ABSTRACT

This study is based on the use of generalized shape as the fundamental variable to characterize the binding properties of antigens, antibodies, receptors, etc. Within this framework the investigation concentrates on a network view of the immune system with particular stress on the effects of cross-reactivity. The schematic model considered postulates cells whose probability of entering a proliferative state depends on the interplay of activating and suppressing factors, which in turn are functions of the degree of binding to corresponding sets of receptors. A one-dimensional shape space is considered. Emphasis is placed on the likelihood that the unstimulated immune system is resistant to perturbation, but not too resistant (control vs. stability tradeoff). Such a “nearly unstable” immune system is constructed with the aid of activation that is relatively short range (i.e., relatively specific in binding) compared to suppression. Surprisingly, it is found that instability, or near-instability, can also occur when activation is relatively long range. Various computer simulations illustrate the properties of the model, for example, focussing or defocussing of the antigenic input.