The cyclic peptide WP9QY (YCWSQYLCY),which was designed to mimic the most criticaltumor necrosis factor (TNF) ÏÏ recognition loop ontype1 TNF receptor, antagonizes the effects ofTNFÏÏ. In this study, we investigated the effects ofWP9QY peptide on collagen- induced arthritis(CIA) mice to evaluate its effects on inflammatorybone destruction. DBA/1J mice were injectedintradermally at the base of the tail with bovine typeII collagen, emulsified in complete Freund’’s adju-vant on day 0 and 21. The three sets of WP9QYpeptide injections (24 mg/kg×8 times per day)were performed before the onset of paw swelling.Mice were sacrificed at day 38 and thereafter, thearthritis scores as well as radiographical and his-tological outcomes were assessed. WP9QY peptideinhibited CIA- induced increase in the arthritisscore. Furthermore, histomorphometric analysis ofthe tibial epiphysis region revealed that WP9QYpeptide inhibited the increase of synovial pannus

infiltration and the decrease of bone volume,which were induced by the CIA. The WP9QY treat-ment prevented the inflammation as well as bonedestruction of the joints in the CIA mice, suggest-ing that the administration of WP9QY peptidemight be useful for developing a drug to preventinflammatory bone destruction.

Key words: WP9QY peptide, TNFÏ, antagonist,inflammation, bone resorption

Introduction

Rheumatoid arthritis (RA) is an autoimmune diseasethat leads to chronic inflammation of multiple joints andsubsequent destruction of cartilage and bone1,2.Proinflammatory cytokines such as tumor necrosisfactor Ï (TNFÏ), interleukin-1 (IL-1), and interleukin-6(IL-6) play a crucial role in the pathogenesis of RA3.Especially, TNFÏ occupies a pivotal position in acytokine cascade that regulates the production ofinflammatory mediators such as IL-1, IL-6 in the syn-ovial membrane culture system4,5.

Collagen- induced arthritis (CIA) is a murine model ofhuman RA which is currently used for its analysis. In

Original Article

Subcutaneous Injections of a TNF-ÏÏ Antagonistic Peptide Inhibit BothInflammation and Bone Resorption in Collagen-Induced Murine Arthritis

Takefumi Kojima1,2, Kazuhiro Aoki2, Kiichi Nonaka2, Hiroaki Saito2, Miyuki Azuma3, Hideyuki Iwai3,Bobby J Varghese4, Hidemi Yoshimasu1, Roland Baron5, Keiichi Ohya2 and Teruo Amagasa1

1) Maxillofacial Surgery, Department of Maxillofacial Reconstruction and Function, Graduate School, TokyoMedical and Dental University2) Pharmacology, Department of Hard Tissue Engineering, Graduate School, Tokyo Medical and DentalUniversity3) Molecular Immunology, Department of Oral Restitution, Graduate School, Tokyo Medical and DentalUniversity4) Developmental Oral Health Science, Department of Orofacial Development and Fanction, Graduate School,Tokyo Medical and Dental University 5) Department of Orthopedics and Cell Biology, Yale University of Medicine, New Haven, CT, USA

J Med Dent Sci 2005; 52: 91–99

Corresponding Author: Kazuhiro AokiDepartment of Hard Tissue Engineering (Pharmacology), GraduateSchool, Tokyo Medical and Dental University, 1-5-45, Yushima,Bunkyo-ku, Tokyo 113-8549, JapanE-mail: kazu.hpha@tmd.ac.jpReceived October 25; Accepted December 3, 2004

the CIA mice, the etiology is associated with local over-production of TNFÏ. The TNFÏ mRNA level wasupregulated before the onset of the swelling of thepaws6. Immunohistochemical observations identifiedTNFÏ protein in the pannus at all stages of the dis-ease7. Thus, TNFÏ seems to be an importantcytokine in the pathogenesis of RA inducing thechronic inflammatory state6,8 and several TNFÏ-blocking therapies ameliorated the development ofinflammation and disease activity in RA9,10.

