Issue Date: July 17, 2006
Pulling Genes' Strings
In an expanding and accelerating connect-the-dots research adventure, scientists of many ilks are converging on a uniquely powerful framework for understanding and manipulating genes. As those in the midst of this research movement see it, this emerging framework, known as epigenetics, is on its way to subsuming even the pharaonic, multi-billion-dollar achievement of sequencing the entire human genome. In the making, say epigenetics researchers, is a next-generation molecular biological schema for asking and answering many of the most vexing questions about life, health, and disease.
How does a single fertilized egg cell end up proliferating and differentiating into the more than 200 cell types that make up a multi-trillion-celled organism like a human being? What molecular players control the pattern of gene activity over the course of a cell's lifetime? If only 3% of the DNA in a person's genome embodies protein-encoding genes, what's the rest of the DNA doing? How do exposures to nutrients, toxins, pollutants, and other environmental agents affect gene expression for better or for worse?
For the growing epigenetics research crowd, answers to such questions are to be found along the many roads that lead to chromatin—the protein-and-DNA stuff of chromosomes. "DNA carries the information that defines who you are," says Allan C. Spradling of the Carnegie Institution of Washington's embryology department in Baltimore. "But the use of that information, as you grow in utero and become a human being or any kind of animal with more than one cell, relies on chromatin."
Dubbed chromatin in 1882 by the German anatomist Walther Flemming, who was among the first to detail the process of cell division, chromatin now is recognized as the exquisitely contorted macromolecular stage on which epigenetic dramas unfold. Famously able to package several feet of DNA into a fantastically crumpled yet functional architecture that fits into a nucleus only micrometers in diameter, chromatin acts as a kind of intracellular microprocessor for gene control.
Instead of electronic signals, chromatin's inputs are an ever-changing wash of enzymes, proteins, RNA, nutrients, toxins, and a growing roster of small and large chemical modifications. These interact with and alter chromatin's gene-controlling microscape in both local and more global ways. Such remodeling controls the relentless activation and silencing of the human genome's estimated 25,000 genes. It is through this epigenetic orchestration that genes—those relatively rare stretches of chromatin's DNA that code for proteins—assemble from the random biological notes they otherwise would be into a lifelong score.
The word epigenetics encompasses all of the layers of genetic control in cells that do not entail changes in DNA sequences. It was first coined by the British embryologist Conrad H. Waddington in the early 1940s, when he defined it as "the interactions of genes with their environment that bring the phenotype into being." Trying to uncover specifics about these interactions, and how they play out in the environment of chromatin, has kept researchers busy ever since, now more so than ever.
They've been finding, for one, that a lot of epigenetic activity involves the most subtle of chromatin modifications. One of the most prevalent modifications is the addition of a single methyl group at those cytosine nucleotides along chromatin's DNA that are right next to a guanine, so-called CpG pairs. Lots of such DNA methylations along a stretch of chromatin usually mean that any genetic sequences in the same stretch are marked as "do not transcribe."
Another subtle but consequential chemical change is the covalent addition of acetyl groups to chromatin's millions of spoollike protein complexes known as nucleosomes. Made of an octet of proteins called histones, nucleosomes amount to DNA-winding-and-binding spools strung all along chromatin like millions of pearls on a vast DNA necklace. Genomic DNA that's wound tightly around the nucleosomes cannot be transcribed into messenger RNA nor translated into protein molecules. With enough histone acetylation, however, the DNA loosens from the nucleosomes, thereby becoming accessible to the enzymatic machinery that starts the gene-to-protein process.
"There is this whole space of chromatin, where it is a protein-DNA complex that's at work," says Richard A. Young of the Whitehead Institute for Biomedical Research in Cambridge, Mass. "It is not just the genetic code that's important because that code is coddled in a blanket???chromatin???and that blanket has all kinds of chemical decorations. And all of it has biological meaning and consequences."
Histones, for one, not only can be acetylated, they also can be phosphorylated, methylated, and otherwise modified with small-molecule and larger-sized chemical marks whose gene-controlling consequences researchers now are figuring out how to interpret.
