Analgesic and Antiinflammatory Effects of Two Novel Κ-Opioid

Home , Binaltorphimine, Dynorphin A, Hemorphin

Anesthesiology 2001; 94:1034–44 © 2001 American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins, Inc. Analgesic and Antiinﬂammatory Effects of Two Novel ␬-Opioid Peptides Waltraud Binder, Ph.D.,* Halina Machelska, Ph.D.,* Shaaban Mousa, Ph.D.,* Thomas Schmitt, M.D.,* Pierre J. M. Rivière, Ph.D.,† Jean-Louis Junien, Ph.D.,‡ Christoph Stein, M.D.,§ Michael Schäfer, M.D.ʈ

Background: This study investigates two new ␬-agonist tet- mediates specific additional responses. For example, eu- rapeptides, FE 200665 and FE 200666, with high peripheral phoria, physical dependence, and respiratory depression selectivity as a result of poor central nervous system are mainly associated with ␮ and ␦ receptors.1 Spinal, penetration. ␮ Methods: Four days after administration of Freund adjuvant supraspinal, and peripheral -receptors also mediate opi- into the hind paw of male Wistar rats, antinociceptive effects of oid-induced gastrointestinal transit inhibition and consti- intraplantar and subcutaneous injection of FE 200665 and FE pation.3 In contrast, opioids acting at ␬-receptors pro- 200666 were measured by paw pressure algesiometry and com- duce dysphoric rather than euphoric effects, which ␬ ␬ pared with the -agonist U-69,593. Peripheral and -receptor limits their dependence liability and produces little or no selectivity was assessed by the antagonists naloxone methio- ␬ dide (NLXM) and nor-binaltorphimine, respectively. Antiin- respiratory depression. Furthermore, -agonists do not flammatory effects were evaluated by paw volume plethysmom- inhibit intestinal transit or induce constipation. Thus, etry and histologic score. ␬-opioids offer advantages in comparison to ␮-opioids Results: Similar to intraplantar U-69,593, intraplantar FE and are currently being pursued as analgesic agents. ␮ ␮ 200665 (3–100 g) and FE 200666 (1–30 g) resulted in signifi- However, centrally acting ␬-opioids can produce psy- cant and dose-related increases of paw pressure thresholds. 4 Higher doses of FE 200665 (0.2–20 mg) and FE 200666 (0.06– chotomimesis and may be aversive in humans. A poten- 6 mg) were required by subcutaneous route to produce similar tially successful strategy to eliminate the adverse side ef- antinociceptive responses, supporting a peripheral site of ac- fects of ␬-opioids is to restrict the access of these tion. nor-Binaltorphimine dose-dependently antagonized this compounds to the central nervous system, as opioid anal- ␬ effect, implying -opioid selectivity. Analgesic effects of subcu- gesia can also be brought about by the activation of opioid taneous FE 200665 and FE 200666 were abolished by intraplantar nor-binaltorphimine, and both subcutaneous and intraplan- receptors on peripheral nerve endings, particularly during 5–7 tar effects were dose-dependently antagonized by subcutaneous inflammatory conditions. Clinical studies have shown NLXM, further demonstrating a peripheral site of action. One to that intraarticular morphine relieves acute postoperative as 6 days after Freund adjuvant inoculation, single and repeated well as chronic arthritic pain.8 The analgesic effect is long- intraplantar injections of FE 200665, FE 200666, and U-69,593 lasting, e.g., up to 2 and 7 days, respectively. This is not significantly reduced paw volume and histologic scores. Both changes were reversed by intraplantar nor-binaltorphimine only explained by an inhibition in pain transmission but 8 and subcutaneous NLXM. also by a local antiinflammatory effect. Conclusion: FE 200665 is a peripherally selective ␬-agonist Previous attempts to design peripherally selective with potent analgesic and antiinflammatory properties that ␬-agonists have not been successful. These attempts may lead to improved analgesic–antiinflammatory therapy were based on small organic molecules. Recently, novel compared with centrally acting opioids or standard nonsteroi- ␬ dal antiinflammatory drugs. and peptidic -ligands were identified by positional scanning of a tetrapeptide combinatorial library screened in 9 OPIOID analgesics, despite their side effects, are unsur- opioid receptor radioligand binding assays. These new ␬ passed in the pharmacologic treatment of severe pain. -ligands have an all D-amino acids sequence and show ␬ They exert their diverse physiologic effects through high selectivity for the -receptor. Furthermore, this ␬ ␬ three distinct membrane-bound receptor subtypes—␮, new class of -ligands are potent -agonists in vitro, ␦, and ␬—in the central nervous system1 and in the potent antinociceptive agents in vivo, and peripherally 2 selective, as shown by the lack of sedative activity in the periphery. 10 In addition to analgesia, each opioid receptor subtype mouse rotarod after peripheral administration. How- ever, these series of new ␬-agonists are short-acting. Chemical optimization has led to the characterization of * Postdoctoral Fellow, § Professor and Chairman, ʈ Associate Professor, Depart- a second generation of tetrapeptide ␬-agonists with high ment of Anesthesiology and Critical Care Medicine, Klinikum Benjamin Franklin, peripheral selectivity and long duration of action.10 In Freie Universität Berlin. † Head of Biology, Ferring Research Institute Inc. ‡ Chief Scientific Officer, Ferring Pharmaceuticals. the current study we examined two of these novel com- Received from the Department of Anesthesiology and Critical Care Medicine, pounds, FE 200665 and FE 200666, which show high Klinikum Benjamin Franklin, Freie Universität Berlin, Berlin, Germany; Ferring affinity subnanomolar and high selectivity for human Research Institute Inc., San Diego, California; and Ferring Pharmaceuticals, Paris, France. Submitted for publication July 31, 2000. Accepted for publication De- ␬-opioid receptor versus ␮-opiod receptor (up to 90,000- cember 1, 2000. Supported by Ferring Research Institute B.V., Hoofddorp, The fold) or ␦-opiod receptor (85,000-fold).10 The antinoci- Netherlands. Address reprint requests to Dr. Binder: Department of Anesthesiology and ceptive and antiinflammatory properties of these poten- Critical Care Medicine, Benjamin-Franklin University Hospital, Freie Universiät tially peripherally selective ␬-opioid peptides were Berlin, Hindenburgdamm 30, D 12200 Berlin, Germany. Address electronic mail ␬ to: [email protected]. Individual article reprints may be purchased assessed and compared with the standard -opioid through the Journal Web site, www.anesthesiology.org. U-69,593.

