Research Grants 1997-1998

Experiments Using a Computer Model of the Immune System

Franco Celada, MD, PhD
Hospital for Joint Diseases-Rheumatology, New York City, New York

A strategic move towards limiting the use of animal experimentation to the strictly indispensable is to set up experimentation in the computer, at a scale that permits a demonstration of both the capabilities and the limitations of the new tool. One example in point is the model of the humoral immune system that we have constructed and refined over half a decade.

Our model is a cellular automaton simulation of the immune system, incorporating the core functions of its components (effector cells, helper cells and antigen presenters). The meeting of cells and molecules is probabilistic, governed by affinity and specificity (a repertoire of binary receptors will match a population of binary determinants), and each cell obeys a realistic set of rules; from the combination of these single actions, systematic phenomena emerge.

For example, immunization by antigen results in processing, presentation to T cells, T-B cell cooperation, clonal growth, antibody formation, affinity maturation by mutation and selection, and memory. Hypotheses are tested by modifying parameters or changing structural or behavioral rules. The verification of the results will still require biological experimentation, but now limited to promising directions, with substantial economy of animal lives.

With the aim of demonstrating and expanding the capabilities of the model, we propose to focus on one central activity of immunologists, the design of vaccines. The perfect vaccine should have the identical immunogenicity of the foreign invader but none of its aggressiveness (toxicity, invasiveness). This goal is never attained because e.g. any treatment of toxins that renders them harmless invariably lowers their degree of crossreaction with the wild type. This can happen by conformational changes or chemical actions on peptides or determinants.

We are going to simulate a number of conditions that can cause this immunogenicity gap, inject model vaccines with a fine degree of difference, study the effect on the titer and affinity of the antibody response and predict counteracting measures, in terms of dose and frequency of vaccine administration. We will also set up experiments to determine the conditions that can favor the outbreak of autoimmune responses by intermolecular help unexpectedly furnished by the antigen. Finally, we will study the critical parameters that may help mutable invaders to escape even a rapidly adapting response by the immune system.