Stalking Infectious Disease

Graphic: Cathleen Heard Law enforcement officials routinely use DNA fingerprinting as a tool to get the guilty punished. Defense lawyers often rely on the same tool to free the innocent. Though their labs may be less dramatic settings than criminal courtrooms, life science researchers also use DNA fingerprinting, but rather than capturing criminals, their goal is to keep tabs on a different sort of culprit: infectious disease. The technology has revolutionized the way diseases are tracked and the way the epidemiologist, molecular biologist, and public health official interact. Infectious disease researchers use DNA fingerprinting--a collection of methods for detecting organisms' unique DNA patterns--to track and confirm the spread of diseases. Fingerprinters usually compare either the genetic sequences of organisms or the size of nucleic acid fragments that have been digested with special enzymes. Fingerprinting has helped in investigating any number of organisms. Researchers use it to study the resurgence of vaccine-preventable diseases such as whooping cough and rubella. They use it to track unusual cases of HIV transmission. And they've employed it to find and stymie disease outbreaks caused by foodborne pathogens, which typically have short incubation periods--sometimes overnight. For other pathogens, particularly tuberculosis (TB)--which can take 10 years to manifest and is passed through the air rather than through food--the technology, though still integral, is part of a more complicated detective story. Foodborne Pathogens After successfully using a DNA fingerprinting method called pulsed-field gel electrophoresis (PFGE) to respond to a 1993 outbreak of foodborne illness caused by Escherichia coli O157:H7-contaminated meat in the western United States, the Centers for Disease Control and Prevention (CDC) decided to set up a national network of public health DNA fingerprinting laboratories to expand its coverage and boost response time. Called PulseNet, the network of laboratories has allowed rapid comparison of bacterial fingerprint patterns among investigators by standardizing fingerprinting methods. In about one day, PulseNet laboratories isolate DNA from the bacteria, cut the DNA with restriction enzymes to a manageable number of large fragments (10-30 fragments ranging in size from 30 to 800 kilobases), and separate the large fragments by PFGE to obtain DNA "fingerprint" patterns. After the patterns are normalized by a computer, CDC scientists monitor the remotely submitted prints to track outbreaks. In the last few years, PulseNet, according to PulseNet principal investigator Bala Swaminathan, chief of the foodborne and diarrheal diseases lab section at the CDC, has decreased the average outbreak size by nearly two-thirds. He notes several successes, including the defusing of a multistate Shigella outbreak in 1998; PulseNet helped to quickly trace the source of the bacteria to parsley grown in Mexico that had been washed with contaminated water. PulseNet now monitors not only E. coli 0157, but also Salmonella, Shigella, and Listeria. The PulseNet server at the CDC is currently being programmed to compare submitted "fingerprint" patterns without CDC intervention. And the CDC plans to bring more states on board. Canada has already started to participate in PulseNet, and several European countries have expressed interest. As for nonfoodborne bacteria, preliminary plans are under way to incorporate fingerprinting efforts on conditions such as meningitis. However, longer incubation periods make it difficult to track and immediately address outbreaks for many such bacteria and viruses. In these cases, DNA fingerprint tools are usually better suited to confirming that a particular strain is on the loose, and tracking the spread of that disease after the fact. Tracing TB Particularly challenging has been tracking TB, a serious health concern in places such as China, India, and Russia, where multidrug-resistant strains continue to spread. In the 1980s, researchers developed a TB fingerprinting method called restriction fragment-length polymorphism (RFLP) based on the insertion sequence IS6110. Two isolates are thought to belong to the same strain if their IS6110 sequences appear the same number of times and in the same number of places. Though the technology is quite helpful, it doesn't usually engender prevention. The bacteria grow slowly--it takes at least two to three weeks before the victim knows he's infected. Even persons who are infected may never come down with the disease. A full one-third of the world's population has the TB bacteria, yet only 5-10 percent of those infected will ever get TB, for reasons not completely understood. Nevertheless, fingerprinting has for years been useful in confirming TB outbreaks in hospitals, where suppressed immune systems and a high-stress environment help awaken dormant bacteria. The technology offers clear