Erythropoietic progenitors capable of colony formation in culture: state of differentiation

The differentiated state of mouse erythropoietic progenitor cells (CFU-E), detected by their ability to form erythropoietin-dependent colonies in vitro, has been investigated. Transfusion-induced plethora was found to reduce the population size of CFU-E in both spleen and femoral marrow, which indicates that a significant number of CFU-E arise by differentiation processes that are themselves erythropoietin-dependent. Individual spleen colonies were found to be heterogeneous in their content of CFU-E, and the numbers of CFU-E per colony were not correlated either positively or negatively with the numbers of granulocyte-macrophage progenitors (CFU-C) present in the same colonies. The absence of a negative correlation between CFU-E and CFU-C indicates that the erythropoietic and granulopoietic pathways of differentiation are not mutually exclusive within individual spleen colonies. The numbers of CFU-E per spleen colony were also found to vary independently of the numbers of pluripotent stem cells (CFU-S) per colony; in contrast, as found previously, the numbers of CFU-C and CFU-S per colony were positively correlated. These results indicate that more randomizing events separate CFU-E from CFU-S than separate CFU-C from CFU-S, and are consistent with the view that CFU-E occupy a position on the erythropoietic pathway of differentiation that is more remote from the pluripotent stem cells than is the corresponding position of CFU-C on the granulopoietic pathway.     

Clonal origin of murine hemopoietic colonies with apparent restriction to granuclocyte-macrophage-megakaryocyte (GMM) differentiation

We characterized murine hemopoietic colonies consisting of granulocytes, macrophages, megakaryocytes, and blast cells and yet lacking erythroid elements. Mouse marrow or spleen cells were cultured in methylcellulose media in the presence of 10% (v/v) pokeweek mitogen-stimulated spleen cell-conditioned medium (PWM-SCM) and 2 units/ml erythropoietin for 8 days. Granulocyte-macrophage-megakaryocyte (GEMM) colonies could be distinguished from granulocyte-erythrocyte-macrophage-megakaryocyte (GEMM) colonies because the former lacked the typical appearance of bursts with red color. Analysis of Y-chromosomes in mixing experiments with male and female marrow cells confirmed the clonal nature of the GMM colonies. Differential counts of GMM colonies revealed varying, but significant, numbers of blast cells in all of the day-8 and day-12 colonies and in seven out of ten day-14 GMM colonies. In general, the percentages of blast cells were inversely related to the length of incubation in culture. Replating experiments confirmed the absence of late erythroid precursors such as CFU-E and normoblasts in all of the 50 day-8 GMM colonies. However, six out of the 50 GMM colonies contained early progenitors capable of erythroid expression, such as BFU-E, CFU-EM, CFU-GEM, and CFU-GEMM. In contrast, the three day-14 GMM colonies which did not reveal blast cells failed to produce secondary colonies. Thus, while the progenitors for the latter colonies are restricted to only granulocyte-macrophage-megakaryocyte differentiation, some of the apparent GMM colonies containing blast cells may have originated in early progenitors close to pluripotent stem cells. Detailed cytological analyses and replating experiments are necessary for characterization of true differentiation potentials of mixed colonies in culture.     

In vitro development of CFU-E and BFU-E in cultures of embryonic and post-embryonic chicken hematopoietic cells

A culture method is proposed for the in vitro development of chicken erythrocytic progenitors. When grown with avian erythropoietin, Colony Forming Unit Erythrocytic (CFU-E) and Burst Forming Unit-Erythrocytic (BFU-E) give rise respectively to erythrocytic colonies and bursts within 3 and 6 days. BFU-E development is greatly enhanced by pokeweed-mitogen-spleen-cell-conditioned medium and requires higher erythropoietin concentrations than for CFU-E. An antigen specific to immature red cells can be detected on CFU-E but not on BFU-E, showing that both progenitors represent distinct entities. BFU-E and CFU-E are found in embryonic marrow and yolk sac. In the young blastoderm BFU-E becomes detectable at the primitive streak stage.     

