In vitro differentiation potential of clones derived from normal individuals and patients with OA.

Representative histological staining of Alizarin Red, Safranin O and Oil Red O from 4 normal (A) and 4 OA clones (E). QPCR data is presented for each of the clones for osteogenic (B,F), chondrogenic (C,G) and adipogenic (D,H) markers. *p<0.05, **p<0.01

OA scoring with and without cell injection in a rat DMM model.

Representative images of rats with sham surgery (A, A’), DMM injury with injection of saline (B, B’), DMM injury with injection of MSCs (C, C’) or non-MSCs (D, D’). The OARSI and Krenn scoring was quantified in each group (E). Scale bars equal 200µm in A,B,C,D and 20µm in A’,B’C’D’. *p<0.05.

Exogenous cell contribution to cartilage repair in a rat DMM model.

Representative images of rats with sham surgery (A) without cell injection (no TdTomato – B) and endogenous PRG4/Lubricin staining (C). In rats injected with saline (D), no TdTomato (E) and disrupted PRG4/Lubricin (F) staining was observed. When normal (G) or OA (M) MSCs where injected into injured rats, TdTomato expression (H.N) was observed. When normal (J) or OA (P) non-MSCs were injected no TdTomato (K,Q) and disrupted PRG4/Lubricin (L,R) staining was observed. Scale bars equal 30µm. Tissue cytometry was employed to quantify the number of tdTomato positive, PRG4 positive and double positive cells within the articular cartilage (S). *=p<0.05, N.S. = not significant.

Contribution of transplanted cells to inflammatory signaling.

Representative images of MSCs (A) and non-MSCs (B) in the synovium of rats post-DMM. In rats injected MSCs, only minimal CCL2 (A”) expression is observed and/or colocalized with tdTomato staining (A’). In rats injected with non-MSCs, intense CCL2 staining (B”) was observed that colocalized with tdTomato staining (B’). this was quantified using tissue cytometry (C). In rats injected MSCs, little Ki67 staining was observed in the cartilage (D-D”). In rats injected with non-MSCs, nearly every transplanted cell in the synovium was also positive for Ki67 (E-E”). This was quantified using tissue cytometry (F). *=p<0.05, n.d. = none detected.

Proteomics analysis of MSC vs. non-MSC populations.

Schematic representation of the quantitative shotgun proteomics workflow (A). Volcano plot demonstrating the distribution of statistically changing proteins identified in MSCs (red) and Non-MSCs (blue) (C). Metascape analysis and heatmap showing pathways regulated by differentially expressed proteins in MSCs vs. non-MSCs (D). STRING-db analysis of protein-protein interaction networks with elevated proteins in MSCs (D).

Analysis of synovial cells expressing CD47.

The gating strategy to identify and sort CD47Hi vs. CD47Lo cell populations (A,B). CD47Hi vs. CD47Lo cells were identified and isolated from normal and OA synovial tissue (B,C) The CD47Hi vs. CD47Lo cells were injected into rats that underwent DMM injury and it was observed that rat which received the CD47Hi cells presented with significantly lower OARSI scores than rats receiving CD47Lo cells. Immunofluorescence analysis demonstrated that the transplanted human CD47Hi cells (tdTomato) integrated into the rat articular cartilage and expressed the chondrocyte marker Sox9 (E). This was not observed in the cartilage of rats that received the CD47Lo cells (F).