96. Stein A, Yassouridis A, Szopko C, Helmke K, Stein C. Intraarticular morphine versus dexamethasone in chronic arthritis. Pain 1999; 83: 525–32.

Pain Reviews 2001; 8: 113–119