Effect of recombinant human interleukin 3, granulocyte-macrophage colony-stimulating factor and granulocyte colony-stimulating factor on human BFU-e in serum-free cultures

The effects of recombinant human hemopoietic growth factors on early and late human erythroid progenitors (BFU-e and CFU-e) were investigated in serum-free cultures. Recombinant human erythropoietin (rhEpo) induced the formation of not only human CFU-e-derived colonies but also human BFU-e-derived bursts. Recombinant human interleukin 3 (rhIL-3) alone did not induce the formation of human BFU-e-derived bursts and human CFU-e-derived colonies. In the presence of rhEpo, rhIL-3 dose dependently increased the number of bursts stimulated by rhEpo, although rhIL-3 did not have the augmentative effect on human CFU-e growth. On the other hand, rhIL-3 did not stimulate the formation of murine BFU-e-derived bursts, and murine IL-3 did not stimulate the formation of human BFU-e-derived bursts. The results indicated that the burst-promoting activity of IL-3 was species-specific between human and murine cells. Recombinant human GM-CSF (rhGM-CSF) or recombinant human G-CSF (rhG-CSF) failed to induce human burst formation and did not augment the effect of rhEpo on human burst formation. The results of the present study suggest that in vitro, IL-3 can stimulate BFU-e in collaboration with Epo, but GM-CSF and G-CSF do not stimulate BFU-e growth in the presence or absence of Epo.     

A comparison of stem cell assays using early or late spleen colonies

Abstract The distribution of spleen colony diameters was determined 5.5, 8.0, 10.5 and 13.0 days after injection of normal bone marrow cells to lethally irradiated recipients. A relative lack of small colonies on day 8.0, as compared with days 5.5, 10.5 and 13.0, argued against a time continuum in colony appearance. The spleen colonies observed after 10 days or more probably represented a mixture of colonies which developed from the originally transplanted CFU-S and those arising from secondary CFU-S. Thus, late appearing spleen colonies may not necessarily identify a different, less mature, population of CFU-S. Administration of increasing amounts of bone marrow cells was used for comparing the linearity of the CFU-S assay for colonies observed after 8 days or after 12 to 13 days. The influence of overlapping colonies on the results was considerably augmented if large spleen colonies were observed after 12 or 13 days. Subsequently the CFU-S assay lost much of its quantitative character. We believe that some previously published data might have been misinterpreted by neglecting the important differences between 'early'and 'late'CFU-S assays.     

Hemoglobin switching in sheep: commitment of erythroid stem cells to expression of the betaC-globin gene and accumulation of betaC-globin mRNA

The switch from HbA (α₂β₂^A) to HbC (α₂β₂^C) synthesis was induced by injection of erythropoietin into a lamb homozygous for HbA. Serial samples of bone marrow were analyzed to detect the initial commitment of erythroid stem cells (CFU-E) to form colonies which made HbC in vitro, and to detect the initial accumulation of β^C-globin mRNA and the onset of HbC synthesis in erythroblasts in vivo. CFU-E-derived erythroid colonies were formed in plasma clot culture at a low erythropoietin concentration, and the relative amounts of β^A- and β^C-globin synthesized were determined after a 24 hr pulse of ³H-leucine, added after 84 hr in culture. RNA was extracted from nuclei and cytoplasm of “early” and “late” populations of bone marrow erythroblasts which had been fractionated by Ficoll-Hypaque density centrifugation. The concentration of β^A- and β^C-globin mRNA was determined by annealing to purified synthetic DNAs (cDNAs) complementary to β^A and β^C mRNA. No β^C-globin was synthesized in erythroblasts or in CFU-E-derived erythroid colonies prior to the injection of erythropoietin. An increase in the concentration of CFU-E in the bone marrow and the appearance of β^C-globin synthesis in CFU-E-derived colonies were detected 12 hr after the erythropoietin injection. In contrast, β^C mRNA was not detected in either “early” or “late” erythroid cells until 36 hr later. The first measurable β^C-globin mRNA was accompanied by the appearance of β^C-globin synthesis in bone marrow erythroblasts. Our results suggest that the accumulation of β^C-globin mRNA is a relatively late event following induction of HbA to HbC switching by erythropoietin. The expansion of the compartment of erythroid stem cells and the commitment of CFU-E to β^C-globin synthesis appear to precede the detectable accumulation of β^C mRNA by 24–36 hr.     