evidence, important when hospital officials don't want to admit that the bacteria had spread from within the hospital. According to Christopher Braden, a medical epidemiologist in the division of tuberculosis elimination at the CDC, only when confronted with DNA fingerprint data--in conjunction with hospital records documenting that patient X had been in the same room with patient Y--would some hospitals accept that they had a problem. Much larger population-based studies have attempted to identify recent or potential upcoming TB outbreaks with mixed results. In the early 1990s, researchers at Stanford University and the University of California, San Francisco, used fingerprinting to type the TB isolate of every TB-infected person in San Francisco over two years. The studies suggested that despite a supposedly efficient tuberculosis-control program, nearly a third of new cases of tuberculosis in San Francisco were the result of recent infection, something that conventional contact tracing had missed. Braden, though, points out that even here, such results could be misleading: The technique may not be specific enough, and because the bacteria can remain latent for so long, investigators may be detecting the remnants of small outbreaks or clusters that occurred years ago. "Now the question is, what proportion of the clusters is due to the fact that the genotype is wrong, is due to an outbreak that happened many years ago, or is due to a recent transmission?" explains Braden. In special cases, though, the technique does paint a clearer picture of what the bacteria has wrought in a large population. A quick-lived multidrug-resistant TB outbreak in New York City in the early 1990s spread via the HIV-infected, whose immune systems are typically weak. According to Barry Kreiswirth, director of the Public Health Research Institute (PHRI) Tuberculosis Center in New York, which was founded in 1992 in response to the outbreak, in places where TB is endemic, the activity of the otherwise dormant bacteria may actually indicate HIV infection. The New York outbreak essentially burned itself out because most of the afflicted died. According to Kreiswirth, though, over the last 10 years investigators have seen its lingering effects, and 150 cases of the same strain have been documented since 1995. DNA fingerprinting has enabled epidemiologists and molecular biologists to trace this unseen world of silent TB transmission not only through time, but through space as well. Strains common in Russia, where TB has reached epidemic proportions, have recently turned up in the United States, often in people of Russian descent--possibly the result of a more open Russian society since the fall of communism. Using fingerprinting, researchers can actually track the disease, and therefore the migration of people, over the last 100 years. The trickling of foreign strains into nonendemic countries could be an indication of big problems to come, particularly if the strains are resistant to multiple antibiotics. "The World Health Organization, the World Bank, and global society have to be concerned," remarks Kreiswirth. He adds that major pharmaceutical companies are hesitant to engage in the costly and laborious task of developing new antibiotics, since the major markets continue to be in cash-poor Third World countries. Says Kreiswirth, "Someone is going to have to somehow give drug companies an incentive." New Technologies More than just new drugs, the key to better treatment includes new diagnostic tools that could supplement or even improve fingerprinting techniques. A technique called spoligotyping could, suggests Braden, help resolve some of the potential pitfalls in population-based fingerprint studies. Spoligotyping--an increasingly popular PCR-based technique that Kreiswirth calls "a poor man's DNA fingerprinting"--is less specific but much quicker and easier to carry out. Some studies have suggested that the method could effectively be used in conjunction with the more specific, more difficult fingerprinting as a secondary typing method--much like the two-step ELISA (enzyme-linked immunosorbent assay)-western blot HIV test. Perhaps even more promising are emerging DNA microarray technologies, which can simultaneously detect the expression patterns of thousands of genes. Among the uses for microarrays: the ability to detect bacterial drug resistance and to determine which vaccines are ineffective and why. Microarray enthusiast Peter Small, an assistant professor of medicine in the division of infectious diseases at Stanford University, and one of the lead investigators in early 1990s population-based TB studies in San Francisco, has recently done comparative genomics studies of Bacille Calmette-Guerin (BCG) vaccines, a live attenuated strain of bacteria used to treat tuberculosis.2 His team is trying to understand why it's not nearly effective as it once was. Microarray studies revealed that five regions representing 38 open