The human CD4 population can be divided into functionally distinct and largely reciprocal subsets based on their differential expression of CD45 isoforms (CD45RA, CD45RO) and the CD29/VLA beta chain. CD4+CD45RO+ CD29high "memory" (helper inducer) cells respond maximally to recall antigens and provide help for B cell IgG synthesis. In contrast, the CD4+CD45RA+ CD29low "naive" (suppressor inducer) population responds poorly to recall Ag, lacks helper function for B cells, but can both induce CD8 cells to suppress B cell IgG synthesis and proliferate preferentially in an autologous mixed lymphocyte response (AMLR). The phenotypic "conversion" after activation and the preferential responsiveness of CD45RA-CD45RO+ CD29high cells to recall antigen led to the view that CD45RA+ cells are "naive" and immature and convert to CD45RA-CD45RO+ "memory" cells after activation. This conversion was believed by many to be unidirectional and irreversible. It has become increasingly clear that the naive-memory concept outlined above is far from settled and that naive CD4+CD45RA+ T cells retain their unique functional program after activation and are distinct from the freshly isolated CD4+CD45RO+ subset. Moreover, CD45RA is not irreversibly lost following activation, but in fact recycles on the cell surface. Given the problems with CD45 isoform expression as a definition of maturational state, we have investigated the possibility that more reliable cell surface molecules are needed which could delineate between the functions of activated CD45RA+ and CD45RA- CD45RO+ cells. We could show that CD31 and CD27 are preferentially expressed on the CD4+CD45RA+ subset of cells and their expressions are stably maintained on these cells.(ABSTRACT TRUNCATED AT 250 WORDS)