We have previously shown that removal of plastic-adherent cells completely abolishes the ability of normal human peripheral blood mononuclear cells (PB MNC) to produce erythroid bursts in serum-free (SF) medium.  Thus, plastic-adherent cells must provide an activity, as yet uncharacterized, that is important for proliferation and/or differentiation of early erythroid progenitor cells.  Since stromal cells of hemopoietic tissue are known to be the sites of production of human stem cell growth factor (SCF, c-kit ligand), we investigated the question of whether SCF would have any effect in reversing the loss of erythroid burst production following removal of cells by plastic adherence.  Recombinant human (rHu) SCF was added to the SF cultures of Ficoll-Hypaque density gradient separated MNC which had been exposed to 1.5 hrs. adherence to plastic.  We found that rHu SCF, in conjunction with either 0.25 mM hemin or 0.8 nM rHu interleukin-3 (IL-3), supported >50% of the maximum burst-formation observed before the adherence step.  No synergism between rHu SCF and rHu IL-3 was observed, but rHu SCF increased the cellularity of the burst-component colonies (BCC).  We conclude that SCF can replace a stimulatory activity of adherent cells and thus partially restore burst-formation.

 Recently, a novel growth factor termed stem cell factor, SCF, has been reported to stimulate primitive pluripotential hemopoietic progenitors in mice (1-2), rats (3) and humans (2-6).  Several groups have isolated SCF (7-11), also termed mast cell growth factor (MGF) (10) and have shown it to be the ligand for the c-kit  receptor protein (7-11).  The product of the c-kit  proto-oncogene has tyrosine kinase activity and maps to the mouse W  locus on chromosome 5 (12-13).  These findings demonstrated that the genetic anemias of W/W v and Sl/Sl d  mice were interconnected, with the former having  defects in the c-kit  receptor and the latter defective production of SCF (11,13-16), the c-kit receptor ligand being the product of the Steel (Sl ) locus (9,11,14).  Besides the soluble forms of SCF, membrane-bound forms have also been reported (10).  Serum-containing (SC) studies of the effects of SCF on highly enriched, adherent cell-deprived, human bone marrow (BM) BFU-E preparations have shown that it can potentiate primitive hemopoietic colony-formation, and have suggested that it synergises with  diverse cytokines, among which are Epo and IL-3 (3-6).  No comparable studies exist for human BFU-E from peripheral blood. 

We have previously reported the development of a novel serum-free (SF) medium capable of supporting the growth of circulating erythroid progenitors (18).  However, even though the medium was capable of supporting burst formation at low cell density, removal of adherent cells from the peripheral blood mononuclear  cell (PB MNC) suspensions prevented burst formation in SF medium (18). This observation contrasted sharply to what was observed when Polycythemia vera  PB MNC were similarly treated and plated under the same conditions: removal of plastic-adherent cells then had no effect (19).  In this report, we have investigated whether or not SCF could replace the putative activity provided by plastic-adherent cells and support normal burst-formation by non-adherent (NA) PB MNC.

Cell preparations

After informed consent, peripheral blood was obtained by venipuncture from healthy volunteer donors and was immediately placed in alpha-minimal essential medium ( a-MEM) containing 2% FCS (#SP80219, Gibco, Grand Island, N.Y.) and 10 U/ml of preservative-free sodium heparin (#820 5077MF, Gibco).  Mononuclear (MNC) cells were separated by Ficoll-Hypaque (Pharmacia, Montreal, P.Q., Canada) density- gradient centrifugation at 400 xg for 40 min. 

Adherent cells were removed by 90 min exposure to the plastic of 50 ml Falcon Tissue Culture Flasks (#3013, Becton Dickinson, N.Y.) in the presence of 2% FCS + a-MEM, without agitating the flask at the time of removal of the cell suspension.  Cell suspensions were washed three times (400 xg for 10 min.), the first wash in the presence of 2% FCS + a-MEM, and the subsequent two washes in a-MEM alone.  Cell counts were made with the Trypan Blue (0.4%, #630-5250, Gibco) dye-exclusion method.