"This is a paradigm with the potential to reshape all of biology," says pediatrics professor Robert Waterland of Baylor College of Medicine, Houston. One way Waterland has been making his mark in epigenetics research is by studying how methyl-donating nutrients-among them folate and vitamin B-12-in the diets of female mice during pregnancy lead to changes in chromatin methylation and gene expression. In one influential study, Waterland and colleague Randy Jirtle of Duke University found that the extra methylation in the chromatin of nutrient-fed pregnant mice silenced the so-called agouti gene. As a consequence, instead of giving birth to fat, yellow pups-the normal results-these mother mice delivered thin, brown pups. Says Waterland: "There are subsets of genes in the genome where subtle differences in nutrition during embryonic development can affect patterns of methylation, and this can get set for life."
One of the most tantalizing signs that epigenetics now is rocketing into prominence in molecular biology and medicine is the Food & Drug Administration's recent approval of the first two drugs that act by eliciting epigenetic changes in chromatin. Both drugs, which work by removing methyl groups from the DNA in chromatin, were developed for treating a family of blood disorders, myelodysplastic syndromes, which leave patients anemic, fatigued, weak, and sometimes at risk for developing leukemia. One of the drugs, 5-azacitidine (Vidaza), has been marketed by its Boulder, Colo.-based developer, Pharmion Inc., since the drug's approval in 2004. Bloomington, Minn.-based MGI Pharma's structurally related drug, decitabine (Dacogen), was approved just two months ago.
A third chromatin-interacting drug, valproic acid, already is sold under several brand names. One of its effects in cells is to inhibit an enzyme that adds acetyl groups to histone proteins. This trait has led to a number of clinical trials now under way to test the safety and effectiveness of valproic acid for treating HIV and for increasing the effectiveness of radiation therapy for certain brain tumors.
"Epigenetics has been hidden in the closet," says Donald F. Corcoran, president and chief executive officer of MethylGene, which now is working with Pharmion, maker of Vidaza, to combine demethylation agents with inhibitors of histone deacetylases (HDACs), a class of enzymes that remove acetyl groups from histones. Researchers have found that HDAC inhibitors can reactivate genes whose normal job is to suppress tumor growth but that have become inappropriately silenced.
"Every major pharmaceutical company now has a program in histone deacetylases," says University of Southern California cancer researcher Peter A. Jones, director of the USC/Norris Comprehensive Cancer Center. Testimony to the industry's commitment to HDAC inhibitor development was Merck's acquisition in 2004 of Aton Pharma and with it the HDAC inhibitor SAHA (suberoylanilide hydroxamic acid). This epigenetics-based agent is now in numerous human clinical trials for patients with a variety of cancers. Last month, the company announced that FDA had accepted the drug for priority review for treating refractory cutaneous T-cell lymphoma. The drug could be approved for that use as soon as October, according to Merck.
Even as epigenetics research begins to show clinical and commercial promise in medicine, Young of the Whitehead Institute has been identifying chromatin landmarks that might have everything to do with the grand question of how multi-trillion-celled organisms develop from single cells. This April, he was among one of three large research teams reporting work that mapped out intriguing patterns of chemical marks in the chromatin of embryonic stem cells. As the researchers see it, these features likely are important parts of the epigenetic mechanisms by which these tabula rasae of cells begin differentiating into more specialized cell types.
"We are trying to understand the genetic regulatory circuitry of human development," Young says. And what could be a better place to start this quest, he asks, than with embryonic stem cells, the handful of so-called totipotent cells in the embryonic, pinpoint-size clutch of cells from whence all other cell types emerge?
For their part, Young and 27 collaborators from research institutions in Massachusetts, Amsterdam, and Yokohama and Tsukuba, Japan, examined the role of so-called polycomb group proteins (Cell 2006, 125, 301). This set of proteins can mix and match to form a family of chromatin-binding and gene-repressing complexes known as PRCs. From previous studies by others, Young's team knew that PRCs wield their genetic control by binding to and gumming up specific locations along chromatin that host DNA sequences associated with so-called gene promoters.
Young's team chose to focus on a polycomb gene repressor complex known as PRC2. PRC2's chemical raison d'être is to bind to chromatin and add a gene-silencing methyl group to a lysine on one of the histone proteins (histone-3) in the nucleosomes in the vicinity. What a difference these minuscule, four-atom methyl groups can make: Collectively, they alter the chromatin in the PRC2-targeted regions in a way that shuts down the gene-coding stretches that cohabit these chromatin regions. The methylation achieves this gene-silencing result by causing the local DNA to tighten around nucleosomes like fishline on reels. With the decreased slack, DNA-reading and -transcribing enzymes cannot gain access to the genetic code.