The WP9QY peptide (Tyr1- Cys2- Trp3- Ser4- Glu5-Tyr6- Leu7- Cys8-Tyr9) has a disulfide bond betweenCys2 and Cys8 and composes a cyclic peptide. It isdesigned to mimic the loop 1 of domain 3 on type 1TNF receptor (TNFR(I)), which is the most critical sitefor TNFÏ recognition. Twenty five ÒM of WP9QY pep-tide inhibits the TNFÏ binding to TNFR(I) as a sameextent as 1 nM of soluble TNF- receptor in a competi-tive radioreceptor assay11. WP9QY peptide protecteddose-dependently from TNFÏ- induced cells death invitro11. Recently Suzuki et al. reported that WP9QYpeptide dose- dependently inhibits inflammatory boneresorption induced by Porphyromonas gingivalisinfection in mice, suggesting that WP9QY peptideworks as a TNFÏ antagonist, in vivo12. Furthermore,WP9QY peptide also prevented the bone loss inovariectomized (OVX) mice, where TNFÏ is involved inthe bone resorption. These results suggested thatWP9QY peptide has an inhibitory effect on TNFÏ-mediated bone resorption in vivo13.

In this report, the effects of WP9QY peptide on theCIA mice were investigated if this peptide has beneficialeffects on inflammatory bone destruction. We foundthat the WP9QY peptide inhibits both inflammation andbone destruction in the CIA mice.

Materials and Methods

MaterialsThe WP9QY peptide was synthesized and its activi-

ty against TNFÏ/TNFR(I) interaction was assayed asdescribed elsewhere11. Bovine type II collagen wasobtained from the Collagen Research Center (Tokyo,Japan) and complete Freund’s adjuvant from Difco(Detroit, MI). Specific pathogen- free 6 week- oldmale DBA/1J mice were purchased from JapanCharles River Breeding Laboratories (Kanagawa,Japan). Prior to the experiment, all mice were allowedto have ad libitum the food (Oriental Yeast Co.,Ltd.,Tokyo, Japan) and water for 7 days for acclimatization.

They were maintained in a 12- hour light and darkcycle. The experimental procedures were reviewed andapproved by the Animal Care and Use Committee ofthe Tokyo Medical and Dental University.

Induction of CIAThe induction and assessment of CIA were per-

formed as previously described14. Briefly, 200Ògbovine Type II Collagen in 0.05 M acetic acid emulsifiedin complete Freund’s adjuvant was intradermallyinjected at the base of the tail to the male DBA/1J mice(7 week- old). Twenty-one days after the primaryimmunization, the mice were boosted in the same way.

Treatment with WP9QY peptide The WP9QY peptide was dissolved in NaOH, neu-

tralized in NaH2PO4, and buffered in PBS. The threesets of vehicle (the mixture of NaOH, NaH2PO4, PBS)and WP9QY peptide (24 mg/kg×8 times per day)injections (70Òl/mouse) were performed subcuta-neously, before the first appearance of clinical signs ofdisease (at day 22, 24, and 26). In our previous study,this dose was proven to have an inhibitory effect onbone resorption induced by a low calcium diet in mice(data not shown)15. The mice were divided into threegroups; 1) non- immunized mice received vehicle(normal- vehicle mice; n = 6), 2) immunized micereceiving vehicle (CIA- vehicle mice; n = 6), and 3)immunized mice receiving WP9QY peptide (CIA-WP9QY mice; n = 6).

Assessment of arthritis The arthritis score of the mice was examined daily,

from the day of the second immunization (day 21).Initial observation of erythema and/or swelling was con-sidered to be the onset of arthritis. Arthritis score weregiven as 0 to 3 to the four paws as previouslydescribed16. The criteria for the grading were as follows;0 = normal; 1 = slight swelling and/or erythema; 2 =pronounced edematous swelling; and 3 = deformedpaw or joint, with ankylosis, joint rigidity.

Radiological assessment of arthritisAt the end of the experiment (day 38), mice were

sacrificed under ether anesthesia, and the hind pawswere collected and fixed in 4% cacodylate- buffered glu-taraldehyde- formalin fixative (pH 7.4) for 2 days,washed with water for a day and later used for radio-logical analysis. Three-dimensional reconstructionimages and coronal images of ankle joints and kneejoints were obtained with a micro focus X-ray computed

T. KOJIMA et al. J Med Dent Sci92

tomography (Micro- CT; SKYSCAN 1072, Aartselaar,Belgium).