Clonal Cell Culture

Serum-free culture of peripheral blood MNC has been previously described by us (18). Assays were performed in flat bottom Nunclon Delta SI 24 well multidishes (#1-43982, A/S Nunc, Roskilde, Denmark). Between 5 x 104 and 1 x 105 PB MNC were plated in 0.5 ml of final culture medium containing  a-MEM, 0.8% of 1,500 centipoise methylcellulose (Methocel A4M, premium grade, Dow Chemical Co., Midland, Ml), 1% fatty acid- and globulin-free crystalized BSA (Sigma) which was subsequently  deionized with analytical grade Ion Exchange Resin (AG 501-X8 D), BioRad Labs, Richmond, CA), 2 x 104 ß-mercaptoethanol (BDH Biochemicals, Poole, England), 270 mg/ml  fully iron-saturated bovine transferrin (Sigma), 7 x 10-7M d-a-tocopherol, Clinic Products, Windsor, Ont.), 3 x 10-8M retinyl acetate (Nutritional Biochemical Corp., Cleveland, OH), 8 mg/ml L-a-phosphatidyl choline dipalmitoyl synthetic (Sigma), 5.6 mg/ml porcine liver cholesterol, grade 1 (Sigma), 10 mg/ml of each of the four deoxy- and ribonucleosides (Sigma), 2 mM L-glutamine (Sigma), 100 U/ml penicillin G and 50 mg/ml streptomycin sulfate (Gibco), plus combinations of several growth factors at the indicated concentrations. Bovine type 1 hemin (ferric chloride protoporphyrin IX) was purchased from Sigma (#H-2250), with approximately 97% purity by spectrophotometric assay and used at a final concentration of 0.25 mM. The multidishes were incubated at 37oC in a humidified atmosphere and 5% CO2 for 14 days.  All erythroid colonies and bursts were scored by in situ observation with an inverted microscope.  All assays were done in triplicate and results are expressed as the means +/- SEM.

Recombinant Growth Factors

E. coli-derived recombinant human somatomedin-C (referred to as rHu IGF-1) having the natural amino acid sequence and the recombinant human preparations of Epo (rHu Epo) and Interleukin-3 (rHu IL-3) were generous gifts from AMGen, Thousand Oaks, CA, courtesy of Dr. J Egrie. Recombinant human SCF was a  gift from Dr. A. Bernstein (S. Lunenfeld research institute, Ont.). The final concentrations of these GFs used in the optimal medium were as follows: 100nM rHu IGF-1, 1.8nM (6.0 U/ml) rHu Epo, 0.8nM  rHu IL-3;  rHu SCF was utilized at the concentrations shown.   For purposes of calculating rHu SCF molarities, a molecular weight of 35 Kd (7) was chosen.

Criteria for Scoring of Erythroid Colonies and Bursts

An erythroid burst is defined as either a single colony or a cluster of (burst-component) colonies, each having =/> 50 hemoglobinized  cells, scored at day 14 of growth.  Even though hemoglobinized  colonies with < 50 cells could also be observed, and often appeared to belong to a burst, their counts were not included unless > 2 such colonies were contiguous. 

Statistical Analysis

Mean +/- standard error (SEM) number of bursts and burst-component colonies (BCC) for each 3 dishes were calculated.  Data sets were compared using the Student two tailed t test.  Curve smoothing (distance weighted least squares) was performed using a SYSTAT 5.0 software programme (SYSTAT Inc., Evanston IL).

 When rHu SCF was added to cultures of untreated PB MNC under SF conditions at concentrations ranging from 30 to 1500 pM (1 to 50 ng/mL) in the absence of IL-3, they showed virtually no change in the number of bursts (Fig. B1, closed circles) or of burst-component colonies (BCC) (Fig. B2, closed circles).  In the presence of 0.8 nM rHu IL-3, no significant increase (t =2, with 2 df and  p>0.1) in burst-formation was observed at 1.5 nM rHu SCF (Fig. B1, open circles).  In contrast, adherent cell-depleted PB MNC showed a significant increase of burst- and BCC-formation upon the addition of as little as 30 pM rHu SCF (Figs. B1 and B2, closed squares), but no increase upon the addition of IL-3 (in the presence of hemin and 1.5 or 3.0 nM rHu SCF). It is unclear whether burst-formation under these conditions has reached a plateau of rHu SCF activity at the highest concentration tested (in the presence of IL-3) of 3nM (Fig. B2, closed squares), inasmuch as the BCC number appeared to plateau at ca. 0.7 nM (Fig. B1, closed squares) but the burst number did not.  The control point shown in parentheses (closed square at 0 rHu SCF in Fig. B2) represents erythroid colonies with <50 cells  (an average of 25) which are promoted by the BPA-like action of hemin alone and were tentatively grouped as clusters, possibly equivalent to undeveloped bursts (Fig. B1, closed square in parentheses).  Some of these small hemin-dependent colonies approached the 50-cell size, and the  effect of rHu SCF, under these conditions (in the absence of IL-3), was to increase the cellularity (up to 200 cells) of the BCC; it also rendered the colonies rounder and more compact. It thus allowed full burst development.

