Pain 90 (2001) 271280

[Link]/locate/pain

Bee venom injection into an acupuncture point reduces arthritis associated edema and nociceptive responses
Young-bae Kwon a, Jae-dong Lee b, Hye-jung Lee b, Ho-jae Han c, Woung-chon Mar d, Sung-keel Kang b, Alvin J. Beitz e, Jang-hern Lee a,*
a

Department of Veterinary Physiology, College of Veterinary Medicine and School of Agricultural Biotechnology, Seoul National University, Suwon 441-744, South Korea b Department of Acupuncture and Moxibustion, College of Oriental Medicine, Kyung-Hee University, Seoul, South Korea c Hormone Research Center, Chonnam National University, Kwang-ju, South Korea d Natural Products Research Institute, Seoul National University, Seoul, South Korea e Department of Veterinary Pathobiology, College of Veterinary Medicine, University of Minnesota, St Paul, MN, USA Received 29 March 2000; received in revised form 9 August 2000; accepted 9 August 2000

Abstract Bee venom (BV) has traditionally been used in Oriental medicine to relieve pain and to treat inammatory diseases such as rheumatoid arthritis (RA). While several investigators have evaluated the anti-inammatory effect of BV treatment, the anti-nociceptive effect of BV treatment on inammatory pain has not been examined. Previous studies in experimental animals suggest that the therapeutic effect of BV on arthritis is dependent on the site of administration. Because of this potential site specicity, the present study was designed to evaluate the anti-nociceptive effect of BV injections into a specic acupoint (Zusanli) compared to a non-acupoint in an animal model of chronic arthritis. Subcutaneous BV treatment (1 mg/kg per day) was found to dramatically inhibit paw edema caused by Freund's adjuvant injection. Furthermore, BV therapy signicantly reduced arthritis-induced nociceptive behaviors (i.e. the nociceptive scores for mechanical hyperalgesia and thermal hyperalgesia). These anti-nociceptive/anti-inammatory effects of BV were observed from 12 days through 21 days postBV treatment. In addition, BV treatment signicantly suppressed adjuvant-induced Fos expression in the lumbar spinal cord at 3 weeks postadjuvant injection. Finally, injection of BV into the Zusanli acupoint resulted in a signicantly greater analgesic effect on arthritic pain as compared to BV injection in to a more distant non-acupoint. The present study demonstrates that BV injection into the Zusanli acupoint has both anti-inammatory and anti-nociceptive effects on Freund's adjuvant-induced arthritis in rats. These ndings raise the possibility that BV acupuncture may be a promising alternative medicine therapy for the long-term treatment of rheumatoid arthritis. q 2001 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved.
Keywords: Bee venom; Anti-nociception; Anti-inammation; Arthritis; Fos immunohistochemistry; Acupuncture

1. Introduction Rheumatoid arthritis (RA) is an autoimmune disorder of unknown etiology that is characterized by progressive joint destruction, deformity, disability and premature death in most patients. Currently, non-steroidal anti-inammatory drugs (NSAIDs) supplemented with steroid hormone remains the major recommended strategy for its treatment (Campbell, 1998). While these drugs transiently suppress inammation and ameliorate symptoms, they do not significantly improve the long-term disease outcome (Scott et al., 1998). Furthermore, long term treatment with NSAIDs may result in serious side effects, such as gastrointestinal ulcer* Corresponding author. Tel.: 182-31-290-2732; fax: 182-31-291-0536. E-mail address: JHL1101@[Link] (J.-h. Lee).

genicity and renal morbidity (Pincus et al., 1992). Because of these shortcomings, a more effective and safe therapeutic strategy is desired to treat RA. As a traditional alternative medicine approach, bee venom (BV) therapy has been utilized to relieve pain and to cure inammatory diseases such as RA in humans (Billingham et al., 1973). In experimental animals, the induction of arthritis is successfully suppressed by long-term BV treatment (Hadjipetrou-Kourounakis and Yiangou, 1984; Somereld and Brandwein, 1988). Interestingly, it has previously been suggested that the local application of a low dose of BV near the inamed site (hind limb) more potently inhibits the development of the Freund's adjuvant-induced RA than the administration of BV into more distant sites (such as an arbitrary site on the back) (Eiseman

0304-3959/01/$20.00 q 2001 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved. PII: S 0304-395 9(00)00412-7

