Developmental Genomics
The normal morphogenesis of organs and structures requires a precise control of gene expression in time and in space during embryogenesis. The molecular processes that orchestrate these genetic programs are thus essential for ontogenesis, and their perturbations can lead to malformations or diseases. In this context, our laboratory focuses on three main lines of research that synergistically aim to characterize the normal and pathological lifespan of cis-regulatory control of developmental genes and its relevance to morphogenesis.
Example of a gene cis-regulatory activity over time during limb development. For a gene to fulfill its function, it must transition from an inactive to a transcriptionally active state, be maintained for a defined period, and be timely shut down (decommissioned). These steps, collectively referred to as a regulatory trajectory, are controlled by enhancers interacting with target promoters and ensure the formation of gene expression domains and the proper patterning of organs and structures during embryogenesis.
Mapping Time- and Cell-Specific Enhancers During Organogenesis
The non-coding genome contains a vast number of transcriptional enhancers that regulate the spatiotemporal activities of genes. However, the precise locations and specific activities of enhancers relevant to organ formation remain largely unknown. Identifying these regions is crucial for interpreting the impact of human genetic variants that disrupt normal organ development.
To map these regions in vivo, we employ a combination of single-cell multiomics and cell sorting of specific cell types at different time points during organ development (e.g., limb, trunk). By analyzing open chromatin profiles and histone modifications, we can produce genome-wide cartography of enhancer regions.
: Chondrocyte-specific enhancers mapped in the limb and trunk using a cell-sorting approach. (A) Fluorescent labeling of the chondrogenic Col2a1 locus. (B) Mapping of 2,704 high-confidence enhancers using ATAC-seq and H3K27ac in sorted chondrocytes. (C) Reporter assay of an Fgfr3 enhancer (hs2696). (D) Transcriptional effect of hs2696 enhancer deletion on Fgfr3 expression.
Integration of Early and Late Enhancer Activities in Expression Pattern Formation
Here, we investigate how distinct enhancer repertoires—early-acting and late-acting—regulate gene expression over time. In previous research (see below), we used a cell-tracking approach to monitor the expression and regulatory control of the limb patterning gene Shox2, with dmCherry marking active expression and EYFP recording cells that have previously expressed the gene. Through this approach, we demonstrated that early- and late-acting enhancers function in a time-constrained manner to initiate and sustain the activity Shox2, respectively.
Our ongoing research explores how early- and late-acting enhancer activities are integrated to shape expression patterns. Specifically, we aim to determine whether late enhancers can function independently of early enhancers and what additional transcriptional specificity they contribute to gene regulation over time.
: (A) Dual fluorophore system tracking Shox2 activity: dmCherry reflects real-time Shox2 transcription, while EYFP records cells that have previously expressed Shox2. (B) The Δearly Shox2 allele, specifically deleting 8 out of 26 early-acting enhancers. (C) Images of control (trac) and Δearly forelimbs at E11.5 and E14.5. Note the reduced dmCherry/EYFP signal at E11.5 and the loss of EYFP signal in distal limbs at E14.5 (white arrows). (D) Shox2 expression levels (RNA-seq) in E10.5 and E14.5 limbs (NS = not significant, ** = p-adj < 0.01).
Non-Coding Variations in Gene Regulation and Congenital Malformations
Here, we model how non-coding variations can drive gene misexpression and congenital malformations. Building on our showing that the ectopic expression of Pitx1 driven by its own enhancer Pen leads to forelimb Pitx1 misexpression and skeletal malformations, we demonstrated that the variable repositioning of the Pen enhancer influences the proportion of ectopically expressing cells, correlating with phenotypic manifestations (see below).
We are currently investigating the impact of structural variations on enhancer-promoter (E-P) connectivity, with a strong focus on the 3D genome. In particular, we aim to understand how these structural changes influence transcription in terms of quantity, spatial distribution, and timing.
: (A). C-HiC analysis of the Pitx1 locus in wildtype, Inv1 (113kb inversion), and Inv2 (204kb inversion). Arrows pinpoint the contacts between Pitx1 and Pen in mESCs in each configuration. (B): Histogram and quantification of EGFP mis-expressing cells in forelimbs. Grey and green areas indicate EGFP- and EGFP+ gating; the red dotted line marks fluorescence limits. (C): Alizarin red/alcian blue staining of E18.5 forelimbs. Skeletal defects are indicated by arrows.