Inhibition of the SDF-1/CXCR4 signaling cascade was evaluated in vitro by using human cartilage explant cultures and human chondrocyte cultures, and in vivo by using the Duncan Hartley guinea pig model of progressive idiopathic knee osteoarthritis.
Blockage of SDF-1/CXCR4 in human cartilage explants
Cartilage explant culture
The study was approved by the Institutional Review Board at Rhode Island Hospital, and informed consent was obtained from each donor. Articular cartilage samples were obtained from patients with OA at time of total knee arthroplasty (n = 3); two women (ages 55 and 76 years) and one man (age 55 years). At harvest, the samples were immediately placed into DMEM culture medium and transported to the laboratory, where 1.5 × 0.5-cm square full-thickness cartilage explants were cut from the normal medial tibia region (Mankin score, 0 to 2) by using scalpels and were placed into 24-well plate in DMEM culture medium containing 10% FCS (Gibco, Grand Island, NY, USA) at 37°C in 5% CO2.
The 1.5 × 0.5-cm explants were cut into three equal parts (0.5 × 0.5 cm) and randomly divided into three treatment groups. Explants in Group 1(n = 3) were incubated with SDF-1 (250 ng/ml) to evaluate the penetration of SDF-1 into cartilage. Explants in Group 2 (n = 3) were incubated in media containing anti-CXCR4 monoclonal antibody (100 ng/ml; R&D systems Inc., Minneapolis, MN, USA) plus SDF-1 (250 ng/ml), to evaluate the effect of receptor blockade. The explants in Group 3 (n = 3) were left untreated as controls. The cartilage explants and culture medium were collected on days 2 and 4.
The cultured cartilage explants from each group were rinsed with HBSS in Tissue-Tek OCT (SakuraFinetek USA, Torrance, CA, USA) and snap frozen in liquid nitrogen. Serial 10-μm sections were cut perpendicular to the cartilage surface. The sections were fixed for 20 minutes at -20°C by using 70% ethanol containing 50 mM glycine. Sections were treated with hyaluronidase (2 mg/ml; Sigma Chemical Co., St Louis, MO, USA) for 30 minutes at 37°C (only for SDF-1 and CXCR4 immunostaining) and permeabilized in 0.2% Triton X-100/PBS for 5 minutes at room temperature (RT). Endogenous peroxidase was quenched, and endogenous biotin and avidin binding sites were blocked by the sequential incubation with avidin and biotin for 15 minutes and a blocking solution for 10 minutes at RT.
Penetration of SDF-1 into cartilage was evaluated with immunostaining with anti-SDF-1 antibody (25 μg/ml; R&D Systems), whereas chemokine receptor CXCR4 was evaluated with anti-CXCR4 antibody (25 μg/ml; R&D Systems). Both antibodies were applied for 1 hour at 37°C, followed by incubation with biotinylated secondary antibodies for 10 minutes at room temperature. After washing with PBS, sections were incubated with a streptavidin-peroxidase conjugate for 10 minutes, followed by a solution containing diamino-benzidine (DAB; chromogen) and 0.03% hydrogen peroxide for 5 minutes. Sections were counterstained with hematoxylin. Photographs were taken with a Nikon microscope. Additional sections were stained with Safranin-O, and the severity of proteoglycan loss and cartilage damage was quantified by using the modified Mankin grading system .
Glycosaminoglycan and MMP-13 release to culture medium
Culture media were collected at the same time as the cartilage explants (days 2 and 4), and the sulfated-glycosaminoglycan was quantified spectrophotometrically by using dimethylmethylene blue dye (DMMB) with bovine chondroitin sulfate as standard controls . The concentration of MMP-13 activity in the medium was quantified with ELISA (catalog no. F13M00; R&D Systems).
Blockage of SDF-1/CXCR4 in cultured human cartilage cells
Chondrocyte isolation and culture
Chondrocytes were isolated from the OA cartilage samples described earlier by using our standard method . In brief, small pieces of cartilage (≈0.5 g) were minced, digested with pronase (2 mg/ml, Boehringer Roche) for 30 minutes at 37°C, and then digested with bacterial collagenase (1 mg/ml; Type IA, Sigma, C 2674) for 6 to 8 hours at 37°C in a shaker. Residual multicellular aggregates were removed by filtration, and the cells were plated in DMEM containing 10% FCS, L-glutamine, and antibiotics. After cells were grown to confluence, they were split once (passage 1) and plated either in eight-well chambers (Nalge Nunc International Corp., Naperville, IL, USA) at 1 × 105 cells/well or in 100-mm-diameter culture dishes (Becton Dickinson Labware, Franklin Lakes, NJ, USA) at 1 × 106 cells/plate. At 90% confluence, the cells were cultured under serum-free conditions overnight and then treated with SDF-1 (250 ng/ml) or transfected with SiRNA CXCR4 for 4 hours before SDF-1 treatment .