With the audacious experimental scope that has become commonplace over the past decade as a result of genome-scale analysis techniques, the researchers devised a protocol to map the genomic locations of a single component of PRC2 in human embryonic stem cells. The technique relies on antibodies to isolate the bits of chromatin where the specific PRC2 component binds and microarrays carpeted with DNA fragments derived from the human genome to determine those bits' location in the genome.
Their data showed that the gene-repressing complex binds suspiciously close to the DNA associated with gene promoters. Results from an extensive series of additional experiments strongly suggested to the researchers that PRC2 has a penchant to shut down those particular genes, which, when expressed in embryonic stem cells, trigger differentiation. In other words, PRC2???and the gene-silencing methyl groups it installs???appears to be important in maintaining a chromatin state that correlates with the cell remaining a stem cell. "This is likely to explain, at least in part, why PRC2 is essential for early development," the researchers suggest in their Cell paper.
In the same issue of Cell, another research team revealed other potentially pivotal elements of epigenetic control during development. In their case, cancer researcher Bradley E. Bernstein of Massachusetts General Hospital, chemist Stuart L. Schreiber of the Broad Institute of Harvard and Massachusetts Institute of Technology, and coworkers at research centers around Boston and in Montpellier, France, identified a unique pattern of methyl marks at tantalizing locations throughout the chromatin of mouse embryonic cells (Cell 2006, 125, 315). The sequences of DNA in these stretches of chromatin do not code for genes, but they are found extensively in the genomes of mammals, whether they be mice or men, and frequently are associated with key developmental genes.
Using some of the same types of genome-scale analysis techniques that Young's team relied on, Bernstein and his colleagues found that many histone-3 proteins in the nucleosomes within these noncoding regions harbor subtle modifications that amount to a traffic light for gene expression. In molecular terms, the researchers found large chromatin regions in which two different lysines on histone-3 are methylated. This dual methylation serves as a red light for genes in the vicinity. Additionally, the researchers found that demethylating one of these lysines switches the red light to a green one. They suggest that these red and green lights work together to "silence developmental genes in [embryonic stem] cells while keeping them poised for activation." Just what signals intervene to send the stem cells down paths of differentiation into other cell types remains unknown.
In yet another related paper published online that same week in April, this one in Nature, Young, his Whitehead colleague Rudolf Jaenisch, and their coworkers heaped on yet more revelations about specific features of stem cell chromatin that are linked to whether the stem cells remain totipotent or begin to differentiate (Nature 2006, 441, 349).
This research effort revealed that two different PRCs occupied the same chromatin regions in mouse embryonic stem cells that also host genes for developmentally important transcription factors. Moreover, they found that nucleosomes at these PRC-decorated chromatin sites had a trimethylated lysine residue in their histone-3 components. This chromatin modification leads to silencing of the associated developmental genes, the researchers found.
The late April week in which these Cell and Nature papers came out was a banner week for uncovering features of the chromatin landscape that likely govern the developmental state of embryonic stem cells. Yet that week's input to the literature is also reflective of what the global epigenetics research wildfire is producing these days. All of this activity, says C. David Allis, head of the Laboratory of Chromatin Biology at Rockefeller University, is creating a positive feedback loop for research.
"When you show the specific epigenetic context that defines how a young embryonic stem cell might be locked as a stem cell or how it might shift to becoming another type of cell, more kinds of crowds will get into epigenetics research," he says. "There will be more developmental biologists coming in," he suggests. And because chromatin-based phenomena undoubtedly play roles in whether stem cells can ultimately be used in such coveted applications as regenerating damaged tissue, whether it will be possible to reverse epigenetic changes underlying cancer and other disease processes, and whether cloning entire organisms from cells can ever become practical and predictable, Allis expects the ranks of epigenetics research to continue rising.
Additional signs that the field is taking on a new self-awareness is the launching this year of the Elsevier journal Epigenetics and a building momentum to organize a Human Epigenome Project modeled somewhat after the multi-billion-dollar Human Genome Project. "There is a need for a large-scale epigenetic mapping of our cells that moves biology to this new century with a fresh face," argues Manel Esteller, director of the Spanish National Cancer Center's Cancer Epigenetics Laboratory, in Madrid. "There is a necessity for a world-scale Human Epigenome Project for the normal and cancer cell." Individual and institutional champions of such a project gathered in Philadelphia to further define their goals as C&EN went to press. On the table, notes USC's Jones, will be to identify which chromatin features and modifications to investigate, in what detail, and in what types of cells.
