The response of T cells to antigen shows an amazing degree of both sensitivity and specificity, with a cell responding to 1-10 peptide-MHC complexes and being sensitive to single amino acid substitutions. Kinetic proofreading or feedback pathways achieve specificity at the level of the receptor, whereas serial engagement of receptors by ligand molecules enhances sensitivity. Crosstalk between receptors, integration of signals and/or tuning of responses is important at the level of the cell. Induction of anergic or regulatory cells by suboptimal stimuli prevents cell activation by multiple encounters with weak ligands. Thus, for optimal sensitivity and specificity, it is necessary to have mechanisms that operate at the level of the receptor, the cell and finally, the population of responding cells.

The ability of the immune system to respond by ridding a pathogen without debilitating the host depends upon the ability of the effector Th (eTh) to make a discrimination between self? and non-self? antigens. This ability is somatically learned and involves the sorting of the somatically generated random repertoire of initial state Th (iTh) into two classes of specificity: one, anti-self, the functional expression of which must be inactivated; the other, anti-non-self, the functional expression of which must be activated. We propose a model for the origin of a sufficiency of eTh anti-non-self and an insufficiency of eTh anti-self based on two postulates. (i) An antigen-independent pathway to a priming level of eTh anti-non-self under conditions where iTh anti-self are effectively deleted by interaction with self. This state is established during a window of fetal development and maintained throughout life because self is persistent. (ii) Associative recognition of antigen (peptide-MHC class II) on an antigen-presenting cell between iTh and primer? eTh that results in the rapid induction of an effective level of helper activity to non-self antigen. A computer simulation is provided that enables evaluation of this model. 10.1093/intimm/dxf078

The problem of protecting computer systems can be viewed generally as the problem of learning to distinguish self from other. We describe a method for change detection which is based on the generation of T cells in the immune system. Mathematical analysis reveals computational costs of the system, and preliminary experiments illustrate how the method might be applied to the problem of computer viruses. 1