Big Ideas, Big Conflicts in Plan to Synthesize a Human Genome
The biggest beneficiary of a plan to fabricate a human genome from scratch could be a Massachusetts startup called Gen9 that has close ties to the authors of the still-secretive proposal.
Two weeks ago, more than 130 scientists, ethicists, and government funding officials met behind closed doors at Harvard University to discuss a follow-up to the Human Genome Project—one that would write a genome rather than read it. The event, named HGP Write, was meant to rally interest around the idea of synthesizing all six billion DNA letters of a human genome and using the results to “boot up” a cell.
That grand goal will require new technology as well lots and lots of DNA. “We are going to be one of the companies that is going to make this possible,” says Kevin Munnelly, the CEO of Gen9, which was started in 2009 to manufacture DNA strands and whose founders include George Church, the visionary Harvard Medical School geneticist at the center of the genome-writing plans. “I don’t think anyone can make as much DNA as we can with six technicians in 15,000 square feet.”
A robot moves fluids at Gen9’s laboratory in Cambridge, Massachusetts. The startup manufactures DNA for biotechnology companies.
Constructing a human genome, analysts say, would require three times the amount of man-made genes than is currently produced globally each year and require at least $90 million worth of DNA.
A visitor entering Gen9’s Cambridge facility immediately faces a life-size poster of Church over a quote that reads, “The best way to predict the future is to change it.” That’s more or less what Gen9 was created to do. By commercializing technology to make DNA more cheaply, Church has said, he hopes to “do for biology what Intel has done for electronics.”
Currently, it’s mostly drug companies and agricultural giants like Syngenta that order precise, made-to-order gene sequences. They’re used to genetically modify plants or bacteria for purposes including production of drugs such as human insulin. But these strands of genetic As, Gs, Cs, and Ts are typically only a thousand letters long.
Better and cheaper ways of writing DNA could make it possible to print entire genomes. Want a microörganism that eats carbon dioxide and makes fuel? Someday you might be able to point and click and get one delivered. In addition to Church, the company’s other scientific founders (who have their own life-size portraits) are engineer Joe Jacobson of MIT and synthetic biologist Drew Endy of Stanford University. “Gen9 is their baby. This is their idea. This is what they have wanted to do for a long time,” Munnelly says of the three founders. “The founding principle is to promulgate synthetic biology by making it easier to adopt.”
Thanks in part to innovations at Gen9, very large-scale DNA fabrication is getting closer. In 2001, it cost $12 for each extra pair of DNA letters chemically strung onto a gene in a lab. But prices have dropped steadily. In March, Gen9 began offering prices as low as just three cents a pair.
The speed at which the price of DNA is falling is comparable to the steady improvement seen in computers under Moore’s Law, the rule of thumb that processing power doubles every two years. Yet radically faster progress is needed if the cost of writing a human genome is ever to match the price of reading one, which has declined to just $1,000 thanks to a series of big improvements in sequencing techniques. At today’s pace, that wouldn’t happen until 2066. That’s far too long for the gray-bearded Church, which helps explain his self-described “fanatic” efforts to improve the technology.
But falling prices also pose an economic dilemma if demand for DNA doesn’t increase quickly. Leading me through offices equipped with dated office furniture, Munnelly says much of the company’s capacity remains idle. In fact, a predecessor company, Codon Devices, shut its doors after selling DNA so cheaply that it lost money.
That’s why an ambitious, possibly government-supported project to make a human genome could help. It would create more demand for DNA and encourage the yet-to-be-invented technologies that will be needed to shape that DNA into complex human chromosomes. “This is going to open up a lot of people’s eyes to what the market can or can’t do,” says Munnelly. “I think those technologies could get a big boost if a project like this is announced.”
A proposal outlining the rationale for synthesizing a genome has been submitted to a journal, most likely Science, by Church and several others. The pending publication, Church says, was the reason for the attempt to keep the HGP Write meeting private and bar the press from reporting on it.
