Cis-Regulation of Gremlin1 Expression during Mouse Limb Bud Development and its Diversification during Vertebrate Evolution

Embryonic development and organogenesis rely on tightly controlled gene expression, which is achieved by cis-regulatory modules (CRMs) interacting with distinct transcription factors (TFs) that control spatio-temporal and tissue-specific gene expression. During organogenesis, gene regulatory networks (GRNs) with selfregulatory feedback properties coordinately control growth and patterning and provide systemic robustness against genetic and/or environmental perturbations. During limb bud development, various interlinked GRNs control outgrowth and patterning along all three limb axes. A paradigm network is the epithelial-mesenchymal (e-m) SHH/GREM1/AER-FGF feedback signaling system which controls limb bud outgrowth and digit patterning. The BMP antagonist GREMLIN1 (GREM1) is central to this e-m interactions as its antagonism of BMP activity is essential to maintain both AER-Fgf and Shh expression. In turn, SHH signaling upregulates Grem1 expression, which results in establishment of a self-regulatory signaling network. One previous study provided evidence that several CRMs could regulate Grem1 expression during limb bud development. However, the cis-regulatory logics underlying the spatio-temporal regulation of the Grem1 expression dynamics remained obscure. From an evolutionary point of view, diversification of CRMs can result in diversification of gene regulation which can drive the establishment of morphological novelties and adaptions. This was evidenced by the observed differences in Grem1 expression in different species that correlates with the evolutionary plasticity of tetrapod digit patterning. Hence, a better understanding of spatio-temporal regulation of the Grem1 expression dynamics and underlying cis-regulatory logic is of interest from both adevelopmental and an evolutionary perspective.
Recently, multiple candidate CRMs have been identified that might be functionally relevant for Grem1 expression during mouse limb bud development. For my PhD project, I genetically analyzed which of these CRMs are involved in the regulation of the spatial-temporal Grem1 expression dynamics in limb buds. Therefore, we generated various single and compound CRM mutant alleles using CRISPR/Cas9. Our CRMs allelic series revealed a complex Grem1 cis-regulation among a minimum of six CRMs, where a subset of CRMs regulates Grem1 transcript levels in an additive manner. Surprisingly, phenotypic robustness depends not on threshold transcript levels but the spatial integrity of the Grem1 expression domain. In particular, interactions among five CRMs control the characteristic asymmetrical and posteriorly biased Grem1 expression in mouse limb buds. Our results provide an example of how multiple seemingly redundant limb-specific CRMs provide phenotypical robustness by cooperative/synergistic regulation of the spatial Grem1 expression dynamics.
Three CRMs are conserved along the phylogeny of extant vertebrates with paired appendages. Of those, the activities of two CRMs recapitulate the major spatiotemporal aspects of Grem1 expression in mouse limb buds. In order to study their functions in species-specific regulation of Grem1 expression and their functional diversification in tetrapods, I tested the orthologous of both CRMs from representative species using LacZ reporter assays in transgenic mice, in comparison to the endogenous Grem1 expression in limb buds of the species of origin. Surprisingly, the activities of CRM orthologues display high evolutionary plasticity, which correlates better with the Grem1 expression pattern in limb buds of the species of origin than its mouse orthologue. This differential responsiveness to the GRNs in mouse suggests that TF binding site alterations in CRMs could underlie the spatial diversification of Grem1 in limb buds during tetrapod evolution.
While the fish fin and tetrapod limb share some homologies of proximal bones, the autopod is a neomorphic feature of tetrapods. The Grem1 requirement for digit patterning and conserved expression in fin buds prompted us to assess the enhancer activity of fish CRM orthologues in transgenic mice. Surprisingly, all tested fish CRMs are active in the mouse autopod primordia providing strong evidence that Grem1 CRMs are active in fin buds and that they predate the fin-to-limb transition. Our results corroborate increasing evidence that CRMs governing autopodial gene expression have been co-opted during the emergence of tetrapod autopod.
Furthermore, as part of a collaboration with Dr. S. Jhanwar, I contributed to the study of shared and species-specific epigenomic and genomic variations during mouse and chicken limb bud development. In this analysis, Dr. S. Jhanwar identified putative enhancers that show higher chicken-specific sequence turnover rates in comparison to their mouse orthologues, which defines them as so-called chicken accelerated regions (CARs). Here, I analyzed the CAR activities in comparison to their mouse orthologues by transgenic LacZ reporter assays, which was complemented by analysis of the endogenous gene expression in limb buds of both species. This analysis indicates that diversified activity of CARs and their mouse orthologues could be linked to the differential gene expression patterns in limb buds of both species.