Robert E. Klein
Let's make one thing clear: Jim Collins won't grow you a house any time soon.
More than a decade ago, Collins, a decorated scientist at Boston University, helped give birth to synthetic biology, which soon grew into arguably the world's hottest and most poorly defined scientific discipline. Its practitioners made big promises: that by harnessing the ideas of engineering and applying them to genetics, they would create cheap, abundant biofuels, customized medicine, even self-growing houses, as one scientist predicted.
The potential for mastering life was so exciting that scientists talked about applications decades away as if they were around the bend. Scholars from the Bay Area and Boston issued forth into industry, promising to reinvent life from the inside out, evolution be damned. Drawing inspiration from electrical circuits and computing, they'd create standardized biological parts. The notion drew easy comparisons to Legos: Snap them together, and soon enough you'd have control of life.
Since 2004 investors have poured at least $1.84-billion into synthetic-biology start-ups; the government has added many more millions in research dollars. But more recently, the hype has died down. Most of those companies have made grinding progress, not breakthroughs. Much potential remains to reinvent manufacturing and medicine, but the road is far longer than some imagined.
There's a simple reason for this problem: The tools have outpaced the knowledge. The cost of genetic sequencing and synthesis continues to plunge, but the functions of many genes in even the simplest forms of life, like bacteria and yeast, stubbornly hold on to their secrets. Genetic networks interact in complex, mysterious ways. Engineered parts take wild, unexpected turns when inserted into genomes. And then evolution, a system that would drive any electrical engineer mad, tiptoes in.
As synthetic biology passes from precocious youth toward maturity, it is returning home to academe. Collins sees an upside to that retrenchment: The science, once a domain of biological amateurs and outsiders, can now inform basic research into life's unending complexity.
Kelvin Ma for The Chronicle Review
Trim and tall, Collins, 47, has had a remarkable rise. A former physicist who, until the late 1990s, knew little about molecular biology, he wears his laurels—MacArthur "genius," Rhodes scholar, Howard Hughes investigator—lightly. If you're outside the field, you'd have little idea he's considered a founder of synthetic biology.
But what is the field, exactly? As one engineer once quipped, ask five people the definition of synthetic biology, you'll get six different answers, because one person is bound to be conflicted. It's a field where most of its practitioners consider its most visible success—Craig Venter and company's wiring together of a microbial genome in 2010—to be an impressive technical feat, but not synthetic biology. The phrase has subsumed whole disciplines. Many scientists who once practiced genetic engineering, methods relatively unchanged, now operate under the grant-friendly halo of synthetic biology.
For Collins, though, the definition is simple. It's "genetic engineering on steroids." Where biotechnology of the past used cut-and-paste tools to introduce single genes into DNA, "we're looking at introducing networks of genes and other elements under regulated control to rewire the internal workings" of the organism. Synthetic biology is not simply replicating the known genome of a bacteria, as Venter's group did; it's more the natural evolution of gene splicing once its data got big, much as theoretical biology has morphed into systems biology, or molecular biology into genomics, metabolomics, and a parade of other -omics.
Like all of biology, synthetic biology remains captive to basic scientific knowledge. All the gene-sequencing and DNA-synthesis tech in the world won't tell you how a genetic network works. Take E. coli, probably the best known organism, at a molecular level, in the world. Its DNA sequence is well mapped, and scientists understand most of what its genes do. But even for this bug, much of its overall biology remains mysterious.
"The hype is a thing that's hurt the field. ... I think it's better to be a bit sober."
The almost irrational complexity of life has been one of the biggest surprises to synthetic biologists, says Tim Lu, a former student of Collins's and now an assistant professor and medical doctor at the Massachusetts Institute of Technology. "Early on, when I came into the field, I thought we could directly translate everything I learned from electrical engineering and computer programming into biology," he says a bit wistfully. "Many people, including myself, now realize that biology, although it can be inspired by engineering, is not exactly the same."
Life is messy. That's been one of the primary lessons learned by Collins and his lab. No genetic circuit stands alone inside an organism. It's far from a silicon wafer. "Everything," Lu says, "interacts inside the cell."
In some way, Collins owes his interest in synthetic biology to Star Wars. Back in the 1990s, he met a promising student named Tim Gardner, who wanted to create bionic arms that would plug into the nervous system, replacing lost limbs. Most potential mentors brushed Gardner off, suggesting he work on their ideas instead. But then he came to Collins, who had long specialized in biomechanics—how human bodies walk or run, and how they can do it better with machine assistance. It was an interest spawned early on by his grandfathers, who never got much technical assistance for their blindness and paralysis. Collins barely blinked before he responded: "Yeah, we can do that!"
A year on, Gardner understood that his idea was impractical. "I decided we would learn how to regrow arms before we could develop arms as good as Luke Skywalker's," he says. As it happened, one of Collins's mentors at BU, Charles Cantor, was prodding him to reverse-engineer the gene networks that the Human Genome Project was uncovering. (Cantor led the project in its early days.) But such reverse engineering seemed unendingly tricky at the time. And so Gardner and Collins thought about how you'd grow an arm. The first step would be finding a way to start and stop cell growth on command. Instead of reverse-engineering a network, they would build it from the ground up.
