The next step is to establish whether the amplified DNA that comes from that organism is present in all cases in which there are similar symptoms of disease. The first unknown organism to be identified in this way was the bacterium that causes a rare intestinal disease called Whipple’s disease (Tropheryma whipplei). Currently, this approach is being used in an attempt to identify bacterial pathogen(s) thought to be responsible for other diseases with unknown causes (i.e., etiology), such as bacterial vaginosis, atherosclerosis, and inflammatory bowel disease.
Insights into Pathogen Evolution
A current preoccupation of many infectious disease researchers interested in deciphering the root causes of pathogen evolution and the dynamics of epidemics is to combine epidemiological and evolutionary knowledge about pathogen virulence gleaned from genomics. Application of mathematical modeling is then used to gain better understanding of pathogen physiology, ecology, and disease transmission. In formulating these models, the two most common assumptions are that pathogens evolve in response to selective pressures placed on them by their environment, namely the host—and, as we are beginning to learn, also the external environment—and that virulence (i.e., the deleterious effects of the pathogen on its host) is directly proportional to its ability to be transmitted.
Recent studies along these lines have provided new insights into the potential effects of imperfect vaccines and the reemergence of certain diseases once thought eradicated (or at least well-controlled) through vaccination, such as whooping cough. It has been a generally accepted premise that infection with one strain of a bacterial pathogen will significantly reduce the susceptibility of the host to subsequent infections with other related bacterial strains due to the host’s acquired immunity. However, these recent studies now indicate that the selective pressure of immunization may also be driving the evolution of pathogens like Bordetella pertussis, the bacterium responsible for whooping cough. Genomic sequence comparisons indicate that Bordetella strains, against which most of the current vaccines were designed and used extensively in developed countries, have now been replaced in the population by novel variant strains lacking key components previously recognized by the immune system to help clear the pathogen from the body. Experiments in mice further suggest that the existing vaccines are less protective against some of these new variants.
Modeling the Host-Pathogen Interaction in Experimental Animals
Studies of disease-causing bacteria growing under laboratory conditions need to be supplemented by studies in animal models. The most familiar type of infectious disease model is the laboratory rodent. Inbred strains of mice and rats are widely used as models for the infection process. The availability of animals with known genetic mutations in their defense systems has increased the utility of these rodent models. Numerous examples of the use of these models will be seen in later chapters of this book, including discussions of when it is appropriate to use animals in experimental schemes to study infectious diseases and the ethics of using vertebrate animals in experiments.
A new wrinkle on the animal model story is the increasing range of “animals” used, from the nematode Caenorhabditis elegans to the fruit fly Drosophila melanogaster and the zebra fish Danio rerio. Certainly, one motivation for using such models is the fact that the complex body of regulations and restrictions that has built up around the use of laboratory rodents and other warm-blooded animals is so complicated and expensive that only the best-funded laboratories can use them. However, more compelling reasons for the use of these new models are that so much is known about their genetics and that they are much more easily and rapidly genetically manipulated than mammals.
In the case of C. elegans, for example, the developmental origin and fate of every cell in the whole organism is known. Drosophila has a long history of use as a model for insect and human genetics, and many characterized mutants are available. In fact, a type of receptor on human neutrophils that is important for responding to bacterial infections (Toll-like receptors) was first discovered in Drosophila (Toll receptors). The zebra fish is a newer infection model, but it also has some of the same attractive features as the nematode and fruit fly models (e.g., small size, easy maintenance, short generation time, and ease of genetic manipulation). Zebra fish have the added advantage that they have somewhat more advanced host defense systems than the nematode and fruit fly and are thus a better model for the human immune system, particularly with regard to aquatic pathogens.
Given the genetic distance between these animals and mammals, a certain degree of care must be used in choosing the experimental questions and interpreting the results. For example, although nematodes, fruit flies, and zebra fish have phagocytic cell defenses that exhibit some similarities to that of humans, the systems are not identical and are evolutionarily distant from mammals. For example, insects and worms lack adaptive immune responses as are found in humans and other mammals, and so cannot be used to study antibody responses and inflammation. They also lack many of the immune signaling components present during infection in mammals. While zebra fish are genetically tractable vertebrate models with complete adaptive immune responses in adults, only a few of the immune components have been functionally studied thus far. Indeed, recent studies have also revealed questions regarding the evolutionary conservation of some of the processing and activation of inflammation in zebra fish. Nonetheless, these simple models can be used to generate hypotheses that can later be tested in laboratory rodents or other animals and, in some cases, humans.
Correlation Studies
Another type of modeling that has been used for a long time in epidemiological studies but is relatively new to pathogenesis studies is the statistical analysis of not just microbial populations, but also human and animal populations. At present, this type of modeling is still rather unsophisticated and based on seeking correlations between traits of the organism and outcomes of disease. In other words, the model can be used to ask whether the production of a particular protein is associated in a statistically significant fashion with various aspects of the disease progression in humans. This approach has the advantage of being easy to do because one merely needs to apply preexisting statistical methods. There are, however, two rather serious problems with this approach.
First, this kind of “modeling” is not modeling in the sense that this term is used in physics or chemistry, in which principles are first expressed mathematically in a way that generates specific predictions about the outcome of an experiment. Instead, correlation studies are usually performed without any clear idea of a theoretical connection between the parameters being tested. As such, finding a correlation does not prove cause and effect. There is an urban legend that illustrates this. A gentleman in California happened to pull down a shade in his apartment just before the onset of a particularly severe earthquake and remained convinced for the rest of his life that pulling down the shade had helped to cause the earthquake.
A second problem is that the items to be checked for correlation are chosen by the researcher, and there may or may not be some theoretical underpinnings to the choice. These problems do not necessarily make the correlation studies inappropriate, but these issues do emphasize the need for scrutiny. If the approach is treated as one for potentially generating hypotheses rather than as a method that provides a proof of cause and effect, then the objections disappear. As more mathematicians, physicists, and bioinformatists are becoming interested in applying their tools to study infectious diseases, more sophisticated modeling approaches are beginning to emerge.
As important as new technologies have been, the most important advance has been a new appreciation for the importance of focusing not just on the properties of a bacterium in a test tube, but also on the myriad ways in which the bacterium interacts with its environment and stimulates responses from the human body. In this book, we will place great emphasis on this bacterium-host interaction. It will become clear very quickly that although considerable progress has been made, there is much to be learned and many opportunities for readers of this book to participate in future research in the area of bacterial pathogenesis.
SELECTED READINGS
Ahmed N, Sechi LA. 2005. Helicobacter pylori and gastroduodenal pathology:
