By David N. Leff

The unsung hero of bioscience research is named Mus musculus.

Those laboratory mice make ideal mammalian stand-ins for modeling human reproduction and disease. They¿re conveniently small, reproduce speedily, inbreed identical copies, and run through their life span from birth to old age in a mere two years.

Yet, for molecular geneticist Frederick Ausubel, that murine model posed a problem. Specifically, for his experiments on bacterial virulence he would need more than 100,000 mice. His solution was to find an animal that modeled mice, to wit the roundworm, or nematode, Caenorhabditis elegans. This soil-dwelling invertebrate, barely a millimeter long, matures from embryo to adult in 3.5 days, and lays 300 eggs at a time.

¿The way this research started,¿ explained Ausubel, a professor of genetics at Harvard Medical School, ¿was with the idea of trying to identify a human pathogen that also had the ability to infect and cause disease in one or more model genetic organisms. So that a variety of experiments could be done in these organisms with the hope they would be relevant to pathogenesis in a mammalian host, specifically a mouse.

¿The work began,¿ he recounted, ¿by identifying a strain of the bacterium Pseudomonas aeruginosa that came from an infected human patient.¿

Ausubel observed, ¿Very little is known about the mechanism by which P. aeruginosa infects humans. It¿s an opportunistic pathogen, which means it doesn¿t normally infect healthy individuals. But people who are immunocompromised ¿ cancer patients, for example, who are receiving chemotherapy ¿ are targets for Pseudomonas infection. Patients who have surgery or any severe trauma, such as burn injury, get serious Pseudomonas infection. And finally, it gives cystic fibrosis victims chronic lung infections.¿

With the human isolate of the P. aeruginosa strain in hand, Ausubel asked the experimental question: Are the same virulence factors that are required for pathogenesis in a human host also pathogenic in a model host ¿ plant, mouse, insect, nematode?

Nematode Choice Allowed For Broader Screen

¿So we did an experiment,¿ he went on, ¿that you simply can¿t do if you¿re only working with a mammalian host, which is to carry out a very broad-based, genome-wide screen for Pseudomonas mutants that were not pathogenic. To do that we randomly mutagenized the bacterium and then tested thousands of individual mutant colonies, one by one, in a model host, to see whether or not those mutants had a pathogenicity defect.

¿You couldn¿t do that with a mouse,¿ Ausubel pointed out, ¿because the total numbers of animals that would be involved in the experiment would be too large ¿ over a hundred thousand mice.¿ Which is why he and his colleagues at Harvard-affiliated Massachusetts General Hospital in Boston did the experiment in the C. elegans roundworm.

They report its result in the current Proceedings of the National Academy of Sciences (PNAS), dated March 2, 1999. Their paper, of which Ausubel is senior author, bears the title ¿Pseudomonas aeruginosa killing of Caenorhabditis elegans used to identify P. aeruginosa virulence factors.¿

To assay the virulence of several selected P. aeruginosa mutants they had generated, the team seeded cultures of them onto Petri dishes, to which they added enumerated populations of the nematodes. Five days later, they counted any worms that had survived.

For wild-type bacteria, the number was zero. The unmutated microbes had killed off all the nematodes. For mutant strains, they observed the opposite result: The nematodes had killed and eaten all the avirulent bacteria. Thousands of living, glutted nematodes teemed in the Petri dishes.

Unknown Virulence Genes Surface

¿The key here,¿ Ausubel pointed out, ¿was the identification of genes that Pseudomonas required for killing the nematode. And it turns out that those are also the same genes involved in mammalian pathogenesis. So by carrying out that very simple screen with a model host on a Petri plate we were able to identify pathogenicity factors which hadn¿t been known previously, and which are required for mammalian pathogenesis.¿

He pointed out that ¿Pseudomonas aeruginosa is the primary and most important pathogen in human burn wounds. People who study human burns and their pathogenic infection develop mouse models for testing their treatment.¿ One of the PNAS paper¿s co-authors,¿ he noted, ¿is Ronald Tompkins, chief of staff at the Shriner¿s Burns Institute in Boston.

¿The model we used to test our bacterial mutations,¿ Ausubel narrated, ¿utilizes a very small burn, less than 5 percent of the animal¿s total body surface. It¿s inflicted by pressing a heated brass rod against the outstretched skin of the mouse to create a very localized injury. Then we inoculated the wound with P. aeruginosa cultures. That simulates the infection that would happen in a human burn patient.

¿In this experiment,¿ Ausubel related, ¿most of the mutants were orders of magnitude less virulent than the wild type. So that in the case of the wild-type bacteria, an inoculation dose of 5,000 to 50,000 cells killed 100 percent of the mice. Whereas with some of the mutants, at that dose there was essentially no lethality.

¿One of the major goals of this research,¿ he emphasized, ¿is to reduce the amount of animal experimentation by finding a suitable model that will simulate a mouse. The reason we needed to do the mouse experiments at all was to demonstrate that in fact mutations we had identified using the nematode were relevant for mouse pathogenesis.¿

¿From a commercial perspective,¿ he continued, ¿this approach is an easy way to identify novel virulence factors in human pathogens, as potential anti-infective drug targets.¿

Ausubel looks forward to ¿a drug that would specifically block pathogenicity, as opposed to killing the pathogen outright. Antibiotics in general,¿ he observed, ¿essentially go after bacterial protein synthesis, DNA replication or cell-wall biosynthesis ¿ factors that the bacterium needs whether or not it¿s being a pathogen.

¿If there are highly conserved pathogenicity factors,¿ he continued, ¿irrespective of the bacterium, there¿s some chance that a drug that targeted them would have a broad range of activity against a large number of pathogens. Then these targets become somewhat attractive, because if you had a drug that targeted such a virulence factor it probably could be designed to be very nontoxic to the human host.

¿The mechanism by which Pseudomonas kills a human host,¿ Ausubel concluded, ¿is generally septic shock. We hope that by studying from a genetic perspective the way the models respond to this bacterial infection, we can address some of the questions about how mammalian hosts also respond and ultimately look at a human septic response.¿ n