As the hoopla over the first draft of the human genome fades, a new, more fundamental endeavor is quietly gearing up in the same Maryland laboratories where much of the mapping of Homo sapiens took place. At The Institute for Genomic Research, a sandy-haired biologist named Scott Peterson and his team are trying to create something nature has not: a single-celled creature with the smallest number of genes necessary to stay alive.
"Some may disagree, but I don't think what I'm doing is creating life," says Peterson. "We're modeling life. We're examining what are the genetic requirements for living cells." The search for such secrets is what drives this nonprofit institute called TIGR (The Institute for Genomic Research). Founded in 1992 by Craig Venter, the pioneer in genetic sequencing who was a coleader of the Human Genome Project, TIGR is dedicated to studying genomes of all stripes--from human to microbe. Peterson's focus, however, is on a particular set of secrets. Although he admits, "I'm motivated by the challenge of doing something nobody's done before," he hopes to understand from this experiment exactly how a cell works.
Predictably, he has chosen to study nature's simplest bacterium, Mycoplasma genitalium. Found in the comfy environs of the human urogenital tract, the needs of this mycoplasma are easily fulfilled, and so, over its long evolutionary history, it has shed thousands of unnecessary genes, becoming the very model of austerity. (The genome of food-poisoning culprit E. coli, considered a basic life-form, is nine times bigger.) By tinkering with mycoplasma's slender set of genes, Peterson is in search of answers to two fundamental questions: How many genes, exactly, does a cell need to live? And which genes are they?
Success will mean more than making history or providing crucial insights into how cells function. Along the way, Peterson's team must grapple with some searching ethical dilemmas. The results could, for example, lead to customized microbes for chewing up toxic waste, but they could also show a clear path to creating bioweapons more deadly than anything nature has dished up. So Peterson is understandably cautious: "I'm not going to name it Dolly or anything like that. Things like that tend to put people off."
The search for the smallest genome stretches back to 1955, when biophysicist Harold Morowitz began collecting a Noah's ark of microbes in his lab at Yale and inspecting each organism's simple circular chromosome. One day he found an impressively runty germ, a species of Mycoplasma, and decided to study it. NASA funded the research, figuring that alien life might resemble something as seemingly primitive and genetically streamlined as mycoplasma. Morowitz supposed that if you knew what each of mycoplasma's genes did, a computer could be used to simulate the system. He foresaw this as a way to study the whole cell, not just one gene here or metabolic pathway there. He imagined the science of understanding genes both in detail and in concert--what Peterson would now call genomics.
Peterson's office has all the cultural trappings of the day: a Pokem—n poster, a lava lamp, a cappuccino machine, a Mac G4. But the icon that testifies to his membership in biology's next generation is a diagram taped to his filing cabinet. With several colored bars spanning several rows, it resembles a small, delicately colored Persian rug. It represents the more than 1,700 genes of Haemophilus influenzae--the first complete sequence of a bacterial genome.
Craig Venter helped inaugurate the genomic revolution with that sequencing project in 1995. But H. influenzae's genetic code was something of a disappointment. More than a third of its genes were completely unknown. Determined to crack a simpler, more manageable genome, Venter's team set their sights on M. genitalium. Three months later, Claire Fraser, now president of the institute, had nailed the 470-gene sequence. Still, mycoplasma's genetic instruction book was too complex for scientists to grasp how the genes work together. Was there a way to make it even simpler?
Craig Venter began looking around for someone who could help. And he found Scott Peterson. Back in the late 1980s, years before the word genome filtered its way into everyday parlance, Peterson had been a budding bacteriologist in graduate school, putting in 14-hour days sequencing sections of mycoplasma DNA. Peterson says he "didn't see a bright future in microbiology." But in 1996 Venter recruited him to launch a long-term exploration of mycoplasma's genes. "Craig had a very alluring scientific pitch," says Peterson. "He outlined a 10-year commitment to learn just how this cell works." Gene-by-gene analysis, chromosome engineering, computer simulations, anything and everything was at his disposal. "Venter said, 'Wow, we could make a chromosome,'" recalls Peterson. Eventually, Venter left TIGR to head up Celera, a private genomics company, leaving the minimal-genome project in his protŽgŽ's hands.
