FLOW CYTOMETRIC ANALYSIS OF HUMAN BONE-MARROW PERFUSION CULTURES - ERYTHROID DEVELOPMENT AND RELATIONSHIP WITH BURST-FORMING UNITS-ERYTHROID

Flow cytometry has been used in recent years to study antigenic and ph ysical changes accompanying hematopoietic cell differentiation. Such s tudies have provided the basis for rapid and objective assays that are potential alternatives to the colony assays currently in widespread u se. In this report, erythropoiesis was examined in growth factor-suppl emented perfused cultures of bone marrow mononuclear cells (BMMNC) usi ng flow cytometric analysis of the transferrin receptor (CD71), glycop horin A (gly A), and CD33. A marked progression of erythroid antigen e xpression occurred with time. Initially, fewer than 10% of the cells w ere CD71(++), but by day 8, 19-34% of the cells were CD71(++)gly A(-). This was followed by the appearance of gly A on 10-60% of the CD71(+) cells. After day 10, CD71 expression on many gly A(+) cells decrease d so that a population of CD71(-)gly A(+) cells (11-54%) accumulated b y day 14. Each phenotype was sorted for morphologic identification and colony assay analysis. CD7l(++)gly A(-) cells were 85% blasts, one-ha lf of which were erythroblasts, and were significantly enriched for bu rst-forming units-erythroid (BFU-E). The time-varying number of CD71(+)gly A(-) cells in these cultures was found to correlate with the num ber of BFU-E present in each of six independent experiments (correlati on coefficient, r = 0.78-0.97). Three-color analysis was next used to examine CD33 expression on BFU-E, and in fresh BM, most were found to be CD33(-). During culture, however, the number of BFU-E recovered fro m CD33(+) populations first increased and then decreased. Therefore, C D33 was not particularly useful for identifying BFU-E. In contrast, CF U-GM were mostly found to be in the CD71(+)CD33(+) population througho ut the culture period. When erythropoietin (Epo) was not added to thes e cultures, the percentage of gly A(+) cells was reduced from 33 to 3. 3%. Further, the omission of Epo caused an 80% decrease in the number of BFU-E and a corresponding 94% decrease in the number of CD71(++)gly A(-) cells, thereby maintaining the relationship between CD71(++)gly A(-) cells and BFU-E. Therefore, flow cytometric analysis was found to be useful in assessing erythroid development, and this approach may b e used to develop flow cytometric assays for other populations of inte rest in hematopoietic cell cultures.