Anesthesiology, V 94, No 6, Jun 2001 1034

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 PERIPHERAL OPIOID ANALGESIA AND ANTIINFLAMMATION 1035

Materials and Methods (2) pannus formation (degree of intrusion of granulation tissue into the joint space); and (3) periosteal reaction Induction of Arthritis (extent of new bone formation). A rating scale from 0 to Experiments were conducted in male Wistar rats 8 was used for each variable (i.e., maximum obtainable weighing 250–300 g. Rats were housed individually and score ϭ 24). This procedure was based on the method- kept in a temperature-controlled room (22°C Ϯ 1) with ology originally described by Ackerman et al.11 a 12-h alternating light–dark cycle. To induce inflammation, rats were sedated by brief halothane anesthesia and Drugs received an intraplantar injection of 150 ␮l Freund com- We used the ␬-opioid agonists U-69,593 (Sigma, De- plete adjuvant (FCA; Calbiochem, San Diego, CA) into isenhofen, Bavaria, Germany), FE 200665 and FE 200666 the right hind paw. Experiments and animal care were (Ferring Research, San Diego, CA), and the opioid antag- conducted in accordance with standard ethical guide- onists nor-binaltorphimine (nor-BNI) and naloxone me- lines (Tierschutz komission, Berlin, Germany). thiodide (NLXM; Sigma). All opioid drugs were dissolved in distilled water and administered intraplantarly Nociceptive Thresholds (100 ␮l), subcutaneously into a skin fold in the neck Between 4 and 6 days after induction of inflammation, (200 ␮l), or intrathecally (10 ␮l). Antagonists were given mechanical nociceptive thresholds were assessed using either concomitantly or 5 min before agonist administra- a paw pressure analgesymeter (Ugo Basile, Comerio, tion. Control animals received distilled water. Italy). The paw pressure threshold (PPT) required to elicit hind-paw withdrawal was determined by averaging Implantation of Intrathecal Catheters three consecutive trials separated by 10 s. The sequence Rats were handled and trained in the test situation for of left and right paws was alternated between animals to 3 days before intrathecal catherization. Anesthesia was avoid bias. The experimenter was blind to the conditions induced and maintained with 2% halothane via a loose- used. Data are represented as percent maximum possi- fitting plastic mask. The intrathecal catheters were pre- ble effect (MPE) with an arbitrary cutoff point of 250 g pared according to a method described previously.12 Ϫ and are calculated by the formula: (PPT treated PPT Briefly, a polyethylene tubing (PE-10; Portex, United Ϫ ϫ pretreated/250 PPT pretreated) 100. Kingdom) was cut in 200-mm lengths and inserted 15 mm in a cervical direction, through a previously made Evaluation of Inflammation incision at the L3–L4 level. The animals were allowed 4 Antiinflammatory effects were assessed by monitoring days to recover from anesthesia and surgery. During this paw volume with a plethysmometer (Ugo Basile) and by time, animals showing any sign of neurologic damage histologic examination. Histologic sections were pre- were discarded from the study. pared as follows. Rats were deeply anesthetized with Drugs were injected in a volume of 10 ␮l followed by halothane and perfused transcardially with 100 ml of 5 ␮l of vehicle to flush the catheter. The effectiveness of 0.1 M phosphate-buffered saline (pH 7.4) and 300 ml cold the catheter was tested by an injection of 10 ␮l lidocaine phosphate-buffered saline containing 4% paraformalde- 2% (Sigma) 24 h before the experiment. Only animals hyde and 14% saturated picric acid (pH 7.4) (fixative experiencing an immediate yet reversible paralysis of solution). The hind limbs were removed, postfixed for their hind limbs were included in the study. After the 42 h at 4°C, and washed for 30 min in distilled water. experiments, in which the animals were used only once, Decalcification was performed with commercially avail- the rats were killed, and the correct position of the able hydrochloric acid–based decalcifying agent (Decal; catheter tip was confirmed at autopsy by an experienced Decal Chemical Corporation, Congers, NY). After decal- investigator who was blinded to the experimental cification, the tissues were washed in water for 30 min results. and cryoprotected overnight at 4°C in phosphate-buffered saline containing 10% sucrose. The tissues were Experimental Protocols then embedded in tissue-Tek compound (O.C.T.; Miles Nociceptive Thresholds. Inc., Elkhart, IN) and frozen at Ϫ20°C. Sections, 6-␮m Timecourse of Drug Action. The antinociceptive thick, were cut on a cryostat, mounted onto gelatin- effects of intraplantar or subcutaneous FE 200665, FE coated slides, and stained with hematoxylin and eosin. 200666, and U-69,593 were examined as a function of The tissues were oriented to allow longitudinal sections time, in both ipsilateral and contralateral paws (n ϭ 6 to be cut so as to show the dorsoventral faces of the per group). Baseline (pre-agonist) PPTs were obtained tarsal, metatarsal, and phalangeal joints, bones, and soft and then reevaluated at 5, 10, and 30 min, and 1, 6, 24, tissue on each slide. Each section was quantitatively and 48 h after opioid administration. Control rats reevaluated by two observers, who were blind to the ceived distilled water. treatment regimen, using the following criteria: (1) peri- Dose–Response Relations. The antinociceptive articular inflammation (density of inflammatory cells); dose–response relation of the opioid agonists FE 200665

