The preventive effects of two nutraceuticals on experimentally induced acute synovitis

Equine Veterinry Journl ISSN 0425-1644 DOI: 10.1111/evj.12629 The preventive effects of two nutrceuticls on experimentlly induced cute synovitis E. VAN DE WATER *, M. OOSTERLINCK, M. DUMOULIN, N. M. KORTHAGEN,P.R.VAN WEEREN, J. VAN DEN BROEK, H. EVERTS, F. PILLE 1 nd D. A. VAN DOORN 1 Deprtment of Surgery nd Anesthesiology, Fculty of Veterinry Medicine, Ghent University, Belgium Deprtment of Equine Sciences, Fculty of Veterinry Medicine, Utrecht University, The Netherlnds Deprtment of Frm Animl Helth, Fculty of Veterinry Medicine, Utrecht University, The Netherlnds Equivdo, Equine Nutrition Consultncy, Utrecht, The Netherlnds. *Correspondence emil: eline.vndewter@ugent.e; Received: 16.02.16; Accepted: 19.08.16 1 These uthors contriuted eqully to this work. Summry Bckground: Nutrceuticls re often used in the mngement of equine osteorthritis, ut scientific evidence of their efficcy is lcking. Ojectives: To study the preventive effects of two new nutrceuticls fter the experimentl induction of synovitis in comprison with positive nd negtive control tretments. Study design: Blinded, controlled, rndomised experiment. Methods: Twenty-four helthy Stndrdred horses were rndomly llocted to supplement AT (multi-ingredient, 28 dys), supplement HP (collgen hydrolyste, 60 dys), meloxicm (4 dys) or plceo (60 dys). Synovitis ws induced in the right intercrpl joint y intr-rticulr injection of 0.5 ng lipopolyscchride (LPS) of Escherichi coli while tretments were continued. Blood nd synovil fluid were smpled efore tretment, immeditely prior to LPS injection, nd t 8, 24 nd 48 h post-injection. Synovil fluid smples were nlysed for totl nucleted cell count (TNCC), totl protein (TP) nd selected iomrkers (prostglndin E 2 [PGE 2 ], interleukin-6 [IL-6], glycosminoglycns [GAGs], type II collgen synthesis [CPII], mtrix metlloproteinse [MMP]). Lmeness ws scored y visul exmintion nd pressure plte nlysis immeditely prior to LPS injection, nd t 8, 24 nd 48 h post-injection. Clinicl exmintions were performed efore tretment, immeditely prior to LPS injection, t 2, 4 nd 6 h post-injection, nd then twice per dy during the test period. Results: Before tretment nd intr-rticulr chllenge, there were no sttisticlly significnt differences mong the tretment groups for ny of the prmeters. After intr-rticulr chllenge, the plceo group showed significntly higher synovil fluid TP, TNCC nd PGE 2 compred with the meloxicm group, lthough the model did not induce relevnt mount of lmeness. Both nutrceuticls resulted in significntly lower synovil fluid TP, TNCC nd PGE 2 compred with plceo. No sttisticl differences in IL-6, GAGs, CPII or MMPs were oserved mong tretment groups. No dverse effects were oserved. Min limittions: Despite evidence of synovitis, lmeness ws too mild to detect. Conclusions: The preventive dministrtion of these nutrceuticls showed nti-inflmmtory effects in this vlidted synovitis model. Therefore, further studies of their clinicl pplicility re wrrnted. The Summry is ville in Chinese see Supporting informtion. Keywords: horse; LPS; rthritis; iomrkers; pressure plte Introduction Nutrceuticls re often used in the mngement of osteorthritis, which is common cuse of chronic lmeness in horses [1]. However, their curtive efficcy remins controversil [2,3] nd the qulity of relevnt studies is generlly low [4,5]. Equine in vitro studies hve suggested tht the comintion of glucosmine nd chondroitin sulphte cn result in reduced crtilge degrdtion [6 8] nd my hve nti-inflmmtory effects [6,9,10]. Although their orl iovilility in horses is reported to e low [11,12], n in vivo study in osteorthritic horses showed significnt clinicl improvement fter tretment with glucosmine nd chondroitin sulphte compred with plceo tretment [13]. Furthermore, methylsulphonylmethne (MSM) is nturl nti-inflmmtory gent [14] tht hs een found to decrese joint pin nd swelling in humn sujects with osteorthritis [15,16] nd to significntly meliorte exercise-relted oxidtive nd inflmmtory lood chnges in jumping horses [17]. Finlly, collgen hydrolyste showed stimultory effect on type II collgen This is n open ccess rticle under the terms of the Cretive Commons Attriution License, which permits use, distriution nd reproduction in ny medium, provided the originl work is properly cited. iosynthesis of chondrocytes in n in vitro model with ovine chondrocytes [18]. Moreover, severl studies showed eneficil effects of collgen hydrolyste on joint pin ssocited with osteorthritis in humn sujects [19,20]. Studies of its efficcy in equine cses re lcking, ut it hs een found to e well sored nd ville for mino cid metolism in horses [21]. Until now, the forementioned nutrceuticls hve only een tested using in vitro studies or in selected ptients from curtive perspective. As