Mutations are the source of evolutionary novelty and phenotypic diversity. However, how genomic sequence encodes complex morphological structures and how variants of genomic sequence affect this process remains largely unsolved. This problem continues to captivate biologists, geneticists, anthropologists and clinicians as it promises to give answers to the question of how specific phenotypes and diseases arise and how they are encoded in the genome. In this project I follow the hypothesis that phenotypic diversity is mainly driven by alterations in gene regulation and thus encoded in the non-coding genome. Furthermore, I propose that extreme cases of evolutionary adaption are ideal to study fundamental aspects of gene regulation in developmental systems. How novel phenotypes/morphology emerge and how they are encoded in the genome will thus help to understand how changes in sequence determine phenotypes. Ultimately, we aim at constructing regulatory landscapes following the theme that an important demonstration of understanding something is to build it from scratch, as famously stated on Richard Feynman's blackboard.
Tetrapod limbs are an ideal model system to study the genomic origin of phenotypic diversity, as they have undergone extensive evolutionary diversification such as the hoof in horses, flippers to swim, or wings to fly. Limbs also represent one of the best studied model system for development. Furthermore, developing limbs display a sequence of well-defined temporal events, such as the formation of the apical ectodermal ridge (AER), the zone of polarizing activity (ZPA), the formation and growth of skeletal elements that are governed by a known set of developmental master genes. To link genomic changes with specific morphology, we focus on an extreme example of evolutionary adaption, the development of wings in bats (Chiroptera). With their forelimbs adapted as wings, they represent the only known example of self-powered flight in mammals. Unlike most quadrupedal animals where the differences between fore- and hindlimbs are relatively small, the bat forelimb has developed into a wing and thus differs dramatically from the hindlimb.
As shown by us previously, phenotype-associated loci can be identified in an unbiased phylogenetic approach based on the comparison of large genetic and epigenetic data sets between species. Such approaches are complemented be a detailed analysis of gene expression on a single cell basis. Single cell technologies allow to dissect gene expression within different cell types/clusters and to follow them through different stages of development. This gives us an unprecedented resolution of gene expression during limb development. The comparison of hindlimb vs. forelimb in mouse and bat reveals bat-specific genes expressed during wing development. Together with chromatin modification data to identify active regions in the genome during limb development, we are able to create a dense map of gene activities of the critical stages of limb development. Integrative data analysis reveals genes and their regulatory landscapes that are involved in bat wing development.
A second major technological breakthrough is in genomic engineering and the de novo synthesis of large (up to 150 kb) DNA fragments using yeast assembly approaches. This makes it possible to re-engineer complex rearrangements and entire regulatory landscapes. Technologies for single copy targeting and CRISPR-Cas9 genome editing allow the insertion of these synthetic DNA fragments into mouse embryonic stem cells (mESCs) to produce mice that carry bat regulatory sequences in their genome. This approach's modularity and versatility will give fundamental insights into gene regulation and how differences in sequence give rise to morphological adaptation and diversity.
In the next step, synthetic DNA production will be used to create artificial de novo regulatory landscapes with designed gene expression. For this purpose, we use AI and machine learning tools to identify the sequence grammar behind regulatory sequences that drive specific gene expression. Dissecting regulatory mechanisms will eventually help to understand how regulatory landscapes work during development and how precise gene expression is achieved. The possibility to produce long artificial DNA sequences will spark a technological revolution in the functional analysis of mammalian genomes, in particular regarding the function of non-coding DNA in human diseases, traits, and evolution. The ability to create and manipulate gene expression has far-reaching consequences, including applications in the medical field such as the design of specific gene expression in gene therapy.
