Watching Bacteria Evolve in the Lab

Using rapid new DNA-sequencing technologies, University of California-San Diego researchers followed evolutionary changes in E. coli grown under stressful conditions. They were able to identify which genes mutated, when, and what the effects were on the bacteria’s growth. The researchers say the technique, called experimental evolution, will help those trying to learn how to genetically engineer bacteria to churn out high concentrations of ethanol and other useful chemicals (see “Bacterial Factories”).

Bacteria such as E. coli evolve relatively quickly: they divide rapidly and sloppily, passing on error-filled copies of their genetic information to the next generation. Using new microarray technology, Bernhard Palsson, a professor of bioengineering, and his colleagues studied this rapid evolution over very short timescales at a high level of detail.

The advance could be particularly helpful to synthetic biologists who are reengineering bacteria to give them novel functions. “This approach will give us new insights into [organisms’] adaptive response to synthetic parts being put in,” such as novel genes or networks of genes, says James Collins, professor of biomedical engineering at Boston University. Researchers are putting in such new “parts” to better control microbes’ synthesis of a particular compound useful to humans. (For example, researchers have engineered yeast that produces a malaria drug; see “Cheaper Malaria Drugs.”)

In one set of experiments, Palsson and his coworkers provided E. coli only with glycerol, which the microbes do not metabolize very well, for nutrition. The cells grew slowly at first, but after 20 days, they grew 150 percent faster, and at 44 days they were thriving. “Those that were more fit for the environment took over the culture,” says Christopher Herring, who worked on the research as a postdoc in Palsson’s lab. “There were dramatic changes in how [well] the cells grew in a short period of time.”

Experimental evolution could prove a powerful tool for researchers working on metabolic engineering, says Herring. The genetic networks of metabolism are complex and include elements that researchers would have a hard time predicting. Herring says that experimental evolution “can show connections between different physiological systems [that we] didn’t know about before.” Herring is now a research scientist at Cambridge, MA-based Mascoma, which hopes to design microorganisms that efficiently convert biomass into ethanol (see “Redesigning Life to Make Ethanol”).

Collins says metabolic engineering is “often done irrationally.” When researchers introduce new parts to bacteria or yeast, they don’t know whether other mutations have been introduced, “let alone how other pathways may be involved.” Comparative genome sequencing could provide this kind of information, allowing researchers to better predict the effects of genetically engineered changes and to rapidly identify which changes lead to favorable attributes.

Gregory Stephanopoulos, a chemical engineer at MIT and a leader in the metabolic-engineering field, is more skeptical about the impact of the San Diego approach. Sequencing genomes and analyzing them to find relevant mutations is not the problem, he says. In Palsson and Herring’s growth experiment, it was obvious that E. coli that grew well were worth resequencing to find the relevant mutations, he says. But when working on a complex problem like improving a microbe’s ethanol-production efficiency, he says, “in some cases you can identify superior strains, but in general it’s not straightforward.”

Still, Stephanopoulos and the others say that comparative genome sequencing can now help researchers attribute changes in microbes’ traits (such as the ability to thrive on glycerol) to changes in genotype. In doing so, the technology could help microbiologists and pharmaceutical companies study how strains of antibiotic-resistant bacteria, a major health problem, emerge, and what mutations are responsible.