That decision, however, came under sharp criticism from Gen9 cofounder Endy, who exposed the proceedings in a critical tweet. Even though Endy owns shares in the company, and strongly agrees that better DNA construction tools are needed, he penned an editorial questioning whether “an enormous moral gesture” like creating a human genome is “an appropriate demand driver” for cheaper DNA methods. He suggested that scientists focus instead on printing bacterial chromosomes.
One concern is that making a human genome could have weighty religious and ethical implications. Although the prospect remains remote, people could in theory be designed on computers and born without parents. Others said organizers did an inadequate job of laying out their ties to startups. “Those companies would have a very large profit opportunity; they stand to benefit tremendously,” says Peter Carr, a synthetic biologist at MIT’s Lincoln Lab.
Gen9 has already been selling large amounts of discounted DNA to the coauthors of the upcoming proposal, including Jef Boeke, a researcher at the NYU School of Medicine in New York. Boeke leads an international consortium now trying to replace the chromosomes of a yeast cell with artificial ones. For its part, Church’s lab on Harvard’s medical school campus has ordered millions of DNA letters in an effort to revise the genetic code of E. coli, work that is also pending publication.
A third organizer of HGP Write, Andrew Hessel, is a futurist at Autodesk, a CAD-software company that attendees said paid for the Harvard meeting. The company, which has been spreading funding around synthetic-biology labs and is developing software for biological design, also collaborates with Gen9 and obtains DNA from it. “I’m up to my eyeballs in nondisclosure agreements with Autodesk,” says Munnelly.
Even without a mega-project, the possibility that DNA synthesis will become a big new industry has been tempting investors. A California company, Twist Biosciences, has raised more than $140 million over the last 18 months. Church says he is shareholder in that company, too.
Automation has caused the price of DNA to fall quickly. In April, Gen9 began selling DNA for $0.03 per pair of genetic letters, for large orders.
Gen9 and Twist have been breaking ground by adopting techniques from the computer industry to mass-produce billions of short pieces DNA. Twist, for instance, makes them in tiny wells micro-machined onto a silicon chip. Gen9 relies on a process of microprinting DNA strands on glass slides. Those are cleaved off and combined to form genes as long as 10,000 letters.
The new mass-production methods are replacing an older, more costly way of making genes, first developed in the 1970s. “The synthesis industry is having a renaissance at the moment, and that’s very exciting,” says Jason Kelly, a cofounder of Ginkgo Bioworks, which this year placed an order for 100 million genes from Twist, the largest purchase of its kind.
Ginkgo is searching for new ways to manufacture valuable fragrances and other chemicals. To do that, it needs copies of plant genes that it can insert into yeast cells to test, often trying thousands of variations. That’s why the price of DNA is still a problem, he says. Imagine that computer programmers, who work with 1s and 0s, had to pay 10 cents every time they sent an instruction. That would limit the number of programs they would try out. “What we really need is the ability to make many, many more individual genes cheaply,” says Kelly. “That’s the pain point; that’s what everyone wants.”
What’s less clear is why anyone would want an entire synthetic genome. “Any synthetic human genome project is going to take many years, if not decades, and any commercial return is way, way off in the future,” says Rob Carlson, a consultant who studies the industry. Without an economic motive, he’s not sure genome-writing will become affordable any time soon. “The commercial biotech use of synthetic DNA is never, ever, going to provide sufficient demand to scale up production to build many synthetic human genomes.”
Church thinks that point of view may miss the mark. “People are stuck in a rut the same way we were stuck in a rut thinking about computers in the 1960s,” he says. “You’d ask people if they want one, and they’d say ‘What’s a computer?’ It fills a whole room and doesn’t do anything.” They couldn’t imagine that transistors would lead to killer applications like Angry Birds or Facebook.
Church says engineers know from nature that there’s vast room for improvement. While chemists take three minutes to extend DNA by a single letter, molecules inside cells write DNA at 1,000 letters a second. A human cell that divides makes a new genome in 24 hours, basically for free.
“There is a growth path here,” says Church. “With each new innovation we get closer.”