272

Y.-b. Kwon et al. / Pain 90 (2001) 271280

et al., 1982; Hadjipetrou-Kourounakis and Yiangou, 1984). These observations suggest that the anti-inammatory effects of BV on arthritis may be dependent on the site of administration in the experimental animals. While there is suggestive evidence for site specicity in the anti-inammatory effects of BV, it is not known whether BV treatment has anti-nociceptive effects on arthritis, or whether this antinociceptive effect is dependent on the site of administration. These issues are addressed in the present study. In clinical practice, both electric stimulation and certain forms of acupuncture have been used to induce analgesia in local as well as widespread areas of the body (Kane and Taub, 1975; Vincent and Richardson, 1986). Therefore, acupuncture has been offered as an alternative treatment for the chronic pain syndromes related to RA and osteoarthritis in humans (Struthers et al., 1983; Berman et al., 1999). Recently, Chen et al. (1998) reported that subcutaneous BV injection into the peripheral cutaneous receptive eld resulted in a protracted, tonic monophase of increase in spike responses of wide-dynamic-range neurons for more than 1 h. Although the precise mechanism of acupuncture is not fully understood, it is possible that BV injection (stimulation) into an acupoint may produce a longer lasting and more potent analgesic effect than traditional acupuncture stimulation. We hypothesize that BV injection into a traditional acupuncture point (Zusanli) will attenuate RAinduced pain and that this anti-nociceptive effect will be greater than that produced by injection of BV into a nonacupoint. Complete Freund's adjuvant has been utilized to induce an arthritic immunopathological disease that displays many of the pathological features of human RA (Colpaert, 1987). Unilateral injection of Freund's adjuvant into the paw induces `primary' inammatory signs and hyperalgesia within hours at the site of inoculation. Subsequently, `secondary' inammation and pro-nociceptive signs appear between the 10th and 15th day post-inoculation, especially in the contralateral paw. The hyperalgesia associated with injection of Freund's adjuvant may persist for 8 weeks after adjuvant injection (Philippe et al., 1997). This adjuvantinduced arthritis model has been used extensively to analyze the anti-inammatory/anti-nociceptive effects of newly developed drugs on chronic arthritis (Colpaert, 1987). In the present study this adjuvant-induced arthritis model was used to evaluate the anti-arthritic effect of BV treatment. Changes in paw volume (edema) were measured as a parameter of inammation evoked by adjuvant injection. The potential anti-nociceptive effect of BV treatment on both thermal and mechanical hyperalgesia was also measured in both the ipsilateral and contralateral hind limbs following induction of chronic arthritis. In addition, the anti-nociceptive effect of BV on adjuvant-induced spinal cord Fos expression was analyzed using a computerized image analysis system as previously described (Abbadie and Besson, 1994). Finally, we investigated whether injection of BV into an acupoint near the inamed site had more

potent anti-nociceptive effects on chronic arthritis than injection into a more distant non-acupoint. 2. Materials and methods 2.1. Experimental animals Experiments were performed on 60 male Sprague Dawley rats (obtained from the Laboratory Animal Center of Seoul National University, South Korea) weighing 130 150 g at the beginning of the experiment. Animals were kept in a 12:12 h light-dark cycle (07:00 h onset) in a temperature controlled room (23 ^ 0:58C). Food and water were available ad libitum. The food was placed on the sawdust in the cage to minimize the need for animals to make potentially painful movements to obtain food. All of the methods used in the present study were approved by the Animal Care and Use Committee at Seoul National University and conform to NIH guidelines (NIH publication No. 8623, revised 1985). All algesiometric assays were conducted under the ethical guidelines set forth by the International Association for the Study of Pain (IASP) (Zimmermann, 1983). 2.2. The induction of arthritis Experimental animals were briey anesthetized with 3% isourane in a mixed N2O/O2 gas. Arthritis was induced by a single subcutaneous injection (50 ml) into the plantar surface of the right hind paw of heat-killed Mycobacterium butyricium (Difco Laboratory, Detroit, MI) suspended in sterile mineral oil (20 mg/ml). Control animals were similarly injected with sterile vehicle. 2.3. Experimental groups and bee venom treatments Experimental animals were divided into four groups: (1) saline-treated non-arthritic animals (Sham, n 10); (2) saline-treated arthritic group (RA-Sal, n 10); (3) BV-treated/Zusanli acupoint arthritic group (RA-BV/Z, n 10); or (4) BV-treated/non-acupoint-back arthritic group (RA-BV/ B, n 10). BV (Sigma, St. Louis, MO; Cat no. V3125; 1 mg/kg) dissolved in saline was administrated subcutaneously and bilaterally into the Zusanli point (RA-BV/Z) or into an arbitrary non-acupoint located near the midline on the back (RA-BV/B). The Zusanli point (ST36) was located 5 mm lower and lateral to the anterior tubercle of the tibia. Sham and RA-Sal animals were injected bilaterally into the Zusanli point with an equal volume of saline. BV treatment was started the day after adjuvant injection and animals were injected daily for a period of 3 weeks. All algesiometric assays were performed beginning 9 days after adjuvant injection at the time of induction of systemic arthritis (Philippe et al., 1997). 2.4. Evaluation of paw volume Paw volume of both the ipsilateral and contralateral hind

