Enucleation of Human Erythroblasts Is a Process of Asymmetric Cytokinesis.

Background. During erythroblast enucleation, nuclei surrounded by plasma membrane separate from erythroblast cytoplasm. Enucleation has been thought to be a process similar to cytokinesis. However, more concrete evidence has been difficult to obtain because of a lack of an ex vivo experimental system capable of confirming cytokinesis. Focusing on the mechanism of cell division, we investigated the redistribution of cytoplasmic proteins and integral membrane proteins during enucleation, using ex vivo generation system of mature human blood cells from hematopoietic stem cells.

Materials and Methods. The highly purified human CD34⁺ cells were grown in the presence of interleukin-3, stem cell factor and erythropoietin (EPO) in a liquid phase. After 7 days of culture, the generated cells (day 7 cells) were replaced in a medium with EPO alone. The cells matured and terminally differentiated into reticulocytes during a 13–15-day culture period. We mainly used non-gravity and non-pipetting system to avoid physical stress that may disrupt the connection between the nucleus and reticulocyte. Day 9 cells, predominantly consisted of polychromatophilic erythroblasts and expressed glycophorin A (GPA) at a purity of 97%, were labeled with DNA-staining dye SYTO21 for the direct monitoring of the enucleation process, using differential interference contrast microscopy. We also cultured day 9 cells until day 14, on 4-well culture slides or on the membrane of a cell culture insert system, and removed culture medium by aspiration without centrifugation and pipetting. The day 14 cells on the slide were analyzed using immunohistochemical staining, whereas the cells on the membrane were embedded in O.C.T. compound for confocal microscopy.

Results. Approximately a half of erythroblasts enucleated until day 14. The monitoring of the enucleation process at day 13 showed autonomous extrusion of SYTO21 positive nucleus from single erythroblast. Some of the expelled nuclei were still connected with reticulocyte through strings that were positive for antibody against tubulin, actin, GPA, band 3 and glycophorin C (GPC). The expelled nuclei were covered by lamin, a protein specific for nuclear membrane, which were surrounded by a substance positive for GPA, band 3, GPC, p55, 4.1R⁸⁰, actin, tubulin, b-spectrin, calnexin and cytochrome C, although the distribution of each proteins were asymmetric between nuclei and reticulocytes. An intense area of GPA, GPC, band 3, 4.1R⁸⁰, actin, tubulin, myosin and b-spectrin was found in the region of the constriction between the extruding nucleus and incipient reticulocyte in enucleating cells. In cells just before enucleation, tubulin and actin formed a radial array around the nucleus. The center of the radial array was positive for centrin and NuMA, indicating that the cenriole formation occurred during an enucleation process.

Conclusion. Our investigations show that a part of human erythroblasts enucleate independent of an interaction with accessory cells. The appearance of cenriole and the asymmetric redistribution of cytoplasmic and integral membrane proteins during enucleation strongly suggest that enucleation of human erythroblasts is a process of asymmetric cytokinesis.