MiR-218 affects hypertrophic differentiation of human mesenchymal stromal cells during chondrogenesis via targeting RUNX2, MEF2C, and COL10A1

Abstract

Abstract Background Human mesenchymal stromal cells (MSC) hold hopes for cartilage regenerative therapy due to their chondrogenic differentiation potential. However, undesirable occurrence of calcification after ectopic transplantation, known as hypertrophic degeneration, remains the major obstacle limiting application of MSC in cartilage tissue regeneration approaches. There is growing evidence that microRNAs (miRs) play essential roles in post-transcriptional regulation of hypertrophic differentiation during chondrogenesis. Aim of the study was to identify new miR candidates involved in repression of hypertrophy-related targets. Methods The miR expression profile in human articular chondrocytes (AC) was compared to that in hypertrophic chondrocytes derived from human MSC by microarray analysis, and miR expression was validated by qPCR. Putative targets were searched by in silico analysis and validated by miR reporter assay in HEK293T, by functional assays (western blotting and ALP-activity) in transiently transfected SaOS-2 cells, and by a miR pulldown assay in human MSC. The expression profile of miR-218 was assessed by qPCR during in vitro chondrogenesis of MSC and re-differentiation of AC. MSC were transfected with miR-218 mimic, and differentiation outcome was assessed over 28 days. MiR-218 expression was quantified in healthy and osteoarthritic cartilage of patients. Results Within the top 15 miRs differentially expressed between chondral AC versus endochondral MSC differentiation, miR-218 was selected as a candidate miR predicted to target hypertrophy-related genes. MiR-218 was downregulated during chondrogenesis of MSC and showed a negative correlation to hypertrophic markers, such as COL10A1 and MEF2C. It was confirmed in SaOS-2 cells that miR-218 directly targets hypertrophy-related COL10A1, MEF2C, and RUNX2, as a gain of ectopic miR-218 mimic caused drop in MEF2C and RUNX2 protein accumulation, with attenuation of COL10A1 expression and significant concomitant reduction of ALP activity. A miR pulldown assay confirmed that miR-218 directly targets RUNX2, MEF2C in human MSC. Additionally, the gain of miR-218 in human MSC attenuated hypertrophic markers (MEF2C, RUNX2, COL10A1, ALPL), although with no boost of chondrogenic markers (GAG deposition, COL2A1) due to activation of WNT/β-catenin signaling. Moreover, no correlation between miR-218 expression and a pathologic phenotype in the cartilage of osteoarthritis (OA) patients was found. Conclusions Although miR-218 was shown to target pro-hypertrophic markers MEF2C, COL10A1, and RUNX2 in human MSC during chondrogenic differentiation, overall, it could not significantly reduce the hypertrophic phenotype or boost chondrogenesis. This could be explained by a concomitant activation of WNT/β-catenin signaling counteracting the anti-hypertrophic effects of miR-218. Therefore, to achieve a full inhibition of the endochondral pathway, a whole class of anti-hypertrophic miRs, including miR-218, needs to be taken into consideration.