Quantitative analysis of the CD8+T–cell response to readily eliminated and persistent viruses

Abstract
The recent development of techniques for the direct staining of peptide–specific CD8+T cells has revolutionized the analysis of cell–mediated immunity (CMI) in virus infections. This approach has been used to quantify the acute and long–term consequences of infecting laboratory mice with the readily eliminated influenza A viruses (fluA) and a persistent γherpesvirus (γHV). It is now, for the first time, possible to work with real numbers in the analysis of CD8+T CMI, and to define various characteristics of the responding lymphocytes both by direct flow cytometric analysis and by sorting for furtherin vitromanipulation. Relatively little has yet been done from the latter aspect, though we are rapidly accumulating a mass of numerical data. The acute, antigen–driven phases of the fluA and γHV–specific response look rather similar, but CD8+T–cell numbers are maintained in the long term at a higher ‘set point’ in the persistent infection. Similarly, these ‘memory’ T cells continue to divide at a much greater rate in the γHV–infected mice. New insights have also been generated on the nature of the recall response following secondary challenge in both experimental systems, and the extent of protection conferred by large numbers of virus–specific CD8+T cells has been determined. However, there are still many parameters that have received little attention, partly because they are difficult to measure. These include the rate of antigen–specific CD8+T–cell loss, the extent of the lymphocyte ‘diaspora’ to other tissues, and the diversity of functional characteristics, turnover rates, clonal life spans and recirculation profiles. The basic question for immunologists remains how we reconcile the extraordinary plasticity of the immune system with the mechanisms that maintain a stable milieu interieur. This new capacity to quantify CD8+T–cell responses in readily manipulated mouse models has obvious potential for illuminating homeostatic control, particularly if the experimental approaches to the problem are designed in the context of appropriate predictive models.