A comparison of the effects of rHu SCF and of the two defined BPA-like activities used in our SF medium (18), hemin and rHu IL-3, upon burst-formation shows that, before the removal of adherent cells, hemin accounted for 62% of the maximum number of bursts scored when all three factors were present (cf open bars 1 and 3, Fig. B3).  Under the same conditions, hemin accounted for 58% of the maximum number of BCC (cf open bars 1 and 3, Fig. B4).  In the absence of hemin, and in the presence of both SCF and IL-3 (Figs. B3 and B4, open bars 4), the number of bursts obtained (37.5% of the maximum) is slightly greater than half the number of bursts obtained with hemin alone, and the number of BCC (32% of the maximum) was  48%  of that obtained with hemin alone (cf open bars 1 and 4, Figs. B3 and B4).  This low efficiency of colony-formation in the absence of hemin seems to be due to a lack of hemoglobinization, for many other colonies morphologically recognizable as erythroid and with >50 cells could be seen.  Optimal conditions for burst-formation appeared to require the presence of all three factors, rHu IL-3, rHu SCF and hemin (Fig. B3, open bar 3), even though the addition of rHu IL-3 to both hemin and rHu SCF did not significantly increase the number of BCC observed (cf open bars 2 and 3, Fig. B4).  Removal of adherent cells appeared to prevent the capacity of hemin to promote erythroid colonies large enough for them to be classified as BCC; the number of erythroid colonies with 20 to 50 hemoglobinized cells (Fig. B4, striped bar 1) and of "undeveloped bursts" (Fig. B3, striped bar 1) were fewer than the number of BCC and bursts formed in the absence of hemin (Figs. B3 and B4, closed bars 4).  Finally, addition of rHu IL-3 to a combination of hemin and rHu SCF did not significantly increase the number of bursts obtained with hemin and rHu SCF alone (cf closed bars 2 and 3, Fig. B3) and only marginally increased the number of BCC obtained with hemin and rHu SCF (cf closed bars 2 and 3, Fig. B4).  

We have previously reported that a SF medium designed for the growth of normal PB BFU-E was unable to support erythroid burst-formation when the MNC suspensions plated were depleted of adherent cells, even though the medium contained plateau concentrations of hemin (0.25 mM), rHu Epo (> 0.9 nM), rHu IGF-1 (100 nM) and rHu IL-3 (0.8 nM) (18).  In the present work, 2 out of 4 normal donors examined showed day 14 erythroid colony formation under these conditions (and with or without Epo), but the colonies had <50 hemoglobinized  cells and did not classify either as BCC or bursts (results not shown). Our results show that, for burst-formation to be detected in SF medium when PB MNC were depleted of adherent cells, SCF was required (30 pM).  SCF had this capability as long as either hemin or IL-3 were also present in the medium (results not shown). Qualitatively, the effect of SCF was one of increasing the cellularity of the BCC and the compactness of their morphology.  

In consonance with the reported ability of SCF to stimulate the growth of primitive, pluripotential hematopoietic progenitors (1-6), our results also suggest that SCF has a BPA-like activity, and that, as others have suggested, it might be directly acting upon the earliest BFU-E (5-6), possibly by recruiting them into cycle.  If this ultimately proves to be the case, SCF may provide the function of the BPA found in fetal calf serum shown to be neither IL-3 nor GM CSF (20), or it may substitute for the BPA that has been shown to be the product of adherent cells (21). 

It should be pointed out that removal of adherent cells was only partially restored by rHu SCF: under optimal conditions, burst number reached only half of that attained without adherent cell removal and the number of BCC was 40% lower.  Three explanations may be given for these findings:

1)  No plateau of rHu SCF activity was reached; it is possible that concentrations of rHu SCF higher than 100 ng/mL are necessary to return burst-formation to normal.  Other workers have also found no plateau for SCF activity, in the presence of rHu IL-3, using concentrations identical to ours, although in SC conditions (5).

2)  It is possible that, unlike BM BFU-E, normal PB BFU-E, which are not actively undergoing DNA synthesis (22-23), may more frequently adhere to plastic; this would result in the recovery of  fewer erythroid progenitors with plastic-adherent PB MNC than with BM MNC.  If so, increasing the concentration of rHu SCF would be ineffective.

3) Lastly, it is possible that factors other than SCF may also be required to fully restore burst-formation. Future work may answer these questions.

1.   Zsebo K, Martin F, Suggs S, Wypych J, Lu H, McNiece I et al (1990) Biological characterization of a unique early acting hematopoietic growth factor.  Exp Hematol 18:703 (abs).