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 1036 BINDER ET AL.

Fig. 1. Time course and dose–response of the antinociceptive effects (% maximum possible effect [MPE]) of FE 200665, FE 200666, and U-69,593 on paw pressure threshold in inﬂamed paws. FE 200665 administered (A) intraplantarly and (B) subcutaneously; FE 200666 administered (C) intraplantarly and (D) subcutaneously; and U-69,593 administered (E) intraplantarly and (F) subcutaneously. *Signiﬁcant difference from controls, P < 0.05, analysis of variance (Dunnett test).

and FE 200666 were determined (n ϭ 6 per group) and oid agonists (400 ␮g/rat, subcutaneously). NLXM was ad- compared with U-69,593. After baseline measurements, ministered subcutaneously 5 min before intraplantar opioid animals received intraplantar injections of water or FE agonists (2.5, 5, and 10 mg/kg). Finally, to confirm periph- 200665 at 3, 10, 30, or 100 ␮g/paw; FE 200666 at 1, 3, eral selectivity, the most potent ␬-opioid peptide, FE 10, 30, or 100 ␮g/paw; and U-69,593 at 25, 50, 100, or 200665, was administered intrathecally in a dose range of 200 ␮g/paw in both hind paws. Equivalent doses were 1–10 ␮g. The most effective dose was then coadministered also administered subcutaneously into a skin fold of the with NLXM (10 mg/kg subcutaneously). In addition, NLXM neck to exclude a central site of action. was administered intrathecally (10 ␮g) together with in- Receptor Specificity and Site of Action. The ␬-opi- traplantar FE 200665 (100 ␮g). oid antagonist nor-BNI, as well as the peripherally selective Antiinflammatory Effects. opioid antagonist NLXM, were administered with FE Time Course and Dose Response. The antiinflam- 200665, FE 200666, or U-69,593 (intraplantarly) to deter- matory effects of FE 200665, FE 200666, and U-69,593 mine the ␬-specificity as well as the site of action (periph- were examined as a function of dose and time in both eral or central) of these opioid agonists (n ϭ 6 per group). ipsilateral and contralateral paws (n ϭ 6 per group). nor-BNI was administered either concomitantly (50, 100, or Baseline paw volumes were obtained and then reevalu- 200 ␮g/paw, intraplantarly) or immediately before the opi- ated according to the time and dose schedule for antino-

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 PERIPHERAL OPIOID ANALGESIA AND ANTIINFLAMMATION 1037

ciceptive effects. In addition, paw volume was also mon- itored daily from the induction of inflammation (day 0) until day 6, when the rats were killed. Drugs were administered either daily or by single injection simulta- neously with FCA on day 0. After the animals were killed, the hind limbs were removed for histologic assessment. Receptor Specificity and Site of Action. The specificity and site of action of the antiinflammatory effects of the three ␬-opioids were assessed by concomitant administration of nor-BNI (200 ␮g/day intraplantarly) or NLXM (10 mg/kg subcutaneously). The antagonists were injected immediately before ␬-agonist administration either daily or by single injection on day 0.

Statistical Analysis Data are represented as mean Ϯ SEM. Dose–response curves and antagonist data were assessed by analysis of variance followed by a linear regression analysis or a post hoc Dunnett test for non–linearly distributed antagonist data. Time course data were analyzed by two-way repeated-measure analysis of variance (treatment ϫ time) followed by a post hoc Dunnett test. All data were speciﬁcally tested for normality (Skewness and Kurtosis normality tests). Differ- ences were considered signiﬁcant at P Ͻ 0.05. All tests were performed using Sigma Stat 2.03 statistical soft-

ware. ED50 values were calculated using Pharm/PCS pharmacological calculations with computer pro- grams (Philadelphia, PA).

