or factors. Prelimary studies were performed to ascertain the role of other HGFs in the growth of BFU-E using fetal blood. It was found that, IL-1, IL-3, IL-6, IL-9 and GM-CSF all had BPA activity, the most potent being IL-3 and IL-9. These studies require confirm ation with purified progenitor to exclude indirect effects. Other factors of interest as potential fetal BPA include SCF and IL-11. It must be pointed out that fetal liver derived progenitors had a lesser requirement for BPA compared to adult cells suggesting either different sensitivity to these factors or that Epo has more widespread effects on fetal cells. There is no firm evidence that fetal blood generally contains detectable amounts of the HGFs generally regarded as potent BPAs in studies of erythroid progenitors. The studies detailed here were hampered by the small specimen volumes available but it seems unlikely that IL-1, IL-3, IL-6 or GM-CSF are present in the fetal circulation. Other candidate HGFs with an established BPA function such as SCF or IL-9 may be present and further studies when assays for these factors become available would be interesting. Fetal blood does contain M-CSF at levels significantly higher than the maternal blood samples collected at the same time. The significance of high levels of M-CSF in the context of the regulation of fetal haemopoiesis is unknown. In pathological states such as Rhesus haemolytic disease of the newborn HGFs including IL-3 may be present in increased concentrations and further studies are merited. A possible methology to explore these areas further would be to examine fetal tissues containing haemopoietic activity using non-isotopic in -s itu hybridization for cytokine mRNAs, together with immunophenotypic identification of the cells involved.
Fetal progenitors were used in studies of signal transduction. In particular Epo was not found to mobilize intracellular calcium on addition to fetal erythroid progenitors although the cells were subsequently shown to be Epo-responsive in proliferation assays or clonogenic assays. Although this is in agreement with the findings in the report by [Imagawa, et al., (1989)] it disagrees with data from [Miller, et ai.,
(1988 )]. In subsequent studies they demonstrated that the response was seen primarily in adult late stage erythroblasts [Miller, et al. , (1989)]. With further refinements to the system, using a digital confocal imaging system Epo was observed to preferentially increase nuclear free calcium in these cells [Yelamarty and Miller, (1 9 9 0 )]. From these data the suggestion is that Epo-mediated changes in compartmentalized intracellular free calcium might be involved in promoting the nuclear involution that accompany erythroid maturation. It would be interesting to repeat the experiments in fetal progenitors using the methodology employed by Miller and colleagues. Another possible explanation for the contradictory findings would be the existence of different signalling pathways depending on the maturational
Stage of the cells tested. Ultimately it will be of interest to contrast the Epo induced kinase signalling events [Witthuhn, et al., (1993)] in fetal and adult progenitor cells.
In the intracellular calcium experiments already discussed, it was possible to prove that the cells tested were functionally competent by exploiting the presence of FcyRII molecules (recognised by moAbs in the CDw32 cluster) on the isolated progenitors. The FcyRII molecules proved to function as signal transduction molecules upon cross-linking thereby modulating the cellular calcium level. The finding of FcyRII on fetal progenitors was unexpected but has now been confirmed by others in adult blast cells and progenitors [Ball, et a!., (1989)]. The functional role of FcyRII receptors on fetal progenitors remains unknown but one intriguing possibility might be the existence of a ligand other than monomeric immunoglobin. It should be remembered that the FcyRII molecule contains elements of the immunoglobulin superfamily as do many HGF receptors. It may be speculated that this alternative ligand could be involved in adherence within the microenvironmental niche.
In the studies of primitive human fetal liver derived erythroid cells removal of erythropoietin decreased thym idine uptake w ithout obviously affecting the percentage of cells in the active phases of the cell cycle, thus it appears that erythropoietin is not acting here as a cell cycle recruiting stimulus, in that the proportion of cells in S/G2 M is not influenced by addition of erythropoietin. These data are compatible either with the view that Epo deprivation results in prolongation of the cell cycle without arrest in a particular phase of the cycle or alternatively that cells may accumulate at the G i phase and deprived of the mitogenic signal provided by Epo they die via apoptotic mechanisms, so that there is no obvious change in the cell cycle profile. Furthermore the possibility that Epo is not functioning as a progression factor but as a maintenance or competence factor is compatible with the fact that the high levels of c-myc protein which are seen in fetal liver proerythroblasts and erythroid progenitor cells ex vivo are maintained in the presence of Epo, where as in the absence of Epo c-myc protein levels fall within 6 - 12 hours. However, the precise role of the c-myc protein in proliferation or differentiation processes remains unknown. It has been suggested that continued relatively high c-myc protein levels are a feature of proliferating cells irrespective of cell cycle status [Evan, et a!., (1992)]. More recently the c-myc protein has been shown to associate with a helix-loop-helix zipper protein termed MAX to form a sequence-specific DNA-binding complex [Blackwood and Eisenman, (1991)]. Also, [Koury and Bondurant, (1990)] have recently suggested that erythropoietin prevents programmed cell death (apoptosis) the mechanism by which Epo appears to