The kinetics of hematopoietic stem cells during and after hypoxia. A model analysis

A previously described mathematical model of the hematopoietic stem cell system has been extended to permit a detailed understanding of the data during and after hypoxia. The model includes stem cells, erythroid and granuloid progenitors and precursors. Concerning the intramedullary feedback mechanisms two basic assumptions are made: 1) The fraction "a" of CFU-S in active cell cycle is regulated. Reduced cell densities of CFU-S, progenitors or precursors lead to an accelerated stem cell cycling. Enlarged cell densities suppress cycling. 2) The self renewal probability "p" of CFU-S is also regulated. The normal steady state is described by p = 0.5, indicating that on statistical average each dividing mother stem cell is replaced by one daughter stem cell, while the second differentiates. Diminished cell densities of CFU-S or enlarged densities of progenitors and precursors induce a more intensive self renewal (p greater than 0.5), such that the stem cell number increases. The self renewal probability declines (p less than 0.5) if too many CFU-S or too few progenitors and precursors are present. The model reproduces bone marrow data for CFU-S, BFU-E, CFU-C, CFU-E, 59 Fe-uptake and nucleated cells in hypoxia and posthypoxia. Although the ratio of differentiation into the erythroid and granuloid cell lines is kept constant in the model, a changing ratio of CFU-E and CFU-C results. The model suggests that stem cells and progenitor cells are regulated by a regulatory interference of erythropoiesis and granulopoiesis.     

氨甲酰胆碱和Impromidine对CFU—S细胞周期的影响

正常情况下处于S期的CFU-S比例低于10％。氨甲酰胆碱(Cach 10~(-13)—10~(-9)mol/L)和Impromidine(Impro 10~(-9)—10~(-4)mol/L)在体外与小鼠骨髓细胞短时培育后,增加了CFU-S对细胞毒剂羟基脲的敏感性。反应最大时,9d和13dCFU-S的减少率分别是32.8和60.6％(Cach)以及38.4和49.5％(Impro)。这种效应可分别被胆碱能N受体阻断剂筒箭毒(10~(-6)mol/L)和组胺H_2受体阻断剂甲氰咪呱(10~(-6)mol/L)所阻断,表明9d和13d CFU-S表面胆碱能N受体和组胺H_2受体的密度或活性存在差别,再次证实了CFU-S是一个不均一的细胞群。     

Hemopoietic precursor cell defects in nonanemic but stem cell-deficient W44/W44 mice

Hematopoietic stem cell deficiencies cause a severe macrocytic anemia in W/Wv mice. W44/W44 mice, on the other hand, are not anemic, but, since they accept marrow implants without prior total body irradiation, they have inherited a stem cell lesion. In an attempt to identify the aberrant stem cell(s), we have determined the concentration in W44/W44 marrow of hematopoietic precursors known to be deficient in W/Wv marrow. The in vitro erythroid burst-forming units (BFU-E), the in vivo spleen colony-forming units (CFU-S), and the cells that repopulate the erythroid compartment of stem cell-deficient mice were examined. The progenitors of 7-day bursts are dramatically reduced in W/Wv marrow but are present in normal concentrations in W44/W44 marrow. W44/W44 marrow CFU-S, unlike W/Wv, generate visible spleen colonies 10 days after injection into lethally irradiated recipients. The colonies are, however, smaller and at least 2 times less numerous than those produced from equivalent numbers of +/+ marrow. An additional defect was the inability of W44/W44 stem cells to compete with genetically marked +/+ cells during erythroid repopulation. An estimate of the number of W44/W44 stem cells needed to compete with +/+ cells was provided by enriching W44/W44 progenitors fivefold. Twice as many enriched W44/W44 marrow cells as unfractionated +/+ cells were required to replace competitor cells. This suggests that there are up to 10 times fewer stem cells somewhere in the W44/W44 erythrogenerative pathway. The data support the conclusion that an erythroid progenitor less mature than the BFU-E is one of the cells most severely affected by expression of the mutant gene.     

Characterization of stem cells and progenitors of hemopoiesis by cell sorting.

Cell sorting has been used as a method for characterizing hemopoietic stem cells and progenitors. Fluorescent antibody-surface labels and changes in fluorescence polarization induced by in vitro stimulation with potential hemopoietic regulators were used. As detected by significant enrichment of CFU-S (pluripotent stem cells) in fluorescence-activated cell sorting, some CFU-S bear 'unique antigens' recognized by rabbit anti-human brain sera, human anti-human sperm sera, and 129 anti-F9 serum, but not A . TH anti-A . TL (Ia) ascites. Significant changes in fluorescence polarization induced by in vitro stimulation of mouse bone marrow with potential hemopoietic regulators were also observed; further, progenitors of human T-lymphocyte colonies were observed to exhibit a significantly decreased mean polarization value after short-term stimulation with PHA-LCM (phytohemagglutinin-stimulated leukocyte conditioned medium).     

Erythropoietin responses and physical characterization of erythroid progenitor cells in Rauscher virus infected BALB/c mice.

