Stroll the streets of San Francisco and you're likely to overhear someone talking about biofuels. It's the latest technology wave to hit the Bay Area, and scientists and investors are swarming toward any startup claiming a better way to make ethanol or biodiesels. Amyris Biotechnologies may actually have found one. Having previously reengineered microbes so that they would produce a malaria drug, the company is now drawing on its expertise at creating efficient bacterial factories to cheaply churn out novel types of biofuels.

Amyris is one of the first companies to spring from the relatively new field of synthetic biology. Unlike the conventional genetic engineering currently used in the manufacture of antibiotics and protein drugs such as insulin, synthetic biology involves hacking the entire metabolic system--changing the structure of some proteins, altering the expression of others, and adding in genes from other organisms--to create an efficient microbial machine. "We think of biological components as parts you assemble and try to get to function as a whole," says Jay Keasling, a bioengineer at the University of California, Berkeley, and one of Amyris's cofounders.

Plants and microbes naturally make small quantities of chemicals called terpenoids, which are the precursors of myriad products, including some pharmaceuticals and fuels. Several years ago, after developing new ways to boost bacteria's production of terpenoids, Keasling and three of his postdoctoral students founded Amyris to commercialize their work.

For its first project, the company selected artemisinin, a potent malaria drug derived from the sweet wormwood tree (see TR10 2005). By tinkering with yeast's metabolic processes, Keasling and his colleagues were able to boost its production of an artemisinin precursor a million-fold. After just two years of work, they are close to meeting their final goal for the drug--producing it in industrial quantities at prices affordable to developing nations. Now, having created microbial factories that can cheaply churn out carbon-based molecules, the group has turned its attention to biofuels.

Making fuel is different from making medicine. In most cases, pharmaceutical companies aren't concerned with how efficiently they make their drugs because they know they can charge premium prices for them. New fuels, on the other hand, must compete in price with petroleum. Rather than trying to find better ways to make ethanol--the aim of most new biofuel efforts--the researchers chose to create entirely novel biofuels, guided by their own ideas about what a fuel might look like if designed from scratch. "We looked at the Merck Index and said, If you could pick any molecule to use as fuel, what would you pick?" says Jack Newman, one of Amyris's cofounders and vice president of research.

The researchers selected several candidate compounds based on their energy content (ethanol has only 70 percent the energy of gasoline), their volatility (an ideal fuel shouldn't evaporate too fast), and their solubility in water (unlike ethanol, a water-insoluble fuel could be piped around the country like petroleum). After narrowing the list by determining which fuels could be both produced in the lab and used in today's engines, they were left with a selection of compounds including replacements for both diesel and jet fuel. "We've tested a lot of fuels with fantastic properties," says Neil Renninger, Amyris cofounder and vice president of development.

Amyris scientists are now designing metabolic pathways that yield these compounds and tinkering with them to make production as efficient as possible. "You have to walk down a cost curve of production," says Renninger. "At the bottom, you get a product so cheap you can burn it."

While the company is still a long way from having a practical biofuel, its progress will be under close watch. As ethanol is being used more and more for transportation fuel, biofuels have captured the attention of investors. Indeed, in 2001, when Keasling and colleagues first thought about making biofuels, Amyris found very little investor interest. That has changed. "We went out with the aim of raising $7 million [during a 2006 round of financing] and ended up with $20 million," says Newman. "We had to turn down multiple investors."