Y.-b. Kwon et al. / Pain 90 (2001) 271280

273

paws was measured in duplicate using a water displacement plethysmometer (UGO BASIL, Italy) every 3 days for 21 days after adjuvant injection and the mean values were recorded. Paw volume measured just prior to adjuvant injection was used as the control volume (day 0). Data were plotted as the change of paw volume versus control volume at each time point. Measurement of paw volume and all behavioral tests were performed blindly. 2.5. Ankle exion test Arthritis-induced hyperalgesia was evaluated by quantifying the total number of vocalizations evoked by ankle exion or ankle extension. The rat was held comfortably and gentle exion and then extension was applied to the ankle joint of each hind limb. The number of vocalizations emitted during both the exion and extension periods was then quantied. A exion and an extension stimulus was repeated every 5 s until a total of ve stimuli each had been delivered to each hind limb. A vocalization rating of 0 (no vocalization) or 1 (vocalization) was given in response to each exion or extension stimulus, depending on whether or not the animal vocalized. Thus, for each animal the vocalization rating ranged from 0 to 10 at each hind limb. For humane reasons, the ankle-exion test series was only performed once in each animal on any given test day. 2.6. Thermal hyperalgesia test (Hargreaves's Method) To assess nociceptive responses to thermal stimuli, paw withdrawal latency was tested using the procedure previously described by Hargreaves et al. (1988). Briey, rats were placed in a plastic chamber with a glass oor and allowed to acclimatize for 5 min before testing. A radiant heat source was positioned under the glass oor beneath the hind paw to be tested. The withdrawal latency of both hind paws was measured to the nearest 0.1 s using a photoelectric cell connected to a digital clock. The intensity of the light source was calibrated to produce withdrawal within 910 s in normal animals. The test was duplicated at 5-min intervals in each hind paw. 2.7. The mechanical hyperalgesia test The Randall-Selitto test was applied to evaluate mechanical hyperalgesia in arthritic animals. A graded mechanical force (g) was delivered through an analgesy meter (LETICA, LE7356) onto the convex surface of the paw. Rats withdrew their hind paw or vocalized when the applied force reached the nociceptive threshold. The test was duplicated at 5 min intervals in each hind paw. The threshold force in normal animals ranged from 140 to 160 g. 2.8. Fos immunohistochemistry One group of adjuvant injected rats was not subjected to any of the above nociceptive behavioral tests (n 5/group), but at the end of the experiments (21 days) were used for Fos

immunohistochemistry. The animals were deeply anesthetized with 5% isourane, perfused transcardially with calcium-free Tyrode's solution, followed by a xative containing 4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer (pH 6.9). All perfusions were performed between 09:00 h and 12:00 h. The spinal cord was then removed immediately after perfusion, post-xed in the same xative for 4 h and then cryoprotected in 30% sucrose in phosphate buffered saline (PBS, pH 7.4) for 48 h. Frozen serial frontal sections (40 mm) were cut through the lumbar L3-L5 spinal cord using a cryostat (Microm, Germany). After elimination of endogenous peroxidase activity with 0.3% hydrogen peroxide in PBS and preblocking with 1% normal goat serum and 0.3% triton X-100 in PBS, the free-oating sections were incubated in polyclonal rabbit anti-Fos antibody (Calbiochem, 1:10,000) at 48C overnight. The sections were subsequently processed using the avidin-biotin-peroxidase procedure previously described (Lee and Beitz, 1993; Kwon et al., 1999). Foslike immunoreactive (FLI) neurons were visualized using a 3-3 diamino-benzidine reaction intensied with 0.2% nickel chloride. 2.9. Imaging analysis of Fos like immunoreactive neurons Tissue sections were examined using darkeld microscopy (Zeiss Axioscope, Germany) to determine the segmental level according to Abbadie and Besson (1994) as well as the gray matter landmarks to dene individual spinal cord laminae. The sections were then examined under a brighteld microscope at 100 to localize FLI neurons. For quantitative analysis of FLI neurons, lumbar spinal cord sections were scanned and the ve with the greatest number of labeled cells at the L35 level were selected from each animal. Individual sections were digitized with 4096 gray levels using a cooled CCD camera (Micromax Kodak 1317; Princeton Instruments, Tucson, AZ) connected to a computer-assisted image analysis system (Metamorph; Universal Imaging, West Chester, PA). In order to maintain a constant threshold for each image and to compensate for subtle variability of the immunostaining, we only counted neurons that were at least 50% darker than the average gray level of each image after background subtraction and shading correction were performed. After adjusting the threshold image, individual neurons were considered to be specically Foslabeled when a total pixel area between 6 and 20 pixels was detected. The microscope illumination and data acquisition settings were xed throughout the entire analysis procedure. All Fos quantitation procedures described above were performed blindly with regard to the experimental condition of each animal. To assess the effect of BV treatment on spinal cord Fos expression, the following four gray matter regions were selected for analysis based on cytoarchitectonic criteria: (1) supercial dorsal horn (SDH, laminae I and II); (2) nucleus proprius (NP, laminae III and IV); (3) neck

274

Y.-b. Kwon et al. / Pain 90 (2001) 271280

(NECK, laminae V and VI); and (4) the ventral horn (VENT, laminae VIIIX). 2.10. Statistical analysis Thermal and mechanical hyperalgesia data were expressed as percent change and compared to that of the Sham group at each time point. Data were expressed as the mean ^ SEM. Repeated measures ANOVA were performed to determine the overall effect. Paired t-tests were then used to determine probability values when repeated measures ANOVAs indicated a signicant drug effect. Throughout, P , 0:05 was considered to be statistically signicant. 3. Results 3.1. Evaluation of paw volume Injection of adjuvant into the right hind paw produced a `primary' swelling of the injected paw that was rst measured 3 days after adjuvant injection (Fig. 1a). A `secondary' response, characterized by swelling of the non-injected contralateral hind paw and tail, was rst evident at 12 days post-injection (Fig. 1b). The incidence ratio of arthritis in the contralateral paw was 100% in the RA-Sal group. In the RA-BV/Z group, a signicant decrease in paw volume was observed in both hind limbs as compared with those of the RA-Sal group (Fig. 1). In addition, the RA-BV/Z group showed a signicant decrease in paw volume of the ipsilateral hind limb in comparison to that of the RA-BV/B group (P , 0:05; Fig. 1a). In the RABV/B group, only the contralateral paw volume was significantly decreased as compared to that of the RA-Sal group (P , 0:05; Fig. 1b). 3.2. Ankle exion test Gentle exion and extension of the inamed ankle elicited vocalizations that were recorded as a measure of hyperalgesia. In the RA-Sal group, the vocalizations caused by exion or extension of the ipsilateral hind limb increased gradually from 9 days post-adjuvant injection through to the end of the experiment (Fig. 2a). The number of vocalizations caused by ipsilateral hind limb exion or extension was decreased in the RA-BV/Z group, whereas this suppressive effect was not evident in the RA-BV/B group (Fig. 2a). In the contralateral limb of the RA-Sal group rats, pain related vocalization increased gradually from 12 days post-adjuvant injection (Fig. 2b). BV treatments signicantly decreased vocalization induced by exion or extension of the contralateral limb in rats from both the RA-BV/Z and RA-BV/B groups (P , 0:05; Fig. 2b). 3.3. Thermal hyperalgesia test The paw withdrawal latency (PWL) of animals in the