Infection of BALB/c mice with Rauscher leukemia virus (RLV) gives rise to pronounced erythrocytopoiesis manifesting in splenomegaly and is associated with progressive development of anemia. In the spleen erythroid colony forming units (CFU-E) increase exponentially up to 800-fold that of normal levels by the third week of infection. In vitro these CFU-E are dependent on erythropoietin for colony formation, their erythropoietin requirements being higher than that of CFU-E from normal mice. Numbers of CFU-E in spleen and degree of splenomegaly in anemic RLV infected mice were also shown to be modified by red blood cell transfusion, but progression of the disease was not stopped. Erythroid burst forming units (BFU-E) were also responsive to erythropoietin. However, a small proportion of cells also formed BFU-E colonies at concentrations which did not support growth of normal marrow BFU-E. When compared to normal, CFU-E found in RLV-infected spleen have similar velocity sedimentation rates. However, buoyant density separation of leukemic spleen cells indicated that CFU-E were more homogeneous (modal density 1.0695 g/cm3) than CFU-E from normal spleen. Analysis of physical properties of CFU-E and the nonhemoglobinized erythroblast-like cells, which accumulate in the spleen showed that they differed mainly in their distribution of cell diameter. Our findings show that erythroid progenitor cells in RLV infected mice are responsive to erythropoietin in vitro. Also in vivo erythropoiesis appears to be under control of erythropoietin but other factors which lead to progression of RLV disease apparently exist. Most proerythroblast-like cells, which are characteristic of this disease, apparently lack the potential to form colonies and may be more mature than CFU-E.     

A comparison of stem cell assays using early or late spleen colonies

The distribution of spleen colony diameters was determined 5.5, 8.0, 10.5 and 13.0 days after injection of normal bone marrow cells to lethally irradiated recipients. A relative lack of small colonies on day 8.0, as compared with days 5.5, 10.5 and 13.0, argued against a time continuum in colony appearance. The spleen colonies observed after 10 days or more probably represented a mixture of colonies which developed from the originally transplanted CFU-S and those arising from secondary CFU-S. Thus, late appearing spleen colonies may not necessarily identify a different, less mature, population of CFU-S. Administration of increasing amounts of bone marrow cells was used for comparing the linearity of the CFU-S assay for colonies observed after 8 days or after 12 to 13 days. The influence of overlapping colonies on the results was considerably augmented if large spleen colonies were observed after 12 or 13 days. Subsequently the CFU-S assay lost much of its quantitative character. We believe that some previously published data might have been misinterpreted by neglecting the important differences between 'early' and 'late' CFU-S assays.     

A model of hemopoietic stress in a lactate dehydrogenase mouse mutant with hemolytic anemia

A codominantly inherited mutation of the lactate dehydrogenase (LDH) in the C3H mouse causes a severe hemolytic anemia in homozygous mutants, whereas viability and fertility are close to normal. Investigation of multipotent hemopoietic stem cells (CFU-S), myeloid (GM-CFC) and erythroid progenitors (BFU-E, CFU-E) in femur and spleen indicates a general shift from bone marrow to splenic hemopoiesis. In terms of total body hemopoiesis, however, the BFU-E pool is 1.4- and the CFU-E pool 19-fold enlarged, whereas CFU-S and GM-CFC show little or no deviation from normal. It is concluded that this mouse mutant is an appropriate model of long-term hemopoietic stress showing that compensation in this severe hemolytic anemia is achieved primarily by an increase of the number of the most mature erythroid progenitors.     

In vitro effect of glucan on mouse hemopoietic committed progenitor cells

In vitro methylcellulose clonal cell culture assays of granulopoietic and erythropoietic colonies were used to study the primary effect of glucan on the hematopoietic stem cells. Addition of glucan to the cultures inhibits the formation of colony-forming units-erythrocytic (CFU-E), enhances the production of burst forming units-erythrocytic (BFU-E) and has no effect on colony-forming unit-culture (CFU-C). These results indicate that glucan has a direct effect on late and early erythroid precursor cells.     

Induction of commitment of murine erythroleukemia cells (TSA8) to CFU-E with DMSO

The commitment of novel mouse erythroleukemic (MEL) cells (TSA8) to colony-forming units of erythroid (CFU-E) by dimethylsulfoxide (DMSO) was investigated. After exposure to the inducer in liquid culture, the cells were transferred to a semi-solid culture to examine their ability to form erythroid colonies which were dependent on erythropoietin. Exposure to DMSO for 2 days is optimum for CFU-E type colony formation and colonies induced in this manner are equivalent to CFU-E. The induction occurred in a synchronous manner. Partly stained colonies appeared prior to CFU-E formation and are thought to be a result of asymmetric cell division. Appearance of these partly stained colonies suggested that the number of erythropoietin receptors is important in the complete responsiveness to erythropoietin. TSA8 cells constitute a suitable model system in which to analyse the mechanism of commitment in early erythropoiesis.     