There are two kinds of kids who become engineers, as Collins puts it. There are the ones who take stuff apart, from the lounge chair to the radio, to see how it works. Those are the reverse engineers; in biology, they'd be systems biologists. The others are tinkerers, the forward engineers, the ones who get lost in erector sets. They're always thinking of new ways to put stuff together, be it bionic arms or vibrating insoles (another of Collins's inventions, used to boost balance). The erector sets for synthetic biology—sequences downloadable off the Web; DNA simulations; a suite of cut-and-paste tools—seemed to be in place in the late 90s. "We realized the field was now in a position that we could take a forward-engineering approach," Collins says. "And really the idea was: Could you put engineering into genetic engineering?"
With Cantor's encouragement, Collins and Gardner leapt into a new, wet world. Neither was versed in the field, and the idea of reading Molecular Biology of the Cell left Collins cold. He couldn't get past the first page; too much detail. They couldn't learn it all, so instead they found a problem to solve: If they wanted to program complex tissues to grow, their cell would need a sort of memory system, just like a computer. Thinking back on his sophomore-year class in electrical-circuit design, Gardner had an idea: Inside a genome, he realized, he could build a flip-flop.
It's as good a point as any to call the inception of synthetic biology. (Other points were occurring at the same time at Princeton University.) A flip-flop is a basic electrical circuit that can be flipped by current between two states: for computers, we call them 1 or 0. The same could be done, Gardner realized, in DNA with two gene promoters, which initiate the reading of DNA, and two genes that naturally turn each other off.
Most of his peers found the idea stupid, Gardner says, but Cantor gave him and Collins the keys to his lab. Gardner pulled the lab primer off the shelf, asking simple questions: What's an agar plate? How do you make that? After four months, he started building the circuit in E. coli bacteria, hoping to light up a green fluorescent protein gene. No luck.
Collins supported Gardner with his own research slush fund, and then sent him down, in 1998, to a meeting at the Office of Naval Research. To this day an important player in synthetic biology—Darpa, the Pentagon's influential research wing, is heavily involved in standardizing the design of microbes for manufacturing—the military hoped to tap into the chatter of the 20 or so engineers and physicists talking about gene circuits and forward engineering.
At the meeting, Gardner aired all his dirty laundry, the troubles of a neophyte new to wet work. It was raw, like a lab talk. And in the middle of it, he was interrupted by Sydney Brenner, a future Nobel laureate, who sketched on the whiteboard everything he was doing wrong. "I remember thinking, 'This is how science works,'" Gardner says. "It's pretty in-your-face." Despite that setback, Collins continued to pester the Navy, calling its office once a month. It finally gave in and financed Gardner's work.
Back in the lab, Gardner continued to try his scheme in a different strain of E. coli. That did it. Something had been off with his first line. The green protein fluoresced. He had built a toggle switch.
Gardner, Collins, and Cantor described their switch in Nature, published early in 2000 back-to-back with a study by Michael Elowitz and Stanislas Leibler at Princeton, which used the same tools and genes, independently concocted, to create a three-gene oscillator. Nature's editors took note. The next week, in their lead editorial, they said something was up.
The Human Genome Project would be done soon. DNA sequencing was getting cheap. Some sort of automated DNA design software, a CAD for life, didn't seem far off. They called it quantitative biology. Soon enough, everyone would call it synthetic.
Since that early success, Collins has steadily built his lab, overtaking much of Cantor's space. He's had many protégés, like Tim Lu, an MIT engineering student who also completed a medical doctorate on the side, becoming enmeshed in synthetic biology as he saw, working the wards, how little doctors understood about the underlying molecular causes of disease. (He was also, early on, accident prone, setting the lab on fire a few times.) Given Collins's interest in human health, Lu's work centered on viruses known to prey on bacteria. They engineered viruses to shatter biofilms, the resistant homes created by microbe colonies—think plaque, or sink gunk. Then they used these same viruses to improve the killing efficiency of antibiotics.
The lab just recently published work describing a gene-circuit "breadboard" to make E. coli more modular, making it much easier to swap in genetic components than in the past. Future work involves engineering probiotics to fight cholera—they just got a grant from the Bill & Melinda Gates Foundation for that—and collaborating with stem-cell researchers on how to deploy their creations.
It's all promising work, even if it won't yet grow you a house.
As Collins's lab has grown, the ideas behind synthetic biology have traveled down different paths, seeking maturity. The notion of interchangeable, predictable genetic parts became the basis of the BioBricks Foundation; it also launched a thousand articles comparing synthetic biology to Legos. College and high-school students became engaged by the iGEM competition, pitting their designer microbes against each other, tournament-style. And venture capitalists began to take notice, as government and private money, spurred by high oil prices, began to pour into companies seeking to take simple synthetic-biology designs and scale them up.
