A panel of B lymphoid-reactive monoclonal antibodies was used to analyze the phenotypic changes that accompany B lymphocyte development in normal human bone marrow. The B lymphoid cells were identified using light scattering and the expression of CD19 on a flow cytometer. Quantitative three-color immunofluorescence was then used to correlate other cell surface antigens on these cells identified as B lymphoid in normal marrow. CD10 and CD20 identified almost exclusive populations and provided a convenient means of discriminating between the less and more mature B lineage cells. The CD10+ cells could be further subdivided using CD34. The population of CD19+, CD10+, CD34+ cells comprised only 0.6% of marrow cells, but these contained the majority of terminal deoxynucleotidyl transferase (TdT+) cells. In the assessment of class II antigens, HLA-DR was expressed on all B lineage cells whereas HLA-DP preceded HLA-DQ in appearance during the developmental process. Among the later antigens expressed on B lineage cells, cell surface IgM, CD20, and HLA-DQ were expressed at essentially the same time. Cell surface CD10 was lost at the time when CD21 and CD22 were acquired on the cell surface. These data illustrate that multiparameter flow cytometry can be used to define a continuous progression of stages of B lymphocyte development based on cell surface antigen expression even though these cells represent a minor fraction of normal marrow cells.

Flow cytometry was used to identify maturational differences of erythroid lineage cells in normal human bone marrow by combining physical characteristics, the expression of multiple cell surface antigens, and nucleic acid content. Normal low-density bone marrow cells could be divided into four populations, based on forward and right-angle light scattering. Erythroid cells, at different maturational stages, were found in three of these four marrow subpopulations. The sequentially correlated expression of three cell surface markers--HLe-1, transferrin receptor, and glycophorin--allowed us to study erythroid maturation from the colony forming cell to the mature erythrocyte. HLe-1 was expressed on the earliest identifiable erythroid cells and was progressively lost as the cells matured. Transferrin receptor began to be expressed at the BFU-E stage and disappeared at the late reticulocyte stage. Transferrin receptor expression preceded glycophorin expression, the latter beginning on morphologically recognizable erythroid precursors just after the CFU-E stage. In contrast to both HLe-1 and transferrin receptor, which were progressively lost during the maturational process, once glycophorin had been maximally expressed on the cell surface, it remained at constant quantities to the mature erythrocyte stage. Although developing nucleated erythroid cells at approximately the normoblast stage had light-scattering properties similar to those of lymphoid cells, these two cell types could be resolved by cell surface antigen expression. Normoblasts were glycophorin positive and HLe negative, whereas lymphoid cells expressed HLe and either Leu 4, Leu 11, or Leu 12. Decreases in cellular nucleic acid content, corresponding first to the extrusion of the nucleus and second to the loss of reticulum, characterized the later stages of erythroid development. These characteristics and instrumentation can be used to purify erythroid cells at various developmental stages.