*2.2. Immunohistochemistry*

Cryosections were brought to room temperature and treated with 4800 U/mL hyaluronidase (Sigma, Merck Life Science UK, Dorset, UK) for 2 h and fixed with 4% formaldehyde for 10 min. Slides were washed 3 times in phosphate buffered saline (PBS) between all steps and all steps were performed at room temperature. Goat and horse serum were used to block non-specific binding of the primary mouse and rabbit antibodies, respectively (30 min). Sections were then incubated with mouse monoclonal primary antibodies against perlecan (clone A74, Abcam, Cambridge, UK), collagen type III (clone FH-7A, Abcam) and a polyclonal rabbit antibody to bovine collagen type VI (kindly gifted by Shirley Ayad, University of Manchester, UK) for 60 min, then incubated with biotinylated goat anti-mouse and horse anti-rabbit secondary antibodies (Vectastain Elite ABC kit, Vector Laboratories, Upper Heyford, UK) for monoclonal and polyclonal primary antibodies, respectively, for 30 min. An isotype-matched IgG was used in place of the primary monoclonal antibodies (R&D, Cat No MAB002) as a negative control and normal rabbit serum (Abcam, Cat no ab7487) for the polyclonal, and 0.3% hydrogen peroxide in methanol was used to block endogenous peroxidase activity (30 min). The Vectastain Elite ABC kit (Vector Laboratories) was used to enhance labelling and the ImmPACT® DAB Peroxidase substrate (Vector Laboratories) was used to reveal staining. The sections were dehydrated in serial solutions of 70%, 90% and 100% isopropanol (2 min each) and cleared in xylene (2 × 5 min). The slides were mounted in Pertex (CellPath, Newtown, UK) before imaging.

A semi-quantitative score was developed to assess the immunolocalisation and degree of staining for perlecan in the superficial, mid, and deep zones of the cartilage biopsies. Each zone was scored separately as 0 = no staining, 1 = pericellular staining, 2 = mixture of pericellular and matrix staining, or 3 = matrix staining. Each sample was then given an overall score which was a summation of the scores for the three zones. A high overall score equates to a more widespread matrix immunostaining, whereas a low score equates to more restricted pericellular staining throughout the tissue. Image analysis was performed using FIJI-ImageJ software (Version 1.5), using the Colour Deconvolution and Threshold plugins to establish the levels of perlecan staining as a percentage of the total area of the section.