Results

Nociceptive Thresholds Time Course of Drug Action. Treatment with FE 200665 and FE 200666 produced a similar time course of drug action (figs. 1A–1D). A peak was observed 5 min after opioid administration for both drugs; thereafter, PPT decreased. Intraplantar FE 200665 produced a significant antinociceptive effect in inflamed paws until 30 min after drug administration. By 1 h, PPT had re- turned to baseline. In noninflamed paws, FE 200665 (100 ␮g) was markedly less potent but still produced a significant effect (33 Ϯ 4.8%). FE 200666 produced a significant antinociceptive effect until 10 min after Fig. 2. Dose–response effects of intraplantar nor-BNI on the antinociceptive action (% maximum possible effect [MPE]) of intraplantar administration in inflamed paws. Similar (A) FE 200665 (100 ␮g), (B) FE 200666 (30 ␮g), and (C) U-69,593 to FE 200665, administration into the noninflamed (100 ␮g). P < 0.05, analysis of variance (linear regression). paw was much less effective (data not shown). The time course produced by U-60,593 was of longer du- Dose–Response Relations. FE 200665, FE 200666, ration. Significant elevation of PPT was observed until and U-69,593 dose-dependently increased PPT in in- 1 h after intraplantar administration to inflamed paws, flamed paws (figs. 1A–1F) when administered intraplan- returning to baseline by 2 h. Maximum antinocicep- tarly. A comparable MPE of approximately 85% was tion occurred at 10 min (fig. 1E). In all cases systemic observed for FE 200665 and U-69,593 in inflamed paws administration resulted in a loss of potency (figs. 1B, (figs. 1A and 1E). However, FE 200666 was markedly less 1D, and 1F). Control animals treated with distilled effective, producing an MPE of only 35% (fig. 1C). water showed no effect. Higher doses of FE 20066 (up to 100 ␮g) produced no

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 1038 BINDER ET AL.

additional effect (data not shown); therefore, its low activity is likely to be a result of decreased efficacy. Systemically administered, much higher doses were nec- essary to produce an equivalent antinociceptive response for all opioids tested (figs. 1B, 1D, and 1F). Indeed, subcutaneous administration of FE 200665 (0.2, 0.6, 2, 6, and 20 mg/rat) and FE 200666 (0.06, 0.2 0.6, 2, and 6 mg/rat) was only effective in inflamed paws at the higher doses tested, greater than 2 and 0.6 mg/rat, respectively (figs. 1B and 1D), and ineffective in noninflamed paws (data not shown). Subcutaneous administration of U-69,593 (200, 400, and 800 ␮g/rat) dose-dependently increased PPT in both paws—maximum 85% ipsilateral (fig. 1F), 42% contralateral (data not shown)—although the dose–response curve was shifted to the right, i.e., an eight times higher intraplantar dose was needed to produce the same MPE (figs. 1E and 1F).

Receptor Specificity The ␬-specificity of intraplantar FE 200665 and FE 200666 was determined using the ␬-opioid antagonist nor-BNI and compared with that of U-69,593. nor-BNI dose-dependently attenuated opioid-induced antinociception for all three drugs in both inflamed and noninflamed paws (dose range, 50–200 ␮g/paw; fig. 2). The highest dose (200 ␮g/paw) completely abolished the antinociceptive response of all opioids tested (figs. 2A– 2C). This indicates that the analgesic actions of these opioids are mediated via ␬-receptors.

Peripheral Selectivity Subcutaneous versus Intraplantar nor-BNI. The antinociceptive response produced by intraplantar FE 200665 and FE 200666 was completely abolished, and that of intraplantar U-69,593 was almost completely abolished, by intraplantar nor-BNI (200 ␮g/paw; figs. 3A–3C). Subcutaneous nor-BNI (400 ␮g/rat) had no effect on the intraplantar agonist response. The antinociceptive effects of subcutaneous FE 200665 (2 mg/rat) and FE 200666 (600 ␮g/rat) in both inflamed and noninflamed paws were also abolished by intraplantar nor- BNI (200 ␮g/paw), indicative of a peripheral mode of action (figs. 3A and 3B). Although intraplantar nor-BNI significantly attenuated subcutaneous U-69,593 (800 ␮g/ rat) in inflamed paws, the antagonism was not complete (fig. 3C). However, higher doses of nor-BNI were com- Fig. 3. Effect of nor-BNI (200 ␮g intraplantarly or 400 ␮g subcuta- pletely able to abolish the antinociceptive effect of neously) on the antinociceptive action (% maximum possible ef- U-69,593 (data not shown). fect [MPE]) of (A) FE 200665 (100 ␮g/paw intraplantarly or 2 mg/ Naloxone Methiodide. The peripherally selective rat subcutaneously), (B) FE 200666 (30 ␮g/paw intraplantarly or 600 ␮g/rat subcutaneously), and (C) U-69,593 (100 ␮g/paw in- systemic opioid antagonist NLXM (2.5–10 mg/kg) dose- traplantarly or 800 ␮g/rat subcutaneously). *Significant differ- dependently attenuated intraplantar FE 200665–, FE ence from agonist-treated animals, P < 0.05, analysis of vari- 200666–, and U-69,593–induced antinociception in in- .subcutaneous ؍ .intraplantar; s.c ؍ .ance (Dunnett test). i.pl flamed paws (figs. 4A–4C). The antinociceptive effects in noninflamed paws, as well as those of subcutaneous administration of these opioid agonists, were also abolished by 10 mg/kg NLXM (figs. 4A–4C).