inflmmtion plys crucil role in the pthogenesis of osteorthritis, reducing the initil inflmmtion cn e seen s the cornerstone of preventive tretment. The efficcy of prevention cn e studied using vlidted synovitis model sed on the intr-rticulr injection of lipopolyscchride (LPS) [22,23]. Therefore, linded, controlled, rndomised study ws designed to test the effects of two nutrceuticls using the vlidted model of cute synovitis descried ove nd comprehensive set of ojective synovil nd clinicl vriles s outcome prmeters. These included quntittive locomotion nlysis using pressure plte nd n rry of synovil iomrkers. The hypothesis ws tht preventive dministrtion of these nutrceuticls would hve eneficil effect on the degree of joint inflmmtion nd lmeness. 532 Equine Veterinry Journl 49 (2017) 532 538 2016 The Authors Equine Veterinry Journl pulished y John Wiley & Sons Ltd on ehlf of EVJ Ltd

E. Vn de Wter et l. Effects of nutrceuticls in synovitis Mterils nd methods Study design The study ws linded, controlled, rndomised experiment in lock design with four tretment groups of six horses per group, ll housed under the sme conditions. Horses were rndomly llocted to one of four dietry tretments: 1) supplement AT (Cvlor ArtiTec [Liquid], contining glucosmine sulphte 2KCL, shrk chondroitin sulphte sodium, MSM, oswellic cid dry extrct 65%, Annsus comosus extrct 2500 GDU, L- glutmine, feverfew dry extrct PE 4:1, hyluronic cid), dministered t 45 ml twice per dy for 28 dys prior to rticulr chllenge nd during the 3-dy test period; 2) supplement HP (Hydro-P, collgen hydrolyste), t 90 g once per dy for 60 dys prior to rticulr chllenge nd during the 3- dy test period; 3) meloxicm (Metcm c ), t 0.6 mg/kg once per dy for 4 dys prior to rticulr chllenge nd during the 3-dy test period (positive control); nd 4) plceo (csein), t 90 g once per dy for 60 dys prior to rticulr chllenge nd during the 3-dy test period (negtive control). The numer of horses (n = 6) per tretment group ws sttisticlly determined using n priori power nlysis test [24] for the primry outcome vrile (prostglndin E 2 [PGE 2 ]), sed on the pulished stndrd error of PGE 2 concentrtion in synovil fluid in horses with synovitis [22], nd men estimted effect of nutrceuticl tretment of 45% PGE 2 reduction [25], t = 0.05; = 0.2, nd n effect size of 1.68. Experimentl procedures strted 61 dys efore the rticulr chllenges (post-injection dy [PID] 61), when lood nd synovil fluid smples were otined to provide seline dt efore supplementtion. Immeditely prior to the rticulr chllenge with LPS of Escherichi coli (PID 0), lood nd synovil fluid were smpled nd git evlution (visully nd y pressure plte nlysis) ws performed to estlish seline vlues efore the induction of rthritis. The ltter procedures were repeted t 8 h (PID 0.3), 24 h (PID 1) nd 48 h (PID 2) post-injection (Fig 1). Horses were locked in groups of four sed on ody weight nd ge. Tretments were rndomly llocted within ech lock nd locks were entered into the study t 1-week intervls. Horses The study included 24 helthy nd cliniclly sound femle French (n = 19), Belgin (n = 4) nd Dutch (n = 1) Stndrdred horses (men s.d. ge: 9 2.8 yers; men s.d. ody weight: 495.9 38.9 kg), otined from locl reeding centre. Horses with known history of lmeness or gstrointestinl prolems were excluded from the study. Feeding nd supplementtion Horses were individully fed stndrdised diet of concentrtes nd highqulity hy tht met their nutritionl requirements [26]. Horses sujected to tretment AT, meloxicm nd plceo received the plceo supplement, nd horses sujected to tretment HP received the HP supplement, topdressed over the morning concentrte rtion. The morning dose AT nd meloxicm were dministered orlly prior to the morning concentrte feed rtion. Intr-rticulr chllenge At PID 0, fter seline pressure plte nlysis nd rthrocentesis, synovitis ws induced in the right intercrpl joint y injection of 0.5 ng Strt HP Strt plceo Dy Strt AT Tretment period Strt meloxicm LPS injection 0............ 33......... 57 58 59 60 61 62 63 64 PID 61 PID 0 PID 1 PID 0.3 PID 2 Fig 1: Timeline of tretments (HP, supplement HP; AT, supplement AT; LPS, lipopolyscchride) nd evlution points, displyed y post-injection dy (PID). LPS from E. coli (L5418 d ) (lot 093M4041V) in 0.8 ml sterile isotonic sline [22]. Two weeks prior to the strt of the study, the initil LPS solution (1 mg/ml) ws septiclly diluted to finl concentrtion of 0.625 ng/ml. The diluted LPS solutions were stored in glss vils t 4 C. Physicl exmintion (respirtory rte, hert rte nd rectl temperture) ws performed immeditely prior to LPS injection, t 2, 4 nd 6 h fter LPS injection, nd then twice per dy during the test period to control for systemic signs of endotoxemi. Smpling Synovil smples were tken t PID 61, PID 0, PID 0.3, PID 1 nd PID 2. If necessry, horses were sedted with detomidine 