In The Lab

A Cell in the Machine: Computerized Cultures Accelerate Product Development
By Jerry W. Kram
Computer modeling rules the world. It helps engineers to create everything from the large, complicated frameworks that support 100-story skyscrapers to small, equally complicated electronic gadgets. Therefore, it was inevitable that Genomatica Inc. in San Diego, Calif., would use the power of the computer to peer into the smallest unit of life itself-the cell.

The cells of bacteria and fungi can make just about any chemical on Earth-directly or indirectly-under the right conditions. However, these organisms don't simply produce these chemicals for humanity's benefit. Therefore, the chemical pathways that lead to the production of ethanol, biobutanol, succinic acid and others are often too inefficient to do the job on an industrial scale.

Genomatica is seeking to move a step beyond the tools that nature has given us by implementing a platform to rapidly develop and implement biofactories. "We are building an integrated metabolic engineering platform to address challenges in the industrial chemical industry," says Mark Burk, senior vice president of research and commercial development for Genomatica. "The essence of the platform and where the innovation comes from is the metabolic modeling and simulation technologies. This provides us the ability to not only understand the metabolic network and the complexity of that network, but it also provides a means for interpreting data as we implement designs and strategies."

Genomatica's approach is to engineer its target organisms inside a computer, or "in-silico." Burk says the company starts by identifying the genes and metabolic processes inside organisms that are related to the production of the desired chemical product. These are fed into the company's trademarked computer modeling software SimPheny.

The company has a library of complete genomes for at least 25 economically important organisms that forms the basis of its operations. "All of our models are built on [genome] sequence data, and if you go back even five years ago, the number of organisms available by sequence and the speed with which you can get genomic sequences really opens up the door for contributions from this kind of modeling," he says.

SimPheny identifies genes for proteins that metabolize or otherwise interfere with the production of the desired product. Genomatica's technicians can analyze the likely consequences of blocking the expression of one or many of those genes. Some genetic changes will increase the desired chemical production, while others may cripple or kill the organism. The model allows Genomatica to concentrate on the former and avoid the latter. "What we are in the process of doing is building out our experimental capabilities that allow us to implement those designs and develop organisms for specific product opportunities," Burk says.

By modeling the modified organisms first, Genomatica can produce 50 or so highly promising strains for evaluation rather than taking a brute force shotgun approach. Information generated from the analysis of the strains is fed back into SimPheny to continue to refine its algorithms.

The microbiological and physiological components of Genomatica's integrated platform take organisms identified by the model and modified by the genetic engineers, and evaluate them for their use as industrial fermenters. Using a process called "adaptive evolution," strains are pushed through the application of selective pressure for high growth rates and production in a continuous fermentation environment.

Burk says Genomatica is a leader in developing the modeling of metabolic processes. The company has worked the past seven years to make its platform a practical tool for industry.

From the identification of a potential product to handing cultures over to a customer for implementation, Genomatica's integrated platform can shave valuable months off the development process, Burk says. "I believe overall we will be able to cut the development time to commercialization in half through the use of computational techniques," he says. "I believe this is a very robust platform, and it is demonstrating and validating itself with our customers right now."