gene logic

Gentle Reader, I presume thou wilt be very inquisitive to know what antic or personate actor this is, that so insolently intrudes upon this common theatre to the world’s view…

- Robert Burton

Welcome to Gene Logic. Your host is Michael A. White, PhD, a biochemist/systems biologist/geneticist/genomicist, currently working as a postdoctoral researcher in the Department of Genetics and the Center for Genome Sciences & Systems Biology at the Washington University School of Medicine in St. Louis.

I look at life from a gene’s eye view. Imagine that you are a stretch of DNA, ensconced in the nucleus of the cell.  The entire external world, both inside of the cell and outside, is represented to you almost exclusively as quantities and activity states of nuclear proteins (and regulatory RNA). From that information alone, you are able to read the state of the world and properly compute the correct transcriptional response.  Amazing.

Read about my research and publications below. My CV is here. Grab your favorite brew and join me for conversation at The Finch and Pea.  See what I’m reading over at the stack.  Follow my latest findings on science and whatever else. And see what we’re doing in the Cohen lab. Contact me via Gmail:  vheissu1974@gmail.com

My research interests in gene logic and cis-regulation

Most important cis-regulatory decisions are governed by the combined action of activators and repressors, but we rarely understand how specific combinations of promoter-bound activators and repressors generate particular transcriptional patterns. In nearly all cases we lack quantitative, systems-level models that explain how patterns of gene expression arise as a consequence of the often non-intuitive interplay between activation and repression. My long-term research goal is to understand how the combined actions of activators and repressors control the spatial and temporal patterns of transcription that are fundamental to nearly every normal and pathological process in biology.

Quantitative, systems-level models are required to fully understand the interplay between transcriptional activators and repressors. These models synthesize experimental results in a formal framework, which is used to design and interpret experiments that reveal how transcriptional patterns are a logical consequence of specific combinations of bound activators and repressors. Working models of the combined action of activators and repressors would accurately predict expression patterns from the sequence of regulatory DNA and would provide a tool for engineering synthetic transcriptional circuits with useful properties. Currently, we cannot engineer transcriptional circuits much beyond what can be achieved by trial and error, and we have almost no power to predict gene expression patterns from regulatory sequence. I am taking an approach that integrates mathematical modeling and quantitative experiments to understand the regulatory logic underlying the early G1-phase transcriptional timer of the S. cerevisiae cell cycle, and the transcriptional decoding of spatial information in the Hedgehog morphogen gradients of developing D. melanogaster. In each case, my goal is to understand how the interplay of both activators and repressors results in the precise control of gene expression.

Understanding the G1-phase transcriptional timer in yeast: A critical biological question is how transcriptional timers control circadian rhythms, key developmental switches, and cell cycle transitions. The most tightly regulated transcriptional timer in the yeast cell cycle controls the set of genes encoding mother-daughter cell separation enzymes that are expressed in a strong pulse in early G1 phase. There is no predictive, mechanistic model that explains how this strong timer works. There are two key reasons for developing a model of this transcriptional timer. First, yeast is experimentally tractable, thus I can dissect the workings of this timer in detail, measure key biochemical parameters, and test predictions produced by my computational model. Second, the regulatory logic of specific transcriptional circuits is often repeated in highly divergent organisms, even when the protein factors involved are not homologous. The regulatory logic controlling the early G1 timer in yeast likely reflects a general strategy that applies to transcriptional timers in other organisms.

I discovered a major component of the early G1 timer. During a genetic screen for new yeast cell cycle TFs, I found that Sfg1 represses early G1-phase genes that are targets of the transcriptional activator Ace2. I showed that Sfg1 affects cell cycle progression through G1 using flow cytometry to assay an sfg1 knockout strain. Using qRT-PCR to measure transcription of Sfg1 target genes during early G1 phase in mutant strains, I showed that the timing of early G1 gene expression is altered by deleting Sfg1, or by abolishing its conserved Cdk phosphorylation sites. My results suggest that the timing of early G1 transcription is regulated by the interplay between an activator, Ace2, and a repressor, Sfg1, both of which are Cdk targets. There is currently no mechanistic model that can explain how the activities of Cdk, Ace2, and Sfg1 produce the strongest transcriptional timer of the yeast cell cycle. 