10 lg/kg nd utorphnol 10 lg/kg i.v. Prior to rthrocentesis, the right intercrpl joint region ws clipped nd septiclly prepred. With flexed crpus, 21 guge, 4-cm needle ws inserted etween the extensor crpi rdilis nd common digitl extensor tendons [27]. Approximtely 3.5 ml of synovil fluid ws withdrwn nd immeditely split into different sterile continers: 1 ml ws collected in n EDTA-coted tue nd the remining mount of fluid ws collected in plin tues. Smples were immeditely stored t 4 C nd nlysed (EDTA) or processed (plin tues) within 1 h of collection. The EDTA liquot ws nlysed for cytology (totl nucleted cell count [TNCC]) nd totl protein ([TP]). The liquot in plin tues ws centrifuged t 600 g for 20 min, liquoted nd stored t 80 C within 2 h fter collection for susequent iomrker nlysis. Blood smples were tken from the left jugulr vein with 21 guge, 4- cm needle nd put in silicone-coted nd EDTA-coted vcutiners. Smples were immeditely refrigerted (4 C), nd susequently trnsferred to the lortory for further nlysis. PID 0 smples were nlysed for hemtology nd serum iochemistry nd compred with PID 61 smples to ssess tretment sfety. The serum iochemistry pnel included ure, cretinine, TP, lumin, -gloulins, -gloulins, c-gloulins, totl iliruin, direct nd indirect iliruin, sprtte minotrnsferse, c-glutmyltrnsferse (GGT), lkline phosphtse, lctte dehydrogense nd cretine kinse. At PID 0.3, PID 1 nd PID 2 only hemtology ws performed to ssess LPS sfety. Synovil iomrker nlysis Prostglndins were determined y high-performnce liquid chromtogrphy (HPLC) tndem mss spectrometry (MS/MS) nlysis following RP-18 extrction of synovil fluid smples [28]. HPLC-MS/MS nlysis ws performed on PerkinElmer LC200 HPLC system e coupled to n electrospry ionistion liner ion trp qudrupole (4000 QTRAP) mss spectrometer f. The instrument ws operted in negtive MRM mode. PGE 2 concentrtions were clculted from pek res reltive to n internl stndrd. Synovil fluid smples were evluted for glycosminoglycn (GAG) content using the 1,9-dimethylmethylenelue ssy, dpted for use in microtitre pltes [28]. Interleukin-6 (IL-6) ws mesured using the GSI Equine IL-6 ELISA Kit g for synovil fluid, in which undiluted synovil smples were found to give the est results. CPII, mrker of type II collgen synthesis, ws quntified using commercil ELISA kits h, s descried in other studies [27,29,30], in ccordnce with the mnufcturer s recommendtions. Generl mtrix metlloproteinse (MMP) ctivity ws mesured using the fluorogenic sustrte FS-6 i.inshort,smpleswerediluted20-foldinmmp uffer (0.1 mol/l Tris, 0.1 mol/l NCl, 10 mmol/l CCl 2, 0.05% [w/v] Triton X-100, 0.1% [w/v] PEG6000, ph 7.5 nd 5 lmol/l FS-6). Smples were dded in triplicte to lck 384-well microplte j nd the fluorescent signl ws monitored continuously for 45 min t 37 C using CLARIOstr microplte reder k. The slope of the resultnt liner curve (reltive fluorescence units/s [RFU/s]) ws clculted s mesure of generl MMP ctivity. A quntity of 5 mmol/l EDTA ws used s negtive control. Clinicl evlution Lmeness ws scored prior to rthrocentesis. At PID 61, this ws performed y routine visul exmintion only using the Americn Equine Veterinry Journl 49 (2017) 532 538 2016 The Authors Equine Veterinry Journl pulished y John Wiley & Sons Ltd on ehlf of EVJ Ltd 533

Effects of nutrceuticls in synovitis E. Vn de Wter et l. Assocition of Equine Prctitioners (AAEP) scle, on which grde 0 represents sound ility nd grde 5 indictes non weight ering cpcity, y two Europen College of Veterinry Surgeons diplomtes, using video-recordings [31]. At PID 0, PID 0.3, PID 1 nd PID 2, oth videorecordings nd pressure plte mesurements were used. Pressure plte mesurements were otined using Footscn 3D 2m-system l s descried y Oosterlinck et l. [32]. The following kinetic vriles were clculted t wlk nd trot: 1) pek verticl force (PVF), in N/ kg; 2) verticl impulse (VI), in Ns/kg; nd 3) stnce time (ST) in ms. For ech set of five trils, ll left forelim (LF) nd right forelim (RF) mesurements were verged nd susequently PVF, VI nd ST rtios etween oth forelims were clculted s: %RF = RF/(LF + RF) 9 100%. Dt nlysis Sttisticl nlysis ws performed using R (lme4 pckge) m. A liner model with rndom horse effects nd with fixed week, time, sedtion, tretment nd the time tretment interction ws used. Akike s informtion criterion (AIC) ws used for model reduction [33]. Week nd sedtion were considered s lock fctors. If fctor ws importnt ccording to the AIC, then 90% ootstrp intervls were clculted for the effect. Residuls were checked for normlity using norml proility plot. TNCC, PGE 2,CPIInd MMP dt were logrithmiclly trnsformed for sttisticl nlysis. Clinicl lmeness scores required logistic regression, ut with exctly the sme model s descried ove. Results Synovil fluid nlysis The results of synovil fluid nlysis re presented in Tle 1. At PID 61 nd PID 