This biofuel mandate came to dominate the field's formative days. Young scientists, including many of Collins's own students, fled the university for the private sector, their innovations disappearing behind a patent-heavy veil. "If you went to the synthetic-biology meetings at that time, the majority of those posters were around bioenergy: modified bugs to convert biomass," Collins says. The tools, the platforms—none of it was ready for the tight time horizons demanded by public companies determined to hit industrial-scale economic efficiencies.
"The hype is a thing that's hurt the field," Collins said. "Some in the community are hyping our capabilities well beyond where we're at. I think it's better to be a bit sober. I think it's fine to speak to emerging applications and emerging capabilities, but to really qualify where we are at and what we can do now, versus what might this field bring decades down the line."
To be sure, Collins is no academic purist. He's advised companies and served on scientific boards. But the uncertainties of entrepreneurship, hopped with the variability of biology, can make a bitter brew. For example, Collins and Lu spun off a company to find a market for their biofilm-busting virus. It started as Novophage, and they hoped to use the bugs for therapeutics, since biofilm-based infections are a common complication of implants. But investors demurred, fearing regulatory and clinical hurdles. Go after industrial biofilms, they said—the crud in HVAC systems and oil pipelines. Chevron invested. But then it became clear the technology was a decade from commercial use. They changed focus again, and are now using engineered phage to quickly detect food pathogens within hours, rather than days. The firm is now called Sample6 Technologies. Its first product is coming soon, they promise.
Collins and Lu remain academic scientists first, but Gardner left his faculty position at Boston University in 2007, joining Amyris, the synthetic-biology start-up famed for its creation of bugs that produce precursors for malaria medication. I reached Gardner on the phone while he was touring Amyris's new fermentation plant in Brazil, where the company is moving into specialty chemicals, while keeping a wary eye on its finances. Over the years, he said, his definition of synthetic biology has broadened beyond the computer analogies of old, the toy systems, as he calls them.
"We don't need to program memory states into bacteria and microbes," he said. At Amyris, all they need to do is build simple genetic switches. Though getting the microbes to work well is another story. "It takes a surprising amount of effort to optimize a single switch," he said.
Then, of course, there's evolution, the bug in the machine of life.
"It changes everything," Gardner said. "It changes the nature of the system you're working on. Biological systems are organically grown. They have a complex structure that isn't necessarily inherent. ... It's a strange and complicated beast. And coupled with that, the system is constantly changing."
The discipline remains far from its dream of rational design.
Collins gladly rattles off its challenges: There's a lack of well-characterized, diverse biological switches and widgets. Mostly, everyone uses about two dozen of the same genetic parts. (Certainly more exist, bottled up in the private sector.) Genome models are not where they need to be to do predictable design at the detailed level comfortable for engineers. And this jumble of components often produces unpredictable exchanges—everything interacting. "That's why the field has developed relatively slowly over this past decade," Collins says.
There are other problems, too. For all of the ink devoted to it, synthetic biology still lacks a textbook—and teachers. Collins wants his former students to go where they're happy, but he laments losing a generation to start-ups. "You had a number of young people rushing into these companies," he says. "And I think it would be better to have more of that talent staying in universities."
If the decline of the biofuel boom has one bonus, it's that some of these stars will return. (Though not Gardner, who could still be leading an industrial revolution.) And if they do, they'll come back to a field that is slowly being courted by some of the more traditional disciplines, and voices, of biology. Scientists like Phillip A. Sharp, Richard J. Roberts, and David Baltimore—all Nobel winners—are increasingly interested in using the field's tools to study their systems in isolation. Sharp, at MIT, compares the field to chemistry. A half century ago, scientists took the step of synthesizing organic molecules. With tinkerers like Collins, synthetic biology seems to be at the same point, Sharp says.
"In the long term, it'll be a source of understanding how systems can exist and be stable and respond to changes, and evolve," Sharp says. "And that's the essence of systems biology. It's the central question in biology."
Perhaps for that reason, Collins sent me down a floor to Caleb Bashor, a languid postdoc who was hanging out at lab bench, spinning bacterial cells in a centrifuge. Bashor wasn't trying to produce biofuels. He wasn't engineering drug-delivery pathways. He was dreaming about the origins of life's complexity.
Bashor was making life a genetic and environmental hell for his bacteria, a variant of E. coli. He had riddled their DNA at random with synthetic genes, each carrying a trigger known to help genes turn on and off. A toxic bath in antibiotics followed. Bashor hoped to find that the surviving cells had used his empty triggers to cling to life.
"It's sort of like using a jacked-up, loaded system to answer a question that's been out there for a while, a really fundamental one," Bashor said. "How do you get more complicated signaling networks? ... There's a feedback between discovery science and synthetic science that I think more people are starting to appreciate now."
It's modest work. The type of research Collins would like to see more often.
"I think the field's made nice progress," Collins said. "It's probably not gone as fast as many thought it would. But I'm comfortable with the pace we've made. It's still a small field. We're at the very early stages. Twelve years is not a long time."