Blockage of CXCR4 with siRNA
Five micrograms of plasmids containing either the pU6RNAi-CXCR4 vector  or the pU6RNAi empty vector  (gifts from Dr. Song, University of Washington), were transfected into these chondrocytes by using a high-efficiency transfection method (Human Chondrocytes Nucleofector Kit; Amaxa Inc., Gaithersburg, MD, USA). For 24 hours after transfection, the cells were incubated in media with or without SDF-1 (250 ng/ml). The total RNA and cell lysates were collected at 36 and 48 hours, respectively. Real-time RT-PCR was carried out to detect the expression of CXCR-4 and MMP-13. CXCR4 protein expression was also evaluated with Western blotting.
Real-time RT-PCR (qPCR)
Total RNA was isolated from chondrocytes with RNeasy isolation kit (cat. no. 74104; Qiagen USA, Valencia, CA, USA), as previously described . The 1 μg of total RNA was transcribed into cDNA by using the iScripTM cDNA synthesis kit (Bio-Rad, Hercules, CA, USA), and 40 ng/μl of the resulting cDNA was used as the template to quantify the relative content of mRNA by using QuantiTect SYBR Green PCR kit (Qiagen) with DNA Engine Opticon 2 Continuous Fluorescence Detection System (MJ Research, Waltham, MA, USA). The primers were designed by using Primers Express software (BioTools Incorporated, Edmonton, AB, Canada), which yielded
CXCR-4 forward (sense) primer, AAA CTG AGA AGC ATG ACG GAC AA,
CXCR-4 reverse (antisense) primer, GCC AAC ATA GAC CAC CTT TTC AG,
MMP-13 forward (sense) primer, TGC TGC ATT CTC CTT CAG GA,
MMP-13 reverse (antisense) primer, ATG CAT CCA GGG GTC CTG GC,
18S rRNA forward (sense) primer, CGG CTA CCA CAT CCA AGG AA, and
18S rRNA reverse (antisense) primer, GCT GGA ATT ACC GCG GCT.
The 18S rRNA was amplified as the internal control. The cycle threshold values for targets genes were measured and calculated with computer software (MJ Research, Waltham, MA, USA). Relative transcript levels were calculated as × = 2-Δ ΔCt, in which Δ ΔCt = ΔCt E - ΔCt C, and ΔCt E = Ctexp-Ct18S, and ΔCt C = CtC-Ct18S.
Total protein was extracted from cells and quantified by using the BAC Protein Assay Kit (Pierce, Rockford. IL, USA) . In brief, 10 μg of total protein was electrophoresed in 10% SDS-PAGE under reducing conditions before being transferred and probed by a human anti-CXCR4 monoclonal antibody (MAB171, 1:1,000 dilution; R&D Systems) and anti-β-actin polyclonal antibody (1:1,000 dilution; Cell Signaling Technology, Danvers, MA, USA). Horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit immunoglobulin G (IgG) (H+L) (1:3,000 dilution, Bio-Rad Laboratories, Richmond, CA, USA) was used as the secondary antibody. Visualization of immunoreactive proteins was achieved by using ECL Western blotting detection reagents (Amersham, Arlington Heights, IL, USA) and subsequent exposure of the membrane to Kodak X-Omat AR film.
AMD3100 blockage of CXCR4 in the Hartley guinea pig OA model
After receipt of IACUC approval, 35 nine-month-old male Duncan-Hartley guinea pigs (0.88 kg ± 0.21 kg) were obtained from Charles River Laboratories (Wilmington, MA, USA). The animals were allocated randomly into three experimental groups: Group 1 was left untreated to serve as a primary OA control (n = 11); Group 2 received continuous infusion of the CXCR4 blocker AMD3100 (Mozobil; Genzyme) via osmotic minipump (n = 13); and Group 3 received PBS via constant infusion osmotic minipump (n = 11). All animals were weighed every other week and euthanized after 3 months (12 weeks) of treatment, at which point, the knees of animals were aspirated, and the hindlimbs were removed en bloc via careful dissection.