0, there were no sttisticl differences mong tretment groups. After LPS injection, there ws mrked increse in TNCC in ll tretment groups, with shrp pek t PID 0.3 (Fig 2). At ll time points fter LPS TABLE 1: Results of synovil fluid nlyses of totl nucleted cell count (TNCC), totl protein (TP), prostglndin E 2 (PGE 2 ), interleukin-6 (IL-6), glycosminoglycns (GAGs), mtrix metlloproteinses (MMPs) nd type II collgen synthesis (CPII) tken t post-injection dys (PIDs) 61, 0, 0.3, 1 nd 2, in the four tretment groups (meloxicm [Melox], supplements AT nd HP, nd plceo [Plc]) PID 61 PID 0 PID 0.3 PID 1 PID 2 TNCC, 910 9 /L medin (rnge) Melox 0.17 (0.12 0.41) 0.11 (0.05 0.15) 61.31 (42.93 78.15) 21.89 (9.05 60.55) 3.59 (1.52 8.43) AT 0.21 (0.12 0.35) 0.16 (0.10 0.29) 45.70 (31.75 61.75) 15.00 (7.88 26.41) 4.59 (1.21 8.16) HP 0.15 (0.09 0.31) 0.20 (0.08 0.41) 55.36 (14.52 103.40) 17.90 (2.83 26.60) 4.95 (1.16 6.67) Plc 0.23 (0.15 0.33) 0.15 (0.11 0.21) 80.67 (68.81 113.37) 26.54 (12.78 80.35) 6.89 (2.55 8.89) TP, g/l men s.d. Melox 15.70 3.88 16.33 4.46 31.33 7.23 36.33 5.85 21.33 1.63 AT 15.50 3.33 18.33 1.97 37.67 8.62 40.67 4.13 24.33 3.88, HP 16.70 1.97 17.67 3.67 37.33 10.17 40.67 8.64 25.33 3.50, Plc 14.20 3.37 18.67 6.53 50.33 3.20 48.00 8.10 29.00 4.15 PGE 2, pg/ml medin (rnge) Melox 27.28 (9.92 143.78) 16.09 (9.02 23.10) 53.22 (19.05 127.70) 33.05 (20.69 79.50) 30.65 (15.45 58.01) AT 29.42 (9.95 107.28) 28.20 (8.78 99.44) 795.06 (415.43 1571.29) 119.72 (96.75 308.46) 130.50 (83.32 264.85),c HP 36.04 (12.42 91.52) 30.14 (13.83 70.14) 2307.82 (234.25 12,517.59) c 268.88 (153.72 439.99),c 154.79 (72.01 443.62),c Plc 36.41 (25.21 157.01) 14.98 (9.10 203.64) 3795.08 (720.28 40,960.19) d 304.56 (101.90 3524.93) c 268.31 (133.20 820.07) c IL-6, pg/ml men s.d. Melox 3.35 8.21 3.79 6.10 29.69 29.82 32.82 44.09 15.16 18.94 AT 14.56 30.10 48.40 106.65 126.5 261.29 118.72 255.30 81.37 188.09 HP 6.85 15.29 8.04 19.70 54.92 81.79 48.40 91.77 25.62 44.61 Plc 5.15 12.62 9.40 19.72 46.69 60.54 29.24 37.01 12.15 13.93 GAGs, lg/ml men s.d. Melox 114.50 42.86 108.20 36.03 180.71 26.99 245.61 78.46 158.17 49.77 AT 106.80 19.78 114.22 13.47 178.13 24.28 241.02 33.56 178.91 25.94 HP 112.50 29.90 115.70 28.30 177.34 39.77 247.95 88.03 205.00 71.36 Plc 125.10 18.82 126.53 17.10 191.41 22.94 291.59 78.04 230.05 67.30 MMPs, RFU/s medin (rnge) Melox 251.73 (157.20 451.47) 215.20 (95.67 396.80) 276.33 (157.20 568.00) 692.70 (295.40 1108.40) 418.13 (239.00 574.27) AT 232.23 (83.20 465.40) 217.72 (152.40 317.80) 695.37 (343.67 1050.60) 929.63 (521.20 1475.87) 511.47 (270.87 703.87) HP 244.13 (118.47 489.13) 222.77 (95.13 365.73) 493.93 (126.13 712.07) 861.80 (466.13 1300.53) 556.47 (224.67 650.33) Plc 285.40 (115.20 440.00) 211.20 (72.80 334.20) 388.93 (33.67 967.67) 908.07 (726.73 1328.13) 592.83 (524.33 700.33) CPII, ng/ml medin (rnge) Melox 1035.91 (264.40 8681.40) 678.86 (521.07 1538.08) 1388.69 (969.47 1662.18) 1732.70 (1085.27 2535.43) 2117.65 (910.16 4948.36) AT 1054.89 (538.99 1569.74) 1031.74 (341.15 1207.49) 1186.43 (253.31 2560.36) 3631.69 (866.49 6320.83) 4379.43 (828.20 6203.69) HP 660.33 (373.05 1075.67) 869.76 (290.59 1094.46) 1058.04 (323.97 2289.94) 6103.99 (1032.04 8639.12) 4283.48 (1278.48 12,303.01) Plc 436.35 (38.12 609.66) 814.47 (33.50 1883.84) 1181.77 (635.77 2874.65) 3583.98 (1117.20 12,341.05) 7653.97 (1591.68 23,051.08) RFU/s, reltive fluorescence units/s. Within time points, tretments with different superscripts show sttisticlly significnt differences. 534 Equine Veterinry Journl 49 (2017) 532 538 2016 The Authors Equine Veterinry Journl pulished y John Wiley & Sons Ltd on ehlf of EVJ Ltd

E. Vn de Wter et l. Effects of nutrceuticls in synovitis ) 3 Totl nucleted cell count (log 10 10 6 /L) 2 1 0 1 PID 61 PID 0 PID 0.3 PID 1 PID 2 Time Meloxicm AT HP Plceo ) 60 50 Totl protein (g/l) 40 30 20,, 10 0 PID 61 PID 0 PID 0.3 Time Meloxicm AT HP Plceo PID 1 PID 2 Fig 2: Logrithmiclly trnsformed ) men totl nucleted cell counts nd ) men totl protein in synovil fluid in the different tretment groups (green: meloxicm; lue: supplement AT; yellow: supplement HP; red: plceo) over time (post-injection dys [PID] 61, 0, 0.3, 1 nd 2). Within time points, outcomes in tretments with different superscripts show sttisticlly significnt differences. injection, meloxicm tretment nd oth supplement tretments resulted in sttisticlly lower TNCC thn plceo tretment. There were no sttisticlly significnt differences etween the meloxicm group nd oth supplement groups. Totl protein showed sustined increse in ll tretment groups t PID0.3ndPID1(Fig2).AtPID0.3ndPID1,TPinthemeloxicm group nd oth supplement groups ws sttisticlly lower thn in the plceo group. At PID 2, TP in the meloxicm group only ws sttisticlly lower thn in the plceo group. Over time, the concentrtion of PGE 2 showed rise t PID 0.3 (Fig 3). IL-6 