Histomorphometric evaluation of arthritisCollected hind paws were embedded in GMA resin

(JB4, Polyscience, Warrington, PA). Standard decalci-fied sections (1.5Òm) were prepared and stained withhematoxylin and eosin (H&E). In order to quantitative-ly assess the arthritis, we analyzed the infiltration rate ofthe synovial pannus of the articular cartilage surface ofthe proximal end of tibia, using H&E stained undecal-cified sections. The calculation formula is as follows: theinfiltration rate of pannus (%) = (the length of the syn-ovial pannus facing the proximal end of tibia) / (the totallength of the proximal end of tibia)×100. To evaluatebone destruction of the knee joints, approximately0.38 mm2 area of the epiphysis, starting 0.5 mm proxi-mal from the growth plate- epiphyseal junction, wasdetermined in order to exclude the cortical bone facingthe epiphyseal growth plate using undecalcified sec-tions with toluidine blue staining. The bone volume(BV/TV) in the epiphysis of proximal tibia was mea-sured using an image analyzing system (KS400, CarlZeiss, Jena, Germany) as previously described17.

Bone histomorphometry of tibial metaphysisTo perform bone histomorphometry in the tibial

metaphysis, sagittal sections of proximal tibia werestained with both tartrate resistant acid phosphatase(TRAP)18 and toluidine blue, and observed under a lightmicroscope with a 20× objective. The measurementswere performed in the secondary spongiosa starting0.3 mm distal from the growth plate to exclude the pri-mary spongiosa as previously described19. A standardhistomorphometric analysis was performed20 usingthe image analyzing system as described above.TRAP- positive cells that formed resorption lacunae atthe surface of the trabeculae and contained two ormore nuclei were designated as osteoclasts.

Measurements of serum anti-Type II collagen(anti- CII) antibodies

At the end of the experiment (day 38), serum sam-ples were obtained and the levels of the serum anti- CIIantibody were assessed using an enzyme- linkedimmunosorbent assay (ELISA) kit (Chondrex, Inc. NERedmond, WA,USA).

Statistical analysis The analysis of arthritis scores was performed by

Mann- Whitney U- test. The other data were first ana-

lyzed by ANOVA. When a significant F ratio (P < 0.05)was identified, groups were compared using Fisher’sprotected least significant difference post hoc test.Tests were carried out using Apple software, StatView 4.1 (Abacus Concepts, Inc., Cupertino, CA,USA). P values less than 0.05 were considered signif-icant.

Results

WP9QY peptide inhibited the disease activity in theCIA mice

To identify the effects of WP9QY peptide on inflam-mation in CIA model, we assessed the development ofinflammation using arthritis score. The onset of arthritiswas not delayed by WP9QY treatment compared tothat in CIA- vehicle mice, appearing on day 26, but theseverity of the disease was less. A significant reductionof the arthritis score appeared from day 28. On day 38,it was still observed in the CIA- WP9QY mice whencompared to the CIA- vehicle mice (Figure 1), indicatingthat the WP9QY peptide had an inhibitory effect on thedisease activity.

The serum levels of anti- CII antibody were signifi-cantly increased in the CIA mice, but the increase of

93THE EFFECTS OF WP9QY PEPTIDE ON MURINE COLLAGEN-INDUCED ARTHRITIS

Figure 1. WP9QY peptide blocks the CIA- induced increase in dis-ease activity.The arthritis score observed in CIA- vehicle mice was inhibited inCIA- WP9QY mice. The day of the first immunization was designat-ed as day 0. The first signs of arthritis onset occurred on day 26. Dataare expressed as mean ±SE of six mice in each group. *P＜0.05, **P＜0.01vs. CIA- vehicle mice.

anti-CII antibody was not reduced by anti-TNF treat-ment16. In accordance with this report, the high level ofanti-CII antibody was observed in the CIA- vehiclemice. The serum levels did not change significantlybetween the CIA- vehicle mice and CIA- WP9QYmice, though the WP9QY peptide lowered the CIA-induced increase in the serum levels (data notshown).

