The Contribution of Model Systems to the Understanding of Infectious Diseases

THE CONTRIBUTION OFMODEL SYSTEMS TO THE UNDERSTANDING OF INFECTIOUS DISEASES JAMES W. MOULDER* When we mean to build, wefirst survey the plot, then draw the model.—Shakespeare , Henry IV (II, i, 3). It is a pity that the student ofinfectious diseases cannot work in such a logical fashion. Instead ofprogressing from model to final edifice, he must start with the imposing structure that is a natural infectious disease and try to figure out the plans on which it was built. Ifhe attempts to draw the architectural plan ofa natural infectious disease in cellular and molecular detail, he will surely fail. Infection ofan intact, multicellular, differentiated host produces phenomena that are too complex to be analyzed by presently available concepts and methodology. The only recourse is to models; and by models I mean host-parasite systems which are simpler than the natural infection yet retain some of its characteristics. A model that is useful in understanding the infectious disease from which it was derived should have these properties: i. Be simpler than the natural infection and therefore more easily manipulable and more likely to yield reproducible results. 2.Retain some of the important features of the natural host-parasite system. 3.Interpret some ofthese features on cellular and molecular levels. 4.Predict some aspects of the behavior of the complex host-parasite system in new situations. Unfortunately there is an unavoidable experimental conflict in achieving the first two properties. As a model is simplified, it is likely to become * Department ofMicrobiology, University ofChicago. This research was supported by a grant (??-1594) from the National Institute ofAllergy and Infectious Diseases, U.S. Public Health Service. The many contributions ofmy students^JenniferJ. Alexander, Robert R. Friis, Hsiu-san Lin, Esther M. Schechter, Mary W. Treuhaft, and Use I. E. Tribby—are gratefully acknowledged. 486 James W. Moulder · Infectious Diseases Perspectives in Biology and Medicine · Spring 1971 more stable and more experimentally amenable, but it will at the same time reproduce fewer and fewer ofthe characteristics of the natural disease . This experimental dilemma is compounded by the realization that it is not at all clear at the onset whether further simplification of a particular model will aid or hinder the attainment of the objectives represented by the third and fourth ideal properties. This is certainly not a new problem for people who study infectious diseases in the laboratory, for they have almost always worked with models. Any infection other than one in which the natural host is infected by the natural route with an infectious dose ofa size encountered in nature is by the definition I havejust given a model infection. Therefore, since almost any laboratory infection is in some sense a model, I have chosen to use one ofthe simplest possible models—a host-parasite system in which the differentiated multicellular host has been replaced by a "host" consisting of a uniform population of single animal cells. The host is an establishedline ofmouse fibroblasts, Earle's L cells, and the parasite is the meningopneumonitis strain ofChlamydia psittaci. Although this model is far removed from natural infectious diseases, I believe it is likely to be pertinent to them because the model has been derived from a disease caused by an intracellular parasite. Such parasites are the etiologic agents of the most widespread, the most devastating, and the most difficult to treat and eradicate of all infectious diseases, and there is good reason to believe that their unhappy successes as agents ofdisease stem in large part from their ability to live and multiply inside the cells oftheir hosts. The L-cell-C. psittaci model is simple; it is relatively easy to handle; and it gives reproducible results. Let us first examine its properties with the aim of deciding whether or not it reproduces significant features of natural host-parasite systems and is able to interpret and predict the behavior ofthese systems. Then, with the L-cell-C. psittaci model as a specific example, let us discuss simple models ofinfectious disease in a more general way: what fundamental problems of infection can profitably be attacked in such a system; when it is likelytofail; and inwhatways model host-parasite systems can be modified to make them more...