nd GAGs showed more sustined rises t PID 0.3 nd PID 1, nd MMPs nd CPII peked t PID 2. The PGE 2 concentrtion ws sttisticlly lower in the meloxicm group thn in the other tretment groups t PID 0.3. Supplement AT resulted in sttisticlly lower PGE 2 concentrtions thn supplement HP nd plceo tretment, nd supplement HP resulted in sttisticlly lower PGE 2 concentrtions thn plceo tretment t this time point. At PID 1, PGE 2 ws sttisticlly lower in the meloxicm group thn in ll other tretment groups. Moreover, supplement AT resulted in sttisticlly lower PGE 2 concentrtions thn plceo tretment t this time point, y contrst with supplement HP. However, there ws no sttisticlly significnt difference etween the two supplement groups t this time point. At PID 2, PGE 2 ws sttisticlly lower in the meloxicm group thn in ll other tretment groups. IL-6, GAG, CPII nd MMP concentrtions showed no sttisticlly significnt differences etween tretment groups t ny time point. Lmeness evlution Lmeness ws very mild (AAEP score 1) nd visully detectle only t PID 0.3 in some horses (four of 24) nd t PID 1 in one horse. Clinicl lmeness scores did not differ sttisticlly etween tretment groups. Even with pressure plte nlysis, no sttisticl differences in PVF nd VI left-toright rtios etween tretment groups could e oserved t ny time point. For ST, very smll nd inconsistent differences etween tretment groups could e oserved t trot fter LPS injection. Sfety of supplements nd LPS injection After supplementtion (t PID 0), serum ctivity of GGT ws incresed in 21 of 24 (88%) horses (men s.d.: 101.6 61.9 U/L; reference rnge: 0 30 U/L) cross ll four groups, including the plceo group, nd hence without ny ssocition with tretment. Other lood prmeters (hemtology nd serum iochemistry) were within norml limits. After LPS injection, lood hemtology nd routine clinicl exmintions reveled no systemic normlities t ny time point. Discussion In vivo equine studies on nutrceuticls re usully clinicl trils [13,34,35], which re suject to n inherent lck of stndrdistion. This is highlighted y the low-qulity pprisl of most nutrceuticl studies, using the score Equine Veterinry Journl 49 (2017) 532 538 2016 The Authors Equine Veterinry Journl pulished y John Wiley & Sons Ltd on ehlf of EVJ Ltd 535

Effects of nutrceuticls in synovitis E. Vn de Wter et l. 5 4 c d PGE 2 (log 10 pg/l) 3 2,c c,c,c c 1 0 PID 61 PID 0 PID 0.3 Time Meloxicm AT HP Plceo PID 1 PID 2 Fig 3: Logrithmiclly trnsformed men prostglndin E 2 (PGE 2 ) concentrtions in synovil fluid in the different tretment groups (green: meloxicm; lue: supplement AT; yellow: supplement HP; red: plceo) over time (post-injection dys [PID] 61, 0, 0.3, 1 nd 2). Within time points, outcomes in tretments with different superscripts show sttisticlly significnt differences. list descried y Person nd Lindinger [4]. To overcome the limittions of clinicl studies, experimentl models cn e used s these provide highly stndrdised circumstnces. Using n IL-1-sed experimentl synovitis model, Person et l. [25] studied the effect of nutrceuticl (composed of mussel, shrk crtilge, lone nd Biot orientlis lipid extrct) on the inflmmtory response in synovil fluid. In susequent study of nother nutrceuticl ( iologicl extrct of high-rosmrinic cid mint) [36], synovitis ws induced y injection of LPS nd, in ddition to synovil fluid nlysis, lmeness ws ssessed s n outcome prmeter, leit sujectively. The current study is the first to show n nti-inflmmtory effect of nutrceuticls on experimentlly induced synovitis using comprehensive set of ojective synovil nd clinicl vriles, including nlysis of inflmmtory nd crtilge iomrkers, nd quntittive evlution of locomotion, which results in n excellent qulity score (>80.0) ccording to the clssifiction y Person nd Lindinger [4]. In the current study, cler synovil inflmmtion ws oserved fter LPS injection, especilly in the plceo group. Moreover, the registered non-steroidl nti-inflmmtory drug meloxicm resulted in significnt reductions in TNCC, TP nd PGE 2 concentrtions compred with the plceo tretment, indirectly confirming the vlidity of the synovitis model used in this study. The resulting lmeness, however, ws too mild to e detectle, either visully or y pressure plte evlution, nd therefore could not e used s prmeter with which to discriminte etween tretments. This contrsts with the findings of de Gruw et l. [22], who found men lmeness score of 3/5 on the AAEP grding scle for the plceo tretment t 8 h post-injection with the sme LPS dose. This discrepncy is most likely ttriutle to differences in LPS ctivity etween lots, ut my lso relte to the use of different hndling procedures to crete the diluted LPS solution, the use of Stndrdred horses in the present study vs. Wrmloods in the study conducted y de Gruw et l. [22], or comintion of these fctors. Breed-dependent effects my e suspected. Person