The inhibitory effects of WP9QY peptide on theinvasion of inflammatory cells and pannus forma-

tion in the knee joints of CIA miceTo confirm the inhibitory effect of WP9QY peptide on

the activity of CIA, a histological assessment of theknee joints was performed. The histological findings inthe CIA- vehicle mice revealed the abnormalities ininvasion of inflammatory cells and bone erosion(Figure 2B), but those in the CIA- WP9QY miceshowed markedly reduced invasion of inflammatorycells and absence of bone erosion (Figure 2C).Proliferative synovitis composed of abundant fibrob-lasts, hypertrophic synoviocytes, and infiltration of

T. KOJIMA et al. J Med Dent Sci94

Figure 2. Histological observations of knee joints in CIA mice.Knee joint in CIA- vehicle mice (B) showed pannus formation (arrows) and bone erosion compared to normal- vehicle mice (A). Kneejoint in CIA- WP9QY mice (C) showed reduction of infiltrating cells and inhibition of bone erosion. The region within the black rectangleof panel (B) represents the entire field of panel (D). The region within the black rectangle shows the invasion of pannus in the tibialepiphysis. Hematoxylin and eosin staining. Bar; 100 Òm.

inflammatory cells composed of neutrophils,macrophages, lymphocytes were observed at theproximal end of tibia in CIA- vehicle mice (Figure 2D). Inorder to quantitatively evaluate the pannus formation,we calculated the infiltration rate of synovial pannus onthe proximal end of tibial epiphysis. This histomorpho-metric assessment revealed that WP9QY peptide sig-nificantly inhibited CIA- induced pannus formation andinfiltration (Figure 3). These results suggest thatWP9QY peptide inhibited the inflammatory destruction

of the knee joints in CIA mice.

WP9QY peptide prevents bone destructionThree- dimensional reconstruction images from the

Micro- CT scanning data (Figure 4A-C) illustratedbone erosion on the articular surfaces of the anklejoints in CIA- vehicle mice (Figure 4B) compared to nor-mal- vehicle mice (Figure 4A), while almost no erosionwas detected in the CIA- WP9QY mice (Figure 4C).Three- dimensional reconstruction images from theMicro- CT scanning data revealed the protectiveeffects of WP9QY peptide against bone destruction inCIA mice. Similarly, X-ray radiography showed severebone destruction of the ankle joints in the CIA micewhile WP9QY treatment prevented this bone destruc-tion in the CIA mice (data not shown).

Three- dimensional reconstruction images from theMicro- CT scanning data (Figure 5A-C) illustratedbone erosion on the articular surfaces of the knee jointsin CIA- vehicle mice (Figure 5B) compared to normal-vehicle mice (Figure 5A), while almost no erosion wasdetected in CIA- WP9QY mice (Figure 5C). Coronalimages of the knee joints by Micro- CT showed that thedecrease of trabecular bone of femur and tibia in CIA-vehicle mice (Figure 5E) compared to normal- vehiclemice (Figure 5D), while the WP9QY peptide preventedthe CIA- induced decrease of trabecular bone (Figure5F).

To further investigate the effects of WP9QY peptideon bone erosion accompanied with CIA at the kneejoints, histological and histomorphometric assess-ments were performed in the epiphysis of proximal tibiaas described in the Materials and Methods. The histo-morphometric measurement showed that bone volume

95THE EFFECTS OF WP9QY PEPTIDE ON MURINE COLLAGEN-INDUCED ARTHRITIS

Figure 3. WP9QY peptide protected knee joints from inflammatorydestruction. WP9QY peptide blocked the infiltration rate of synovial pannus on thearticular cartilage surface of the proximal end of the tibia. The infil-tration rate of pannus in normal- vehicle mice was not detectable(N.D). The infiltration rate of pannus was calculated as described inMaterials and Methods. Data are expressed as mean ±SD for eachgroup. *P＜0.05 vs. CIA- vehicle mice.