Sham group did not change signicantly throughout the 21 days of the experiment. In contrast, both the ipsilateral and contralateral PWL of animals in the RA-Sal group decreased signicantly and remained lower throughout the experiment. There was a signicant decrease in the PWL of the ipsilateral hind paw of all the arthritis-induced groups compared to the Sham group, beginning 9 days after adjuvant injection (Fig. 3a). However, the ipsilateral PWL of the RA-BV/Z group showed a signicantly higher thermal threshold as compared with those of the RA-Sal and RABV/B groups (P , 0:05; Fig. 3a). On the contralateral side, the PWL of the RA-BV/Z group remained similar to that of the Sham group (Fig. 3b). The contralateral PWL of the RABV/B group initially decreased (similar to the RA-Sal group), but gradually returned to the level of the Sham group by 21 days (Fig. 3b). 3.4. Mechanical hyperalgesia test There was a signicant bilateral decrease in the hind paw mechanical pain threshold of animals in the RA-Sal group as compared to those in the Sham group (P , 0:05; Fig. 4a,b). In the ipsilateral paw, the pain threshold of the RA-

Fig. 1. Effect of bee venom (BV) on the change in paw volume in nonarthritic animals (Sham group, n 10), saline-treated arthritic animals (RA-Sal, n 10) and BV-treated arthritic animals (RA-BV). BV (1 mg/ kg) was administrated into either the Zusanli acupoint (RA-BV/Z, n 10) or an arbitrary non-acupoint site on the back (RA-BV/B, n 10). (a) Depicts the change in the paw volume of the ipsilateral right hind limb. (b) Depicts the change in the paw volume of the contralateral left hind limb. *P , 0:05: signicantly different from RA-Sal group. 1P , 0:05: signicant difference between the RA-BV/Z and RA-BV/B groups.

Y.-b. Kwon et al. / Pain 90 (2001) 271280

275

ipsilateral NP, NECK and VENT regions of the RA-BV/Z group were signicantly reduced compared to the RA-Sal group (P , 0:05; Figs. 5a and 6C). Interestingly the number of FLI neurons was signicantly reduced in all regions of contralateral spinal cord of animals in the both the RA-BV/ B and RA-BV/Z groups as compared with those in the RASal group (P , 0:05; Fig. 5b). 4. Discussion In this study, unilateral Freund's adjuvant injection into the right paw was found to evoke paw edema initially in the ipsilateral limb and subsequently in the contralateral limb. The appearance of edema followed a time course similar to that previously described in other studies (Eiseman et al., 1982; Philippe et al., 1997). The development of arthritic associated thermal and mechanical hyperalgesia follows the induction of systemic arthritis beginning at approximately 9 days post-adjuvant injection as previous reported by Philippe et al. (1997). Hyperalgesia was found to persist up to at least 3 weeks post-injection as previously reported

Fig. 2. Effect of bee venom (BV) on the ankle exion/extension score in the saline-treated arthritic group (RA-Sal, n 10) and BV-treated arthritic groups (RA-BV). BV (1 mg/kg) was administrated into either the Zusanli acupoint (RA-BV/Z, n 10) or an arbitrary site on the back (RA-BV/B, n 10). (a) Mean score from the ipsilateral right hind limb. (b) Mean score from the. contralateral left hind limb. *P , 0:05: signicantly different from the RA-Sal group.

BV/B group was not statistically different from that of the RA-Sal group (Fig. 4a). In contrast, the pain threshold of the ipsilateral paw was signicantly increased in the RA-BV/Z group as compared to the RA-Sal group (P , 0:05; Fig. 4a). Finally, the pain threshold of the contralateral paw in both the RA-BV/Z and RA-BV/B groups remained similar to that of the Sham group (Fig. 4b). 3.5. Fos immunohistochemistry Very few FLI neurons were observed in the lumbar spinal cord of the Sham group (Figs. 5a,b and 6A). In contrast, a signicantly higher number of FLI neurons were observed in both the ipsilateral (Fig. 6B) and contralateral sides of the L3-L5 lumbar spinal cord segments in rats from the RA-Sal group 21 days after unilateral Freund's adjuvant injection. Approximately 50% of FLI neurons were distributed in the NECK region of the dorsal horn at this timepoint, whereas smaller numbers of FLI neurons were observed in the SDH, NP and VENT regions of the spinal cord (Figs. 5a,b and 6B). The number of FLI neurons in the ipsilateral spinal cord of animals in the RA-BV/B group was not signicantly different from that of animals in the RA-Sal group (Figs. 5a and 6D). In comparison the number of FLI neurons in the