et l. [36] reported tht no more thn three of eight Stndrdred horses showed grde 1 lmeness on the AAEP scle 12 h fter LPS injection in the intercrpl joint nd tht none of them showed ny lmeness t 24 h, lthough they used slightly lower dose (0.3 ng LPS of E. coli O55:B5). Positive spects of this issue re tht horses suffered less, which is good from n equine welfre perspective, nd tht the sutle lmeness induced resemles the clinicl situtioninosteorthritistogreterdegreethnthemorefulminnt inflmmtory rection descried y de Gruw et l. [22]. Nevertheless, slightly higher degree of lmeness would hve permitted quntittive git nlysis. The less pronounced lmeness in the present study in comprison with tht reported y de Gruw et l. [22] ws mirrored y lower pek TNCC nd lower mximl CPII concentrtion thn descried in the erlier pper nd reflects slightly lower degree of inflmmtion. However, the effectiveness of the experimentl model is illustrted y the sttisticlly relevnt increses in cell counts nd inflmmtory iomrkers in the plceo group nd the sttisticlly significnt differences etween the meloxicm nd plceo tretments for TNCC nd TP. The mximl concentrtions of PGE 2 nd GAG were similr to those descried y de Gruw et l. [22], ut y contrst with the study conducted y de Gruw et l. [23], the present study did not revel sttisticlly significnt difference in GAG content etween the meloxicm nd plceo tretments. The sttisticlly lower TNCC, TP nd PGE 2 concentrtions fter LPS injection in oth supplement groups compred with the plceo tretment group prove tht oth supplements hve nti-inflmmtory effects. As inflmmtion plys n importnt role in the pthogenesis of osteorthritis [37,38], it is plusile tht these supplements my e of enefit during the developmentl stge of osteorthritis. Unfortuntely, the cute, temporry synovitis model does not llow for the drwing of conclusions out possile long-term effects. Aprt from n increse in serum GGT ctivity, no chnges in lood hemtology or serum iochemistry were oserved fter supplementtion. Incresed GGT ctivity ws oserved in ll tretment groups, including the plceo group, which suggests it ws not ssocited with the supplementtion; therefore, the present uthors conclude tht the periods of supplementtion with the products descried here (28 dys for AT nd 60 dys for HP) cn e considered sfe. The reson for the incresed GGT ctivity in ll groups remins uncler. The horses did not show ny clinicl symptoms of liver disese or ny other normlities. There ws no dietry chnge during the study. The limittions of the current study re tht the clinicl evlution of horses locomotion ws sed on video-recordings, nd tht pressure plte nlysis ws not performed prior to rticulr chllenge. There re inherent limittions in the evlution of locomotion using video-recordings compred with full clinicl exmintion, ut the use of video-recordings voided oserver is t individul time points. Moreover, pressure plte nlysis confirmed the sence of relevnt symmetry in horses locomotion. At PID 61, mesurements were performed t reeding frm t which lrge herd of horses ws ville, wheres the induction of synovitis nd pressure plte nlysis were performed t the university hospitl. Pressure plte equipment ws not ville t the reeding frm. However, the im of the study ws not to compre locomotion efore nd fter supplementtion with nutrceuticls, ut to evlute the effects of the preventive dministrtion of nutrceuticls on the cute synovitis induced with LPS. 536 Equine Veterinry Journl 49 (2017) 532 538 2016 The Authors Equine Veterinry Journl pulished y John Wiley & Sons Ltd on ehlf of EVJ Ltd

E. Vn de Wter et l. Effects of nutrceuticls in synovitis Conclusions The nutrceuticls investigted in this study meliorted experimentl joint inflmmtion in vlidted synovitis model. Therefore, the clinicl enefits of these products in ptients with vrious degrees of joint pthology should e evluted. Authors declrtion of interests D.A. vn Doorn ws previously employed y Nutriquine NV nd hired in his role s n equine nutrition consultnt y the funders of this project (Nutriquine NV, Drongen, Belgium nd Sonc BV, Son, the Netherlnds) to co-ordinte the execution of the project. He does not receive ny roylties relting to the products tested in this tril. The funding compnies did not prticipte in nlysis or the decision to pulish. All uthors declre they hd full utonomy nd independency in reserch nd pulishing. Ethicl niml reserch The study ws pproved y the ethicl committee of the Fculty of Veterinry Medicine of Ghent University (no. 2013/165). Sources of funding This study ws funded y Nutriquine NV (Drongen, Belgium) nd Sonc BV (Son, the Netherlnds). Acknowledgements The uthors would like to thnk Dr Gy Vndele nd Dr Hilde Vndele of the Keros Breeding Centre for their co-opertion in the study, nd Dr Jnny de Gruw for her dvice concerning the lipopolyscchride