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 PERIPHERAL OPIOID ANALGESIA AND ANTIINFLAMMATION 1039

Fig. 5. Antinociceptive effects (% maximum possible effect [MPE]) of intrathecal [i.t.] FE 200665. (A) Dose–response relation and (B) antagonism by naloxone methiodide (NLXM; administered subcutaneously and intrathecally). P < 0.05, analysis of ؍ .subcutaneous; i.pl ؍ .variance (linear regression). s.c intraplantar.

trathecal FE 200665 (10 ␮g). Similarly, when NLXM was administered intrathecally (10 ␮g), it was completely ineffective against peripherally administered FE200665 Fig. 4. Dose–response effects of naloxone methiodide (NLXM, (100 ␮g; fig. 5B). administered subcutaneously) on the antinociceptive action (% maximum possible effect [MPE]) of (A) FE 200665 (100 ␮g/ paw intraplantarly or 2 mg/rat subcutaneously), (B) FE 200666 Antiinflammatory Effects (30 ␮g/paw intraplantarly or 600 ␮g/rat intraplantarly), and (C) Effect of Dose and Time Schedule. Paw volume U-69,593 (100 ␮g/paw intraplantarly). P < 0.05, analysis of measurements were used to assess the antiinflammatory ؍ ؍ variance (linear regression). i.pl. intraplantar; s.c. ␬ subcutaneous. effects of all three -opioids. A significant decrease in ipsilateral (inflamed) paw volume was observed at 24 Intrathecal Administration. FE 200665 produced a and 48 h after intraplantar and subcutaneous treatment dose-dependent antinociceptive response when admin- with a single injection of FE 200665 (30 and 100 ␮g/paw istered intrathecally (fig. 5A). The peripherally selective intraplantarly and 2 mg/rat subcutaneously; fig. 6A). FE opioid antagonist NLXM (10 ␮g/kg) was completely 200666 and U-69,593 also produced a reduction in paw without effect when administered systemically with in- volume at 24 and 48 h at 30 ␮g/paw intraplantarly and

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 1040 BINDER ET AL.

peripherally selective antagonist NLXM (10 mg/kg) significantly antagonized the antiinflammatory effects of FE 200665, FE 200666, and U-69,593 judged by paw volume ( figs. 8A–8C) or histologic score (fig. 9). Histologic assessment concurred with paw volume measurements and showed an approximate 50% reduction in arthritic parameters for all three ␬-opioids (fig. 9). Thus, the antiinflammatory as well as the antinociceptive effects of these ␬-opioids are mediated via ␬-receptors in the periphery.

Discussion

The current study found the novel ␬-opioid peptide FE 200665 to be a potent analgesic agent, with an antinociceptive effect comparable to that of U-69,593. Both FE compounds showed no overt central effects such as sedation or respiratory depression. Rivière et al.10 previously showed via the rotarod test in mice, that central nervous system–mediated sedation occurred at 643- and 84-fold higher doses in comparison to analgesic doses, for FE 200665 and FE 200666, respectively. All three opioids dose-dependently increased PPT (figs. 1A–1D), although FE 200666 was markedly less effective (33% MPE) in comparison to FE 200665 (85% MPE) and U-69,593 (86% MPE). The duration of the antinociceptive effects produced by these agents was noticeably less for the two opioid peptides than for U-69,593. Although the FE compounds maintained an analgesic effect for only 30 min, this is considerably longer than other ␬-peptides, which has been attributed to the rapid degradation rate of peptides in vivo.13 The FE compounds, shown here to be peripherally selective, as well as U-69,593, also show greater antinociception in inflamed than in normal paws (fig. 2), consistent with current literature. The efficacy and potency of peripheral opioid analgesics in general is greatly enhanced during inflammatory conditions.7,14 This phe- nomenon has been extensively examined during conditions such as neuropathic pain,15 visceral pain,16 bone damage,17 and inflammation of subcutaneous tissue, vis- ␬ cera, or joints, induced by FCA, formalin, carrageenan, Fig. 6. Effect of intraplantar -opioids on ipsilateral paw volume 5–7 4 days after administration of Freund complete adjuvant, as a prostaglandin, or neurogenic inflammation. function of time. (A) FE 200665, (B) FE 200666, and (C) U-69,593. The receptor specificity of FE 200665, FE 200666, and *Significant difference from controls, P < 0.05, analysis of vari- U-69,593 were investigated with the ␬-antagonist nor- ance (Dunnett test). BNI. All three opioids administered intraplantarly were 100 ␮g/paw intraplantarly, respectively (figs. 6B and dose-dependently antagonized by intraplantar nor-BNI 6C). No effect was observed in noninflamed paws. The (figs. 2A–2C), indicative of ␬-opioid selectivity. The pe- antiinflammatory effects of these ␬-agonists became evi- ripheral selectivity of these opioids was then assessed by dent at a much later time points than their antinocicep- the peripherally selective antagonist NLXM, a general tive effects. Early and continuous treatment resulted in opioid antagonist whose access to the brain is restriced an increased effectiveness of all three opioids (fig. 7). via the incorporation of a quarternary ammonium ion, as There was little difference between a 6-day treatment well as subcutaneous (distant) versus intraplantar (local) course and treatment given only once on initiation of treatment with nor-BNI. NLXM completely abolished the inflammation. effect of FE 200665 and FE 200666 and markedly atten- Receptor Specificity and Site of Action. Concomi- uated that of U-69,593. When these opioid agonists were tant administration of ␬-antagonist nor-BNI (200 ␮g) and administered subcutaneously, the antinociceptive effect

