UC Berkeley

The same two meters of DNA is carefully packed into the nucleus of nearly every cell in a human’s body, where it encodes essentially all of the complex information required to build a complete human being. However, DNA by itself cannot give rise to life; it must be decoded and maintained by specialized macromolecules, including proteins that read, regulate, replicate, recombine, and repair DNA. Mapping where and how these life-giving proteins interact with DNA can provide key insights into how they function or malfunction in healthy and diseased cells.

High-throughput DNA sequencing technologies form the basis of several powerful methods for mapping protein-DNA interactions across the genome, but they often require researchers to blend together many thousands or millions of cells to provide enough material to make an accurate measurement. Due to this blending, these bulk methods cannot capture the dynamic and heterogeneous nature of protein-DNA interactions as they regulate the genome in individual cells. While newer methods are beginning to enable protein-DNA mapping in single cells, they are incompatible with high-resolution microscopy, which can provide rich orthogonal information about nuclear organization and other complex phenotypes in single cells. Furthermore, existing protein-DNA mapping approaches fail almost completely within highly repetitive DNA sequences, which constitute roughly 5-10% of the human genome and play indispensable roles in maintaining genome stability.

In this body of work, I have developed two new technologies to address each of these limitations in turn. Firstly, I designed an integrated microfluidic platform (μDamID) that combines high-resolution imaging and sequencing information in the same single cells, allowing for the joint analysis of the nuclear localization, sequence identity, and variability of protein-DNA interactions in single cells. Secondly, I worked collaboratively to develop DiMeLo-seq (Directed Methylation with Long-read sequencing), which uses cutting-edge DNA sequencing technologies to map protein-DNA interactions on long, single molecules of DNA that retain endogenous DNA methylation marks and can be mapped to highly repetitive regions of the genome. Together, these new methods expand the toolkit available to researchers to study the fundamental processes that regulate the genome, with the potential to enhance our understanding of embryo development, stem cell differentiation, and diseases resulting from genome misregulation.