model nd iomrkers. Authorship E. vn de Wter, M. Oosterlinck, F. Pille, M. Dumoulin nd D.A. vn Doorn contriuted to the study design nd execution, dt nlysis nd interprettion, nd the preprtion of the mnuscript. N.M. Korthgen, P.R. vn Weeren, J. vn den Broek nd H. Everts contriuted to the study design, dt nlysis nd interprettion, nd the preprtion of the mnuscript. All uthors pproved the finl mnuscript. Mnufcturers ddresses Nutriquine NV, Drongen, Belgium. Sonc BV, Son, the Netherlnds. c Boehringer Ingelheim Vetmedic, Ingelheim/Rhein, Germny. d Sigm-Aldrich Corp., St Louis, Missouri, USA. e PerkinElmer, Inc., Wlthm, Msschusetts, USA. f Applied Biosystems Interntionl, Inc., Bleiswijk, the Netherlnds. g Genorise Scientific, Inc., Glen Mills, Phildelphi, USA. h IBEX Phrmceuticls, Inc., Queec, Cnd. i Brchem AG, Buendorf, Germny. j Greiner Bio-One GmH, Frickenhusen, Germny. k BMG Ltech, Ortenerg, Germny. l RSscn Interntionl NV, Pl, Belgium. m R Foundtion for Sttisticl Computing, Vienn, Austri. References 1. Neundorf, R.H., Lowerison, M.B., Cruz, A.M., Thomson, J.J., McEwen, B.J. nd Hurtig, M.B. (2010) Determintion of the prevlence nd severity of metcrpophlngel joint osteorthritis in Thoroughred rcehorses vi quntittive mcroscopic evlution. Am. J. Vet. Res. 71, 1284-1293. 2. Trumle, T.N. (2005) The use of nutrceuticls for osteorthritis in horses. Vet. Clin. N. Am.: Equine Prct. 21, 575-597. 3. Brien, S., Prescott, P., Bshir, N., Lewith, H. nd Lewith, G. (2008) Systemtic review of the nutritionl supplements dimethyl sulfoxide (DMSO) nd methylsulfonylmethne (MSM) in the tretment of osteorthritis. Osteorthritis Crtilge 16, 1277-1288. 4. Person, W. nd Lindinger, M. (2009) Low qulity of evidence for glucosmine-sed nutrceuticls in equine joint disese: review of in vivo studies. Equine Vet. J. 41, 706-712. 5. Vndeweerd, J.-M., Coisnon, C., Clegg, P., Cmier, C., Pierson, A., Hontoir, F., Segermn, C., Gustin, P. nd Buczinski, S. (2012) Systemtic review of efficcy of nutrceuticls to llevite clinicl signs of osteorthritis. J. Vet. Intern. Med. 26, 448-456. 6. Orth, M.W., Peters, T.L. nd Hwkins, J.N. (2002) Inhiition of rticulr crtilge degrdtion y glucosmine-hcl nd chondroitin sulphte. Equine Vet. J. 34, Suppl. 34, 224-229. 7. Dechnt, J.E., Bxter, G.M., Frisie, D.D., Trotter, G.W. nd McIlwrith, C.W. (2005) Effects of glucosmine hydrochloride nd chondroitin sulphte, lone nd in comintion, on norml nd interleukin-1 conditioned equine rticulr crtilge explnt metolism. Equine Vet. J. 37, 227-231. 8. Neil, K.M., Orth, M.W., Coussens, P.M., Chn, P.S. nd Cron, J.P. (2005) Effects of glucosmine nd chondroitin sulfte on meditors of osteorthritis in cultured equine chondrocytes stimulted y use of recominnt equine interleukin-1. Am. J. Vet. Res. 66, 1861-1869. 9. Chn, P.S., Cron, J.P. nd Ros, G.J.M. (2005) Glucosmine nd chondroitin sulphte regulte gene expression nd synthesis of nitric oxide nd prostglndin E 2 in rticulr crtilge explnts. Osteorthritis Crtilge 13, 387-394. 10. Byron, C.R., Stewrt, M.C., Stewrt, A.A. nd Pondenis, H.C. (2008) Effects of cliniclly relevnt concentrtions of glucosmine on equine chondrocytes nd synoviocytes in vitro. Am. J. Vet. Res. 69, 1129-1134. 11. Meulyzer, M., Vchon, P., Beudry, F., Vinrdell, T., Richrd, H., Beuchmp, G. nd Lverty, S. (2008) Comprison of phrmcokinetics of glucosmine nd synovil fluid levels following dministrtion of glucosmine sulphte or glucosmine hydrochloride. Osteorthritis Crtilge 16, 973-979. 12. Welch, C.A., Potter, G.D., Gis, P.G. nd Eller, E.M. (2012) Plsm concentrtion of glucosmine nd chondroitin sulfte in horses fter n orl dose. J. Equine. Vet. Sci. 32, 60-64. 13. Gupt, R.C., Cnerdy, T.C., Skggs, P., Stocker, A., Zyrkowski, G., Burke, R., Wegford, K., God, J.T., Rohde, K., Brnett, D., de Wees, W., Bgchi, M. nd Bgchi, D. (2009) Therpeutic efficcy of undentured type-ii collgen (UC-II) in comprison to glucosmine nd chondroitin in rthritic horses. J. Vet. Phrmcol. Ther. 32, 577-584. 14. Eisuzki, K. (2003) Aspirin nd methylsulfonylmethne (MSM): serch for common mechnisms, with implictions for cncer prevention. Anticncer Res. 23, 453-458. 15. Ush, P.R. nd Nidu, M.U. (2004) Rndomised, doule-lind, prllel, plceo-controlled study of orl glucosmine, methylsulfonylmethne nd their comintion in osteorthritis. Clin. Drug Invest. 24, 353-363. 16. Dei, E.M., Agr, G., Fichmn, G., Ziv, Y.B., Krdosh, R., Hlperin, N., Elz, A., Beer, Y. nd Dei, R. (2011) Efficcy of methylsulfonylmethne supplementtion on osteorthritis of the knee: rndomized controlled study. BMC Complement. Altern. Med. 11, 50. 17. Mr~non, G., Mu~noz-Escssi, B., Mnley, W., Grcı, C., Cydo, P., Snchez de l Muel, M., Olrri, B., Leon, R. nd Vr, E. (2008) The effect of methylsulphonyl methne supplementtion on iomrkers of oxidtive stress in sport horses following jumping exercise. Act Vet. Scnd. 50, 45. 