Rowley Lab
Looping, insulation, and transvection are three features of chromatin organization of interest to my lab.
Chromatin is deliberately arranged in the 3D nucleus so that genomic loci can influence each other long-range. The organization of chromatin is altered during development, in response to hormonal signaling, and after heat shock, indicating that the nucleus is a dynamic environment. Abnormal 3D chromatin organization is implicated in diseases such as aberrant limb formation, cancer, and embryonic lethality. My overall goal is to discover mechanisms driving the establishment and dynamics of 3D chromatin organization. Specifically, I aim to elucidate how looping, insulation, and transvection occur.
Example of a Hi-C map in mammals. CTCF loops appear as punctate spots, while compartments are displayed as a checkerboard of interactions. Image from Rowley et al., Nat Rev Genet 2018.
CTCF loops are a prominent feature of mammalian chromatin organization and are thought to alter enhancer – promoter interactions. The current model is that loops are formed via cohesion-mediated extrusion that is blocked by CTCF. How mammalian CTCF is able to block extrusion is unknown. Interestingly, Hi-C maps in Drosophila melanogaster do not display CTCF loops despite nearly identical DNA binding of the CTCF protein. Understanding why these differences occur will be a major focus of the lab.
Profiles of hd-pairing signal at loci with different numbers of architectural proteins. Image from Rowley et al., Cell Rep 2019.
Transvection occurs via pairing of homologous chromosomes enabling enhancers on one chromosome to affect the expression of genes on the homologue. In Drosophila, homologous pairing of chromosomes occurs throughout interphase and is an important part of nuclear organization. Transvection is broadly affected by the presence or absence of architectural proteins, although it is unknown how. Pairing has been typically examined using low resolution FISH, but during my post-doctoral work I developed a method to examine homologous pairing at high resolution (~250 bp) using Hi-C data. I found that pairing occurs in “buttons” only a few kb in size, called hd-pairing sites. While condensin II was known to broadly inhibit pairing, I discovered that condensin II binds to hd-pairing sites partially unpairing them. By genome-wide analysis of hd-pairing sites, I also found that clustering of architectural proteins correlates with homologue pairing. My lab is interested in how these proteins affect pairing and/or transvection.
Actual Hi-C (top left) compared to a simulation created using only Gro-seq data (bottom left). Circles indicate examples of discrepancies. Image from Rowley et al., Mol Cell 2017.
3. Predicting 3D chromatin organization.
As we gain an understanding of 3D chromatin organization, my lab will use that data to improve algorithms to predict Hi-C maps from other data types. This will be useful to form new hypotheses on regions that fail prediction in wild-type cells, but also could be used to predict changes in the 3D chromatin landscape in response to various treatments.
Principal Investigator
Associate Professor, Department of Genetics, Cell Biology, and Anatomy
Director, Bioinformatics & Systems Biology PhD Program
