Chromosomal DNA is packaged into protein/DNA structures called chromatin. Certain chromatin structures silence genes, whereas others facilitate gene expression. Thus, chromatin is a remarkably dynamic material; structures formed to prevent gene expression can be remodeled to enable expression in response to cellular signals during growth and development. As many genes important for cell proliferation and differentiation are regulated at the chromatin level, it is important to understand how packaging by chromatin influences transcription, to characterize the protein machines that facilitate the transitions between chromatin states, and to understand how these machines are regulated. These are the goals of my laboratory.
The basic unit of chromatin structure is the nucleosome, which consists of a "bead" composed of eight histone proteins, around which 147 base pairs of DNA are wrapped. Nucleosomes display three dynamic properties in vivo: compositional changes, covalent modifications by enzymes, and repositioning (Figure 1). In regard to composition, nucleosomes can be reconstructed from canonical histones or from special histone variants. These variant nucleosomes assist with biological processes such as centromere construction or gene activation. Between nucleosomes is a small segment of free DNA (about 40 bases), leading to a view of the chromosome that resembles beads on a string. The wrapping of DNA around nucleosomes blocks the access of other protein factors to this DNA, and this can silence gene expression and other chromosomal processes. Another mode of gene silencing involves the methylation of the DNA itself—a covalent modification that recruits gene-silencing factors. Thus, DNA wrapping and modification work together to promote gene silencing and must be reversed for gene activation.
To enable access to the underlying DNA, nucleosomes must be mobilized, restructured, or ejected. As nucleosomes themselves have limited mobility, their dynamic movement requires the action of nucleosome-remodeling complexes (remodelers). Remodeler complexes restructure, reposition, and eject nucleosomes, thus providing access to the DNA. In addition, gene activation requires the removal of DNA methylation marks, which may involve the recruitment of a DNA demethylation system.
We are addressing several central questions in chromatin biology. First, we want to understand how remodeler complexes move and eject nucleosomes, and how they know which nucleosome to move. Second, we want to understand how genes are selected to receive DNA methylation (causing silencing) and how the methylation process can be reversed by demethylation to promote activation. We are especially interested in how this process is regulated in germ cells and during embryo development, where genes turn on and off to change cell fate. Third, we study a particular class of genes—those transcribed by RNA polymerase III (Pol III)—to understand how gene expression is controlled both by chromatin and by cellular and environmental conditions. Fourth, we are using advanced array and DNA-sequencing technologies to develop methods to analyze the relationship between chromatin structure and gene expression genome-wide.
To understand the mechanism that remodelers use to move or eject nucleosomes, we purify remodeler complexes and examine their action on purified nucleosomes in vitro. The primary remodeler that we study is RSC (remodels the structure of chromatin), which consists of 15 proteins in one large complex. We have shown that the remodeler "engine" is an ATPase subunit that functions as a DNA translocase that pumps DNA around the surface of the nucleosome in the form of DNA waves (Figure 2), resulting in the movement of the histone octamer relative to the DNA. We have collaborated with Eva Nogales (HHMI, University of California, Berkeley) to determine the structure of the entire RSC complex at moderate resolution, which revealed a large flexible protein machine that contains a large pocket of nucleosome dimensions. We have collaborated with Carlos Bustamante (HHMI, UC Berkeley) to determine the speed and force of DNA translocation on individual RSC-nucleosome complexes. Kaede Hinata and Cedric Clapier in my lab are studying how the ATPase subunit is regulated. Hinata has shown that two actin-related proteins (ARPs), Arp7 and Arp9, bind directly to the ATPase subunit of the RSC remodeler and regulate the activity of the nucleosome-remodeling reaction. Clapier has developed new assays to monitor DNA translocation in a single-molecule format; these assays are providing important details on how the translocase domain pumps DNA and how ARPs impact that mechanism.
A central question is how remodelers select particular nucleosomes to remodel. Margaret Kasten is using a set of protein motifs termed bromodomains to study Rsc4, an RSC subunit that binds to the histone H3 tail when it is covalently modified (acetylated). Our collaboration with Christopher Hill's lab (University of Utah) has yielded the crystal structure of the Rsc4 bromodomains bound to the H3 tail. Because RSC contains six additional bromodomains, it may serve as an important model complex for understanding how a remodeler can "read" combinations of acetylation modifications on histone tails, thereby selecting which nucleosome to remodel. Timothy Parnell is now determining whether RSC ejects nucleosomes bearing the special histone variant Htz1 to expose gene promoters and activate transcription.
Another key question is how signaling pathways communicate environmental status to the transcription machinery to determine whether a gene should be transcribed. To address this, we are analyzing the genes transcribed by RNA Pol III. We conducted whole-genome occupancy analyses to define all the genes in the yeast (Saccharomyces cerevisiae) genome that are transcribed by Pol III (the Pol III transcriptome) and to understand their regulation. We also showed that RSC occupies virtually all genes transcribed by RNA Pol III, and Parnell's recent work has shown that RSC removes nucleosomes from Pol III genes, allowing Pol III to access these genes. Recently, we determined the entire RNA Pol III transcriptome in human cells and have revealed the close collaboration between RNA Pol II (which transcribes mRNAs) enhancer-binding factors and Pol III: enhancer-binding factors open up the chromatin to allow Pol III access to Pol III genes, which reside in these enhancers. Furthermore, we have characterized the key regulator of Pol III transcription, Maf1. Our work has revealed the basic strategy that allows Maf1 to repress Pol III: starvation and stress lead to Maf1 dephosphorylation, its migration to the nucleus, and its attachment to (and inhibition of) Pol III. Andrew Oler has recently identified the phosphatase that is primarily responsible for Maf1 dephosphorylation, and he is now determining how it is regulated. We also have new evidence for a second mode of generating dephosphorylated Maf1, involving changes in translational efficiency.
An emerging area of interest in our laboratory is how chromatin modifications help guide early development. We are interested in how chromatin and chromatin modifiers help germ cells (sperm and eggs) and early embryos regulate genes important for controlling cell fate. Kunal Rai, Itrat Jafri, Magdalena Potok, and Shan-Fu Wu are using the zebrafish model system to study the dynamic chromatin structure of germ cells, zygotes, and early embryos. A major new effort is the characterization of DNA methylation and demethylation enzyme systems in zebrafish (with clear human homologs), which may be utilized at many points in the germline and embryo to alter transcriptional competency. One such mechanism involves the use of DNA repair enzymes and cytosine deaminases to remove methylated cytosine and to replace it with unmethylated cytosine. We are currently setting up the Center for Zebrafish Epigenetics and Chromatin (CZECH), which will be a resource for many labs interested in viewing, storing, analyzing, and interpreting genomics data from zebrafish. Furthermore, Sue Hammoud (in collaboration with Douglas Carrell, University of Utah) is studying the chromatin status of human sperm to better understand its contributions to embryo development and human fertility. Remarkably, we have found that genes of developmental importance are packaged in a distinctive manner in the germline (sperm), a packaging that is likely to promote their proper regulation in the embryo. We are investigating how these structures are established, maintained, and altered in the early embryo, and we are using the zebrafish to better understand this system.
Finally, we have developed new methods for genomics analysis, such as strand-specific transcriptome determinations (an effort led by Natalie Dutrow), to understand chromatin-transcription relationships and to discover more about the functions of noncoding RNAs in chromatin and transcriptional regulation. We are now characterizing 90 noncoding RNAs that we discovered in the model organism Schizosaccharomyces pombe.
This work is supported in part by grants from the National Institutes of Health.
As of May 30, 2012