Figure 4. The effects of WP9QY peptide on bone erosions of ankle joints in CIA mice. Three-dimensional micro-CT images of ankle joints of normal- vehicle mice (A), CIA- vehicle mice (B), and CIA- WP9QYmice (C) were taken on day 38. WP9QY peptide protected the CIA- induced joint destruction. Arrows in panel B indicatebone erosions. Representative radiographs from each group are shown. Bar; 1 mm.

of the tibial epiphysis (BV/TV at epiphysis) in CIA- vehi-cle mice was significantly reduced compared to that ofnormal- vehicle mice, and this decrease in the bone vol-ume was significantly inhibited by WP9QY treatment(Figure 6). These results suggest that WP9QY peptideprevented bone destruction in the knee joints in CIAmice.

Bone morphometry in the tibial metaphysisSince it was reported that the induction of arthritis

causes bone loss in other parts of the bones as well asthe joint destruction21, a histomorphometric analysiswas performed in the tibial metaphysis. The bone vol-ume (BV/TV) was significantly decreased in CIA-

vehicle mice compared to normal- vehicle mice, and thedecreased BV/TV was significantly blocked in CIA-WP9QY mice (Figure 7A). The number of osteoclasts(N. Oc/BS) was significantly increased in CIA- vehiclemice compared to normal- vehicle mice. This increasein N. Oc/BS was significantly reduced in CIA- WP9QYmice (Figure 7B). This result indicates that WP9QYpeptide could inhibit the cancellous bone in the meta-physis of the tibia as well as bone erosion of the artic-ular surface of the knee joints.

T. KOJIMA et al. J Med Dent Sci96

Figure 5. Micro-CT reconstruction images of knee joints illustrate the inhibitory effects of WP9QY peptide onarticular bone erosion at knee joints.Three-dimensional micro-CT images of knee joints in the normal- vehicle mice (A), the CIA- vehicle mice (B)and the CIA- WP9QY mice (C). Coronal images of knee joints in normal- vehicle mice (D), the CIA- vehiclemice (E) and the CIA- WP9QY mice (F). Arrows in panel B indicate articular bone erosions. Arrow heads inpanel E indicate reduction in trabecular bone. The WP9QY treatment prevented CIA- induced knee jointdestruction and trabecular bone resorption. Representative three-dimensional images from each group areshown. Bar; 1 mm.

Discussion

The present study demonstrates that WP9QY pep-tide inhibits both inflammation and bone destruction incollagen- induced arthritis (CIA) mice, a chronicautoimmune model of human RA.

TNFÏ, IL-1 and IL-6 are the key mediators in the per-petuation of synovitis and bone destruction and areupregulated in the RA synovium and synovial fluid3. Infact, blockage of these proinflammatory cytokinesameliorated the inflammation of joints in the CIAmice22-24. Moreover, anti- TNFÏ treatment such asneutralizing antibody (Infliximab) or a solubleTNFR:Fc fusion protein (Etanercept) reduced theinflammation in patients with persistently active RA9,10.These reports indicated that TNFÏ is important for thedevelopment of inflammation in RA. The presentobservation is consistent with these reports. It seemsthat the WP9QY peptide, a TNFÏ antagonist, reducesinflammation in the CIA mice by interfering with theTNFÏ / TNFR (I) interaction.

The present study has confirmed that osteoclastsplay a pivotal role in RA-mediated joint destruction. Theproliferation and differentiation process of osteoclasts

97THE EFFECTS OF WP9QY PEPTIDE ON MURINE COLLAGEN-INDUCED ARTHRITIS

Figure 6. Histomorphometric analysis reveals the inhibitory effectsof WP9QY peptide on bone volume in the proximal end of tibial epi-physis. The measurement areas were determined as described inMaterials and Methods. BV/TV in CIA- vehicle mice was significant-ly reduced compared to normal- vehicle mice. This decrease inBV/TV was significantly inhibited in CIA- WP9QY mice. Data areexpressed as mean ± SD for each group. #, P＜0.05 vs. normalmice; *, P ＜ 0.05 vs. CIA- vehicle mice.

