Proteoglycan 4 downregulation in a sheep meniscectomy model of early osteoarthritis

Twelve four-year-old female pure-bred Merino sheep were used for this study. Six of the sheep underwent open lateral meniscectomy of both stifle (knee) joints as described previously [12], and the other six underwent a sham operation identical in all aspects except that the lateral meniscus was not excised. After recovery from surgery, the animals were maintained in an open paddock for three months before being killed. The protocol used for this study was approved by the animal ethics committee of Murdoch University (AEC 832R/00).

Full-depth articular cartilage was harvested from the medial tibial plateau (MTP) and lateral tibial plateau (LTP) in regions normally covered (COV) and uncovered (UNCOV) by menisci from either the right or left stifle joint, randomly selected (Figure 1a). Care was taken not to sample tissue from the joint margins or osteophytes. Tissue samples were snap frozen in liquid nitrogen before storage at -80°C until required. Coronal full-thickness osteochondral slabs (5 mm) were prepared through the mid weight-bearing region of the tibial plateau from the contralateral joint of each animal (Figure 1a). The specimens were then fixed, decalcified and processed before staining with toluidine blue and fast green as described previously [12]. Histological slides were subsequently evaluated by two independent observers with a modified Mankin scoring scheme, previously developed in our laboratory for this ovine model [12]. The modified Mankin score has a range of 0 to 29, the value increasing with severity of cartilage degeneration. In each compartment the worst score evident for the region examined was used to calculate the mean score (n = 6 for each group).

To avoid the necessity for decalcification, articular cartilage spanning the entire MTP or LTP was micro-dissected as a single piece from the underlying subchondral bone after formalin fixation of osteochondral slabs of two representative sham-operated and meniscectomised animals, and these full-thickness cartilage specimens were then embedded in paraffin and 4 μm sections were deparaffinised in xylene and graded ethanols. Sections were digested with Proteinase K (code no. S3020; DakoCytomation, Glostrup, Denmark) for six minutes at room temperature followed by incubation in Protein Block Serum Free (code no. X909; DakoCytomation) for ten minutes at room temperature. Primary antibody incubations were performed overnight at 4°C with a 1:600 dilution of 06A10, a protein A purified rabbit polyclonal anti-PRG4 antibody generated by immunisation with a truncated form of recombinant human PRG4 (generously provided by Wyeth Pharmaceuticals, Boston, MA, USA). This antibody has previously been used to specifically detect PRG4 in the superficial zone of bovine articular cartilage [13]. Secondary antibody incubation and colour development were performed as described previously. The intensity and number of positively stained cells were evaluated across the width of the MTP and LTP by two observers [14].

About 100 mg of frozen cartilage samples was fragmented in a Mikro-Dismembrator (Braun Biotech International, Melsungen, Germany). Total RNA was isolated with the RNeasy Mini Kit (Qiagen, Valencia, USA) and quantified with a fluorimeter (Perkin Elmer, Beaconsfield, UK) with the use of SYBR^® Green II colour reagent (Cambrex Bio Science, Rockland, USA). The quality and integrity of total RNA were assessed on 2% (w/v) agarose gels stained with ethidium bromide. For all samples total RNA (1 μg) was reverse transcribed into cDNA with the Omniscript kit (Qiagen, Hilden, Germany) in accordance with the manufacturer's instructions. Real-time PCR was performed with a Prism 7000 Sequence Detection System from Applied BioSciences (Foster City, CA, USA). Primers were designed to bovine PRG4 (ovine sequence 99% homologous) (forward, 5'-CTGCCCAACATCAGAAAACCC-3' ; reverse, 5'-TTCCTTCGCCCATCAGTCTAAG-3') (Genbank accession no. AF056218) and generated a single PCR product in sheep, which was confirmed by sequencing (SUPAMAC, Sydney, Australia). cDNA (1 μl, corresponding to the reverse transcription of 25 ng of total RNA), 0.5 μl of each primer (10 μM), 0.5 μl of SYBR green (Molecular Probes, Sydney, Australia) and the Platinum Plus Taq (7 μl; Invitrogen, Sydney, Australia), with ROX (6-carboxy-X-rhodamine) (0.1 μl) used as an internal control. The thermal profile was as follows: 50°C for five minutes, 95°C for five minutes, and 40 cycles of 95°C for 15 seconds, 58°C for 30 seconds and 72°C for 30 seconds. Melt curves were also determined to demonstrate the specificity of the amplification. Standard curves were generated from plasmids (pGEM Teasy; Promega, Sydney, Australia) containing the PCR products, and the linear amplification range for both plasmid DNA and sample cDNA was determined. Analysis of the curves showed that the sample diluted in a parallel manner to the plasmid and that the cycle threshold (Ct) for all unknown samples fell within the linear range, allowing quantification of gene copy number relative to the plasmid concentration. Samples were analysed in triplicate and values were normalised to total RNA as recommended for experiments in vivo involving tissue specimens [15].

Comparisons of non-parametric data from the modified Mankin histological scoring of the stained tissue sections were assessed with the Mann-Whitney U test. Statistical evaluation of significant differences in expression levels was undertaken with the unpaired Student t test with Benjamini-Hochberg correction for multiple comparisons.

Lateral meniscectomy resulted in macroscopic joint changes characteristic of OA with cartilage fibrillation and erosion, particularly in the lateral compartment (Figure 1b). The most severe lesions were confined to the COV region of the LTP, with surface fibrillation and variable loss of the characteristic superficial zone cells. Histological grading of the cartilage specimens confirmed and quantified the regional histological observations. Modified Mankin scoring was significantly increased in the LTP COV (4.4 ± 1.2 to 17.6 ± 2.7; mean ± SD; p < 0.01) and LTP UNCOV (3.1 ± 1.1 to 6.7 ± 1.9; p < 0.01) regions after lateral meniscectomy; however, it remained unchanged in MTP COV (2.3 ± 0.9 to 4.3 ± 2.6) and MTP UNCOV (2.6 ± 1.2 to 4.3 ± 1.6) regions.

PRG4 was immunolocalised to chondrocytes in the superficial zone of normal cartilage (Figure 2a,c,e,g; positive PRG4 indicated by red-brown colour) and little or no extracellular matrix staining was detected. Additionally, positive immunostaining cells were more prominent in the COV regions of normal cartilage than in the UNCOV regions. After meniscectomy, extensive loss of PRG4-positive cells was observed in the superficial zone of LTP specimens, both COV and UNCOV regions (Figure 2b,d), corresponding to areas of degenerative change. No obvious change in PRG4 immunostaining was observed in the MTP COV and UNCOV cartilage regions after meniscectomy (Figure 2f,h).

To evaluate whether topographical variation existed in the expression of PRG4 in normal joints, the ratio of mRNA levels in the COV versus UNCOV cartilage regions in sham-operated sheep was evaluated. PRG4 expression was found to be increased (1.8-fold, p < 0.05) in COV compared with UNCOV cartilage in the MTP and similarly increased (1.6-fold, p < 0.05) in COV compared with UNCOV cartilage in the LTP (Figure 3a).

After lateral meniscectomy, PRG4 mRNA levels were found to be significantly decreased compared with sham-operated controls in cartilage from the covered region of the LTP (7.0 fold, p < 0.01) (Figure 3b). PRG4 expression was also found to be decreased in the UNCOV region of the LTP (2.4 fold, p < 0.05) and in the COV region of the MTP (3.8 fold, p < 0.05).