A series of recent high profile reports suggested that a simple positive selection approach utilizing SLAM cell surface markers results in enrichment of long-term repopulating stem cells to the same degree as much more complex multi-step negative and positive selection schemes. These studies, performed in the murine model, also reported that for the first time immunohistochemistry could be used to identify HSCs in tissue sections, allowing analysis of HSCs in their niches. Investigators hoped that the same markers would translate to human HSCs, but in a series of experiments we have shown that unfortunately SLAM markers are unlikely to be relevant in larger animals and humans. SLAM-purified rhesus macaque and human mobilized peripheral blood, bone marrow and cord blood CD34+ cells were transplanted into immunodeficient mice, and there was no enhancement of engraftment in the SLAM+ versus SLAM- populations. These results are importantsince investigators are already utilizing SLAM purification of human cells without supportive data documenting HSC enrichment. In collaboration with the laboratory of Jennifer Lippincott-Schwartz, we have identified a polarized membrane domain on primitive hematopoietic stem and progenitor cells that is critical for interactions with osteoblasts, central functional components of the hematopoietic niche in vivo. These domains allow attachment to osteoblasts, contain a number of proteins known to be important in organizing membrane domains, including tetraspanins, integrins, and prominin, and have a central lipid component. Cholesterol depletion or disruption of actin polymerization disrupts the domains, and prevents attachment to osteoblasts. Following attachment to osteoblasts, we have documented membrane transfer from the HSCs to the osteoblasts, and initiation of a signaling cascade in osteoblasts, resulting in increased expression of SDF-1, a primary mediator of stem cell mobilization and engraftment. We have now demonstrated these domains to be important for in vivo homing and engraftment of human CD34+ cells, in murine xenograft model. We have found these domains to be disrupted during the S/G2/M phases of the cell cycle, and investigated the link between the loss of these membrane domains and the previous observation from our group and others that HSCs in active cycle do not home or engraft as efficiently as cells in the Go phase of the cell cycle. We are exploring the critical protein components in the domain, studying whether siRNA knockdown or full knockout of tetraspanins, specifically CD82, impacts on domain formation or HSC behavior. Blocking CD82 blocks adhesion and engraftment. Conversely, we are searching for compounds that increase domain formation and tetraspanin expression, in hopes that homing and engraftment of cells exposed to these agents in vitro might be improved. IBD increases expression of the tetraspanin CD9 however this increased expression does not result in higher engraftment. We have begin to compare these domains in primitive normal versus primary human leukemic blasts, and found that the domains are disrupted in leukemic cells. We are now using competitive repopulation studies and multicolor 3D confocal and 2 photon imaging to ask whether leukemic and normal stem cell compete for the same in vivo bone marrow niche. Most recently, we have begun to explore the relationship between oxygen concentration and the localization and behavior of HSCs. Marrow cells sorted for the lowest concentration of reactive oxygen species have characteristics of more primitive long-term engrafting cells, are not in active cell cycle, and purification based on ROS concentration may provide a simple one-step HSC enrichment process. Once cells are mobilized to the peripheral blood, outside of their normal protective niche, ROS concentration no longer correlates with primitive stem cell phenotype and function. We are carrying out gene expression profiling and plan to study epigenetic markers within ROS low and ROS high marrow stem cell populations. We will also ask whether cells cultured under ROS low conditions better maintain HSC properties, and can be genetically transduced and expanded, compared to cells under standard ex vivo culture conditions. Finally we have recently developed a very novel and powerful technique combining combinatorial transduction with 3-5 lentiviral vectors expressing fluorescent proteins with confocal and 2 photon imaging of intact bone marrow and other hematopoietic organ architecture, to be able to localize and map hematopoiesis at a clonal level in the bone marrow environment.
Showing the most recent 10 out of 28 publications