Fig. 3. Effect of bee venom (BV) on paw withdrawal latency (PWL) produced by noxious heat stimuli in the non-arthritic group (Sham, n 10), the saline-treated arthritic group (RA-Sal, n 10) and the BVtreated arthritic groups (RA-BV). BV (1 mg/kg) was administrated into either the Zusanli acupoint (RA-BV/Z, n 10) or an arbitrary site on the back (RA-BV/B, n 10). (a) Percentage inhibition of arthritic PWL compared to the Sham PWL in the ipsilateral right hind paw. (b) Percentage inhibition compared to the Sham PWL in the contralateral left hind paw. *P , 0:05: signicantly different from RA-Sal group. 1P , 0:05: signicantly different from the RA-BV/B group.

276

Y.-b. Kwon et al. / Pain 90 (2001) 271280

Individual components of BV have only been tested to verify the anti-inammatory effect of BV treatment on RA. BV consists of a variety of different peptides including melittin, apamin, adolapin and mast cell degranulating (MCD) peptide (Billingham et al., 1973). Although adolapin (20 mg/kg) and puried MCD peptide (1 mg/kg) have antiinammatory activity (Martin and Hartter, 1980; Shkenderov and Koburova, 1982; Koburova et al., 1985), these substances are present in very small quantities (12%) in whole BV. It is possible, however, that these peptides play a minor role in the anti-inammatory effect of BV at the dose used in the present study. Recently, Saini et al. (1997) reported that melittin is a major component of BV (50% of dry weight) and that it binds to secretory phospholipase A2 (PLA2) and inhibits its enzymatic activity. Because PLA2 is a major inammatory trigger (i.e. it causes arachidonic acid release) whose activity is enhanced in RA, it is possible that the formation of melittin-PLA2 complex by BV injection is able to suppress some of the symptoms associated with the development of arthritis. However, other investi-

Fig. 4. Effect of bee venom (BV) on mechanical threshold (Randall-Selitto test) in non-arthritic animals (Sham, n 10), saline-treated arthritic group (RA-Sal, n 10) and BV-treated arthritic groups. BV (1 mg/kg) was administrated into either the Zusanli acupoint (RA-BV/Z, n 10) or an arbitrary non-acupoint site on the back (RA-BV/B, n 10). (a) Percentage inhibition versus Sham mechanical threshold in the ipsilateral right hind paw. (b) Percentage inhibition versus Sham mechanical threshold in the contralateral left hind paw. *P , 0:05: signicantly different from RA-Sal group. 1 P , 0:05: signicantly different from the RA-BV/B group.

by other investigators (Abbadie and Besson, 1992; Philippe et al., 1997). Since the development and persistence of edema and hyperalgesia found in the present study parallel that reported by other investigators, we believe this animal model of persistent arthritis is useful for evaluating the antinociceptive/anti-inammatory effect of BV therapy. 4.1. Anti-inammatory effect of bee venom on adjuvantinduced arthritis Several investigators have previously reported that BV is a useful therapeutic agent to manage arthritis-induced edema in experimental animal models. For example, the induction of arthritic inammation by adjuvant injection into the base of the tail is successfully suppressed by bilateral intramuscular BV injection (0.5 mg/kg) into the hind limb (Hadjipetrou-Kourounakis and Yiangou, 1984). In addition, Eiseman et al. (1982) reported that BV treatment reduced paw edema in a unilateral adjuvant injected arthritis model. In the present study, we also observed that paw edema evoked by adjuvant injection was signicantly suppressed by BV therapy.

Fig. 5. Histograms showing the number of Fos-like immunoreactive (FLI) neurons in non-arthritic animals (Sham, n 5), saline-treated arthritic group (RA-Sal, n 5) and bee venom (BV)-treated arthritic groups at three weeks post-adjuvant injection. BV (1 mg/kg) was administrated into either the Zusanli acupoint (RA-BV/Z, n 10) or an arbitrary non-acupoint site on the back (RA-BV/B, n 5). (a) The number of FLI neurons in the gray matter on the ipsilateral side (right) of the spinal cord. (b) The number of FLI neurons in the gray matter on the contralateral (left) side of the spinal cord. 1P , 0:05: signicantly different from Sham group. *P , 0:05: signicantly different from RA-Sal group. SDH, supercial dorsal horn (laminae III); NP, nucleus proprius (laminae IIIIV); NECK, neck of dorsal horn (laminae VVI); VENT, ventral horn (laminae VIIIX).