Figure 7. The effects of WP9QY peptide on cancellous bone in tib-ial metaphysis. BV/TV (A) in the CIA- vehicle mice was significantlydecreased compared to normal- vehicle mice (open bar) and thisdecrease was inhibited in the CIA- WP9QY mice. The number ofosteoclasts (N. Oc /BS) (B) in the CIA- vehicle mice was significant-ly increased, compared to normal- vehicle mice and this increase wassignificantly inhibited in the CIA- WP9QY mice. Data are expressedas mean ± SD for each group. #, P ＜ 0.05 vs. normal mice; *, P＜0.05 vs. CIA- vehicle mice; **, P ＜ 0.01 vs. CIA- vehicle mice.

requires cytokines such as macrophage colony-stimu-lating factor (M-CSF) and receptor activator of NF-ÔBligand (RANKL) expressed on osteoblast, in combina-tion with RANK expressed on osteoclast. In the synovialinflammatory tissue, activated T cells express RANKLand regulate the osteoclast formation which results inthe bone destruction of RA25-27. Activated T cells pro-duce interferon (INF)-Î, a strong inhibitor of osteoclastformation and therefore the balance between RANKLand IFN-Î may be skewed in favor of the RANKLexpression in RA28,29. Recently, other studies havereported that the proinflammatory cytokine, TNFÏ hasintrinsic osteoclast- induced properties mediatedthrough signaling which is independent of theRANKL/RANK system30,31. Therefore, the inhibitoryeffects of WP9QY peptide in the CIA- induced boneresorption of knee joints and tibial epiphysis might bemainly due to the interference with TNFÏ/TNFR(I)interaction.

However, there is a report that TNFÏ itself fails toinduce differentiation of osteoclasts in the absence ofRANKL priming32. Moreover, TNFÏ- treated RANKL-/-mice showed the rare appearance of TRAP andcathepsin K- positive osteoclasts on bone surfaces andfailed to induce hypercalcemia33, Since many reportsconfirmed that the synergism between TNFÏ andRANKL promotes bone resorption by enhancing the dif-ferentiation32 and the activity34, this synergism seems tolead to the severe inflammatory bone destruction in RA.In view of these reports, it is still not clear whetherWP9QY peptide inhibited osteoclastogenesis in the CIAmice only by antagonizing TNFÏ. It could be speculat-ed that WP9QY peptide inhibited inflammatory bonedestruction of the joint by blocking TNFÏ/TNFR(I)and/or synergism of TNFÏ and RANKL in the CIA mice.In addition, similarities between the ligand recognitionsite on TNFR(I) and the cognate receptor RANK, indi-cate the possibility that WP9QY peptide directlyinhibits the RANKL/RANK interaction, therebydecreasing the osteoclast activity in the CIA mice.Since it is reported that higher serum levels of solubleRANKL are present in RA patients than in healthy35, thefindings that WP9QY peptide also inhibited the sys-temic reduction of trabecular and cortical bone ofproximal tibia, which is not a site for RA, in this studyfurther support the inhibitory effects of the WP9QY pep-tide on RANKL/RANK interactions.

The molecular size of the WP9QY peptide that weused in this study is 1226, so the WP9QY peptide isexpected to overcome the disadvantages of largemacromolecules such as antibodies and soluble

forms of receptor, which include poor bioavailability, sta-bility, expense for peptide synthesis, and the risk ofsevere and occasionally life- compromising sideeffects11. Furthermore, WP9QY peptide apparentlyhad no side effects, as no systematic abnormality or nopathological changes were observed in the tissueseven under high dose of peptide administration.Moreover, we found that the half-life of the WP9QYpeptide is very short, which could be one of the rea-sons why this peptide could not delay the onset ofarthritis as do anti-TNF agents and why multiple injec-tions were necessary during 24 hours in the experi-ment. Strategies to overcome these disadvantagesshould be addressed for further use of this peptide.

In conclusion, the present result showed thatWP9QY peptide inhibited both inflammation and bonedestruction of the joints in CIA mice, possibly by amechanism that interferes with the interactionbetween ligand and receptor pairs, TNFÏ/TNFR(I)and RANKL/RANK, by the reduction of synergism ofTNFa and RANKL. Therefore, the development ofimproved type of WP9QY peptide, that prevents moreefficiently the interaction of the critical loop on the TNFreceptor (I) with its binding site on TNFa may be usefulfor the amelioration of inflammatory bone destructionsuch as RA and periodontitis.