Biosynthesis of globin chains in fetal liver and adult marrow cultures : Comparative analysis of individual colonies derived from early, intermediate or late erythroid progenitors

The relative synthesis of globin chains (α,β,^Gγ,^Aγ) has been comparatively evaluated in erythroid colonies from 26 fetal livers (7–15 gestational week) and 13 ‘normal’ adult marrows. Clusters deriving from erythroid colony-forming units (CFU-E) were analysed either individually or in pools of –20 colonies. Bursts deriving from earlier erythroid progenitors (erythroid burst-forming unit, ‘primitive’ or ‘mature’, P-BFU-E or M-BFU-E, respectively) were always analysed individually. Since γ-globin synthesis peaks earlier than β-chain production in both the fetal and the adult erythroblastic pathway, the globin synthetic pattern has been comparatively evaluated, in so far as possible, in colonies at an homogenous, advanced stage of hemoglobinization.In fetal liver cultures, the relative β-synthesis in CFU-E clusters, M- and P-BFU-E bursts constantly shows low, fairly uniform values. In adult marrow cultures, the relative γ-production in the corresponding three classes of colonies is characterized by low, rather homogeneous levels (except for more elevated γ-synthetic values occasionally observed in pooled CFU-E clusters comprising a majority of poorly-hemoglobinized colonies). A gradual decrease of relative γ-production has never been observed in colonies deriving from progressively more differentiated erythroid progenitors of both fetal and adult origin.These results suggest that fetal and adult BFU-E are endowed respectively with a program for prevailing HbF or HbA synthesis, which is not substantially modulated at the level of erythroid progenitors under standard culture conditions. By implication, it is postulated that, in fetal and more particularly adult age, modulation of globin synthesis is mediated via mechanism(s) acting at the level of erythroblasts, i.e. at the level of the early γ- and the late β-synthesis in their maturation pathway. The Hb switch (i.e. the switch from prevailingly HbF to HbA synthesis program) is possibly dependent on the ontogenic ‘maturation’ of BFU-E (and/or stem cells), which peaks in the perinatal period.     

A fractionation procedure of mouse bone marrow cells yielding exclusively pluripotent stem cells and committed progenitors

The cell surface phenotype of pluripotent hemopoietic stem cells (CFU-S) and committed progenitors (CFU-C1, CFU-C2, BFU-E) of mouse bone marrow was analyzed with respect to their binding of wheat germ agglutinin (WGA) and two monoclonal antibodies, anti-GM-1.2 and anti-PGP-1. Stained cells were fractionated on the basis of differences in fluorescence and light scatter intensity using a light-activated cell sorter. The 6% of the cells that bound most WGA and that also had a relatively high forward light scatter (FLS) and low perpendicular light scatter (PLS) contained nearly all stem cells (CFU-S) and progenitors. Anti-GM-1.2 stained only mature myeloid cells, not CFU-S or the in vitro colony-forming cells. Anti-PGP-1 stained all bone marrow cells in varying intensities: lymphoid cells were dull, CFU-S were intermediate, CFU-C2 were brighter, and mature myeloid cells very bright. Enrichment of progenitor cells was performed by a two-step sorting procedure. First, the 6% most WGA-binding cells with high FLS and low PLS were sorted out. A 10-15-fold enrichment of progenitors and CFU-S was obtained. Next, these cells were restained with anti-GM-1.2 or anti-PGP-1 and again fractionated on the FACS. The GM-1.2-negative cells were then another four- to sevenfold more enriched for stem cells and progenitors. Of the cells in this fraction, 95% could be assigned to a colony-forming unit. With anti-PGP-1, CFU-C2 could be partly separated from more early cells such as CFU-S and BFU-E.     