18. Oesser, S. nd Seifert, J. (2003) Stimultion of type II collgen iosynthesis nd secretion in ovine chondrocytes cultured with degrded collgen. Cell Tissue Res. 311, 393-399. 19. Clrk, K.L., Sestinelli, W., Flechsenhr, K.R., Aukermnn, D.F., Mez, F., Millrd, R.L., Deitch, J.R., Sherondy, P.S. nd Alert, A. (2008) 24-week study on the use of collgen hydrolyste s dietry supplement in thletes with ctivity-relted joint pin. Curr. Med. Res. Opin. 24, 1485-1496. 20. Bruyere, O., Zegels, B., Leonori, L., Rend, V., Jnssen, A., Bourges, C. nd Reginster, J.-Y. (2012) Effect of collgen hydrolyste in rticulr pin: 6-month rndomized, doule-lind, plceo-controlled study. Complement. Ther. Med. 20, 124-130. Equine Veterinry Journl 49 (2017) 532 538 2016 The Authors Equine Veterinry Journl pulished y John Wiley & Sons Ltd on ehlf of EVJ Ltd 537

Effects of nutrceuticls in synovitis E. Vn de Wter et l. 21. Coenen, M., Appelt, K., Niemeyer, A. nd Vervuert, I. (2006) Study of geltin supplemented diet on mino cid homeostsis in the horse. Equine Vet. J. 36, Suppl. 36, 606-610. 22. de Gruw, J.C., vn de Lest, C.H.A. nd vn Weeren, P.R. (2009) Inflmmtory meditors nd crtilge iomrkers in synovil fluid fter single inflmmtory insult: longitudinl experimentl study. Arthritis. Res. Ther. 11, R35. 23. de Gruw, J.C., vn de Lest, C.H.A., Brm, P.A.J., Rmgs, B.P.B. nd vn Weeren, P.R. (2009) In vivo effects of meloxicm on inflmmtory meditors, MMP ctivity nd crtilge iomrkers in equine joints with cute synovitis. Equine Vet. J. 41, 693-699. 24. Ful, F., Erdfelder, E., Buchner, A. nd Lng, A.G. (2009) Sttisticl power nlyses using G*Power 3.1: tests for correltion nd regression nlyses. Behv. Res. Methods 41, 1149-1160. 25. Person, W., Orth, M.W. nd Lindinger, M.I. (2009) Evlution of inflmmtory responses induced vi intr-rticulr injection of interleukin-1 in horses receiving dietry nutrceuticl nd ssessment of the clinicl effects of long-term nutrceuticl dministrtion. Am. J. Vet. Res. 70, 848-861. 26. Centrl Veevoederureu (2004) Het EWp en VREp system. CVB documenttion report no. 31, CVB, Lelystd. 27. Bssge, II, L.H. nd Ross, M.W. (2011) Dignostic nlgesi. In: Dignosis nd Mngement of Lmeness in the Horse, 2nd edn., Eds: M.W. Ross nd S.J. Dyson, Sunders Elsevier, London. pp 100-135. 28. de Gruw, J.C., Brm, P.A.J., Wiemer, P., Brommer, H., Brm, P.A., vn de Lest, C.H.A. nd vn Weeren, P.R. (2006) Crtilge-derived iomrkers nd lipid meditors of inflmmtion in horses with osteochondritis dissecns of the distl intermedite ridge of the tii. Am. J. Vet. Res. 67, 1156-1162. 29. Lverty, S., Ionescu, M., Mrcoux, M., Boure, L., Doize, B. nd Poole, A.R. (2000) Altertions in crtilge type-ii procollgen nd ggrecn contents in synovil fluid in equine osteochondrosis. J. Orthop. Res. 18, 399-405. 30. Billinghurst, R.C., Brm, P.A., vn Weeren, P.R., Knowlton, M.S. nd Mcllwrith, C.W. (2003) Significnt exercise-relted chnges in the serum levels of two iomrkers of collgen metolism in young horses. Osteorthritis Crtilge 11, 760-769. 31. Americn Assocition of Equine Prctitioners (1991) Guide for Veterinry Service nd Judging of Equestrin Events, 4th edn., AAEP, Lexington, Kentucky. 32. Oosterlinck, M., Pille, F., Bck, W., Dewulf, J. nd Gsthuys, F. (2010) Use of stnd-lone pressure plte for the ojective evlution of forelim symmetry in sound ponies t wlk nd trot. Vet. J. 183, 305-309. 33. Burhm, K.P. nd Anderson, D.R. (2002) Model Selection nd Multimodel Inference: A Prcticl Informtion-Theoretic pproch, 2nd edn., Springer Verlg, Berlin. 34. Hnson, R.R., Smlley, L.R., Huff, G.K., White, S. nd Hmmd, T.A. (1997) Orl tretment with glucosmine-chondroitin sulfte compound for degenertive joint disese in horses: 25 cses. Equine Prct. 19, 16-20. 35. Cyzer, J., Hedderley, D. nd Gry, S. (2012) A rndomized, douledlinded, plceo-controlled study on the efficcy of unique extrct on green-lipped mussel (Pern cnliculus) in horses with chronic fetlock lmeness ttriuted to osteorthritis. Equine Vet. J. 44, 393-398. 36. Person, W., Fletcher, R.S. nd Kott, L.S. (2011) Orl rosmrinic cidenhnced Menth spict modultes synovil fluid iomrkers of inflmmtion in horses chllenged with intr-rticulr LPS. J. Vet. Phrmcol. Ther. 35, 495-502. 37. Smith, M.D., Trintfillou, S., Prker, A., Youssef, P.P. nd Colemn, M. (1997) Synovil memrne inflmmtion nd cytokine production in ptients with erly osteorthritis. J. Rheumtol. 24, 365-371. 38. Benito, M.J., Vele, D.J., FitzGerld, O., vn den Berg, W.B. nd Bresnihn, B. (2014) Synovil tissue inflmmtion in erly nd lte osteorthritis. Ann. Rheum. Dis. 64, 1263-1267. Supporting Informtion Additionl Supporting Informtion my e found in the online version of this rticle t the pulisher s wesite: Summry in Chinese. 538 Equine Veterinry Journl 49 (2017) 532 538 2016 The Authors Equine Veterinry Journl pulished y John Wiley & Sons Ltd on ehlf of EVJ Ltd