Acknowledgements

This work was partly supported by grants-in-aid forscientific research to K. Ohya (#16209053), K. Aoki(#14370589 and #16390530), H. Yoshimasu(#14571881) and T. Amagasa (#15209069) from theMinisrtry of Education, Culture, Sports, Science andTechnology, Japan.

References1. Chen E, Keystone EC, Fish EN. Restricted cytokine expres-

sion in rheumatoid arthritis. Arthritis Rheum 1993;36(7):901-10.

2. Pincus T. Long-term outcomes in rheumatoid arthritis. Br JRheumatol 1995;34 Suppl 2:59-73.

3. Choy EH, Panayi GS. Cytokine pathways and joint inflamma-tion in rheumatoid arthritis. N Engl J Med 2001;344(12):907-16.

4. Brennan FM, Chantry D, Jackson A, et al. Inhibitory effect ofTNFÏ antibodies on synovial cell interleukin-1 production inrheumatoid arthritis. Lancet 1989;2(8657):244-7.

5. Feldmann M, Brennan FM, Maini RN. Rheumatoid arthritis.Cell 1996;85(3):307-10.

6. Thornton S, Duwel LE, Boivin GP, et al. Association of thecourse of collagen-induced arthritis with distinct patterns of

T. KOJIMA et al. J Med Dent Sci98

cytokine and chemokine messenger RNA expression.Arthritis Rheum 1999;42(6):1109-18.

7. Mussener A, Litton MJ, Lindroos E, et al. Cytokine productionin synovial tissue of mice with collagen-induced arthritis(CIA). Clin Exp Immunol 1997;107(3):485-93.

8. Feldmann M, Brennan FM, Maini RN. Role of cytokines inrheumatoid arthritis. Annu Rev Immunol 1996;14:397-440.

9. Lipsky PE, van der Heijde DM, St Clair EW, et al. Infliximaband methotrexate in the treatment of rheumatoid arthritis. Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis withConcomitant Therapy Study Group. N Engl J Med2000;343(22):1594-602.

10. Weinblatt ME, Kremer JM, Bankhurst AD, et al. A trial of etan-ercept, a recombinant tumor necrosis factor receptor:Fcfusion protein, in patients with rheumatoid arthritis receivingmethotrexate. N Engl J Med 1999;340(4):253-9.

11. Takasaki W, Kajino Y, Kajino K, Murali R, et al. Structure-baseddesign and characterization of exocyclic peptidomimeticsthat inhibit TNF ? binding to its receptor. Nat Biotechnol1997;15(12):1266-70.

12. Suzuki Y, Aoki K, Ohya K, et al. TNFÏ antagonist inhibits boneresorption induced by Porphyromonas gingivalis in mice. JapAssoc Den Res 2003; 51st Annual Meeting:Suppl 81.

13. Saito H, Aoki K, Murali R, et al. TNF-Ï antagonistic peptideinhibits bone loss induced by estrogen deficiency. Bone2003; 32:S215.

14. Iwai H, Kozono Y, Hirose S, et al. Amelioration of collagen-induced arthritis by blockade of inducible costimulator-B7homologous protein costimulation. J Immunol2002;169(8):4332-9.

15. Aoki K, Suzuki Y, Shibata T, et al. A TNF-Ï related peptideinhibits bone resorption in vivo and osteoclastogenesis in vitroby interfering with RANK/RANKL interaction. J Bone MinerRes 2001;16:S177.

16. Williams RO, Feldmann M, Maini RN. Anti-tumor necrosis fac-tor ameliorates joint disease in murine collagen-inducedarthritis. Proc Natl Acad Sci U S A 1992;89(20):9784-8.

17. Seto H, Aoki K, Kasugai S, et al. Trabecular bone turnover,bone marrow cell development, and gene expression ofbone matrix proteins after low calcium feeding in rats. Bone1999;25(6):687-95.

18. Glowacki J, Rey C, Glimcher MJ, et al. A role for osteocalcin inosteoclast differentiation. J Cell Biochem 1991;45(3):292-302.

