Born in West Hollywood, Mike Levine says he grew up in a tiny house in the "boring suburbs" with a family he affectionately describes as dysfunctional. It was there, however, he's pretty sure he became interested in science. "My only escape was the backyard," says the professor and codirector of the University of California at Berkeley's Center for Integrative Genomics. "I'd run back there and tear open bugs and look at their guts under the microscope. I don't know if it was biological curiosity or just getting away from my mother. But my fondest memories as a kid are of playing in that backyard ... withdrawing into my own world, with my little microscope, looking at the guts of different bugs that I killed and maimed."
Even getting stung by a dead bee he was dissecting did not dissuade Levine from pursuing a career in biology. As an undergraduate in genetics at UC, Berkeley, in the mid-1970s, Levine washed glassware in the lab of evolutionary biologist Allan Wilson. Wilson had compared the protein-coding sequences of chimps and humans and found them so similar, he knew there had to be another explanation for our differences. The theory was that these differences must be due to differences in regulatory DNA. "I think that subverted my thinking ... and affected a lot of the stuff I do," says Levine, who has since devoted his career to dissecting the molecular basis of gene regulation and the role that regulatory networks play in development and evolution.
"Mike's work has done for animal development what the work on the lac operon and phage lambda did for understanding gene regulation in simpler organisms," says Sean Carroll of the University of Wisconsin in Madison. In addition to his contributions to the identification of the homeobox, a DNA sequence found in many genes involved in controlling embryogenesis, says Carroll, "Mike led the discovery of the modular organization of the regulatory regions of developmental genes." Those "two big discoveries," he says, "had a very large conceptual significance for developmental biology and by extension for evolutionary biology."
Levine has worked to extend these studies from Drosophila - his main squeeze when it comes to model organisms - to a variety of other creatures, helping to shore up the concept that development unfurls in a similar manner in a host of organisms. "That's probably the most important discovery in biology in the past century," says Kees Murre, a former colleague from UC, San Diego. "And Mike was a major part of that."
It could have gone a different way. Before becoming a biomedical scientist, Levine toyed with the idea of going to medical school. "Coming from a modest background, particularly a Jewish family, the pressure to become a doctor was intense," he says. Levine even interviewed at Case Western, wearing, he says, "a hideous suit that looked like I was being bar mitzvahed again." Both Levine and his interviewer "realized this was not my calling. So I came to my senses and went to grad school."
At Yale University in the late 1970s, Levine joined the lab of Alan Garen, who was studying Drosophila development in the department of biochemistry and biophysics. It was a very hard-core department, says Levine, which "was painful at the time. But being introduced to that kind of reductionist thinking - trying to take the most complex phenomena down to the most basic principles - had a really good impact on my training and my future research activity," he says. "Later on I felt perfectly comfortable in thinking about DNA-binding proteins, and proteins interacting with each other cooperatively on enhancers, even though I was dealing with incredibly complicated aspects of embryonic patterning."
From Yale, Levine headed to Switzerland to work with Walter Gehring. In the early 1980s, Levine says, "this was literally one of two labs in the whole world using the tools of modern molecular biology to study these really remarkable, mystical genes that control where body parts are put on an adult fly, whether you have an antenna or a leg or a wing." He became interested in Antennapedia, a gene that when mutated transforms antennae into legs, and he hooked up with student Ernst Hafen to develop the in situ hybridization techniques they needed to see when and where Antennapedia was expressed. "There were six months at least where we were just playing with salt conditions," he says. But then, they looked under the microscope and saw a perfect ring of expression around the middle of the embryo, just where geneticists predicted it should be. "That was a great moment," he says. It led Levine, along with Hafen and their colleague William McGinnis, all in Gehring's lab, to the homeobox.
According to Levine, it happened like this: After learning that Ultrabithorax, a gene that specifies the development of wings, showed a localized pattern of expression similar to that of Antennapedia, they decided to revisit the classic papers of Ed Lewis. In 1978, Lewis had proposed that all these homeotic genes (the ones that tell animals where to put a wing and where to put a leg and so on) arose from a common ancestral gene. So McGinnis carved up the Antennapedia gene and, using those pieces as probes, the trio identified eight genes, which turned out to be the eight homeotic genes in flies. "That pissed off a lot of people," says Levine. "The homeotic genes were the trophies of the Drosophila genome. And we got 'em all. I mean, we got 'em all!" Far from being humble, Levine says, "We were like, 'We kicked your ass pretty good, didn't we, baby!' Those were the days."