By proceeding through cycles of model refinement and experimental  testing, I hope to understand how the combined effects of repression, activation, and Cdk activity produce the early G1 transcriptional timer.

Understanding the transcriptional decoding of spatial information in Hedgehog morphogen gradients: The combinatorial action of transcriptional activators and repressors is critical during development, when morphogen gradients create spatial patterns of gene expression. Positional information in such morphogen gradients is converted into a transcriptional pattern through the interactions between signal-responsive TFs and the cis-regulatory elements of target genes. In collaboration with Dr. Scott Barolo (U. Michigan), I built a computational model to explore the cis-regulatory logic underlying the response to Hedgehog (Hh) morphogen gradients in Drosophila. 

Dr. Barolo’s lab discovered that, contrary to the expectations of the prevailing model of Hh gradient response, many major Hh target genes in Drosophila and mammals are regulated by low-affinity TF binding sites, including dpp, a crucial Hh target gene expressed in the developing Drosophila wing.  Hh target genes are regulated by a spatial gradient of a transcriptional activator, which is opposed by a reciprocal gradient of a transcriptional repressor. Thus, the key to explaining the puzzling observation that low-affinity sites are required for the proper dpp response to Hh is to understand how the interplay of activators and repressors determines dpp expression. To approach this problem I trained a statistical thermodynamic model of activator and repressor binding to the dpp enhancer using quantitative, in vivo dpp expression data collected by the Barolo lab. We discovered a regulatory mechanism based on activator-repressor competition and cooperative repression that explained the role of low-affinity sites in the dpp response to Hh. I used this cooperative repression model to predict the transcriptional pattern expected when the number of TF binding sites in the dpp enhancer was reduced from three to one. My model made the counter-intuitive prediction that the loss of two binding sites from the dpp enhancer would cause an increase in dpp activation. The Barolo lab conducted the experiment, and the result quantitatively matched the model prediction. Our work shows that low-affinity binding sites are necessary to achieve the correct balance between cooperative repression and non-cooperative activation. This result can explain the widespread presence of evolutionarily conserved low-affinity (non-consensus) sites in Hh-responsive genes, and possibly represents a more general regulatory strategy used by morphogens that act through competing activators and repressors.

With a successful model of Hh response in hand, we are now testing the generality of our cooperative repression model by applying it to other contexts. The results of this work will allow us to determine how generally the cooperative repression mechanism applies to Hh responses outside of the wing disc.

The unifying goal of my two interests described above is to develop a quantitative, biophysical understanding of how combinations of transcriptional activators and repressors, binding to particular arrangements of cis-regulatory sequences, generate the spatial and temporal patterns of transcription that lie at the core of important biological processes. In both projects I follow a similar strategy: I use quantitative experiments to build a computational model, and I use the computational model to devise new experiments that test the model and lead to deeper insight into logic of cis-regulation. This integrated approach that seamlessly combines mathematical modeling with quantitative experiments will be necessary to make sense of the extensive data on cis-regulation emerging from genome projects, to understand the role of programs of gene expression in normal development and disease, and to design synthetic transcriptional circuits that can produce new and useful transcriptional patterns in engineered organisms.

Recent publications

** Co-first authorship

Parker, DS**, White, MA**, Ramos AI, Cohen BA, Barolo S. The cis-regulatory logic of Hedgehog gradient responses: key roles for Gli binding affinity, competition, and cooperativity. Science Signaling 4, ra38 (2011). Link

See accompanying Perspective: Whitington T, Jolma A, Taipale J.  Beyond the balance of activator and repressor. Science Signaling 4, pe29 (2011). Link

White MA, Riles L, Cohen BA. A systematic screen for transcriptional regulators of the yeast cell cycle. Genetics 181, 435-446 (2009). PMC2644938  Link