When Constellation Pharmaceuticals first called to recruit venture capitalist Mark Goldsmith to be its chief executive in 2009, he was sceptical. Although Goldsmith was looking to change careers, he worried that the young biopharmaceutical company was heading into murky waters. The firm in Cambridge, Massachusetts, was focusing on epigenetics — the study of heritable changes in gene expression that are not due to changes in DNA sequence. It planned to create cancer treatments that correct the abnormal patterns of epigenetic DNA modifications seen in tumours. “I took some convincing,” he says. “This was not an easy class of targets to go after; they were all unprecedented targets with incompletely understood biology.”

Despite his qualms, Goldsmith took the helm and, nearly three years later, epigenetics has become a hot topic in oncology drug discovery. In January, biotechnology giant Genentech in South San Francisco, California, added its own vote of confidence to the field by investing US$95 million in a partnership with Constellation, which is now hoping to add another 10 scientists to its current roster of 70.

Computer reconstruction of a cancer cell on a DNA autoradiogram. Credit: PASIEKA/SPL

Epigenetics, and cancer epigenetics in particular, is a bright spot in an otherwise stark biomedical-research funding and jobs landscape. Those with the right skills and background — computational and bioinformatics training, familiarity with and interest in translational research, and an intimate knowledge of molecular biology and cancer research techniques — have plenty of opportunities from which to choose. In particular, computational skills are so sought after that they alone could be a bridge to the sub-discipline. “It's a really hot field,” says Benjamin Garcia, a chemist at Princeton University in New Jersey. “I wouldn't be surprised if in five to ten years, you're going to see a lot of universities with epigenetics departments.”

Coming of age

Genetic mutations are not the only way to alter gene expression and protein function. Methyl groups added to DNA can silence a gene, as can chemical changes made to proteins called histones, which package the DNA in chromosomes. The modifications are exquisitely complex: the effect of one epigenetic change can be shaped by other modifications found nearby, and the epigenetic state of a cell will vary depending on the cell's identity and maturity.

By the 1990s, researchers knew that the epigenetic state of a cancer cell was often in disarray. DNA methylation, for example, was markedly reduced in some tumours, unleashing gene-expression programs that were normally kept under lock and key. “The cancer genome was grossly different,” says Susan Clark, an epigeneticist at the University of New South Wales in Australia. “It was an amazing discovery.”

An example of one of Epizyme's inhibitors interacting with an epigenetic enzyme. Credit: L. JIN/EPIZYME

But many in the field needed further convincing before accepting that this epigenetic 'chaos' promoted changes in gene expression and, ultimately, led to cancer. That scepticism was not limited to industry; academics also worried that a career in the field would bring funding struggles and rejections from high-impact journals. “Clearly it was a risk, even ten years ago, to somebody's career to dedicate themselves to an area that seemed to have a lot of hand-waving,” says Clark. “Now that's changed; it is certainly a growth area for young scientists.”

As mounting evidence pointed to the importance of epigenetic changes in cancer, government funders began making significant investments in the field. The US National Cancer Institute (NCI) in Bethesda, Maryland, has several programmes dedicated to epigenetics, including the Epigenetic Approaches in Cancer Epidemiology programme, which funds about 30 projects at a total of $45 million. In 2011, the US National Institute of Environmental Health Sciences awarded about $11 million in grants for epigenetics-related research. The institute has a strong interest in the environment's effect on epigenetics and how that influences diseases such as cancer, notes Edward Kang, a spokesman for the institute, which is based in Research Triangle Park, North Carolina.

Government investment has also fuelled the shift to large, genome-wide epigenomics studies. In October 2011, the European Commission launched its €39.9-million (US$52.1-million) BLUEPRINT project, which brings together 41 institutes and companies to generate at least 100 reference epigenomes from healthy and leukaemic cells. Just over €2 million of that is still to be doled out, says project coordinator Henk Stunnenberg of Radboud University in Nijmegen, the Netherlands. The project's team hopes to recruit at least five more groups from academia and industry. The European Commission support of epigenetics research helped to woo Manel Esteller, an epigenetics researcher at the Bellvitge Biomedical Research Institute in Barcelona, Spain, back to his home country from the United States. Esteller now participates in the BLUEPRINT project and coordinates CURELUNG, another programme funded by the European Commission, which unites 11 institutions and companies and has analysed DNA methylation in nearly 1,000 human lung tumours thus far. “The European Commission offered the opportunity to apply for different grants that were able to complement local funding,” he says. “This extra help has been critical in the success of my projects.”

