Thymocyte subpopulations: an experimental review, including flow cytometric cross-correlations between the major murine thymocyte markers

Abstract

Many markers have been used to define thymocyte subpopulations. The literature gives discordant values on their relative proportions, and suggests that many thymocytes must have phenotypes intermediate between those of the dominant subsets. To clarify these issues, a reassessment of murine thymus subpopulations has been made, using internally consistent, quantitative correlations of most of the established markers. Peanut agglutinin (PNA) receptor, Thy 1, H-2, TL, Ly 1 and Ly 2, have been examined and correlated with each other, and with cell size, physical parameters, cortisone sensitivity and anatomical location. The analysis utilised mainly monoclonal antibodies and flow cytometry but also included differential complement-mediated cytotoxicity. The results show that there are two clearly defined major subpopulations, medullary cells (15% of the total) and cortical cells (85% of the total). These are most clearly distinguished by the combined use of PNA and Thy 1 markers, medullary cells (like peripheral T cells) being low for both, cortical cells being high for both. Cortisone-resistant cells represent about 25% of all medullary cells, which indicates that most medullary cells, as well as all cortical cells, are cortisone sensitive. Amongst the non-dividing thymocytes there are few cells that can be considered of intermediate phenotype when a multiparameter approach is used. The situation for the dividing blast cells is more complex, with some displaying apparent 'intermediate' marker combinations. However, three major subdivisions of blasts stand out, namely medullary blasts, cortical Ly 1+2+ blasts and cortical Ly 1-2- blasts. The cortical population in general contains only Ly 1+2+ and Ly 1-2- cells, but no (or few) Ly 1+2- cells. In contrast, the medullary population contains both Ly 1+2- and Ly 1+2+ cells, in similar proportions to peripheral T cells. An important conclusion of the study is that although no single marker can give an adequate definition and clean separation of subpopulations, quantitative multiparameter analysis can achieve these objectives.