YOU AIN'T SEEN NOTHIN' YET
Before leaving Basel, Levine secured a faculty position at Columbia University. "He came and gave a seminar here when he was a graduate student," says Columbia's Jim Manley. "Everybody was really impressed, so we offered him a job before he'd even done his postdoc."
When Levine came to New York in 1984, he and his lab conducted a broader search for homeobox genes that might be involved in other aspects of embryonic patterning. One of the genes they isolated was even-skipped (eve). This gene is expressed in seven stripes along the length of the developing fly embryo and is key to dividing the embryo into segments. By dissecting the regulatory region flanking the eve gene, Levine and his coworkers determined that separate enhancers were involved in producing each stripe. They then focused on stripe 2, and by whittling down its enhancer and then fishing around for the proteins that bound to it, they learned that the production of eve in that second stripe was controlled by the binding of a set of activators and a set of repressors, which together carve out the boundaries of the stripe. The repressors, it turns out, are highly localized in the embryo, shutting down only the enhancer to which they bind. That leaves the enhancers that have escaped repression free to specify other eve stripes.
"Before Levine's studies of even-skipped stripe 2, it wasn't clear how you generated spatially restricted patterns of gene expression from initially broad crude gradients of morphogens," says former student Joseph Corbo of Washington University in St. Louis. "I think that the even-skipped stripe 2 studies were the defining studies that showed how an organism can interpret those gradients and turn them into specific patterns of gene expression. To me that's Mike's crowning achievement."
"I remember Mike telling me when they had in hand the element that drove stripe 2, he said, 'I'm gonna take this apart like a genetic switch. Like phage lambda or lac operon,'" says Carroll. "He wanted that level of resolution. And that's what's made eve stripe 2 a textbook model of gene regulation." The studies of embryo segmentation also garnered Levine tenure in 1988 at the age of 33, just four years after he joined the faculty at Columbia.
ONTO SEA SQUIRTS, AND OTHER ENHANCERS
In 1991, Levine moved to UCSD, where he began to work on the sea squirt, Ciona intestinalis, a simple animal whose embryos resemble a stripped down version of a vertebrate tadpole. With only 1000 or so cells, the Ciona embryo provides an ideal system for studying the basic body plan shared by all higher life forms. Levine and his team are using it, among other things, to explore "the cellular basis of embryonic development." They would like to use information about the signaling and regulatory molecules present in a cell at any given moment to predict what that cell is doing: Is it migrating from the tail to the head, or is it flattening out and hanging tight to its neighbor? "I don't know the answer," says Levine, "but that's what it'll take to go from a DNA code to a higher organism."
His other major interest is to learn enough about the composition of enhancer elements and the proteins that bind to them to be able to decipher a regulatory code. Such knowledge, he says, would allow scientists to look at a genome sequence and know when and where every single gene will be expressed - and perhaps to reconstruct the animal in question. He hasn't yet built a trilobite, but the mathematical models he's assembled with his colleague, Dmitri Papatsenko, have allowed him to make good predictions about gene expression patterns in Drosophila, and to identify enhancers that behave similarly in honeybee and mosquito. "I think it's pretty amazing," says Don Rio, a colleague at Berkeley. "Especially because a lot of these rules and ideas have come from work in his own lab."
All these rules and ideas and results derive from the fact that Levine is, by all accounts, a hard-driving scientist with an insatiable hunger for data. "He needs to have his data-fix," says Steve Small of New York University, a former postdoc who remembers Levine wearing a t-shirt that read, "Don't talk to me unless you have data." At least he was up front about it, laughs Small. "He used to come into lab and say, 'Stevie, what have you done for Michael S. Levine today?'"
"He is unbelievable when he's on fire," agrees Corbo.
SETTING THE LAB AFIRE
Then there are the times when Levine sets other people on fire. Literally. "The most famous thing I ever did is I torched one of my postdocs," says Levine. Corbo was there at the time. "Mike got a squirt bottle of ethanol, unbeknownst to this hapless postdoc who was sitting at his bench minding his own business," says Corbo. Levine shot a ring of ethanol around the young man's seat and trailed a wick into the hallway. Then he lit it. "So this tongue of flame snaked into the lab and encircled this postdoc," says Corbo.
"My technique was a little off and I put a little too much ethanol around his bench. So it's true, he was temporarily enveloped in a curtain of fire," says Levine. "But the fire receded and he was ok."
"People heard I did this because the guy wasn't making any data," he adds. "That's not true. It was just a joke." Perhaps it was, but undoubtedly the postdoc got back to work to produce more data for Michael S. Levine.