Peterson's first step was to disrupt mycoplasma's genes in various places to figure out which were crucial. To do this, he attacked the mycoplasma genome with bits of DNA called transposons that sneak their way into chromosomes. The invading transposons landed at random within the mycoplasma gene sequence, wreaking havoc. By looking at the cells that died from the attack, Peterson could see where the invading transposon had landed and thereby pinpoint genes essential for the bacterium's life. After this meticulous screening, he and Clyde Hutchison, a colleague from the University of North Carolina at Chapel Hill, identified a list of 300 or so essential genes. Without any one of these genes, mycoplasma would die.
Yet that turned out not to be the sought-after minimal set. If the roughly 300 genes were strung together and slipped into a mycoplasma cell, the most likely result was one pathetically dependent bacterium, if it survived at all. Some genes, like basketball players on a team, tend to work together in cells. The transposon research showed who the team's best players were, but the analysis missed bit players whose teamwork was crucial. In order to approach a real minimal set of genes, Peterson says, you'd have to take genes out a few at a time, a technically challenging proposition. Therefore, "the way to prove that you've got a minimal cell is to make it," he says. But that approach means creating an organism that is utterly new to the face of the Earth.
"If I limit my thinking to the science at hand, it simply represents a challenge," Peterson says. "Where have you gone too far? That's a difficult question. It's one that I haven't properly resolved fully." Although the biotech industry has been altering organisms--from plants to transgenic animals--for more than 20 years, Peterson could see that creating a bacterium with a custom-made, artificially assembled set of genes would be controversial. "In a nutshell, when you are faced with a power unlike anything you've really used before," says Peterson, "you have to stop and ask: Am I using this power appropriately or not?" In 1999, his team commissioned a panel of religious figures and ethicists to discuss the implications. After meeting several times over a year, the panel concluded that the project's basic goals were in keeping with the tradition of sound scientific inquiry. "We found no intrinsic reason in religious or secular ethics that you shouldn't [continue]," says bioethicist Art Caplan of the University of Pennsylvania, who headed the team. "There were some pretty serious religious types who were doubtful when we started and wound up saying it depends what you're going to do with it."
The concern was noteworthy because the institute's team is altering an organism uniquely suited to colonize the human body. "It's not hard for me to imagine a sinister application, and that's frightening," says Peterson. Terrorists, he notes, could use these procedures to hide malevolent genes in other organisms. "If you really want to be sophisticated, you have to cloak what you're doing, like put the genes that make anthrax lethal into more innocent bacteria."
But scientists who make their living in the biotech trade are persistently optimistic about their ability to control the genies they let out of bottles. "Certainly we're interested in the ethical issues around engineering organisms," adds Michael Brasch, Peterson's commercial collaborator. Brasch developed the technology that Peterson's team is using to assemble the artificial chromosome, and he, too, may face public opposition. In an exchange over lunch, Peterson kids his colleague: "You know, they're going to compare you to the gun companies." Brasch half laughs. "I usually hear us compared to Microsoft."
That's because Brasch's firm, Life Technologies, a division of biotech giant Invitrogen Corp., is selling a new scheme as the next killer app of the genome age--the very technique Peterson is using. Dubbed Gateway Cloning Technology, it mimics the way some viruses slip their genes into a host cell's DNA by exploiting genetic tags called recombination sites--regions of DNA that allow bacteria to swap genes with one another. Brasch has developed his own recombination sites that permit him to cut and paste genes with ease, and he's hoping Peterson's work will be a public coming-out party for the system. Before learning about Gateway, Peterson says, his team had been unable to assemble a chromosome. "With available technology, copying the essential genes one by one is easy," Brasch explains. "But linking them together to rebuild the entire genome technically couldn't be done."
The Gateway system should overcome this problem, says Peterson. By including the recombination sites on each gene as it's copied, Gateway will connect mycoplasma's DNA in proper order. Once the chromosome is complete, the recombination sites can be used to identify where genes begin and end. That will make it easy for Peterson to pare the chromosome by removing several genes at a time. But even if Gateway solves the copying problem, Peterson faces another hurdle: choosing which genes to string together. Almost certain to make the cut, he says, will be genes that instruct the cell to make proteins, genes that help build DNA, and genes that are crucial for the cell's replication. The team believes they have properly identified more than 200 crucial genes, including ones for eating, metabolism, and structure. But they have no clue what another 100 of mycoplasma's most essential genes do. "One bad choice could kill the whole thing," Peterson says.