19. Aoki K, Didomenico E, Sims NA, et al. The tyrosine phos-phatase SHP-1 is a negative regulator of osteoclastogenesisand osteoclast resorbing activity: increased resorption andosteopenia in me(v)/me(v) mutant mice. Bone1999;25(3):261-7.

20. Parfitt AM, Drezner MK, Glorieux FH, et al. Bone histomor-phometry: standardization of nomenclature, symbols, andunits. Report of the ASBMR HistomorphometryNomenclature Committee. J Bone Miner Res 1987;2(6):595-610.

21. Nishida S, Tsurukami H, Sakai A, et al. Stage-dependentchanges in trabecular bone turnover and osteogenic capacityof marrow cells during development of type II collagen-induced arthritis in mice. Bone 2002;30(6):872-9.

22. Wooley PH, Whalen JD, Chapman DL, et al. The effect of aninterleukin-1 receptor antagonist protein on type II collagen-induced arthritis and antigen-induced arthritis in mice.Arthritis Rheum 1993;36(9):1305-14.

23. Wooley PH, Dutcher J, Widmer MB, et al. Influence of arecombinant human soluble tumor necrosis factor receptor FCfusion protein on type II collagen-induced arthritis in mice. JImmunol 1993;151(11):6602-7.

24. Sasai M, Saeki Y, Ohshima S, et al. Delayed onset andreduced severity of collagen-induced arthritis in interleukin-6-deficient mice. Arthritis Rheum 1999;42(8):1635-43.

25. Romas E, Bakharevski O, Hards DK, et al. Expression ofosteoclast differentiation factor at sites of bone erosion in col-lagen-induced arthritis. Arthritis Rheum 2000;43(4):821-6.

26. Gravallese EM, Manning C, Tsay A, et al. Synovial tissue inrheumatoid arthritis is a source of osteoclast differentiation fac-tor. Arthritis Rheum 2000;43(2):250-8.

27. Lubberts E, Oppers-Walgreen B, Pettit AR, et al. Increase inexpression of receptor activator of nuclear factor kappaB atsites of bone erosion correlates with progression of inflam-mation in evolving collagen-induced arthritis. Arthritis Rheum2002;46(11):3055-64.

28. Kong YY, Feige U, Sarosi I, et al. Activated T cells regulatebone loss and joint destruction in adjuvant arthritis throughosteoprotegerin ligand. Nature 1999;402(6759):304-9.

29. Takayanagi H, Ogasawara K, Hida S, et al. T-cell-mediatedregulation of osteoclastogenesis by signalling cross-talkbetween RANKL and IFN-?. Nature 2000;408(6812):600-5.

30. Azuma Y, Kaji K, Katogi R, et al. Tumor necrosis factor-Ïinduces differentiation of and bone resorption by osteoclasts.J Biol Chem 2000;275(7):4858-64.

31. Kobayashi K, Takahashi N, Jimi E, et al. Tumor necrosis factorÏ stimulates osteoclast differentiation by a mechanism inde-pendent of the ODF/RANKL-RANK interaction. J Exp Med2000;191(2):275-86.

32. Lam J, Takeshita S, Barker JE, et al. TNF-Ï induces osteo-clastogenesis by direct stimulation of macrophages exposed topermissive levels of RANK ligand. J Clin Invest2000;106(12):1481-8.

33. Li J, Sarosi I, Yan XQ, et al. RANK is the intrinsic hematopoi-etic cell surface receptor that controls osteoclastogenesis andregulation of bone mass and calcium metabolism. Proc NatlAcad Sci U S A 2000;97(4):1566-71.

34. Fuller K, Murphy C, Kirstein B, et al. TNFÏ potently activatesosteoclasts, through a direct action independent of andstrongly synergistic with RANKL. Endocrinology2002;143(3):1108-18.

35. Ziolkowska M, Kurowska M, Radzikowska A, et al. High levelsof osteoprotegerin and soluble receptor activator of nuclearfactor kappa B ligand in serum of rheumatoid arthritispatients and their normalization after anti-tumor necrosis fac-tor ? treatment. Arthritis Rheum 2002;46(7):1744-53.

99THE EFFECTS OF WP9QY PEPTIDE ON MURINE COLLAGEN-INDUCED ARTHRITIS