My laboratory studies the structure-function relationships of the B-ZIP class of sequence-specific DNA binding dimeric proteins. Over 50 B-ZIP genes have been identified in the mammalian genome. In the most general terms, B-ZIP proteins both activate and repress gene expression in response to physiological changes, growth factors (FOS), stress (ATF2), neuronal signaling (CREB), or metabolic changes (CEBP). Previously, we developed structural rules that determine B-ZIP leucine zipper dimerization stability and specificity. Currently, the focus of the laboratory is to understand several aspects of TF:DNA interactions, in particular, the effects of cytosine methylation on DNA binding. The effects of methylation on sequence-specific binding of TFs has far reaching implications on our understanding of gene regulation, genome function, and disease. Methylation patterns are dynamic, can vary between different cell types, young and old cells, and healthy and diseased cells. We have been examining how methylation of cytosine in CG dinucleotides changes the DNA binding specificity of TFs. Previously, we found that C/EBP family members preferentially bind methylated consensus sequences which are critical for cellular differentiation. Recently, we modified a protein binding microarray (PBM) technology which allows for determination of the binding specificity of TFs to thousands of DNA sequences containing methylated CG dinucleotides. Using this method, we showed that the B-ZIP domain ATF4'C/EBPB heterodimer preferentially bound methylated CGATGCAA both in vitro and in vivo (Mann et al. 2013, Genome Research). We extended this technology in collaboration with Jussi Taipale (Karolinska Institute, Sweden) to comprehensively and systematically examine the effect of CG dinucleotide methylation on the DNA binding specificity of a wide range of TFs and identified many that preferentially bind methylated sequences (Yin et al. 2017, Science). The topic of sequence-specific DNA binding of TFs has become more complex with two recent findings: (1) 5-methylcytosine (5mC) is oxidized to 5-hydroxymethylcytosine (5hmC), 5-formyl cytosine (5fC), and 5-carboxy cytosine (5caC) by the ten-eleven-translocation (TET) family of dioxygenases; and (2) 5mC can occur outside of CG dinucleotides. We have since modified the PBM protocol to generate double stranded DNA sequences with 5mC or 5hmC on one strand, mimicking what occurs in several cell types. We are observing dramatic changes in sequence specific DNA binding with these modifications for TFs from the basic helix-loop helix (Syed et al., 2016, Integrative Biology) and B-ZIP families (Syed et al., 2016, Biochemistry; Tillo et al., 2017 Biochemistry). The relationship between individual amino acids of a DNA binding domain and binding particular nucleotides including modified cytosines is not well characterized for B-ZIP TFs. We have examined several mutations of in the DNA binding domains of Zta, an BZIP protein encoded in the Epstein Barr Virus that binds many methylated DNA sequences (Ray et al., BBRC, 2018), and find how alterations of single amino acids can change binding to methylated DNA. Currently we are examining additional mutants and performing molecular modelling techniques to further understand the physical basis for our observed binding data.
Showing the most recent 10 out of 37 publications