The prospect of big-science epigenetics research like this is creating a demand for new analytic techniques that can detect specific chromatin changes and histone markings. Chemical biologist Neil L. Kelleher of the University of Illinois, Urbana-Champaign, for example, and his colleagues are developing mass spectrometry protocols that enable them to profile the number, type, and locations of modifications on the different histone proteins of nucleosomes.
With an eye on keeping the epigenetics research momentum going in the longer term, Allis has teamed up with Thomas Jenuwein of the Research Institute of Molecular Pathology, in Vienna, and Danny Reinberg of the Robert Wood Johnson Medical School, Piscataway, N.J., to edit what Allis says will be the first epigenetics textbook. It's due to be published next year by Cold Spring Harbor Laboratory Press.
"When I teach undergraduate biochemistry, there are only a few pages and only one figure on epigenetic mechanisms in the textbook," notes University of Florida biochemist Keith D. Robertson, whose research focuses on protein interactions underlying the methylation of DNA. Rather than amounting to an aside in a typical molecular biology or genetics course, chromatin and epigenetics could become, with teaching tools like this epigenetics book, the focus of entire courses, Allis suggests. "We hope this book will become the go-to bible on epigenetics," he says.
Allis has long been one of the more influential researchers in the quest to synthesize the thousands of scientific pixels that chromatin and epigenetics researchers have been churning out into a coherent picture. His ascent into this role quickened in 1996 when a research group that he was leading and another research team headed by Schreiber independently published a pair of pivotal findings about how chromatin can be chemically modified by enzymes in gene-controlling ways. At that time, Allis' group became the first to identify a histone acetylase (HAT), an enzyme that adds an acetyl group to nucleosomes' histone proteins. A month later, Schreiber's group reported the ying to that yang, a histone deacetylase. "The field has not been the same since," says Allis.
"What is happening now is that loads and loads of players are being found," says epigenetics researcher Adrian Bird of the University of Edinburgh, in Scotland. Among these epigenetics players are enzymes that methylate, demethylate, acetylate, deacetylate, and otherwise modify numerous chromatin locations. They also include proteins, protein complexes, and enzymes that home in on these chromatin modifications.
The point of this "rich vocabulary" of chromatin alterations, says Bird, is that "you can differentiate one piece of a genome from the rest, according to the array of marks that you have there." With his use of "you" Bird is referring to the perspective of, say, chromatin-interacting proteins and other molecules that contribute to the overall choreography of genetic activation and inactivation.
So rich is this growing repertoire of chromatin markings that Allis and Jenuwein floated an idea in Science in 2001 that it all could be thought of as a "histone code" in which the effective readout of the chromatin markings is the expression or silencing of genes, in sickness or in health (Science 2001, 293, 1074). Other epigenetics scientists have suggested different metaphors, but most of them center on the notion that all of these chemical marks amount to an indexing system in the chromatin landscape that records the developmental "decisions" each cell has taken.
"If you have 1,500 genes that are active in a liver cell, you want the same genes to be active when that cell replicates," says Spradling. "Otherwise it would have to somehow learn to be a liver cell all over again."
To build upon Spradling's example, the chromatin landscape of a liver cell might be indexed on a genomic scale with methyl groups, acetyl groups, and other chemical signposts that collectively orchestrate the activation of the specific subset of the 25,000 genes that make a liver cell a liver cell. From a chemical point of view, all of these modifications change the molecular crowding, electrostatic environment, and other pushes and pulls, thereby changing the very shape of the chromatin. Those changes, in turn, influence which gene-activating or gene-repressing factors are able to approach different parts of the chromatin landscape. For a neuron, the distribution of signposts would be different from that in a liver cell or in any other cell type. If researchers could artificially index chromatin, they might gain the ability to change any cell type into any other cell type or to replace a cancer-inducing index into a health-inducing one.
"At the end of the day, how chromatin works is a structural problem," Allis says. How it packs into dense, genetically quiet regions and looser genetically active ones, for example, has everything to do with structure. And that is where all of this indexing and marking with chemical groups must come in, he says.
"We are all trying to study and understand the rules of chromatin," Schreiber says. This magnificent puzzle is at the heart of life's molecular choreography, he notes.
"We are in the detail-gathering phase, and that is always confusing," Bird adds. "You dream of unified theory that puts it all together. That hasn't come around yet."
- Chemical & Engineering News
- ISSN 0009-2347
- Copyright © American Chemical Society