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 PERIPHERAL OPIOID ANALGESIA AND ANTIINFLAMMATION 1041

Fig. 7. Effect of daily treatment and a single injection of (A) FE 200665, (B) FE 200666, and (C) U-69,593 on paw volume as a function of time. *Signiﬁcant difference from controls, P < 0.05, analysis of variance Freund complete ؍ Dunnett test). FCA) adjuvant.

of U-69,593 was not completely abolished by intraplan- ␮-opioid agonist morphine can inhibit carrageenan-in- tar nor-BNI, in contrast to the FE compounds (ﬁg. 3). duced paw swelling in the rat. Moreover, the effects of

Thus, U-69,593 appears to have both a peripheral and a morphine were dose-dependent (ED30, 1.6 mg/kg) and central site of action, depending on the dose and mode partially antagonized by naloxone. In addition, opioid of administration. This is in agreement with other studies peptides such as dynorphin A and hemorphin-7 have that have also observed a peripheral and central compo- also been shown to modulate inflammation.20,21 Most nent in the site of action of this opioid in a model of studies indicate that ␬-opioid agonists are able to atten- neuropathic pain.18 The peripheral selectivity of the uate experimental arthritis; they have been shown to more potent ␬-opioid peptide FE 200665 was addition- attenuate the severity of adjuvant arthritis in a dose- ally examined via the coadministration of subcutaneous dependent, stereoselective, and antagonist-reversible NLXM with intrathecal FE 200665. The peripherally se- manner.22–25 Recently, U-69,593 was shown to reduce lective NLXM was no longer able to antagonize intrathe- paw and ankle edema in a model of carrageenin-induced cal FE 200665 despite the fact that this antagonist com- inflammation.26 High doses of morphine are also able to pletely abolished the effect of intraplantar FE 200665. attenuate adjuvant arthritis,27 but the dosages are much This provides further evidence that FE 200665 is unable higher than used systemically in humans. to cross the blood–brain barrier. In our study, all three opioid agonists displayed potent The antiinflammatory effects of opioids have been ex- antiinflammatory effects assessed by paw volume and tensively described. Gyires et al.19 showed that the histologic parameters. Administration of FE 200665, FE

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 1042 BINDER ET AL.

Fig. 8. Effect of nor-BNI (administered intraplantarly) or naloxone methiodide (NLXM; administered subcutaneously) on the antiinﬂammatory action of (A)FE 20065, (B) FE 200666, and (C) U-69,593. *Signiﬁcant difference from antagonist- treated groups, P < 0.05, analysis of vari- -Freund com ؍ ance (Dunnett test). FCA plete adjuvant.

200666, and U-69,593 at various time schedules showed cells. For example, a possible reason for this delayed that single treatment on initiation of inflammation was effect may be that these drugs do not affect the initial almost as effective as continuous treatment over 6 days plasma extravasation. The literature in this regard is (fig. 7) and was sustained well beyond the administration controversial. Some studies, using the Evans blue dye of the opioids. This implies that the later stages of hind- leakage technique, have shown that ␬-opioids are unable paw inflammation are dependent on very early factors, to affect plasma extravasation,28 whereas others have e.g., modulation of substance P release from primary found the opposite.29,30 Opioids have been shown to afferent neurons or on the migration or activation of modulate the release of substance P, a potent mediator immune cells, induced rapidly on the initiation of inflam- of neurogenic inflammation.31–33 This may be a mecha- mation and inhibited by a single ␬-opioid treatment. nism, at least in part, by which opioids produce antiin- When treatment was delayed until inflammation was flammatory effects. Second, opioids could act via the established, i.e., 4 days after FCA injection, a significant immune system through the activation of opioid recep- reduction in paw volume was also observed with all tors that have been identified on various immune cells.34 ␬-opioids tested (figs. 5A–5C). Intriguingly, the antiin- Opioids can modulate immune cell proliferation, chemo- flammatory response appeared to be delayed by 24–48 h taxis, superoxide and cytokine production, and mast cell after injection of the ␬-opioids. This may be a result of degranulation.35 Opioids may diminish the synthesis or differential effects of these opioids on plasma extravasa- release of cytokines from macrophages and mast cells35 tion, edema, neurogenic inflammation, and immune or suppress T-lymphocyte function,36 downregulate the

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 PERIPHERAL OPIOID ANALGESIA AND ANTIINFLAMMATION 1043

showed marked antiinflammatory effects and no sign of hyperalgesia in FCA-treated animals. In summary, FE 200665 shows potent antinociceptive and antiinflammatory effects that are mediated via peripheral ␬-opioid receptors. The magnitude of the response produced by the peripherally selective FE 200665 was comparable to that of the standard ␬-opioid U-69,593, whereas FE 200666 was only weakly effective. Our findings suggest that ␬-opioid peptides such as FE 200665 may lead to improved analgesic–antiinflammatory therapy compared with centrally acting opioids or standard nonsteroidal antiinflammatory drugs.

