While most attempts to engineer biofuel-producing microbes have focused on well-known organisms such as yeasts and E. coli, scientists also hope to co-opt the unique metabolic functions of some of the microbial world’s less-studied creatures. Anthony Sinskey and his team at MIT have been cataloguing the genomic secrets of Rhodococcus bacteria, soil-dwelling microbes known to eat a variety of toxic compounds. The goal is to make a biodiesel-producing organism that can use a variety of sources as fuel. “We have done a lot of the basic chemistry and biology,” says Sinskey. “Now we need to figure out how to maximize yields.”
The strain of bacteria that Sinskey is working on, Rhodococcus opacus, is related to the type that causes tuberculosis, but it has two particularly appealing qualities. The bacteria have a flexible appetite, with the ability to eat a number of sugars and toxic compounds–in fact, the microbes were originally isolated from contaminated soil, where they were breaking down petroleum waste products. In addition, R. opacus are one of just a few types of bacteria that naturally produce a type of lipid called tryacylglycerols, which can be chemically converted into biodiesel. “Its life is focused around lipid metabolism, eating weird lipids and making more of them,” says Jason Holder, a postdoctoral researcher in Sinskey’s lab. “The trick is to engineer them to make it more efficiently, using waste streams of carbon.”
The research is part of a larger effort to develop biofuels that, unlike ethanol made from corn or sugarcane, do not rely on food sources or agricultural land. Some companies, such as Synthetic Genomics, Amyris, LS9, and Joule Biotechnologies, are using synthetic biology techniques to engineer bacteria to more efficiently produce desirable metabolic products that can be used for biofuels.
Sinskey’s team has recently sequenced R. opacus’s genome and mapped its 9,000 genes into various metabolic pathways. Understanding these pathways allows scientists to boost or inhibit specific reactions, which can in turn increase the microbe’s efficiency at creating a particular fuel or end product. The researchers have also developed a microarray for Rhodococcus–a genomics tool that allows scientists to quickly assess patterns of gene expression–and are using it to study these metabolicnetworks. “It will allow us to predict other bacteria that might do the same thing,” says Sinskey, “and it will help us identify genes important in the assembly process.” They plan to publish the genome soon.
The researchers have already created a strain of Rhodococcus that can eat a mix of two types of sugars, glucose and xylose. Once scientists have found a way to break down cellulosic biomass into simpler sugars, the ability to use more than one will simplify the production process. “They are not like wimpy E. coli that can’t use different sugars simultaneously,” says Sinskey. “These bacteria gobble them up.” The researchers have also engineered strains that can feed on glycerol, which is a waste product in the production of biodiesel.
Sinskey and his team hope to develop better ways of isolating the lipids from bacteria at a commercial scale, perhaps via additional genetic engineering. For example, altering production of a specific protein encourages the lipids to aggregate into balls, called lipid bodies, which makes the molecules easier to recover. “Ideally, we want to develop a way to make the lipid body pop out of the cell,” says Sinskey.
It’s not yet clear how long it will take to create a process that is efficient enough for commercial production. “I don’t think I’m far behind lots of companies that have lots of publicity in this area,” says Sinskey. “I think in two to three years I will have a robust process.”
Sinskey previously developed a way to make polymers from bacteria, founding a bioindustrial company called Metabolix in the early 1990s. A $300 million plant that will produce the company’s biodegradable plastic is slated to begin operations later this year in December, as part of a joint venture with agricultural giant ADM.