Functional and phenotypic characterization of cord blood and bone marrow subsets expressing FLT3 (CD135) receptor tyrosine kinase.

The class III receptor tyrosine kinase FLT3/FLK2 (FLT3; CD135) represents an important molecule involved in early steps of hematopoiesis. Here we compare cell-surface expression of FLT3 on bone marrow (BM) and cord blood (CB) cells using monoclonal antibodies (MoAbs) specific for the extracellular domain of human FLT3. Flow cytometric analysis of MACS-purified BM and CB cells showed that 63% to 82% of BM CD34+ and 88% to 95% of the CB CD34+ cells coexpress FLT3. Clonogenic assays and morphological characterization of FACS-sorted BM CD34+ cells demonstrate that colony-forming unit-granulocyte-macrophage (CFU-GM) and immature myelo-monocytic precursor cells are enriched in the subpopulation staining most brightly with the FLT3 MoAb whereas the majority of the burst-forming units-erythroid (BTU-E) and small cells with lymphoid morphology are found in the FLT3- population. In contrast, statistically indistinguishable proportions of CFU-granulocyte-erythrocyte-megakaryocyte-macrophage (CFU-GEMM) and more primitive cobblestone area forming cells (CAFC) were detected in both fractions, albeit the FLT3+ fraction consistently showed more CAFC activity than the FLT3- fraction. Although in both, BM and CB the majority of CD34+CD117+ (KIT+), CD34+CD90+ (Thy-1+), and CD34+CD109+ cells coexpress FLT3, three-color phenotypic analyses are consistent with the functional findings and suggest that the most primitive cells defined as CD34+CD38-, CD34+CD71low, CD34+HLA-DR-, CD34+CD117low, CD34+CD90+, and CD34+CD109+ express low levels of cell-surface FLT3 and were therefore not enriched to a statistically significant extent with the bright versus negative sorting scheme. Thus, clear segregation of the most primitive progenitors from BM CD34+ cells was confounded by low apparent levels of FLT3 cell-surface expression on these cells, whereas myeloid progenitors unambiguously segregated with the FLT3 brightest cells and erythroid progenitors with the FLT3 dimmest. Additional phenotypic analyses using MoAbs against progenitor/stem cell markers including the mucinlike molecule MGC-24v (CD164), the receptor tyrosine kinases TIE, FMS (CD115), and KIT (CD117) further illustrate the differences in surface antigen expression profiles of BM and CB CD34+ cells. Notably, CD115 is rarely detected on CB CD34+ cells, whereas 20% to 25% of the BM CD34+FLT3+ cells are CD115+. Furthermore, 80% to 95% of the CB CD34+CD117+ but only 60% to 75% of the BM CD34+CD117+ cells coexpress FLT3. Only a negligible amount of CD34+CD19+ are detected in CB, while in BM 20% to 30% of CD34+CD19+ presumed pro/pre-B cells coexpress FLT3. In contrast, the majority of the CD34+CD164+ and CD34+TIE+ subsets in both CB and BM coexpress FLT3. Analysis of unseparated cells showed that FLT3 expression is not restricted to CD34+ subsets. About 65% to 70% of lymphocyte-gated BM CD34-FLT3+ cells are positive for the monocytic marker CD115 whereas 25% to 30% of these cells consist of CD10 expressing B-cell precursors. Finally, CD34- monocytes in BM, CB, and PB express FLT3 whereas granulocytes are FLT3-. Our data show that detectable FLT3 appears first at low levels on the surface of primitive multilineage progenitor cells and disappears during defined stages of B-cell development, but is upregulated and maintained during monocytic maturation.