Computer Analysis Accelerates Microbiology

The new tools of genome science have compelled many biologists to become, to some degree, computational biologists. At the University of Washington, bioinformatics -- the use of information systems to statistically analyze large biological data sets -- is central to investigations of the bacteria Pseudomonas aeruginosa and Agrobacterium turnefaciens.

In studying any pathogenic microorganism, two powerful new approaches provide starting points for microbiologists: genomics and proteomics. These approaches attempt to identify the genes expressed and the protein content; and determine the function of each gene and protein and how each changes under different conditions.

Bioinformatics dispenses with much formerly onerous experimental groundwork. Researchers can learn much, for example, simply by cutting and pasting a segment of a microorganism’s genome sequence into a Web browser and performing a search. Number crunching will, in minutes, detect if the snippet is similar to a sequence in some other organism where the function is well understood.

“Bioinformatics is needed to integrate the vast data sets produced by genomics and proteomics,” said Dr. Samuel Miller, professor of microbiology and of medicine, “and to correlate the results with clinical information to produce something applicable to patient therapy.”

Miller works on the Pseudomonas bacterium in collaboration with a multi-departmental consortium. His group uses functional genomics to identify changes in the Pseudomonas genes over the course of infection. In addition, using mass spectroscopy as a proteomics tool, they have defined the protein content of Pseudomonas and how it changes under conditions reflective of the environment in the airways of cystic fibrosis patients. Computers compare a database of proteins predicted by the genomics to the database of physical proteins.

Miller’s group has identified Pseudomonas glycolipids that are major determinants of lung inflammation in cystic fibrosis patients and is beginning to define the clinical stages of the infection based on shifting characteristics of the bacteria in the lungs.

Although Agrobacterium is not a human pathogen, its study nevertheless has medical implications. That’s because, in producing crown gall tumors in plants, this bacterium crosses kingdom boundaries in gene transfer.

“It’s the only known case in nature in which a bacterium’s genes are transferred to a plant and integrated into the plant chromosome so that they function as plant genes,” said Dr. Eugene Nester, professor of microbiology.

As a result, Agrobacterium has become the basis of plant genetic engineering and the creation of many genetically modified foods. Microbiologists at UW are using expression arrays and bioinformatics to identify those Agrobacterium genes required for virulence and to study the fundamental mechanism by which the genes are transferred.

Crucially, the same mechanism is used by a number of important human pathogens in the transfer of proteins, including Brucella, the bacterium that causes undulant fever; Bordetella, which causes whooping cough; Helicobacter, which causes gastric ulcers; and Legionella, which cause Legionnaires’ Disease.

UW teams played lead roles in the collaborative efforts that produced the complete genome sequences of both P. aeruginosa and A. turnefaciens. Bioinformatics is enabling researchers to translate the meaning of fast-accumulating data.