Y.-b. Kwon et al. / Pain 90 (2001) 271280

277

Fig. 6. Effect of bee venom (BV) treatment on the Fos expression in the ipsilateral (right) lumbar spinal cord induced by unilateral adjuvant injection into the plantar surface of the right hind paw (3 weeks post-injection). (A) Photomicrograph of a representative spinal cord section illustrating Fos expression in a saline-treated normal animal. The number of Fos immunoreactive cells was analyzed in each of four regions (SDH, NP, NECK and VENT) as described in the materials and methods. (B) Spinal cord Fos expression in a saline treated arthritic animal. (C) Spinal cord Fos expression in a BV (Zusanli acupoint)-treated arthritic animal. (D) Spinal cord Fos expression in a BV (arbitrary non-acupoint site on the back)-treated arthritic animal. Scale bar, 200 mm. SDH, supercial dorsal horn (laminae III); NP, nucleus proprius (laminae IIIIV); NECK, neck of dorsal horn (laminae VVI); VENT, ventral horn (laminae VIIIX).

gators report that melittin injection into mouse paw elicits paw edema at 60 min after injection (Hartman et al., 1991). Furthermore, whole BV injected into hind paw also

produces local inammation that lasts for more than 48 h (Lariviere and Melzack, 1996). It is clear at this point in time that further investigation is required to understand the

278

Y.-b. Kwon et al. / Pain 90 (2001) 271280

mechanisms underlying the anti-inammatory effect of BV in RA. One of the inevitable side effects of BV therapy is a potential allergic reaction that has an incidence of about 3% in the general population (Golden, 1989). However, it has also been demonstrated that long-term BV treatment induces T-cell hypo-responsiveness to the allergen and modulation of cytokine secretion in humans (Kammerer et al., 1997). Furthermore, long-term BV therapy for 18 days reduces the population of splenocytes that produce interleukin-1 and interleukin-2 in both normal and arthritic rats (Hadjipetrou-Kourounakis and Yiangou, 1988). Thus, hyposensitization induced by BV treatment is currently recognized as a successful long-term therapy for BV allergy in humans (Valetine, 1992). Since the side effects of long-term BV therapy are negligible, it may be one of the safer therapeutic approaches for the treatment of arthritis. Interestingly, whole BV, but not its individual components, produces an early and persistent increase in plasma corticosterone concentration (Zurier et al., 1973). Since RA appears to be evoked by an autoimmune reaction against cartilage self-antigens, modication of cytokine and corticosteroid levels by BV therapy could inhibit immune cell attack of the synovial cartilage. However, the precise effect of BV on immune and endocrine function and the relationship of these effects to the alleviation of arthritic symptoms remain to be dened. 4.2. Anti-nociceptive effect of bee venom on adjuvantinduced arthritis In this study, we also demonstrated that the arthritic pain score was normalized to that of non-arthritic animals by BV therapy. Moreover, the thermal and mechanical hyperalgesia induced by adjuvant injection were also dramatically alleviated by long-term BV therapy. The anti-nociceptive effect of BV therapy was further conrmed by examining Fos immunohistochemistry in the spinal cord. It was previously reported that arthritis-induced nociception signicantly increased the number of FLI-neurons in the lumbar spinal cord 3 weeks after adjuvant injection (Abbadie and Besson, 1992). This increased number of spinal cord FLI-neurons is remarkably reduced by treatment with NSAIDs, such as aspirin and acetaminophen (Abbadie and Besson, 1994). In the present study, we observed that BV treatment also dramatically suppressed the increase in the number of spinal cord FLI neurons induced by arthritic pain. The decrease in the number of FLI neurons induced by BV treatment paralleled the increase in anti-nociception indicated by the behavioral data. Collectively these data suggest that BV therapy has a potent anti-nociceptive effect on adjuvant-induced arthritic pain. The anti-nociceptive effect of whole BV can be partially reproduced by some of its individual components, such as adolapin. It was previously reported that adolapin can exert a potent analgesic effect in both the acetic acid-induced

writhing assay and in the Randall-Sellito test (Shkenderov and Koburova, 1982). On the other hand, the anti-nociceptive effects of BV may also be explained by the process of counter-irritation, i.e. when noxious stimuli are applied to body regions, these stimuli increase the pain thresholds and reduce pain-rating scores throughout the body. For centuries, pain has been relieved by counter-irritation methods, including moxibustion (burning herbs to stimulate acupuncture points) on an arthritic limb or cauterizing above the hip for sciatica (Wand-Tetley, 1956). In addition, capsaicin, an extract of red, hot chili peppers now known to stimulate vanilloid receptors, produces itching, pricking and burning sensations caused by the excitation of nociceptors. Since repeated application of capsaicin is followed by a prolonged period of hypoalgesia, usually referred to as desensitization or nociceptor inactivation, it has been traditionally used as a counter-irritant to treat certain types of pain (Fusco and Giacovazzo, 1997; Nolano et al., 1999). It has been reported that unilateral subcutaneous injection of BV into the plantar surface of the hind paw prolongs tonic behavioral responses similar to that of capsaicin (Lariviere and Melzack, 1996) and induces Fos expression in the ipsilateral dorsal horn of the spinal cord (Luo et al., 1998). In addition, blockage of the sciatic nerve with lidocaine resulted in a complete suppression of the BV-induced neuronal ring in the spinal dorsal horn, suggesting that the central neuronal changes following subcutaneous BV are peripherally-dependent (Chen et al., 1998). Thus the initial irritation caused by BV administration may eventually lead to nociceptor inactivation or hypoalgesia similar to that reported for capsaicin. This is supported by the fact that the anti-nociceptive effect of BV was not observed until 11 days after BV therapy was initiated (12 days post adjuvant injection, see Fig. 3). Although it remains to be determined whether peripherally administered BV activates endogenous anti-nociceptive pathways, it is possible that the antinociceptive effect of BV is partially mediated by peripheral counter-irritation. 4.3. The effect of bee venom injection into the Zusanli acupoint It was previously reported that the local administration of certain anti-arthritic drugs has a more potent therapeutic efcacy than that achieved by systemic injection (Wilson et al., 1996). Similarly, administration of a low dose of BV at the inamed site more potently inhibits the development of the Freund's adjuvant-induced arthritis than the administration of BV into a site at some distance from the site of inammation (Eiseman et al., 1982; Hadjipetrou-Kourounakis and Yiangou, 1984). In the present study we show that BV injection into the Zusanli acupoint produced a much more potent anti-nociceptive/anti-inammatory effect on the `primary site' of inammation than BV injection into a non-acupoint site on the midline of the back. However, it was interesting that the site of BV injection (Zusanli