Attempts at computer-modeling life haven't shed much light on the problem. A Japanese group called E-Cell tried in 1997 to create a digital minimal cell. Their 127-gene, less-than-minimal model of a mycoplasma cell was able to simulate life, but not replicate it. The barrier was science's murky sense of how, among other things, mycoplasma divides. "In this particular area," says E-Cell leader Masaru Tomita, "we have to wait for the science to catch up." It's possible Peterson's reductionist strategy will find no definite answers to understanding a cell.
Nevertheless, researchers will begin to understand the inner lives and histories of bacterial genomes. And that alone may be worth the effort. More than 30 genomic sequences now exist for different bacteria, each able to be picked apart. These data mines have spawned a new field in which biologists contrast the genomes of different organisms for clues.
Equally powerful is the prospect of editing genomes to tackle a variety of questions. Take, for example, bacterial disease. Streptococcus pneumoniae, for example, is a microbe that kills more than a million people a year in underdeveloped countries. But why are some strains of the bacterium so deadly and some not even able to inhabit the lungs? Genomic engineering with a man-made chromosome in these bacteria could allow scientists to test a slew of ideas. "We can take a gene out, change it, mutate it, do whatever we want," Peterson says, "and put it back in, leaving everything the same, and ask: What effect did that change have?" Some changes will undoubtedly render the bacteria less dangerous, allowing scientists to identify new drugs that will fight these small terrors.
Mix-and-match chromosome construction could also prove a powerful weapon for tackling questions of evolution. M. genitalium's closest relative is M. pneumoniae, which can cause a bad cough. By comparing the siblings' sequences, it appears that M. genitalium evolved directly from its older brother by discarding 210 genes. Imaginative chromosome reengineering could allow a researcher to replay the divergence of the two species in a frame-by-frame reverse slow motion. "One could start to add the 210 genes back sequentially to M. genitalium," says Peterson, "and ask questions about the evolution."
Ironically, the one question genomic engineering may not be able to answer is which genes are absolutely essential for life. One issue is how to define life--the life-support-machine dilemma, on the most basic level. Normally, says Peterson, M. genitalium replicates itself in about 12 hours. A minimal creature, enfeebled by a bare-bones set of genes, could take much longer, perhaps a month. "And it's so sick that I have to feed it and nurture it. Is that life?" Peterson asks. Even genes designated dispensable may be long-term evolutionary investments. During the first round of experiments, researchers found that the bacteria could live without the gene they believe encodes RecA, a protein that can repair genetic errors. Does that make RecA dispensable? Without the recA gene, mycoplasma cells may survive in the lab, says Peterson, but "in a million years you might suspect they won't be around anymore."
Experiments have also shown that the amount of sugar available determines which bacterial genes are crucial for metabolism. So which would be crucial for a minimal cell? And would the minimal set of mycoplasma's genes that just managed to make do in the cushy lab environment suffice when the bacteria live in mice, where an immune system might go after them? "There's a constant debate over nature or nurture--they're inseparable," says Craig Venter. "I naively thought that we could have a molecular definition for life, come up with a set of genes that would minimally define life. Nature just refuses," he says softly, "to be so easily quantified."
Peterson stands in his lab, holding a FLASK of drifting mycoplasma cells up to the light. "They're not healthy," he says, "and they're starting to suffer." For eight weeks the bacteria have endured yet another trial by fire, this time with a chemical mutagen. Now the genes that allow them to cling to the flask appear to be damaged. Still, the cells are getting by. Losing genes is old hat for Mycoplasma, but the losses over the eons have made its members pathetically reliant. They cannot make raw materials for proteins, DNA, or their cell membranes. So in a lab they demand a diet of ground-up cow hearts, blood serum, and other delicacies. "They're high maintenance," says one assistant.
Without question, Peterson's project--like so many in biotech research--will further our understanding of how genes work together. And it could someday lead to the creation of an entirely artificial single-celled organism, assembled from off-the-shelf components like DNA, proteins, lipids, and sugars. "I think someday science will be in that position," says Peterson, "where we will have to ask: Should we or shouldn't we?" Some restless innovator may build on Peterson's science and take that next uncertain step. It could even be genomics pioneer and entrepreneur Craig Venter, his old boss.
Venter, who maintains informal ties with the institute, says he has no interest in that project. "Right now, the only way you can get life is from life itself," Venter says. "We're working in that direction, but we're a long way away from making the decision to go ahead and do that experiment." Peterson isn't so sure: "I wouldn't put it past him."
The Institute for Genomic Research's Web site describes its varied projects: www.tigr.org.