Miniosmotic pump implantation and drug delivery
The guinea pigs were anesthetized with a solution of 0.2% (vol/vol) xylazine (Rompun; Bayer Pharmaceuticals, Brussels, Belgium) and 1% (vol/vol) ketamine (Ketalar; Parke-Davis, Bornem, Belgium) in PBS. The Mini-osmotic pumps (model 2006; Alza Corporation, Mountain View, CA, USA) were inserted into small subcutaneous pockets over the dorsolateral thorax, created by blunt dissection after a small incision (~1 cm). Before insertion, the 200-μl pump reservoirs were filled with 44.44 mg/ml AMD3100 in PBS (Group 2) or PBS alone (Group 3). At an average pumping rate of 0.15 μl/per hour, each animal in Group 2 received 160 μg AMD3100 per day. Because the pumping duration of the Alzet osmotic pump was 6 weeks, the pumps were exchanged once during the course of treatment.
Synovial fluid collection and analysis
At death of the animals, 100 μl of isotonic saline was injected into both knees of each animal, and the knees were flexed and extended 10 times before aspiration . This technique typically yielded 160 to 180 μl of saline/synovial fluid from each animal. The synovial fluid was centrifuged at 2,000 g for 10 minutes to remove cells and debris and then was frozen at -80°C until analysis. Five markers of articular cartilage metabolism were measured in the synovial fluid samples by following the manufacturer's instructions. SDF-1(catalog no. DSA00), pro-MMP-1 (catalog no. SMP100), and active MMP-13(catalog no. F13M00) were measured by using Quantikine ELISA kits from R&D Systems, whereas IL-1β was measured by using an IL-1β ELISA kit from Invitrogen (catalog no. KMC0011C). Colorimetric density of the developed plates was determined by using a microplate reader set to 450 nm (model BF10000; Packard Bioscience, Meridian, CT, USA). All ELISA assays were performed in duplicate. Glycosaminoglycan (GAG) concentration was measured by using a dimethylmethylene blue dye (DMMB) assay .
Blood collection and serum analysis
Blood (5 ml) was collected by cardiac puncture immediately after the animals were killed. The blood was centrifuged at 1,800 g for 10 minutes, and the separated serum samples were then stored at -80°C until analysis. The level of IL-1β in the serum was measured by using the same IL-1β ELISA kit and plate-reader settings used for synovial fluid testing (catalog no. KMC0011C; Invitrogen). As with the synovial fluid samples, all of the serum samples were run in duplicate.
On explantation, gross morphologic lesions on the tibia plateau were visualized with India-ink staining . The explanted tibiae were then fixed in 10% formalin for 72 hours, followed by decalcification in 10% EDTA solution. The tibiae were hemisected on the mid-sagittal plane, and each half was embedded in a single block of Paraplast X-tra (Fisher, Santa Clara, CA, USA). Serial 6-μm-thick sections were cut at intervals of 0 μm, 100 μm, and 200 μm and collected on positively charged glass slides (Superfrost Plus; Fisher Scientific). The sections were stained with Safranin-O/fast green. Cartilage degradation was quantified by using the modified Mankin grading system . Three independent and blinded observers scored each section, and the scores for all of the sections cut from the medial and lateral tibial plateaus were averaged within each joint.
Analysis of variance (ANOVA) was used in the in vitro studies to compare the three groups in terms of the concentrations of GAG, MMP-13, the relative MMP-13 mRNA levels, and. in the in vivo studies, to evaluate the concentrations of SDF-1, GAG, pro-MMP-1, active MMP-13, and IL-1β. The weights of the guinea pigs were adjusted by using an analysis of covariance (ANCOVA), and a two-way mixed absolute intraclass correlation coefficient (ICC) for the modified Mankin score was calculated. Follow-up pair-wise comparisons between multiple experimental groups were carried out with orthogonal contrasts by using the Scheffé test (α = 0.05) and a test of homogeneity. Adjusted P values for the multiple comparisons were reported. Differences were considered significant at P < 0.05. Statistics were performed by using SPSS software (SPSS Inc.).