Y.-b. Kwon et al. / Pain 90 (2001) 271280

279

acupoint versus non-acupoint) made no difference with respect to BV's anti-arthritic effects on the `secondary site' of inammation (the contralateral limb). It is unclear why BV shows a different therapeutic efcacy between the primary and secondary inammation sites in this unilateral adjuvant injected arthritis model. We assume that BV stimulation of the Zusanli acupoint is involved in the anti-arthritic effect of BV at the primary inammation site. The fact that both sites of BV administration were effective in reducing edema at the secondary inammation site is puzzling and implies that BV may have a possible systemic effect on either immune or endocrine function that affects the contralateral hind limb. In clinical practice, electric stimulation and certain types of acupuncture have been used to induce analgesia in local and distant areas of the body (Kane and Taub, 1975; Vincent and Richardson, 1986). Recently, acupuncture has been used successfully to alleviate the chronic pain symptoms associated with the development of RA and osteoarthritis in humans (Struthers et al., 1983; Berman et al., 1999). Since acupuncture points appear to be sites of neuronal innervation, it is assumed that stimulation of an acupoint may trigger the endogenous pain inhibitory system, or modify the local release of neurotransmitters and neuropeptides, such as substance P (Chan et al., 1998; Sprott et al., 1998; Ulett et al., 1998). While acupuncture is traditionally applied using electrical or mechanical stimulation, the results of the present study suggest that BV administration directly into an acupoint may also be an effective means of stimulation. While these results indicate that BV injection into the Zusanli acupoint is effective in suppressing the nociception associated with the development of arthritis, it is clear that further research is required to determine the mechanisms underlying the antinociception produced by BV stimulation. In summary, the present study demonstrated that longterm BV therapy signicantly reduces arthritis-induced paw edema. More importantly BV administration into the Zusanli acupoint produced a signicant anti-nociceptive effect on arthritis-induced inammatory pain symptoms including thermal and mechanical hyperalgesia. Further supporting data for an anti-nociceptive effect of BV was obtained using Fos immunohistochemistry which demonstrated that arthritis-induced Fos expression in the lumbar spinal cord was signicantly decreased in response to BV administration. Based on the present results, we propose that BV may be a promising candidate for long-term treatment of RA-induced pain and inammation. In addition, we show that BV administration directly into an acupoint provides a potent strategy for reducing the pain and inammation associated with arthritis. Acknowledgements This study was supported by grants from the 1999 Orien-

tal Medicine 2010 R&D project (HMP-99-O-01-0001) from the Ministry of Health and Welfare. The publication of this manuscript was also supported by a Research Fund from the Research Institute for Veterinary Science (RIVS) in the College of Veterinary Medicine, Seoul National University, as well as the Brain Korea 21 project.