Robert Gould: "Personalized treatment for cancer is rapidly gaining ground in industry." Credit: M. MCKEE/EPIZYME

Many of the biggest investments in epigenomics directly fund the larger sequencing centres rather than individual investigators, but smaller laboratories have capitalized on the steady stream of data and new technologies emerging from the programmes. In 2008, the US National Institutes of Health (NIH) launched a $200-million, ten-year Roadmap Epigenomics Project to develop mapping centres and technologies that would allow researchers to survey epigenetic changes on a genome-wide scale. Although the project focuses on the epigenetics of normal, non-cancerous tissue, the technological advances and large data sets have helped cancer research as well. Many cancer genome sequencing projects, including the NCI's The Cancer Genome Atlas (TCGA) programme, include a partial focus on cataloguing epigenetic changes. Kenna Shaw, director of the TCGA programme office in Bethesda, says that the programme has funded around 200 jobs. The bulk of the funding for these large-scale programmes is already dedicated to the larger sequencing centres, but smaller teams are using the data from these projects to generate individual-investigator grant applications, Shaw adds.

These data have helped to persuade investors in industry that epigenetic abnormalities in cancer could provide a wealth of new drug targets. The finding that mutations in epigenetics-related genes may be driving some cancers offers the tantalizing possibility of taking a personalized approach to cancer treatment, a tack that is rapidly gaining ground in industry, says Robert Gould, chief executive of Epizyme, an epigenetics focused biotechnology firm based in Cambridge, Massachusetts. This evidence, plus the successful approval of a first generation of drugs intended to target epigenetic pathways, has convinced almost every major drug company to invest in cancer epigenetics, says Mukesh Verma, a programme officer at the NCI. For example, Novartis, a pharmaceutical firm with its headquarters in Basel, Switzerland, has more than 200 employees working in epigenetics, most of them in cancer, says En Li, head of China Novartis Institutes for Biomedical Research, based in Shanghai. Last year, GlaxoSmithKline in London, in addition to funding its own epigenetics team, paid $20 million to partner with Epizyme in a deal in which Epizyme could ultimately receive as much as $630 million. “GSK's group is partnering with us and is also competing with us on other programmes,” says Epizyme's chief scientific officer, Robert Copeland. “It makes for an interesting dynamic.”

With so much excitement, competition in the field can be fierce. Data from large government projects can be a boon to smaller labs, says Clark, but individual investigators and those new to the field need to carve their own niche. “In the face of those big initiatives, smaller labs have the challenge of asking smaller and more unique questions as to the basic mechanisms underlying these epigenetic changes,” she says. Christopher Vakoc, an epigenetics researcher at Cold Spring Harbor Laboratory in New York, notes that the “tiny” lab he started in 2008 directly competed with several big pharmaceutical companies to discover a role for Brd4 — a 'reader' protein that binds to certain modified histones and modulates gene expression — in acute myeloid leukaemia (J. Zuber et al. Nature 478, 524–528; 2011). After his team's paper was published, Vakoc heard rumours that ten companies were racing to capitalize on the results.

There is also an intense demand for talent. In particular, epigenetics companies and individual labs need bioinformaticians as sequencing projects continue to dump terabytes of data into public databases (see Nature 482, 263–265; 2012). Although this is an opportunity for job hunters with computational training, it creates challenges for those opening labs for the first time, says Jun Song, a computational biologist who opened his lab at the University of California, San Francisco, in 2009. Song has struggled to compete with bigger labs to recruit graduate students and postdoctoral researchers, who often prefer the proven track-record and extensive connections offered by a well-established principal investigator. “We battle to get a talented bioinformatician,” says Clark. “Everybody wants their own.”

Ultimately, Song looked outside biology to recruit three postdocs, two of whom he lured away from high-energy particle physics and the third from applied mathematics. Song himself was trained as a physicist, and says that epigenetics and epigenomics offer a range of challenging computational questions that can entice researchers from other fields. “It would be great to have someone already trained in both biology and computation,” he says. “But as biology becomes more quantitative as a field, I also believe that it's important to bring in new computational scientists and train them in biology.”

The opportunity for cross-disciplinary training in epigenetics can be an advantage for bioinformaticians and molecular biologists alike, says Garcia. “It makes you a more well-rounded scientist,” he says. “And that's what you need these days to compete in the job market.”