C60 Poster Presentations 90 AN ACUTE MODEL OF CYTOKINE-STIMULATED ARTICULAR CARTILAGE DEGRADATION IN THE DOG M.A. Pratta 1 , S.J. Hoffman 1 , F.L. Wang 1 , T. Newman-Tarr 1 , R.W. Coatney 2 , J.B. Morris 2 , C.A. Capriotti 1 , J.L. McLane 1 , K.S. Kilgore 1 , S. Kumar 1 1 GlaxoSmithKline Pharmaceutical Co., Collegeville, PA; 2 GlaxoSmithKline Pharmaceutical Co., King of Prussia, PA Purpose: To develop an acute model of articular cartilage degra- dation in the dog for progression of pharmaceutical agents into long term chronic models. Methods: Canine cytokines were cloned and expressed as de- scribed in a recent publication. All procedures were approved by the GSK Institutional Animal Care and Use Committee. Articular cartilage derived from various joints from Beagle dogs was iso- lated and was either used in cartilage explant assays in organ culture or as a source of primary chondrocytes. Chondrocytes were evaluated in a cell-based aggrecanase assay based on a published method. For in vivo studies, beagle dogs (n=4-6) were anesthetized for both intra-articular (IA) injection and synovial fluid (SF) collection of the stifle joint. Human or canine cytokines were injected in one knee and vehicle into the contralateral. At designated times, SF was collected by arthrocentesis; SF col- lected 7 days pre-injection represented baseline control for each animal. Total aggrecan content (GAG) of the SF was measured using an alcian blue assay and aggrecanase-generated aggrecan fragments (ARGSVIL) were measured by ELISA. Dogs were pro- vided veterinary oversight during the in-life portion of the studies. Results: Stimulation of canine articular chondrocytes and canine cartilage explants with canine IL-1 resulted in induction of aggre- canase activity and aggrecan degradation, respectively, whereas IL-1α and β of human origin were ineffective. Interestingly, both human and canine OSM induced aggrecanase activities and car- tilage aggrecan degradation. Based on these in vitro findings, the effects of cytokines on cartilage aggrecan degradation following IA injection into the knee joints of beagle dogs were evaluated. IA injection of canine IL-1β at doses between 1-200 ng resulted in 3-legged lameness that was dose-dependent in severity and duration. Therefore, the use of IL-1 was discontinued for ethical concerns. However, IA injection of both human and canine OSM at doses up to 1000 ng resulted in no detectable lameness and joint effusion over that of vehicle alone. In addition, OSM injec- tion caused increases in both GAG and the ARGSVIL aggrecan fragment, and based on dose response studies, the optimal dose of canine OSM was determined to be 200 ng/joint. Evaluation of the time course following IA injection of 200 ng of OSM re- sulted in GAG and the ARGSVIL-fragment levels that reached a peak relative to the contralateral controls at 24 hours, and then returned to vehicle levels by 48 hours post-injection. Because of the acute nature of the cartilage aggrecan loss of this model, repeated use of the same colony of dogs has been possible. Conclusions: We have established an acute model of OSM- stimulated cartilage degradation in the dog. This model has utility to for screening of multiple pharmaceutical agents prior to their evaluation in more long term chronic models in the dog or other species. 91 A PRECISE AND SCALABLE METHOD OF MICE KNEE SYNOVIAL FLUID COLLECTION FOR USE IN BIOMARKER ANALYSES D.R. Seifer, B.D. Furman,F. Guilak, S.A. Olson, V.B. Kraus Duke University, Durham, NC Purpose: To develop methods for accurate recovery of the approximate 1-μL of synovial fluid (SF) present in a mouse knees that would permit OA-related biomarker analyses. Methods: Through an investigation into unique materials, poly- acrylate beads (PAB) and the calcium sodium alginate compound (CSAC) Melgisorb (Tendra, Goteborg) were identified as poten- tial SF recovery vectors. The SF recovery efficiency of these novel materials were compared against that of Whatman paper (WPR), which has previously been used in larger (≥3-μL) appli- cations. Once SF is absorbed by the PAB, it is recovered by the addition and subsequent removal of 100-μL of saturated NaCl solution. The optimal time course for maximal SF recovery using the PAB recovery vector was determined through the mapping of efficiency curves. In the case of CSAC, the addition of 35μL of flavobacterium-derived alginate lyase in H 2 O and 15μL 1.0M Sodium citrate dissolved the alginate fibers. The refinement of a small-volume measurement technique, termed “pipette-dialing”, allows for the change in volume (dV) to be quantified in both methods. Each method was assessed for human SF percent retrieval of liquid volume (pipette dialing), total protein (Bradford assay), and cartilage oligomeric matrix protein (COMP, ELISA with monoclonal antibody 12C4). Based on the results of these analyses, the CSAC method was tested in vivo. Upon sacrifice, synovial fluid was collected from both knees of twelve adult male C57BL/6 mice, six of which had been given left knee articular fractures. All of these mice were raised on a high fat (60%) diet. COMP was measured in these samples using the 12C4 ELISA, and all data were analyzed using SPSS statistical software. Results: PAB volume recovery peaked at 5 minutes. Average volume recovery ratios (V recovered /V absorbed ) were found to be 0.987 + 0.367 and 1.00 + 0.502 for PAB and CSAC respec- tively. With the WPR method, it was not possible to accurately measure dV in sample sizes ≤3-μL. Lightly cross-linked PABs (PSL), medium cross-linked PABs (PSM), and CSAC were more efficient (62.7%, 47.1%, and 27.1%) at recovering protein in vitro from 1-μL of human SF than the highly cross-linked PABs (PSH), which recovered only 20.59% of the total protein in 1-μL of human SF; however, the PSL and CSAC procedures were significantly worse at protein recovery than the Whatman paper method, which achieved >95% protein recovery in vitro (p<0.01) (see Figure). CSAC was chosen for testing in vivo due to: 1) the ability to precisely and accurately quantify the volume of liquid recovered, 2) a lower, non-specific background in the 12C4 ELISA assay compared with PAB recovered samples, and 3) the fact that it is easier than polyacrylate to physically manipulate during in vivo trials. The mean COMP concentration of the frac- tured left knees trended higher than that of the corresponding non-fractured right knees in the same animals (3444 + 1515-ng/μL versus 2924 + 1302-ng/μL, p=0.075 in a Wilcoxon Signed-Rank Test). The concentration ratio ([COMP right knee ]/[COMP left knee ]) was significantly higher in the mice subjected to articular fracture (p=0.026, Mann-Whitney U Test). Conclusions: Even though the Whatman paper method was most efficient at protein recovery, its inability to measure dV ≤1-μL would prevent accurate biomarker determinations for in vivo situations of limited (<3 μl) SF availability; therefore, the