References
Abbadie C, Besson JM. c-fos expression in rat lumbar spinal cord during the development of adjuvant-induced arthritis. Neuroscience 1992;48:985993. Abbadie C, Besson JM. Chronic treatments with aspirin or acetaminophen reduce both the development of polyarthritis and Fos-like immunoreactivity in rat lumbar spinal cord. Pain 1994;57:4554. Berman BM, Singh BB, Lao L, Langenberg P, Li H, Hadhazy V, Bareta J, Hochberg M. A randomized trial of acupuncture as an adjunctive therapy in osteoarthritis of the knee. Rheumatology 1999;38:346354. Billingham ME, Morley J, Hanson JM, Shipolini RA, Vernon CA. Letter: an anti-inammatory peptide from bee venom. Nature 1973;245:163 164. Campbell SM. Rheumatoid arthritis: current strategies. Hosp Med 1988;34:2932. Chan WW, Weissensteiner H, Rausch WD, Chen KY, Wu LS, Lin JH. Comparison of substance P concentration in acupuncture points in different tissues in dogs. Am J Chin Med 1998;26:1318. Chen J, Luo C, Li HL. The contribution of spinal neuronal changes to development of prolonged, tonic nociceptive responses of the cat induced by subcutaneous bee venom injection. Eur J Pain 1998;2:359376. Colpaert FC. Evidence that adjuvant arthritis in the rat is associated with chronic pain. Pain 1987;28:201222. Eiseman JL, von Bredow J, Alvares AP. Effect of honeybee (Apis mellifera) venom on the course of adjuvant-induced arthritis and depression of drug metabolism in the rat. Biochem Pharmacol 1982;31:11391146. Fusco BM, Giacovazzo M. Peppers and pain: the promise of capsaicin. Drugs 1997;53:909914. Golden DB. Epidemiology of allergy to insect venoms and stings. Allergy Proc 1989;10:103107. Hadjipetrou-Kourounakis L, Yiangou M. Bee venom and adjuvant induced disease. J Rheumatol 1984;11:720. Hadjipetrou-Kourounakis L, Yiangou M. Bee venom, adjuvant induced disease and interleukin production. J Rheumatol 1988;15:11261128. Hargreaves K, Dubner R, Brown F, Flores C, Joris J. A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia. Pain 1988;32:7788. Hartman DA, Tomchek LA, Lugay JR, Lewin AC, Chau TT, Carlson RP. Comparison of antiinammatory and antiallergic drugs in the melittinand D49 PLA2-induced mouse paw edema models. Agents Actions 1991;34:8488. Kammerer R, Chvatchko Y, Kettner A, Dufour N, Corradin G, Spertini F. Modulation of T-cell response to phospholipase A2 and phospholipase A2-derived peptides by conventional bee venom immunotherapy. J Allergy Clin Immunol 1997;100:96103. Kane K, Taub A. A history of local electrical analgesia. Pain 1975;1:125 138. Koburova KL, Michailova SG, Shkenderov SV. Further investigation on the antiinammatory properties of adolapin: bee venom polypeptide. Acta Physiol Pharmacol Bulg 1985;11:5055. Kwon YB, Yoon YS, Han HJ, Lee JH. Effect of 7-nitroindazole, a selective neuronal nitric oxide synthase inhibitor, on parvalbumin immunoreactivity after cerebral ischaemia in the hippocampus of the Mongolian gerbil. Anat Histol Embryol 1999;28:325329.

280

Y.-b. Kwon et al. / Pain 90 (2001) 271280 strategies for improving clinical effectiveness. Br J Rheumatol 1998;37:546554. Shkenderov S, Koburova K. Adolapin: a newly isolated analgetic and antiinammatory polypeptide from bee venom. Toxicon 1982;20:317 321. Somereld SD, Brandwein S. Bee venom and adjuvant arthritis [letter]. J Rheumatol 1988;15:1878. Sprott H, Franke S, Kluge H, Hein G. Pain treatment of bromyalgia by acupuncture. Rheumatol Int 1998;18:3536. Struthers GR, Scott DL, Scott DG. The use of `alternative treatments' by patients with rheumatoid arthritis. Rheumatol Int 1983;3:151152. Ulett GA, Han S, Han JS. Electroacupuncture: mechanisms and clinical application. Biol Psychiatry 1998;44:129138. Valetine MD. Insect venom allergy: diagnosis and treatment. J Allergy Clin Immunol 1992;89:11891195. Vincent CA, Richardson PH. The evaluation of therapeutic acupuncture: concepts and methods. Pain 1986;24:113. Wand-Tetley JI. Historical methods of counter-irritation. Ann Phys Med 1956;3:9098. Wilson JL, Nayanar V, Walker JS. The site of anti-arthritic action of the kappa-opioid, U-50, 488H, in adjuvant arthritis: importance of local administration. Br J Pharmacol 1996;118:17541760. Zimmermann M. Ethical guidelines for investigations of experimental pain in conscious animals [editorial]. Pain 1983;16:109110. Zurier RB, Mitnick H, Bloomgarden D, Weissmann G. Effect of bee venom on experimental arthritis. Ann Rheum Dis 1973;32:466470.

Lariviere WR, Melzack R. The bee venom test: a new tonic-pain test. Pain 1996;66:271277. Lee JH, Beitz AJ. The distribution of brain-stem and spinal cord nuclei associated with different frequencies of electroacupuncture analgesia. Pain 1993;52:1128. Luo C, Chen J, Li HL, Li JS. Spatial and temporal expression of c-Fos protein in the spinal cord of anesthetized rat induced by subcutaneous bee venom injection. Brain Res 1998;806:175185. Martin W, Hartter P. Basic peptides in bee venom. VI. Structure-activity studies on the anti-inammatory effects of derivatives and fragments of the MCD- peptide. Hoppe Seylers Z Physiol Chem 1980;361:525535. Nolano M, Simone DA, Wendelschafer-Crabb G, Johnson T, Hazen E, Kennedy WR. Topical capsaicin in humans: parallel loss of epidermal nerve bers and pain sensation. Pain 1999;81:135145. Philippe L, Gegout-Pottie P, Guingamp C, Bordji K, Terlain B, Netter P, Gillet P. Relations between functional, inammatory, and degenerative parameters during adjuvant arthritis in rats. Am J Physiol 1997;273:R1550R1556. Pincus T, Marcum SB, Callahan LF. Longterm drug therapy for rheumatoid arthritis in seven rheumatology private practices. II. Second line drugs and prednisone. J Rheumatol 1992;19:18851894. Saini SS, Peterson JW, Chopra AK. Melittin binds to secretory phospholipase A2 and inhibits its enzymatic activity. Biochem Biophys Res Commun 1997;238:436442. Scott DL, Shipley M, Dawson A, Edwards S, Symmons DP, Woolf AD. The clinical management of rheumatoid arthritis and osteoarthritis: