The ability of pathogens to adapt to changing environments has made them a tenacious opponent. Worldwide, infectious diseases remain the single greatest cause of death, killing more than one-third of all human beings. Even in the developed world, newly emergent diseases, growing antibiotic resistance and the ever-present threat of bioterrorism haunt populations that have otherwise come to expect good health.

Along with a new recognition of the power of microbial adaptation, however, has come a great optimism among researchers. "By understanding how microorganisms such as the tuberculosis bacteria have evolved ways to persist," says William R. Jacobs, an HHMI investigator at the Albert Einstein College of Medicine, "we should be able to develop new therapies against them."

A Dangerous Versatility
Infectious diseases have not always been treated with the respect they deserve. In the 1960s, new antibiotics and vaccines seemed about "to close the book on infectious disease," as U.S. Surgeon General William Stewart famously predicted. His confidence was woefully misplaced. The sudden appearance of AIDS and Legionnaire's disease in the final three decades of the 20th century served dramatic notice that the age of infectious diseases was far from over.

Meanwhile, research advances revealed the ability of pathogens to live in a wide range of environments. Think about a Salmonella enterica bacterium that has just traveled on an undercooked piece of chicken into someone's stomach. First it must survive the stomach's extreme acidity. Then it must endure the low oxygen levels and digestive juices of the intestines. Next it must penetrate the wall of a cell in the intestine and keep itself from being dissolved by the cell's lysosomes, which are constantly on the lookout for bacterial invaders. Finally, it must exude its progeny from the cell so that they can infect new hosts.

Any organism with such a complex life cycle must be quite a sophisticated survivor. "Microbes can't anticipate," says Eduardo A. Groisman, an HHMI

investigator at the Washington University School of Medicine. "They react to the environment. Those like Salmonella that live in multiple environments have the means to gather information about changes in their environment and use this information to activate or repress certain subsets of genes at the right time."

Microbes that spend their lives in less varied environments do not require such elaborate genetic controls, Groisman points out. "If you're just eating soup, you only need a spoon—not a knife and a fork." Many of these organisms have shed parts of their genomes over time and rely on their hosts for essential metabolites, while organisms that live in more complex environments have larger genomes and a greater percentage of their genomes is devoted to regulatory functions. "This trend toward more complex regulation goes hand in hand with versatility," says Groisman.

Many pathogens also have evolved an ability to manipulate their hosts to maintain a favorable environment. HHMI investigator Ralph R. Isberg and his colleagues at the Tufts University School of Medicine have been studying how the bacterium Legionella pneumophila, which causes Legionnaire's disease pneumonia, avoids the normal defense mechanisms in the lung. They've found that the bacterium produces a set of proteins that allow it to live within human macrophages and not be exposed to the antimicrobial agents with which macrophages normally kill bacteria.

L. pneumophila's ability to survive in different environments has contributed greatly to its pathogenicity, Isberg says. Until humans began constructing machines that harbored dark, warm reservoirs of water, the microbe was probably a harmless freshwater dweller. The man-made reservoirs gave the bacteria a new place in which to grow, and when water from these reservoirs was sprayed into the air, the microbe found its way into human lungs. "Now most sources of infection are from aerosols from plumbing systems that are inadequately cleaned or air-conditioning systems that aren't working properly," says Isberg.

Beware of the Bacterial Internet
Microbes have a critical asset, a genetic flexibility unknown in multicellular organisms, that they can use in colonizing new environments. Because bacteria multiply very quickly—once every 20 minutes for the common laboratory microbe E. coli in optimal conditions, compared to once every 20 or so years for humans—they must repeatedly copy their DNA. Mistakes in the copying process introduce genetic changes that can alter the structure or expression of bacterial proteins. As a consequence, the microbes are continually spinning off new genetic variants that have multiple copies of genes, genetic deletions or altered patterns of gene expression.

These variants then are exposed to an especially stringent form of natural selection. If a particular variant is disadvantaged in competition with others, it will die and that genetic lineage will eventually be lost; but if the variant has an advantage in that environment, it quickly begins to make copies of itself. Furthermore, when a microbe reproduces, it does not have to dilute its genes with those of a partner, as do organisms that reproduce sexually. It can keep making copies of itself until an environmental niche is filled, at which point it can begin looking for new niches.

Many microbes combine the ability to generate new microbial mutants with another key asset: They can acquire DNA directly from other organisms. Sometimes this DNA travels in viruses or across cytoplasmic bridges between bacteria. In other cases, a bacterium simply laps up DNA that lies exposed in the environment, whether from a decomposing microbe or even from a plant or animal cell. This "horizontally acquired" DNA can range in size from short snippets of genetic material to what are called pathogenicity islands—long stretches of DNA containing many genes that encode potent virulence factors. "Bacteria have their own Internet," says the University of British Columbia's Finlay. "They can download each other's genetic sequences."

These downloads can have dire consequences. In 1982 an outbreak of severe bloody diarrhea sometimes accompanied by kidney failure was traced to undercooked hamburger from a particular fast-food outlet, and examination of meat samples revealed the culprit: a new and highly virulent form of E. coli dubbed O157:H7. The new strain differed from the laboratory strain of E. coli in two main ways. "One was that it had a classic pathogenicity island," explains Finlay. "The other change was the acquisition of a stretch of DNA that codes for a Shiga toxin, found in Shigella dysenteriae. The Shiga toxin kills cells that then plug up the kidneys, so the victim gets kidney hemolytic uremic syndrome. This was a bacterium that had put two preexisting virulence factors together to yield a new pathogen."