Effects of five recombinant hematopoietic growth factors on enriched human erythroid progenitors in serum-replaced cultures

Erythroid progenitors from normal human marrow were purified by a two-step immune panning method permitting both the enrichment of erythroid progenitors (plating efficiency up to 10%) and the separation of CFU-E from BFU-E. The purified erythroid progenitors were grown in serum-replaced conditions; in some experiments at an average of one cell per well. Human recombinant granulocyte-macrophage colony stimulating factor (GM-CSF), interleukin 3 (IL3), erythroid potentiating activity (EPA), and human erythropoietin (Epo) either recombinant or homogenous native were tested for their effect on CFU-E growth. Epo was an absolute requirement for CFU-E growth and was sufficient to obtain colony formation at the unicellular level whereas GM-CSF and IL3 did not further increase the plating efficiency. EPA potentiated the effect of Epo on this progenitor only in experiments performed at unicellular level. Human recombinant GM-CSF, IL3, Interleukin 1 alpha (IL1 alpha), and Epo were subsequently tested for their ability to promote BFU-E growth. GM-CSF and IL3 supported the growth of erythroid bursts in the presence of Epo, even at the unicellular level. However, IL3 promoted a higher number of bursts than GM-CSF under all conditions tested. These two growth factors have no or very small additive effects when tested in combination. IL1 alpha added to Epo alone had no effect on the growth of BFU-E whereas it potentiated the combined action of IL3 and GM-CSF on the primitive BFU-E. In conclusion, this study confirms at the unicellular level and under serum-free conditions that erythroid progenitors are regulated by multipotential growth factors in early phases of erythropoiesis and become sensitive only to Epo in later phases of differentiation.     

A stochastic model of self-renewal and commitment to differentiation of the primitive hemopoietic stem cells in culture

We recently identified a murine hemopoietic stem cell colony which consists of undifferentiated (blast) cells and appears to be more primitive than CFU-GEMM in the stem cell hierarchy. The progenitors for the colony which we termed “stem cell colony” possess an extensive self-renewal capacity and the ability to generate many secondary multipotential hemopoietic colonies in culture. We replated a total of 68 stem cell colonies from cultures of murine spleen cells and analyzed the number of stem cell–and granulocyte(neutrophil)-erythrocyte-macrophage-megakaryocyte (GEMM) colony-forming cells in individual stem cell colonies. Of the 68 stem cell colonies, 35 contained progenitors (abbreviated as “S”-cells) for stem cell colonies. The distributions of S-cells and CFU-GEMM in individual stem cell colonies were extremely heterogeneous. Neither the frequency distributions of S-cells nor CFU-GEMM in stem cell colonies could be fitted well by Poisson distribution. Rather, the frequency distribution of the s-cells could be approximated by a geometric distribution and that of CFU-GEMM by an exponential distribution, both of which are variates of the gamma distribution. Our observations are in agreement with those on the distributions of CFU-S in individual spleen colonies and provided support for a stochastic model for stem cell self-renewal and commitment in culture. Application of the theory of the branching process to the distribution of S-cells revealed a distributional parameter “p” of 0.589 which is also in agreement with the earlier report on the p value for reproduction of CFU-S.     

Colony formation in agar by adult bone marrow multipotential hemopoietic cells

Erythroid colony formation in agar cultures of CBA bone marrow cells was stimulated by the addition of pokeweed mitogen-stimulated spleen conditioned medium (SCM). Optimal colony numbers were obtained when cultures contained 20% fetal calf serum and concentrated spleen conditioned medium. By 7 days of incubation, large burst or unicentric erythroid colonies occurred at a maximum frequency of 40–50 per 10⁵ bone marrow cells. In CBA mice the cells forming erythroid colonies were also present in the spleen, peripheral blood, and within individual spleen colonies. A marked strain variation was noted with CBA mice having the highest levels of erythroid colony-forming cells. In CBA mice erythroid colony-forming cells were mainly non-cycling (12.5% reduction in colony numbers after incubation with hydroxyurea or 3H-thymidine). Erythroid colony-forming cells sedimented with a peak of 4.5 mm/hr, compared with CFU-S, which sedimented at 4.25 mm/hr. The addition of erythropoietin (up to 4 units) to cultures containing SCM did not alter the number or degree of hemoglobinisation of erythroid colonies. Analysis of the total number of erythroid colony-forming cells and CFU-S in 90 individual spleen colonies gave a correlation coefficient of r = 0.93 for these two cell types. In addition to benzidine-positive erythroid cells, up to 40% of the colonies contained, in addition, varying proportions of neutrophils, macrophages, eosinophils, and megakaryocytes. Taken together with the close correlation between the numbers of CFU-S in different adult hemopoietic tissues, including individual spleen colonies, the data indicate that the erythroid colony-forming cells expressing multiple hemopoietic differentiation are members of the hemopoietic multipotential stem cell compartment.