References Fig. 9. Histologic assessment of the antiinflammatory effects of ␬ ␮ 1. Law PY, Loh HH: Regulation of opioid receptor activities. J Pharmacol Exp -opioid treatment and antagonism by nor-BNI (200 g in- Ther 1999; 289:607–24 traplantarly) or naloxone methiodide (NLXM; 10 mg/kg). His- 2. Minami M, Statoh M: Molecular biology of the opioid receptors: Structures, tologic scores judged by periarticular inflammtion, pannus, and functions and distributions. Neurosci Res 1995; 23:121–45 Signif- 3. Rivière PJM, Junien JL: Opioid receptors: Targets for new gastrointestinal* .(24 ؍ new bone formation (maximum obtainable score icant difference from controls, P < 0.05, analysis of variance drug development, Drug Development. Edited by Gaginella T, Guglietta A. St. (Dunnett test). Louis, Humana Press, 2000, pp 203–38 4. Horwell DC: Kappa opioid analgesics. Drugs Future 1988; 13:1061–70 5. Barber A, Gottschlich R: Opioid agonists and antagonists: An evaluation of expression of adhesion molecules, and reduce the mi- their peripheral actions in inflammation. Med Res Rev 1992; 12:525–62 37 6. Barber A, Gottschlich R: Central & peripheral nervous systems: Novel gration of immune cells into the injured tissue. Knock- developments with selective, non-peptidic kappa-opioid receptor agonists. Exp out mice provide unequivocal evidence for the involve- Opin Invest Drugs 1997; 6:1351–68 7. Stein C, Schäfer M, Cabot PJ, Carter L, Zhang Q, Zhou L, Gasior M: ment of opioids in immunosuppression. This has been Peripheral opioid analgesia. Pain Rev 1997; 342:321–4 38 8. Schäfer M: Peripheral opioid analgesia: From experimental to clinical stud- shown by Gaveriaux-Ruff et al., who found that the ies. Curr Opin Anaesthesiol 1999; 12:603–7 effects of morphine on immune responses were com- 9. Dooley CT, Ny P, Bidlack JM, Hougton RA: Selective ligands for the mu, ␮ delta, and kappa opioid receptors identified from a single mixture based tet- pletely absent in -opioid receptor–deficient mice. rapeptide positional scanning combinatorial library. J Biol Chem 1998; 273: Little difference was observed in the antiinflammatory 18848–56 10. Rivière PJ-M,Vanderah TW, Porreca F, Houghton R, Schteingart C, Trojnar effects of FE 200665, FE 200666, and U-69,593, assessed J, Junien J-L: Novel peripheral peptidic kappa agonists (abstract). ACTA Neuro- either by paw volume or histology (infiltration of inflam- biol Exp 1999; 59:181–270 11. Ackerman NR, Rooks WH, Shott L, Genant H, Maloney P, West E: Effects matory cells, new bone and pannus formation). The of naproxen on connective tissue changes in the adjuvant arthritic rat. Arthritis antiinflammatory effects of all the ␬-opioids tested were Rheum 1979; 22:1365–73 12. Bahar M, Rosen M, Vickers MD: Chronic cannulation of the intradural or completely reversed by nor-BNI and NLXM, indicating extradural space in the rat. Br J Anaesth 1984; 56:405–10 ␬ 13. Benter IF, Hirsh EM, Tuchman AJ, Ward PE: N-terminal degradation of low that these effects are mediated via -receptors in the molecular weight opioid peptides in human cerbrospinal fluid. Biochem Phar- periphery. This is in agreement with results of Wilson et macol 1990; 40:465–72 23 ␬ 14. Stein C: The control of pain in peripheral tissue by opioids. N Engl J Med al., who showed that the centrally acting -opioid 1995; 332:1685–90 U-50,488H had greater antiinflammatory effects when 15. Kayser V, Lee SH, Guilbaud G: Evidence for a peripheral component in the enhanced antinociceptive effect of a low dose of systemic morphine in rats with administered locally rather than centrally, and that this peripheral neuropathy. Neuroscience 1995; 64:537–45 effect was reversed by peripherally selective antagonists. 16. Burton MB, Gebhart GF: Effects of kappa-opioid receptor agonists on responses to colorectal distension in rats with and without acute colonic inflam- The ␬-opioid peptides investigated in this report repre- mation. J Pharmacol Exp Ther 1998; 285:707–15 sent a clear advancement compared with other currently 17. Houghton AK, Valdez JG, Westlund KN: Peripheral morphine administration blocks the development of hyperalgesia and allodynia after bone damage in available peripherally selective ␬-agonists, such as asim- the rat. ANESTHESIOLOGY 1998, 89:190–201 18. Catheline G, Guilbaud G, Kayser V: Peripheral component in the en- adoline. Asimadoline (EMD 61753) was previously inves- hanced antinociceptive effect of systemic U-69,593 a ␬-opioid receptor agonist in tigated in our laboratory, and although this ␬-agonist mononeuropathic rats. Eur J Pharmacol 1998; 357:171–8 19. Gyires K, Budavar I, Fürst, S, Molnar I: Morphine inhibits the carrageenan showed promise in early animal studies, our investiga- induced edema and the chemoluminescence of leucocytes stimulated by zymo- tions found that it was ineffective in patients who un- san. J Pharm Pharmacol 1985; 37:100–4 20. Wie ET, Thomas HA, Gjerde EA: Dynorphin A (6-12) analogues suppress derwent knee surgery. Indeed, patients tended to report thermal edema. Peptides 1998; 19:767–75 an increase in pain.39 In addition, asimadoline was 21. Sanderson K, Nyberg F, Khalil Z: Modulation of peripheral inflammation by locally administered hemorphin-7. Inflamm Res 1998; 47:49–55 shown to produce non–opioid-mediated hyperalgesic 22. Walker JS, Howlett CR, Nayanar V: Anti-inflammatory effects of kappa- and proinflammatory effects in FCA-treated rats.39 opioids in adjuvant arthritis. Life Sci 1995; 57:371–8 23. Wilson JL, Nayanar V, Walker JS: The site of anti-arthritic action of the Clearly, more effective peripheral ␬-opioid agonists are kappa agonist, U50488H: Importance of local administration. Br J Pharmacol required. In contrast, the current study found that FE 1996; 118:1754–60 24. Binder W, Walker JS: The peripherally selective kappa-opioid agonist, 200665 was not only a potent analgesic agent but also asimadoline, attenuates adjuvant arthritis. Br J Pharmacol 1998; 124:647–54

