The Gremlin1 cis-regulatory landscape: a paradigm to study enhancer cooperation in regulation of transcription dynamics

The transcriptional regulation of developmental genes expression patterns in time, space and levels, is governed by cis-regulatory modules (CRMs). The activity of CRMs is controlled by transcription factor complexes that act as downstream mediators of signaling inputs. CRMs are associated with their target genes in chromatin domains with enhanced contact frequency, the so-called topologically associating domains (TADs). The incoming signaling cues are integrated into specific transcriptional outputs, which orchestrate development and differentiation. Limb bud development is one of the main molecular and cellular paradigms to study the roles of gene expression regulation during embryonic development. Limbs are external organs, easily accessible, largely dispensable for embryonic and postnatal survival and have adapted to numerous specific functions during vertebrate evolution, resulting in the high level of morphological diversity among vertebrates.
The molecular pathways and morphogenetic events that govern limb patterning are largely conserved, reflecting their crucial roles in gene regulation during limb development. Our group previously identified and functionally analyzed the SHH/GREM1/AER-FGF epithelial-mesenchymal (e-m) self-regulatory signaling system that controls early limb bud outgrowth and patterning. The BMP antagonist Gremlin1 (Grem1) is one of the functionally most essential nodes in this system. Its spatio-temporal expression is regulated by the converging trans-acting inputs of the major limb bud signaling pathways. These inputs are integrated into the dynamic regulation of Grem1 expression by its 310 kb cis-regulatory landscape.
For my Ph.D. research, I used the mouse Grem1 cis-regulatory landscape as a paradigm to study gene transcriptional regulation in the context of embryonic limb bud development. I identified and genetically analyzed the functionally relevant Grem1-associated CRMs. To this end, I initially used reporter assays in transgenic mouse embryos to assess their potential enhancer activity. CRMs with established enhancer activities were then functionally studied by generating CRISPR/Cas9-engineered loss of enhancer function mutant mice. This, in combination with molecular analysis, was used to assess their role(s) in the Grem1 transcriptional regulation. In addition, I used 4C-seq assays to study the physical interactions among CRMs and the Grem1 promoter, in wild-type and mutant mouse limb buds. I also addressed the question of the downstream consequences of enhancer deletions on limb bud development by tracking apoptosis and quantifying limb buds’ cellular proliferation.
My studies revealed that the enhancer redundancy and diversity that regulates the Grem1 expression dynamics during mouse limb bud development was much more complex than the one-to-one correlation often described by others. None of the CRMs characterized was essential on its own for limb development. The transcriptional activities of different CRMs were additive in levels and partially redundant in regulating the spatial and temporal dynamics of the Grem1 expression. The spatio-temporal changes in Grem1 expression levels, caused by the loss of different enhancers alone, were not sufficient to explain the observed phenotypes. Therefore, additional mouse strains lacking several CRMs were generated and analyzed. In light of these results, I performed a comparative molecular analysis of key genes in the self-regulatory SHH/GREM1/AER-FGF signaling system, which provided a better molecular understanding of how these cis-regulatory alterations affect the limb bud outgrowth and patterning. This analysis showed that the cis-regulatory alterations affecting levels and spatio-temporal kinetics of the Grem1 expression are accompanied by specific changes in the self-regulatory feedback loops in mutant limb buds. In addition, I investigated potential effects on the structure of the Grem1 TAD and revealed that alterations in the interactions among CRMs and the Grem1 promoter contributed to the transcriptional regulation of Grem1 expression.
This extensive genetic analysis led to the following major conclusion: the control of transcript levels by the Grem1-associated CRMs is additive, while they function in a cooperative manner to regulate the spatial dynamics of the Grem1 expression in mouse limb buds. In particular, deleting several of the CRMs that regulate spatial aspects of the Grem1 expression disrupts this cooperativity. This, in turn, weakens the robustness of the limb patterning system and results in the loss of pentadactyly. It appears that the observed limb skeletal deformity phenotypes strongly correlate with reduced cell proliferation. Structural analyses reveal that intra-TAD rearrangements play a major role in the robustness of the Grem1 expression.