From birth, everyone has a single genome. However, growing up in a particular environment affects that genome in unique and sometimes unforeseen ways that can't be duplicated in another environment.
"I think there are great epidemiologists who don't understand the molecular biology and there are great bench scientists who have no feel for population studies. We need to start having more people who can speak the same language and bridge those gaps." -- Carmen Marsit
Carmen Marsit has found that much the same can be said of scientific careers -- or, at least, his scientific career. His research focuses on how environmental exposures and lifestyle choices alter gene expression through so-called epigenetic changes to the genome. Now an assistant professor of pathology and lab medicine at Brown University, Marsit says his career has been defined by a combination of his innate interests, his personality, and his exposure to new scientific ideas.
"I think I can apply my interest in epigenetics and epigenetic outcomes into many areas. Not everyone is comfortable with that approach, and I think that's a combination of who I am and my training experience where I got to work with all sorts of people," says Marsit, who is 31. "It's funny how your career can change in ways you don't expect."
Early environmental exposure
Like many biological science majors, Marsit started his college career assuming he was headed to medical school. He completed coursework, wrote papers, and -- like many of his peers at Lafayette College in Easton, Pennsylvania -- participated in summer research. As he studied parasites in frogs and snails and researched how chemicals in water affected animals, he became fascinated with how the environment impacts health. He didn't realize at the time that this exposure to research was fundamentally altering his career path.
The 29 October issue of Science contains a special section on epigenetics. It includes a series of Reviews and Perspectives that look at what defines an epigenetic system and how epigenetic processes influence biology, a Perspective about the possibility of using stem cells to study epigenetic perturbations, and a News Focus story on drugs that reverse the abnormal epigenetic patterns in tumors.
When it came time to compose a personal statement for his medical school applications, he found that he had nothing to say. "I figured if I can't write a simple statement saying why I wanted to be a doctor, it should be a sign that I'm not doing the right thing," Marsit says. "That really forced me to think about what I wanted to do and to follow my interests. And I knew I wanted to do research."
He decided to do a Ph.D. in the Biological Sciences Division at the Harvard School of Public Health. Typically, students in that division focus heavily on basic science and explore in vitro and model systems. But, he says, "When I started, I was really more interested in population science and human health." So he took additional coursework in epidemiology and biostatistics to get a firm understanding of population science. "Too often, epidemiologists specializing in molecular epidemiology don't actually learn laboratory skills, and basic scientists do not understand the epidemiology. I felt like I needed to do both."
Boston's medical research scene includes an abundance of top cancer researchers. That environment exposed Marsit to cancer biology and epigenetics and prompted him to focus on how environmental exposures alter the pattern of chemical tags attached to DNA in lung cancer tumors, work he did in the lab of Karl Kelsey, now the director of the Center for Environmental Health and Technology at Brown. "[Lung cancer] was a good exposure-related disease, and I wanted to pursue how exposures lead to epigenetic changes in lung tumors," he says.
Mining the epigenetic landscape
Epigenetics is the study of mechanisms, other than alterations in the underlying sequence of DNA, that control a gene's expression. Marsit's work, for example, focuses on DNA methylation, which occurs when small methyl groups adorn DNA and block gene expression; and microRNAs, small RNAs that shut off gene expression by rendering messenger RNA unreadable. Environmental exposures such as cigarette smoke, alcohol, and nutrition affect the pattern of DNA methylation and expression of microRNA in ways that are highly personal and change over time.
During his Ph.D. research, Marsit found that lung tumors could be classified into distinct groups depending on which epigenetic changes occurred in critical molecular pathways. That "subclassification may hold tremendous clinical utility because it could one day help define patients' response to various therapies," he says.
Marsit sought a bigger challenge as a postdoctoral fellow. While continuing to work in Kelsey's lab and to collaborate with Margaret Karagas, an epidemiologist at Dartmouth Medical School, he focused on bladder cancer, a disease for which much less is understood about the effects of environmental exposure. "We don't really know about all of the exposures that are contributing to the development of bladder cancer. So it is a much more interesting problem to look at in terms of epigenetics," he says.
He's continuing his bladder cancer collaboration at Brown, where he took a faculty position in 2007 in the university's Department of Pathology and Laboratory Medicine. There, most of his colleagues are experimental pathologists who work in animal model systems. "They brought me in to connect what is happening in model systems to what we see in humans," he says, noting that he had considered a position in a traditional epidemiology department but couldn't bear to give up having a laboratory.
A collaborative environment
When he went to Brown, Marsit visited several departments to talk about his work, including the pediatrics department at the Women and Infants Hospital. "He gave this great talk about genes and environment in cancer," says James Padbury, a professor of pediatrics and director of neonatal-perinatal medicine at Women and Infants. "And I asked him whether he wanted to know what the most robust gene-environment interaction was with the narrowest windows of action and the shortest time to onset of a significant effect. He said, 'Yeah!' And I told him, 'Well, that is the intrauterine environment in the latter portion of pregnancy.' "
"He basically told me that he loved my work but I needed to shift my focus to development," Marsit says.
As a neonatologist, Padbury has seen premature infants survive when born as early as 23 or 24 weeks. He has also seen those preemies struggle with neurodevelopmental difficulties. "These kids are supposed to be in the womb," Padbury says. "Instead, they are in a very different environment that can have expansive, long-term neurodevelopment effects. Optimizing that environment is critical to giving these kids their best chance."
Padbury says that although his team has been able to show that the placenta plays an active role in determining the intrauterine environment, they've been stymied in fully describing how it does so, in part because much of the effect appears to be from epigenetic effects. That's where Marsit's expertise comes in handy, Padbury says. Since their initial conversation, Marsit and Padbury have been collaborating to uncover those epigenetic changes.
"This is really a dynamite collaborative opportunity," Padbury says. "It's an example of a very interdisciplinary, very contemporary, and very translational collaboration."
The third key player in the collaboration is Barry Lester, a developmental psychologist and director of the Brown Center for the Study of Children at Risk, who has developed a scale for predicting developmental delays in early childhood. Using Lester's Neonatal Intensive Care Unit Network Neurobehavioral Scale -- a measure of things such as the ability to follow visual and auditory stimuli, lethargy, and levels of arousal and stress -- the collaborators have begun to see evidence that methylation of the gene for the glucocorticoid receptor alters an infant's ability to respond to visual and auditory stimuli.
"We know there are a ton of exposures and lifestyle characteristics that contribute to a healthy pregnancy, healthy birth, and long-term health in the child," Marsit says. "But it is really hard to pinpoint all of them. So instead, we look at epigenetic profiles and try to think about how these regulatory mechanisms might be more useful as markers of health and disease."
Marsit is also starting to work with Stephen Sheinkopf, an assistant professor in psychiatry and human behavior at Brown and a member of the Center for the Study of Children at Risk. Sheinkopf has been examining distinct characteristics of infants' cries to see if they can identify children at risk for autism. Marsit is interested in seeing if those cries correlate to epigenetic marks. "I'm learning something new every day in neurodevelopment," Marsit says.
Bridging the gap
Marsit's unique training in both molecular biology and epidemiology has enabled him to form these and other collaborations and to ask broad questions in epigenetics. That ability to speak the languages of different disciplines is "becoming more and more important with the push now for translational medicine and moving things from the bench to the bedside," Marsit says.
"I think there are great epidemiologists who don't understand the molecular biology and there are great bench scientists who have no feel for population studies," he adds. "We need to start having more people who can speak the same language and bridge those gaps."
Lisa Seachrist Chiu is a science writer in Washington, D.C.