In vitro cell surface markers are insufficient to identify in vivo/in situ multipotent synovial mesenchymal stem cells isolated from normal or osteoarthritic knees

Abstract

Purpose: Osteoarthritis (OA) is a chronic disease that results in the degeneration of the articular cartilage, but also effects all tissues present in the joint. Current treatments for OA patients focus on pain/symptom relief, however, the core persisting issue of cartilage degeneration remains unmodified. Synovial mesenchymal stem cells derived from the synovium (sMSCs), are an attractive source for articular cartilage regeneration since they demonstrate robust chondrogenic capacity. However, it is now becoming accepted that MSC populations can demonstrate significant phenotypically heterogeneity, which in turn, may affect their potential therapeutic efficacy. Furthermore, it remains unknown if this heterogeneity is natural (e.g. multiple MSC sub-populations present) or if MSCs are influenced by factors in vivo (such as the disease stage of the individual) and/or in vitro (such as culture microenvironment). Additionally, the methods used to define MSCs, such as marker profile could be influenced by varying laboratory/cell culture conditions suggesting that cell surface markers present on MSCs in vitro could be distinct from those that are present in vivo. Therefore, it may be possible that in vitro defined MSC markers may be unreliable to use as a standard to isolate MSCs directly from in vivo/in situ which would be necessary to understand and characterize MSC heterogeneity. A more complete understanding of MSC heterogeneity would allow us to identify and purify only sMSCs that have both self-renewal capacity as well as tri-lineage potential. To address this gap in knowledge we re-assessed the cell surface markers in vivo that are widely used to identify and purify MSCs in vitro. To accomplish this aim and also address questions on MSC heterogeneity, we examined the cell surface marker profiles (in vivo/in situ vs. in vitro) between clonal sMSCs isolated from OA and normal individuals. Additionally, we correlated the cell surface profile in situ with the tri-lineage potential to determine if the differentiation potential of sMSCs can be predicted based on their in situ cell surface expression profile. Methods: Synovium tissue was collected from knees of late stage OA patients (n = 15) and normal controls (n = 21). These biopsies were digested with collagenase type IV and the cell suspension was stained for: MSCs markers (CD90, CD44, CD73, CD271, CD105), a macrophage marker (CD68), and a viability marker (FVS510). An unbiased index sorted approach was utilized where 1 cell/well (per 96 well plate) was sorted based on the absence of CD68 and FVS510. Expanded clones were differentiated into adipocytes, chondrocytes and osteoblasts. Differentiation was analyzed via qPCR and histology. Each in vitro expanded clonal population was also re-examined for surface marker expression. Results: Figure 1 displays the differentiation capacity of 4 clones isolated from a single OA patient. Regardless of the in situ marker profiles expression, all sMSC clones, when cultured in vitro, converged to a unified marker profile of: CD90+, CD44+, CD73+ and CD105+. The in situ marker profile that was correlated with multipotent differentiation capacity of sMSCs was found to be similar between OA and normal clones (CD90+, CD44+, and CD73+). A second cohort of patients was recruited to validate the in situ MSC profile, however, it was found that isolation of CD90+, CD44+, and CD73+ triple positive cells didn't guarantee multipotent differentiation potential as tested by qPCR and histology analysis. Conclusions: This study has demonstrated that using in vitro defined MSC markers to select sMSCs populations directly from human synovial tissue is not specific and the markers currently used by the field are not capable of distinguishing MSCs with or without full differentiation capacity if employed in situ/in vivo. Furthermore, after reviewing in situ vs. in vitro surface marker expression data, it was found that all sMSC sub-populations present in situ (based on marker expression) converge to an indiscriminate profile after culturing. Overall, this suggests that at least some of the MSC markers employed are likely artifacts of the cell culturing process. Thus, the results of this project raise a potentially larger question regarding the identification of MSCs with these standard markers. Do these markers truly describe MSCs? This question should be thoroughly addressed before MSC therapies are continued to be delivered to patients suffering from OA.