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016 1044 BINDER ET AL.

25. Binder W, Carmody J, Walker J: Effect of gender on anti-inﬂammatory and 33. Binder W, Scott C, Walker JS: Involvement of substance P in the anti- analgesic actions of two kappa-opioids. J Pharmacol Exp Ther 2000; 292:303–9 inﬂammatory effects of the peripherally selective ␬-opioid asimadoline and the

26. Catheline G, Le Guen S, Besson J-M: Effects of a k-opioid receptor agonist, NK1 antagonist GR205171. Eur J Neurosci 1999; 11:2065–72 on carrageenin-induced peripheral oedema and Fos expression in the rat spinal 34. Sharp BM, Roy S, Bidlack JM: Evidence for opioid receptor on cells cord. Eur J Pharmacol 1999; 370:287–96 involved in host defence and the immune system. J Neuroimmunol 1998; 83: ␮ 27. Walker JS, Chandler AK, Wilson JL, Binder W, Day RO: Effect of opioids 45–6 morphine and buprenorphine on the development of adjuvant arthritis in rats. 35. Bryant HU, Holaday JW: Opioids in immunologic processes, Handbook of Inflamm Res 1996; 45:557–63 Experimental Pharmacology: Opioids II. Edited by Herz A. Berlin, Springer- 28. Haley J, Ketchum S, Dickenson A: Peripheral k-opioid modulation of the Verlag, 1993, pp 361–92 formalin response: An electrophysiological study in the rat. Eur J Pharmacol 36. Thomas PT, Bhargava HN, House RV: Immunomodulatory effects of in 1990; 191:437–46 vitro exposure to morphine and its metabolites. Pharmacology 1995; 50:51–62 29. Barber A, Bartosyk GD, Bender HM, Gottschlich R, Greiner HE, Harting J, Mauler F, Minck KO, Murray RD, Simon M, Seyfried CA: A pharmacological 37. Wilson JL, Walker JS, Antoon, JS Perry MA: Intercellular adhesion mole- profile of the novel, peripherally-selective kappa-opioid receptor agonist, EMD cule-1 expression in adjuvant arthritis in rats: Inhibition by kappa-opioid agonist 61753. Br J Pharmacol 1994; 113:1317–27 but not by NSAID. J Rheumatol 1998; 25:499–505 30. Chahl LA, Chahl JS: Plasma extravasation induced by dynorphin- (1-13) in 38. Gaveriaux-Ruff C, Mathes HWD, Peluso J, Kieffer BL: Abolition of mor- rat skin. Eur J Pharmacol 1986; 124:343–7 phine-immunosuppression in mice lacking the ␮-opioid receptor gene. Proc Natl 31. Jessell TM, Iverson LL: Opiate analgesics inhibit substance P release from Acad SciUSA1998; 95:6326–30 rat trigeminal nucleus. Nature 1977; 268:549–51 39. Machelska H, Pflüger M, Weber W, Piranvisseh-Völk M, Daubert JD, De- 32. Yaksh TL: Substance P release from knee joint afferent terminals: Modu- haven R, Stein C: Peripheral effects of the ␬-opioid agonist EMD 61753 on pain lation by opioids. Brain Res 1988; 458:319–24 and inflammation in rats and humans. J Pharmacol Exp Ther 1999; 290:354–61

Anesthesiology, V 94, No 6, Jun 2001

Downloaded From: http://anesthesiology.pubs.asahq.org/pdfaccess.ashx?url=/data/Journals/JASA/931230/ on 07/08/2016