BME PhD Dissertation Defense - Suma Jaini

12:00 pm on Thursday, August 14, 2014
44 Cummington Mall, Room 203
Title: "Methods for functional characterization of transcription factor binding sites in bacteria"

James Galagan, Associate Professor, BME (Advisor)
Ahmad Khalil, Assistant Professor, BME (Chair)
Daniel Segre, Associate Professor, Bioinformatics
Igor Kramnik, Associate Professor, Medicine

Understanding gene regulation is necessary to gain insight into and model important cellular processes including disease. Current inability to combat many diseases is partly because of incomplete understanding of gene circuitry. Regulation mechanisms of Mycobacterium tuberculosis, the causative agent of Tuberculosis are not properly understood. Transcriptional regulatory network (TRN) is a network comprising transcription factors (TF) and their targeted genes that provide a powerful framework to analyze the complete regulatory system. Chromatin immunoprecipitation followed by next generation sequencing (ChIP-Seq) is becoming the method of choice to identify genome wide TFBS. Therefore, we use ChIP-Seq on known transcription factors to re-construct the TRN of Mycobacterium tuberculosis (Mtb) and other bacteria. ChIP-Seq reveals various transcription factor binding sites (TFBS) but doesn't provide any information on the mechanism of regulation of the genes by their corresponding TF's. Techniques to gain more insight into the mechanisms include microarray, knock out studies and qPCR. But, these techniques provide a static view of network. Also, they provide information at RNA level and mask the regulation happening at protein level. Therefore, in order to understand both the mechanism of regulation at protein level as well as to capture the network dynamics, we built a synthetic gene circuit in Mycobacterium smegmatis and defined input-output relationships between key TFs and their targeted promoters. We validated this system on kstR, a TF which is a known repressor. KstR regulates genes involved in cholesterol degradation and is shown to de-repress itself and its regulon genes in the presence of cholesterol as well as in hypoxia, where there are no exogenous lipids. We explored the possibility of other by-products that may be responsible for the de-repression of kstR and its regulon. The data suggests that propionyl-coA, a by-product from degradation of cholesterol, odd numbered fatty acids as well as branched chain amino-acids is causing the de-repression of kstR and its regulon.
ChIP-Seq data on transcription factors in M.tb as well as E.coli shows that many TFBS are located immediately upstream of open reading frame start sites, consistent with our understanding of prokaryotic gene regulation. However, the data also suggests that many TFBS are located inside and also downstream of open reading frames. One of our hypotheses is that these novel TFBS might be indirect binding sites that mediate chromatin looping. Therefore, we developed a method 3C (Chromosome Conformation Capture) to understand the regulation in the third dimension by analyzing the chromosomal interactions. We optimized the protocol in E.coli and validated using a known interaction mediated by a repressor GalR. We then identified two regions, 20 kbp apart, containing TFBS of StpA, a nucleoid associated protein, which are not directly involved in gene regulation of their downstream genes. The data from a 3C experiment on an E.coli strain with inducible StpA suggests that these two regions interact by an unknown mechanism. However, the interaction was not lost when a similar experiment is done in StpA knock out strain suggesting that StpA may not be a sole TF responsible for this interaction. Lastly, we developed Hi-C method on E.coli genomic DNA to identify long range interactions in a genome wide and unbiased manner.