Fig 2: CD248 expression by synovial MSCs from arthritis patients. (A) OA and RA synovial MSCs/fibroblasts (RASF or OA SF) were grown on glass coverslips, acetone-fixed, and stained for CD248 using P1A6 (OA) or rabbit anti-CD248 (RA), followed by FITC-conjugated donkey anti-mouse or TRITC-conjugated donkey anti-rabbit respectively. Staining was visualized by confocal microscopy. Magnification ×400. Synovial fibroblasts were double-stained by flow cytometry with affinity-purified rabbit anti-CD248 and mouse anti-CD90 (Thy-1), followed by RPE-conjugated goat anti-rabbit and FITC-conjugated goat anti-mouse secondary antibodies. The staining was analyzed on a FACSCalibur instrument. The percentage of gated cells in each region is indicated. (B) soluble CD248 forms (110 kDa major band and two further cleaved products) were also detected in the supernatants of RA fibroblast cultures (passage 4) in response to IL1-β (24 h treatment). However, an MMP-3-specific inhibitor failed to control CD248 cleavage. (C) CD248 was immunoprecipitated from the synovial fluid using P7C5 bound to protein G Sepharose. The immunoprecipitated protein was detected by Western blot probing with affinity-purified rabbit anti-CD248, revealing a band of 110 kDa representing soluble CD248 (lane 1). A Western blot of lysates prepared from the synovial tissue of an OA and RA patient revealed a band of >120 kDa.

Fig 3: Cell surface and intracellular distribution of CD248 in tumor cells and skin-derived MSCs (HFFFs). (A) RNA was extracted from a series of cell lines and CD248 expression examined by RT-PCR using CD248-specific primers. CD248-specific bands (711bp) were observed in the IMR-32 neuroblastoma and HFFF cell lines. Actin was used as a control reaction. (B) CD248 expression at the protein level was determined by flow cytometric staining of non-permeabilized cell lines with affinity-purified rabbit anti-CD248 followed by an RPE-conjugated secondary antibody. (C) CD248 localization within the cell was determined by double staining of permeabilized (acetone fixed) HFFFs on glass coverslips with affinity-purified rabbit anti-CD248 and a series of anti-organelle markers. TRITC-conjugated donkey anti-rabbit and FITC-conjugated donkey anti-mouse secondary antibodies were used. Expression was localized to the endosomal compartment by double staining with EEA1 and observing colocalization under a confocal microscope (white arrows). Magnification ×630.

Fig 4: CD248 expression by HFFFs in response to cytokines and growth factors: release of a cleaved soluble form. Quiescent HFFFs were treated for 24 h with 20 ng/mL rhVEGF165, 10 ng/mL bFGF, 1 ng/mL TGF-β1, 200 U/mL IL1-β, 50 ng/mL PMA, or were left untreated. (A) Flow cytometric staining of cells with affinity-purified rabbit anti-CD248, and expression of the results as fold changes in the mean of fluorescence over non-treated cells. (B) Western blot of whole cell lysates with affinity-purified rabbit anti-CD248 and comparison of the intensity of the CD248-specific band (160–170 kDa) following treatment with that of the non-treated cells showed no significant change in total cell CD248 expression following any treatment (p > 0.05). Unequal loading was accounted for by blotting for α-tubulin (50 kDa band). The blots are representative of three separate experiments. The graph represents the mean change in intensity compared with non-treated cells ± SEM, n = 3. (C) TCS was collected from treated cells, freeze-dried and subjected to Western blotting with affinity-purified rabbit anti-CD248. Bands of 110 kDa were observed and the intensities of the bands following treatment were compared to the intensity of the band in non-treated cells. Histograms show the analysis of the band intensity of the WB (n = 1). The blot and graph are representative of three separate experiments. The bars represent the fold increase in intensity over non-treated cells.

Fig 5: Perivascular synovial MSCs (pMSCs) express CD248 (endosialin) and may remain at the perivascular level through the multimerin–CD93 anchoring complex. In inflammatory conditions (e.g., RA), interleukin-1 beta produced by macrophages and stimulating the release of MMP will contribute to the shedding of CD248. Membrane CD248 is known to control the PDGF-induced proliferation of MSCs (switching to an aggressive phenotype involved in pannus formation) but the role of sCD248 in this process or as a decoy receptor remains to be ascertained. Microscope image is shown at ×400.