Biology branches out

Evolutionary biologist Rob DeSalle works to build Darwin's theoretical Tree of Life.

By Venkat Srinivasan
Photography courtesy Rob DeSalle

In a City College of New York classroom, Rob DeSalle, AB'77, instructs a group of high-school science teachers how to better teach classification methods in biology. The invertebrate zoology curator from New York City's American Museum of Natural History opens a bag of candy bars, asking the teachers to sort them and construct a hypothetical hierarchy of species, first categorizing by ingredients, and then separately by the wrappers' appearance. DeSalle also uses this exercise to explain the work he does as a systematist, a scientist who studies relationships among species, current and extinct. "Biology starts to make sense when you compartmentalize things," he says.

With more than 1,000 whole genomes sequenced and new ones being processed every day, DeSalle believes scientists can construct a complete Tree of Life.

Darwin proposed the Tree of Life, a structure showing that all life forms systematically evolved from common ancestors, more than 150 years ago. But there are still a lot of questions about how species are related. For each connection, a scientist needs multiple forms of evidence—often DNA data, in addition to anatomical and behavioral relationships.

Part of DeSalle's job is to help add to the Tree of Life, looking at genetic data for clues to the origins of species. Working in the museum's molecular laboratory, the Sackler Institute for Comparative Genomics, he collaborates with scientists who use the lab's equipment for DNA sequencing. They send gene samples from around the world: an unknown carnivore's excrement from Afghanistan, for example, and tiny skin samples from humpback whales off the coast of Gabon. Technicians plug DNA into the lab's gene sequencer, which spits out streams of the genetic alphabet—A, T, C, and G—into databases. Using the sequencer, DeSalle and his researchers deduce a species' internal structure, and then they find different ways to organize and analyze the sequences. For instance, scientists use information on whales' genetic diversity to understand how different the population might have been in the past, valuable knowledge for conservationists seeking to quantify whale loss over centuries.

DeSalle's interest in genetics started by chance. During an undergraduate trip to Chicago's Field Museum of Natural History he fell in love with narwhal whales. His adviser, James Shapiro, a professor in the University's Department of Biochemistry and Molecular Biology, suggested he take up genetics as a way in to studying whale biology. After completing undergraduate work in genetics, DeSalle graduated with a biological-sciences degree from the College and in 1984 earned a PhD in evolutionary biology from Washington University in St. Louis. There, he studied under Alan Templeton, famous for demonstrating the lack of genetic diversity across human races—the basis for the idea that races are a social, not a biological, construct. In 1986 DeSalle joined Yale University's biology department as an assistant professor, where he stayed until 1991, when he joined the American Museum of Natural History.

One of DeSalle's big projects is trying to prove that scientists can create a huge Tree of Life explaining all of evolutionary history. There is a debate right now between scholars, like DeSalle, who believe that this tree can be built, and others, mostly microbiologists, who believe that smaller, individual trees analyzing single genes are the only possible way to study relationships among organisms. Microbiologists argue that horizontal-gene transfer—a process by which organisms like bacteria incorporate genetic material from a life form that isn't its parent or ancestor—would interrupt the tree's record. "Microbes are really slutty," says DeSalle, who has done phylogenies of multicellular animals, plants, and bacteria in consort with other organizations like New York's Cold Spring Harbor Laboratory. "They just pass stuff back and forth to each other." They don't reproduce sexually; instead, bacteria exchange genes by taking in DNA from outside, or in some cases, transferring genetic material between cells.

DeSalle believes the question should not be whether one can construct the Tree of Life, but rather how one should assemble evidence for building that tree. This assembly of evidence—genetic data is just one type—is what he and other systematists call the "supermatrix."

Best articulated, says DeSalle, by biologist John Gatesy of the University of California, Riverside (with whom DeSalle has published several papers), the supermatrix theory includes every piece of data—external shape, gene sequences, proteins, characteristic traits, anatomy—that describes a species and is, therefore, a clue to its origins. Building a supermatrix means simultaneously collecting and comparing all genetic evidence across all species. DeSalle argues that despite horizontal-gene transfer, the supermatrix provides enough crucial information to understand how species evolve, and chart out a relevant relationship between them. "The matrix is meaningful," he insists. "It doesn't put bacteria with hippos and hippos with plants."

The debate, he says, is indicative of a paradigm shift in evolutionary thought. Ten years ago, it would have been impossible to imagine the possibility of creating a supermatrix. "We only had a handful of whole genomes at that point," but now more than 1,000 whole genomes have been sequenced. There's been a "data overflow," he says, with new bacterial genomes being processed every day. With this influx of data comes the opportunity for new methods of analysis, like the supermatrix approach, which essentially links sequences from 10 or 15,000 genes and allows scientists to construct a huge tree from that information, instead of looking at a single gene or two.

The vast amount of data inspires DeSalle, who admits that a tree made using a supermatrix approach "may not be the one true set of relationships that these organisms have, but it is at least the most logical and data-rich hypothesis that we will ever have about these organisms."


Return to top