Problems in the physiology of class I and class II MHC molecules, and of CD45.

1. Co-processing of alloantigens suggests that epitope-loaded MHC class I molecules may pass from tissue cells to dendritic cells. 2. Antigen-presenting cells in the thymus need some special trick in order to load their MHC class II molecules with epitopes from "intermediate concentration" self-proteins in order to induce self-tolerance in developing cells. 3. Cell-cell interactions may transmit signals simply by rearranging surface glycoproteins and thus locally perturbing a phosphorylation equilibrium. 4. The CD45 and STB1 phenotype of most cells in the thymus may be characteristic of a doomed cell.