2.   Broxmeyer HE, Cooper S, Lyman SD, Williams DE (1990) Characteristics of a murine ligand for c-kit (MGF): a stimulating/enhancing factor for early murine and human bone marrow hemopoietic progenitor cells.  Blood 76:134a (abs).

3.  Migliaccio G, Migliaccio  AR, Valinsky  K et al.  (1990) Recombinant rat stem cell factor (rrSCF) induces proliferation and differentiation of primitive hematopoietic progenitor cells (HPC) in serum-deprived cultures.  Blood 76:156a (abs).

4.  McNiece I, Langley K, Zsebo K (1990) Recombinant human stem cell factor (rHSCF) synergises with CSFs and Epo to stimulate colony formation of myeloid and erythroid cells. Blood 76: 154a (abs).

5.  McNiece I, Langley K, Zsebo K (1991) Recombinant human stem cell factor synergises with GM-CSF, G-CSF, IL-3 and Epo to stimulate human progenitor cells of the myeloid and erythroid lineages.   Exp Hematol 19:226.

6.  Broxmeyer HE, Cooper S, Lu L et al.  (1991)  Effect of murine mast cell growth factor (c-kit proto-oncogene ligand) on colony formation by human marrow hematopoietic progenitor cells.  Blood 77: 2142.

7.  Zsebo K, Wypych J, McNiece IK,  et al. (1990) Identification, purification,  and biological characterization of hematopoietic stem cell factor from buffalo rat liver-conditioned medium.  Cell 63:195.

8.  Martin FH, Suggs SV, Langley KE et al.  (1990) Primary structure and functional expression of rat and human stem cell factor DNAs.  Cell 63:203.

9.   Zsebo K, Williams DA, Geissler EN et al.  (1990) Stem cell factor is encoded at the SI locus of the mouse and is the ligand for the c-kit tyrosine kinase receptor.  Cell 63:213.

10.  Anderson DM, Lyman SD, Baird A et al.  (1990) Molecular cloning of mast cell growth factor, a hematopoietin that is active in both membrane bound and soluble forms.  Cell 63:235.

11.  Huang E, Nocka K, Beier DR et al. (1990)  The hematopoietic growth factor KL is encoded by the SI locus and is the ligand of the c-kit receptor, the gene product of the W locus.  Cell 63:225.

12.  Chabot B, Stephenson DA, Chapman VM et al. (1988) The proto-oncogene c-kit  encoding a transmembrane tyrosine kinase receptor maps to the mouse W  locus.  Nature 335:88.

13.  Geissler EN, Ryan MA, Housman DE, (1988) The dominant-white spotting (W ) locus of the mouse encodes the c-kit proto-oncogene.  Cell 55:185.

14.  Copeland NG, Gilbert DJ, Cho BC et al. (1990) Mast cell growth factor maps near the Steel locus on mouse chromosome 10 and is deleted in a number of Steel alleles.  Cell 63:175.

15.  Williams DE, Eisenman J, Baird A et al. (1990) Identification of a ligand for the c-kit  proto-oncogene.  Cell 63:167.

16. Witte ON  (1990) Steel locus defines new multipotent growth factor.  Cell 63:5.

17.  Flanagan JG, Leder P (1990) The kit ligand: a cell surface molecule altered in Steel mutant fibroblasts. Cell 63:185.

  18. Correa PN, Axelrad AA (1991) Production of erythropoietic bursts by progenitor cells from adult human peripheral blood in an improved serum-free medium: role of IGF-I. Accepted for publication by Blood.

19.  Correa PN, Axelrad AA (1991) Circulating erythroid progenitors in Polycythemia vera are hypersensitive to IGF-1. In preparation.

20.  Sonoda Y,  Yang Y, Wong GG et al.  (1988) Erythroid burst-promoting activity of purified recombinant human GM-CSF and Interleukin-3: studies with anti-GM-CSF and anti-IL-3 sera and studies in serum-free cultures.  Blood 72:1381.

21.  Dainiak N, Kreczko S, Cohen A et al. (1985) Primary human marrow cultures for erythroid bursts in a serum-substituted system.  Exp Hematol 13:1073.

22.  Eaves AC, Eaves CJ (1984) Erythropoiesis in culture. Clin in Haematol 13:371.

23.  Shekhter-Levin S, Amato D, Karrass L, Axelrad AA (1985) Heterogeneity of buoyant density and proliferative state of circulating erythropoietic progenitor cells (BFU-E) in man. Exp Hematol 13:1138.