*3.2. Perlecan and Collagen Types III and VI Have a Diffuse Immunolocalisation in Repair Cartilage Tissues*

Perlecan was localised in a discrete manner in the pericellular matrix around chondrocytes in healthy cartilage (Figure 2A,B). However, in naturally and cell therapy repaired cartilage staining for perlecan was seen in a pericellular location in some biopsies, diffusely throughout the matrix in others or both patterns within others. Where fibrocartilage was more abundant, perlecan was more diffuse in the cartilage matrix with some strong staining around chondrocytes, which was strikingly different to healthy cartilage as illustrated in Figure 2C,D, showing donors 10 and 9, respectively. In both natural and cell therapy repaired tissues where hyaline cartilage was visible, perlecan was mostly localised in the pericellular regions, but more prominently than in normal cartilage (Figure 2E, showing donor 2). The more elongated cells within fibrocartilaginous repair tissue were generally weak or moderately stained for perlecan, compared to the more rounded chondrocytes in hyaline cartilage (both repair and normal cartilage) which had strong pericellular perlecan immunostaining. Disorganised fibrous tissue was associated with weak matrix perlecan staining. Isotype controls are shown in Figure S1.

μm. **Figure 1.** Representative histology images of cartilage repair biopsies from the same donor. Natural repair (**A**) and cell therapy repair (**B**) cryosections were stained with haematoxylin and eosin (H&E) to assess general morphology and toluidine blue (TB) to assess proteoglycan content; both samples demonstrated good to excellent matrix metachromasia. Polarised light was used to assess collagen fibre orientation and determine tissue morphology. The natural repair cartilage demonstrated a mostly hyaline (h) morphology whilst the cell therapy repair cartilage was mostly fibrocartilage (f). Scale bars 500 µm.

The perlecan immunohistochemistry scores were similar between the two repair tissues, with no noticeable trend when comparing individual donor-matched samples (Figure 3A). Image analysis of the percentage of perlecan staining in the tissues showed that naturally repaired and cell therapy repaired cartilage had significantly more staining than the healthy tissues (*p* = 0.017 and *p* = 0.018, respectively, Figure 3B). Interestingly, an increase in the perlecan score significantly correlated with a better-quality cell therapy repair, as defined by the ICRS II 'overall score' parameter (r = 0.75, *p* = 0.03, Figure 3C). Perlecan was also strongly localised around small blood vessels that were visible in 6 of the 10 naturally repaired. No blood vessels were observed in either the cell therapy repaired cartilage samples, or the healthy cartilage.

**Figure 2.** Immunohistochemistry of perlecan. Monoclonal antibodies (A74) were used to detect the presence of perlecan in cryosections of core biopsies. (**A**,**B**) Heathy cartilage (*n* = 5) from cadavers all showed distinct pericellular staining for perlecan with a typically hyaline morphology. (**C**,**D**) Naturally repaired cartilage (*n* = 10) from the harvest site of autologous cell therapy donors showed heterogenous staining patterns, some having both widespread matrix and pericellular staining ((**C**), donor 10), whilst in others there was diffuse matrix staining throughout ((**D**), donor 9). (**E**,**F**) Cell therapy repaired (CT, *n* = 9) cartilage also showed a heterogenous localisation for perlecan, similar to the naturally repaired tissues. The sample depicted in (**E**) (donor 2) shows pericellular staining for perlecan in repair tissue with hyaline cartilage morphology, but not as discretely as in the healthy tissues. The sample depicted in (**F**) (donor 7) shows predominantly matrix immunolocalisation of perlecan. Scale bars show 300 µm for low magnification images and 100 µm for high magnification inserts. Isotype controls found in Supplementary Figure S1A–C.

therapy repair, as defined by the ICRS II 'overall score' parameter (r =

–

−0.4 **Figure 3.** Analysis of tissue morphology and perlecan staining. (**A**) The perlecan immunohistochemistry score gives a general idea of the localisation (pericellular, non-pericellular, mixed) of perlecan in the deep, middle and superficial zones of cartilage. The zones were scored as 0 = no staining, 1 = pericellular staining, 2 = mixture of pericellular and matrix staining, or 3 = matrix staining. The final perlecan score shown here is the summation of the scores for the three zones in each sample. No difference was found between the donor-matched natural and cell therapy repaired tissues. Data show the median with interquartile range. (**B**) Threshold image analysis confirmed a higher percentage of perlecan staining in naturally repaired and CT repaired cartilage than in health cartilage. Perlecan was significantly more prominent in the repair tissues compared to controls. (**C**) Regression analysis showed a positive correlation between the ICRS score and perlecan immunohistochemical score (*r* = 0.75, *p* = 0.03) for cell therapy repaired, but not naturally repaired tissues (*r* = −0.4, *p* = 0.25).

– – Collagen types III and VI generally exhibited a diffuse staining pattern throughout the interterritorial matrix, covering 94.3 ± 8.9% (range 70–100) and 95.2 ± 7.1% (range 80–100) of the section area, respectively (Figure 4C,D). However, where there was hyaline cartilage present in the repair tissues (Figure 4B), the staining pattern in these regions for both collagen types III and VI was similar to what is typically observed in healthy cartilage (Figure 4A) [28,29], with the pericellular matrix being immunonegative for collagen type III and immunopositive for collagen type VI and the territorial matrix being immunopositive for collagen type III and immunonegative for collagen type VI.

**Figure 4.** Immunohistochemistry of collagen types III and VI. Monoclonal and polyclonal antibodies were used to detect the presence of collagens type III and VI, respectively, in cryosections of core biopsies. For the repair tissues, two donor-matched samples of natural and cell therapy (CT) repaired cartilage are shown as representative examples (B + C = donor 7, D + E = donor 9). (**A**) Healthy cartilage showing interterritorial staining for collagen type III and pericellular staining for type VI (donor 15). (**B**,**C**) In this instance of hyaline-like cartilage in naturally repaired cartilage, the collagen type III was localised in the interterritorial region while collagen type VI was localised in the pericellular matrix. (**D**,**E**) Both collagens type III and VI are diffused in the matrix of fibrocartilage. Scale bar = 500 µm. Isotype controls found in Figure S1D,E.
