Spatial and Disparate Regulation of Gdf5-Associated Regulatory Region by Fox, Sox and Lmx1B Transcription Factors

Abstract Body : Osteoarthritis (OA), is a degenerative joint condition characterized by inflammation, cartilage erosion and bone spurs. Functional disruption of growth differentiation factor 5 (Gdf5), a protein that regulates bone and synovial joint development, can lead to joint abnormalities including increased susceptivity to OA. Hence, an understanding of how Gdf5 is regulated during joint development will enable a better understanding of OA susceptibility as well as provide therapeutic targets. We previously identified a Gdf5-associated regulatory region (GARR) that contains predicted binding sites to Lmx1b, Osr2, Fox, and the Sox transcription factors, all of which have previously been shown to be involved in cartilage, bone and joint development. We hypothesized that Fox and Sox are necessary for GARR enhancer activity whereas Lmx1b and Osr2 will change the pattern of the enhancer activity.

To examine this hypothesis, we compared the pattern of transcription factor expression to Gdf5 expression using in-situ hybridization. We also analyzed previously published single cell RNA sequencing (scRNA-Seq) data to determine how these factors are related to Gdf5. We performed site-directed mutagenesis to factor-specific binding sites in a GARR-reporter construct and tested them both in-vivo (in developing chicken limbs) and in-vitro (in micromass cultures). Finally, since Sox and Lmx1b have been shown to be involved in OA and Gdf5 regulation, we co-expressed them in micromass cultures with GARR-reporter constructs to determine how they affect activity.
In-situ hybridization and scRNA-Seq data showed overlap in Gdf5 and the transcription factors of interest. Additionally, with the exception of Lmx1b, all other transcription factors were significantly increased in Gdf5 expressing cells; we suspect that this correlation is due to factor-specific upregulation of Gdf5 through GARR.

Mutation of Lmx1b binding sites in GARR caused a reduction in enhancer activity both in vitro and in-vivo (in the dorsal limb). Overexpression of Lmx1b caused a significant increase in GARR activity which was blocked with disruption of Lmx1b binding sites. This suggests that Lmx1b binds to GARR at these sites to upregulate Gdf5 expression. Mutation of Osr2 binding sites caused no change in activity, suggesting that Osr2 regulation of Gdf5 requires additional factors.

Mutation of two binding sites common to Sox and Fox transcription factors caused a significant decrease in enhancer activity, whereas mutation of two Sox only binding sites caused a significant increase in activity. This result indicates that while Fox and Sox transcription factors may positively regulate Gdf5 through GARR, Sox transcription factors may also have a more nuanced role in regulating Gdf5 expression through GARR. In line with this, overexpression of Sox transcription factors caused a decrease in enhancer activity.

Together, these results indicate that Sox and Fox transcription factors exhibit both positive and negative regulatory effects on Gdf5 expression through GARR, whereas Lmx1b activates GARR to upregulate Gdf5 expression. It is still unclear if Sox inhibition of GARR activity occurs though the Sox only sites, or if both Fox and Sox transcription factors can upregulate Gdf5 through shared binding sites. Future studies will help to resolve these possibilities.