Obligatory Asymmetric Cell Division Regulates Self-Renewal In Hematopoietic Progenitor/Stem Cells

Abstract 571

Asymmetric cell division, a proposed mechanism by which hematopoietic progenitor/stem cells (HPSC) maintain a balance between self-renewal and differentiation, has rarely been observed. Here we report the surprising finding that cultured mouse primary HPSC routinely generate pairs of daughter cells with 2 distinct phenotypes after a single round of cell division. Mouse bone marrow cells were cultured on chamber slides in the presence of stem cell factor (SCF). BrdU was added overnight to label dividing cells, and the cells were examined by immunofluorescence microscopy on day 2–4 of culture. In each BrdU+c-Kit+ divided cell doublet, c-Kit was invariably expressed in only 1 of the 2 daughter cells. In contrast, the other daughter cell was negative for c-Kit but positive for the asymmetric cell fate determinant Numb and mature myeloid markers Mac1, Gr1, M-CSFR and F4/80. Similarly, in each BrdU+Sca1+ cell doublet, 1 daughter cell was positive for the stem cell markers Sca1, c-Kit, CD150 and CD201, whereas the other cell was negative for these markers but positive for Numb and the mature myeloid markers. Analysis of 400 such doublets showed that the probability of HPSC undergoing asymmetric division was 99.5% (95% confidence interval 98–100%), indicating that asymmetric division in HPSC is in fact not rare but obligatory. In other model systems, it has been shown that activation of the atypical protein kinase C (aPKC)-Par6-Par3 cell polarity complex and realignment of the microtubule cytoskeleton precede asymmetric cell division. We asked whether similar steps are involved in the asymmetric division of HPSC. We found that c-Kit receptors, upon stimulation by SCF, rapidly capped at an apical pole next to the microtubule-organizing center, followed by redistribution to the same pole of the aPKC-Par6-Par3 complex and microtubule-stabilizing proteins APC, β-catenin, EB1 and IQGAP1. Strikingly, after cell division, the aPKC-Par6-Par3 complex and other polarity markers all partitioned only into the c-Kit+/Sca1+ daughter cell and not the mature daughter cell. The acetylated and detyrosinated forms of stabilized microtubules were also present only in the c-Kit+/Sca1+ cell, as were the Aurora A and Polo-like kinases, 2 mitotic kinases associated with asymmetric cell division. To understand how c-Kit activation triggers downstream polarization events, we studied the role of lipid rafts, cholesterol-enriched microdomains in the cell membrane that serve as organization centers of signaling complexes. These are enriched in phosphatidylinositol 4,5-bisphosphate and annexin 2, putative attachment sites for the aPKC-Par6-Par3 complex. We found that SCF stimulation led to coalescence of lipid raft components at the site of the c-Kit cap, and treatment with a wide range of inhibitors that blocked lipid raft formation abrogated polarization of the aPKC-Par6-Par3 complex and division of the c-Kit+/Sca1+ cells. Because obligatory asymmetric division in cultured HPSC would prevent a net increase in their number, we sought a way to bypass its mechanism. We tested whether inhibition of protein phosphatase 2A (PP2A), a physiological antagonist of aPKC, would enhance aPKC activity and promote self-renewal of HPSC. Treatment of cultured HPSC with okadaic acid or calyculin, 2 well-characterized PP2A inhibitors, increased the percent of c-Kit+/Sca1+ cells undergoing symmetric division from 0% to 23.3% (p<0.001). In addition, small colonies comprised of symmetrically dividing cells uniformly positive for Sca1, c-Kit, CD150 and CD201 were noted in the culture. To functionally characterize the effect of PP2A inhibition, mouse bone marrow cells were cultured in the absence or presence of PP2A inhibitors and transplanted into irradiated congenic mice in a competitive repopulation assay. At 4–8 weeks post-transplant, the donor engraftment rate increased from ∼1 in mice transplanted with untreated cells to >30% in mice transplanted with PP2A inhibitor-treated cells. This dramatic increase indicates that PP2A inhibition can effectively perturb the mechanism of asymmetric cell division and promote the self-renewal of HPSC. In summary, our data showed that obligatory asymmetric cell division works to maintain a strict balance between self-renewal and differentiation in HPSC and pharmacological manipulation of the cell polarity machinery could potentially be used to expand HPSC for clinical use.