Finlay has been studying how O157 subverts the normal functions of human cells to its own ends. It uses a "type III secretion system" to inject proteins, including its own receptor, into the epithelial cells of the intestine. As the bacterial proteins hijack the human cell, they cause it to erect pedestals on which the pathogenic microbes sit, spewing toxins. Surprisingly, these type III secretion systems are common not only in human pathogens but in animal and even plant pathogens, suggesting that they have been passed from microbe to microbe through horizontal gene transfer.

Horizontal transfers of DNA were recognized decades ago, in part because antibiotic resistance was spreading among microbial strains far too quickly for that resistance to be evolving de novo each time. But only in the past few years have microbiologists come to recognize the extent to which horizontal gene transfer has shaped the microbial world. "We used to think that evolution was a slow, constant process—you'd change a base at a time and one protein would slowly segue into another," says Finlay. "What really turned the tide was the recognition of pathogenicity islands, these huge chunks of DNA with virulence factors lined up side by side. Slow evolution is still occurring as these systems are tried and improved in various places before being transferred around. But the recognition of large-scale chromosomal changes gave us a new view of how evolution works."

Researchers are now in a position to gauge the degree of horizontal gene transfer through the use of a powerful new tool: the ability to sequence the complete genomes of microbes. They have found such transfers to be surprisingly extensive. For example, Pascale Cossart, an HHMI international research scholar at the Pasteur Institute in Paris, has devoted her career to studying Listeria monocytogenes, a foodborne pathogen that causes hundreds of deaths and miscarriages in the United States each year. Recently, she and her colleagues compared the complete genome of L. monocytogenes with that of the related strain L. innocua, a harmless microbe also found in food and the environment. The pathogenic form of the microbe, they discovered, has 270 genes that the nonpathogenic form does not, clustered in 100 islands scattered across the genome. The nonpathogenic L. innocua, in contrast, has 149 genes that are not shared with its pathogenic cousin. "Multiple episodes of gene acquisition and deletion" must have occurred to produce this genetic collage, Cossart concludes.

Complete genome sequences also are revealing the much deeper evolutionary relationships among microbes. For example, Washington University's Groisman and researchers Howard Ochman of the University of Arizona and Jeffrey Lawrence of the University of Pittsburgh have been studying the evolutionary process by which various microbes have descended from a common ancestor. "People have been saying that we can't trust phylogenies because of lateral gene transfer," Groisman says, "but I don't think that's the case. If you look at conserved genes, you can still derive phylogenies that are pretty accurate."

These evolutionary relationships are of more than academic interest, researchers say. Uncovering the historical links among pathogens is likely to produce many new ways to diagnose, treat and prevent infectious diseases. At the Albert Einstein College of Medicine, for example, Jacobs has been seeking to exploit the evolutionary history of Mycobacterium tuberculosis, the respiratory pathogen responsible for more deaths worldwide than any other microbe. "The amazing thing about tuberculosis is that it has evolved the ability not only to grow in the lungs but to hang out once it encounters an immune response," Jacobs says. "If you're a pathogen, you don't want to be wiping out all the humans in the world, because there goes your host. A more effective strategy is to go in, dance with the immune system and stay for a lifetime.

"I had an uncle, 84 years old, who died last week of TB. He probably had been infected for 50 years. Then he got a lymphoma that wiped out his immune system and, boom, his tuberculosis was reactivated. Once the immune system starts to go, the pathogen knows that it has to find somewhere else to hang out. So it causes an overt infection that leads to coughing, and someone nearby is infected."

The tricks that microbes use to infect hosts and exert their effects often suggest new treatments, Jacobs says. For example, he and his colleagues have developed a mutant strain of M. tuberculosis that lacks the ability to construct the elaborate surface molecules the pathogen uses to confuse the human immune system. "By understanding how TB keeps itself from being killed by an immune response, we should be able to design novel interventions—vaccines or therapies—that will be able to eliminate TB."

Furthermore, the therapeutic approaches developed through such research may be widely applicable. Because many pathogens have shared their genetic secrets through horizontal gene transfer, new treatments may work for multiple pathogens. For example, Finlay has been examining the type III secretion system as a possible target for controlling many different kinds of infectious agents. "Ten years ago this field was in chaos because we thought that different bacteria did things completely differently," he says. "Now we've come to realize that pathogens often use common processes."

Still, no one expects that we'll win the war against infectious diseases anytime soon, if ever. Evolution will continue to produce new genetic variants—and surprises. "We'll never get rid of infectious diseases," says Finlay. "There will always be new ones."

Photos: (From top) Brett Finlay and Leigh Knodler, ©CNRI/Phototake

Illustration: Nature 405:299-304. fig4.(2001) ©MacMillan Publishers Ltd.

Download this story in Acrobat PDF format.
(requires Acrobat Reader)

Reprinted from the HHMI Bulletin,
March 2002, pages 26-29.
©2002 Howard Hughes Medical Institute