reading frames were present in Mycobacterium bovis but absent from some or all of the BCG strains used in clinical trials. Now his group is using the arrays to look at clinical strains of TB to try to detect variability within the organism. "The central hypothesis of the current work is that there is strain-to-strain variability in the organism that accounts for some of its clinical and epidemiologic behavior," explains Small. "It's a whole new perspective on tuberculosis pathogenesis." PulseNet investigators at the CDC are also looking at new, faster, more discriminating typing technologies, including microarrays; Swaminathan emphasizes that the methods must be broadly applicable, robust, and backward compatible with previous technologies. However, as with the development of new antibiotics, cost is an issue, especially in the case of large population-based studies. Microarray chips are expensive and require labs with high technical capacity, but Small remembers when the CDC didn't think large-scale DNA fingerprinting was feasible either: "The big question is, does this type of analysis yield important information? And if it does, then there are ways to get the prices down." Eugene Russo can be contacted at erusso@the-scientist.com. References 1. P.M. Small et al., "The epidemiology of tuberculosis in San Francisco: a population-based study using conventional and molecular methods," New England Journal of Medicine, 330:1703-9, 1994. 2. M.A. Behr et al., "Comparative genomics of BCG vaccines by whole-genome DNA microarray," Science, 284:1520-3, May 28, 1999. New Technologies, New Ways of Thinking Though DNA fingerprinting and other technologies have revolutionized the way scientists track diseases, no technology will ever replace old-fashioned fieldwork, says Christopher Braden, a medical epidemiologist in the division of tuberculosis elimination at the Centers for Disease Control and Prevention. An intense cluster investigation based on interviews and a review of records is still essential. And if the efficiency and effectiveness of these technology-assisted investigations is to be maximized, epidemiologists, molecular biologists, and public health officials will have to accept new roles. New technologies are challenging long-standing conventions of epidemiology. Bala Swaminathan, chief of the foodborne and diarrheal diseases lab section at the CDC, suggests that PulseNet, a network of labs tracking foodborne pathogens using DNA fingerprinting, has both encouraged and necessitated better collaboration and data exchange between the lab scientist and the epidemiologist. "The technology is so new that we are crossing other people's turf," says Barry Kreiswirth, a molecular biologist and the director of the Public Health Research Institute (PHRI) Tuberculosis Center in New York. He emphasizes that these tools are useful in understanding the social side of disease transmission. In a recent study, he and his colleagues used fingerprinting tools to track similar TB strains in a defined population in one area of New Jersey,1 but public health official etiquette says contacting those persons directly is a big no-no, a clear breach of confidentiality. Scientists have to somehow intertwine "hard" molecular biology data with epidemiologists' contact tracings. "We now know what questions to ask with this population, but we're not allowed to," says Kreiswirth. "There's been quite loud battles [asking] what to depend on." "When you get epidemiologists and molecular biologists together, they ... deterioriate into a discussion of why the discrepancies reflect problems in one [investigation vs. another]," claims Peter Small, an assistant professor of medicine in the division of infectious diseases at Stanford University, and one of the lead investigators in early 1990s population-based TB studies in San Francisco. "Or they get into these questions of how can we exploit these discrepancies to better understand where I might be wrong." He attributes his group's success to the integration of conventional epidemiology and molecular biology from the outset and the close cooperation of San Francisco's public health officials. "We had these fantastic meetings where ... the topic of conversation would go from 'Well, I don't really think that's an extra band, I think it's a partial digest' to 'Where do gay transsexual male prostitutes congregate in San Francisco?'" Investigators are still, it seems, trying to keep up with the far-reaching implications of the new technology. "The ultimate point ... is to lessen the burden of disease," comments Kreiswirth. "And I think that's sometimes what we lose sight of." --Eugene Russo 1. P.J. Bifani et al., "Identification of a W variant outbreak of Mycobacterium tuberculosis via population-based molecular epidemiology," JAMA: The Journal of the American Medical Association, 282:2321-7, Dec. 22, 1999.

Stalking Infectious Disease

Interested in reading more?

Become a Member of