The growth of haematopoietic cells in tissue culture conditions is a well established practise. Haematopoietic stem cells require direct interaction with the components of a supportive microenvironment for optimal maintenance, proliferation and differentiation. The mechanisms underlying the homing and lodging of stem cells in the bone marrow and other haematopoietic organs are mediated by membrane interactions of stem cells with stromal cells and with components of the extracellular matrix. Fibronectin is involved in diverse adhesive interactions with a variety of cell types including haematopoietic progenitor and stem cells (Aizawa et al 1991).
Erythroid progenitor cells have been demonstrated to adhere to fibronectin- coated substrate, but this adhesion is lost upon differentiation (Coulombel 1991,1992; Vuilet-Gaugler et al 1990). Restriction of this expression to colony forming units-erythroid (CFU-E) and the first subsequent cell divisions strikingly correlates with the migratory capacity of these cells. These findings emphasise the necessity of direct interactions between haematopoietic cells and stromal cells, however it has not been possible to accurately quantify the specific attachment of haematopoietic cells to stroma.
Liesveld et al (1989) showed that CD34+ haematopoietic cells could bind to fibroblasts just as effectively as they bound to stromal cells, although the stromal cells were better at supporting haematopoiesis. Thus specific binding of haematopoietic cells to stromal cells may be required in haematopoiesis. This could explain why haematopoietic cells home to specific sites.
Lineage and stage specific adhesion of haematopoietic cells to stroma or ECM has been reported. Gordon et al (1985), found blast progenitors would adhere to cells grown in the presence of methylcellulose but not to cells grown in its absence. More mature cells showed no preference to adhere to either layer. Bearpark & Gordon (1989) reported that CFU-S would readily adhere to stromal layers, while GM-CFC populations would not. Neuraminidase has no effect on the adhesion of blast cells, it can however increase the adhesion of GM-CFC. This suggests that the cell adhesion molecules alter with lineage and maturation. Gordon et al (1990) found that heparan sulphate proteoglycan (HS-PG) was essential for binding of blast-CFC to stroma. They suggested that haematopoietic cells of different lineages at different stages in development recognise and occupy specific microenvironments in the bone marrow. AcSDKP, which is a known inhibitor of entry into cell cycle, increased
adhesion of CFU-S to stromal layers but had no effect on adhesion of GM-CFC to the layers (Lenfant et al 1989). Aizawa et al (1992), have suggested that AcSDKP improves binding through activation of stromal cells. Verfaillie et al (1990,1992) found that early progenitor cells had the capacity to adhere to pre established irradiated stroma while more committed progenitor cells lacked this capacity.
Aizawa & Tavassoli (1988) reported that homing of intravenously transplanted haematopoietic progenitor cells to the bone marrow occurs through a
recognition system with galactosyl and mannosyl specificities. Both galactosyl and mannosyl BSA restricted homing of haematopoietic cells to bone marrow
but had no effect on the homing of stem cells to the spleen. This may reflect a difference in the surface characteristics of stromal cells in the bone marrow and spleen and different molecular bases for the binding of progenitor cells in the two sites.
Houssaint & Hallet (1988) grafted a precolonised liver rudiment under the kidney capsule of an adult mouse. Hepatocytes developed as normal, however the rudiment was never colonised by HSC. These results are in accordance with those of Johnson & Moore (1975), who suggested a host with normal haematopoiesis had no requirement for extra-medullary haematopoietic sites. Houssaint & Hallet proposed that ontogenic maturation of HSC occurs around the time of birth, which would alter their homing capacity. The grafted liver rudiment would be receptive to embryonic HSC only. In contrast Fleischman, Custer & Mintz (1982), injected normal bone marrow cells into the placental circulation of an 11 day murine fetus. The adult HSC were able to seed the fetal liver environment, proliferate and subsequently colonise the bone marrow, indicating that adult HSC capable of colonising fetal liver, in the appropriate host environment, are present in adult bone marrow (Houssaint & Hallet 1988).
Riley & Gordon (1987) and Gordon et al (1990), reported selective adhesion of Bl-CFC to adherent stroma. They found that marrow derived Bl-CFC would adhere to adult derived marrow stromal layers, but not to fetal liver or fetal marrow derived layers. Fetal derived HSC would not produce blast colonies when seeded onto adult derived stromal layers.
In vitro studies by Zanjani, Ascensao & Tavassoli (1993), suggest a hierarchy of homing sites, in which the bone marrow has the highest affinity for homing HSC. Before the bone marrow is haematopoietically active, transplanted HSC home to the liver and the spleen. Once the marrow has developed, its homing potential superseeds that of the other organs, despite the liver and spleen
remaining the major haematopoietic sites till birth. From the initial stage of its development, the bone marrow is capable of engrafting circulating stem cells, but does not support their differentiation and proliferation. Thus while HSC can pool in the fetal bone marrow, the liver remains the main haematopoietic organ till birth. Recent information has emphasised the adhesive capacity as an important property of haematopoietic cells and of their microenvironment. The adhesive properties can determine the distribution of cells and influence their exposure and response to growth factors.