Main

Umbilical cord blood is a rich source of the hematopoietic progenitor cells(1–3) that present a potential alternative to the use of bone marrow progenitors for hematopoietic reconstitution(4, 5). However, the biologic and functional characteristics of the cord blood progenitors are less well characterized than those of adult bone marrow cells.

Although the formation of colonies of GM in response to purified CSFs has been compared between cord blood and bone marrow(6, 7), the cooperative effects of CSFs between the two cell sources is controversial. Thus, although Gabutti et al.(6) showed that the use of SCF in combination with GM-CSF and IL-3 induced a greater increase in the cloning efficiency of cord blood cells than in that of bone marrow cells. Cairo et al.(7), using the same combination of CSFs, failed to detect a significant difference in colony formation between the two cell sources. This discrepancy may be attributable to a masking of the synergistic effects of multiple CSFs that act on multiple cell lineages.

G-CSF acts on the later stages of granulopoiesis and is widely used clinically. SCF, the ligand for the c-Kit receptor, promotes the proliferation of multiple hematopoietic lineages at the earliest stage of hematopoiesis(8). SCF also acts synergistically with other growth factors to induce colony formation by human cord blood cells(9, 10) and bone marrow(11, 12) cells. To clarify the relative responsiveness of cord blood and bone marrow progenitors to CSFs, we have now compared the synergistic effects of G-CSF and SCF on GM colony formation between cord blood and bone marrow CFU-C under the same culture conditions.

METHODS

Collection and fractionation of umbilical cord blood and bone marrow cells. Heparinized cord blood samples were collected from the umbilical cord and placenta immediately after normal delivery. Bone marrow samples were obtained from healthy donors for allogeneic transplantation after informed consent had been obtained. Within 24 h of harvesting from the donor, all specimens were subjected to centrifugation on Ficoll-sodium metrizoate(Muto Pure Chemicals, Tokyo, Japan) for 30 min at 400 × g. Isolated low density (<1.077 g/mL) mononuclear cells were diluted in Iscove's modified Dulbecco's medium (Life Technologies, Inc., Grand Island, NY) containing 10% heat-inactivated fetal bovine serum (ICN Pharmaceuticals, Costa Mesa, CA) and were depleted of monocytes by adherence to plastic flasks for 3-24 h at 37 °C. Hematopoietic progenitor cells that expressed the CD34 antigen were purified according to a two-step standardized procedure(13, 14), which consisted of enrichment by the elimination of cells that attached to covalently immobilized soybean agglutinin (SBA CELLector flask; Applied Immune Science, Menlo Park, CA) followed by the positive selection of CD34+ cells on a polystyrene surface coated with the ICH3 MAb to CD34 (CD34 CELLector flask; Applied Immune Science).

Hematopoietic growth factors. All growth factors were recombinant human CSFs. G-CSF was obtained from Chugai (Tokyo, Japan); SCF, GM-CSF, and IL-3 were from Kirin-Amgen (Tokyo, Japan) and Sankyo (Tokyo, Japan). The specific activities of G-CSF, GM-CSF, and IL-3 were 1.28 × 108 U/mg of protein, 1.91 × 108 U/mg, and 1.0 × 108 U/mg, respectively.

Hematopoietic progenitor cell assay. Myeloid progenitor cells(CFU-C) were assayed by a modified version of a semi-solid agar method originally described by Pike and Robinson(15) and performed in 35-mm plastic dishes. Isolated CD34+ cells (1 × 103) were plated in 1 mL of Iscove's modified Dulbecco's medium containing 0.3% agar (Bacto agar; Difco, Detroit, MI), 20% heat-inactivated fetal bovine serum, 1% deionized BSA (Sigma Chemical Co., St. Louis, MO), 50μmol/L 2-mercaptoethanol (Nacalai Tesque, Kyoto, Japan), and various combinations and concentrations of CSFs. All CSFs were added at the start of the cultures. Cells were incubated at 37 °C in a humidified incubator with an atmosphere of 5% CO2. After 14 d, the cells in the agar dishes were fixed and subjected to esterase double staining. The numbers of colonies(aggregates of ≥40 cells) were recorded from duplicate or triplicate plates. For determination of colony size, 1.2% methylcellulose (MS-1500; Shin-etsu, Tokyo, Japan) was used instead of agar, and after 14 d of incubation, the colonies were individually plucked, and the number of cells in each colony were counted.

Immunologic staining and flow cytometry. Purified progenitor cells expressing the surface antigen CD34 were identified by immunofluorescence with an anti-HPCA-2 MAb(16) (clone 8G12; Becton Dickinson, San Jose, CA), which recognizes a site on CD34 distinct from that recognized by the anti-CD34 MAb used for positive selection. Labeled cells were analyzed with an EPICS Elite flow cytometer(Coulter Electronics, Hialeah, FL). Cells expressing the surface antigen c-Kit were stained with phycoerythrin-conjugated antibodies to this protein(Immunotech, Marseilles, France). The purity and viability of HPCA-2+ cells were >80 and >90%, respectively.

Statistical analysis. Results are expressed as means ± SD. Statistical significance was assessed by Mann-Whitney's U test for unpaired data and the Wilcoxon signed-rank sum test for paired data. A p value of <0.05 was considered statistically significant.

RESULTS

For both bone marrow and cord blood progenitors, SCF alone induced a markedly smaller effect on colony growth than G-CSF, IL-3, or GM-CSF, whereas GM-CSF or IL-3 showed an additive effect and SCF showed a synergistic effect on GM colony formation when combined with G-CSF (Fig. 1). The maximal colony growth from cord blood was apparent with the combination of G-CSF and SCF. The synergism of G-CSF and SCF was therefore investigated further to compare the responsiveness of the two cell sources to CSFs.

Figure 1
figure 1

Effects of G-CSF, GM-CSF, IL-3, and SCF on GM colony formation from isolated CD34+ progenitors from cord blood(left) and bone marrow (right). CD34+ cells(103/mL) were plated in agarose culture medium containing the indicated combinations of CSFs, each at 100 ng/mL. Data are means ± SD of the number of colonies per 103 cells for triplicate culture dishes. Similar results were obtained in four additional experiments.

Full size image

G-CSF alone induced the formation of GM colonies from both cord blood and bone marrow CD34+ cells in a dose-dependent manner, with the maximal effect apparent at 100 ng/mL (Table 1). In the presence of optimal (100 or 500 ng/mL) or suboptimal (1 or 10 ng/mL) concentrations of G-CSF, the number of GM colonies was significantly higher for bone marrow than for cord blood. SCF (100 ng/mL) alone had virtually no effect on colony growth from bone marrow, but it induced the formation of a significant number of colonies from cord blood. Combining SCF with various concentrations of G-CSF resulted in a marked increase in colony formation from both cell sources; the SCF-induced increase in colony number was significantly higher for cord blood than for bone marrow in the presence of G-CSF at concentrations of ≥1 ng/mL.

Table 1 Effects of G-CSF concentration on colony formation from cord blood and bone marrow progenitors in the absence or presence of SCF
Full size table

The G-CSF dose-response data were also expressed as a percentage of the maximal number of colonies formed (Fig. 2). In the presence of G-CSF alone at 10 ng/mL, the percentage of colonies formed was significantly higher for bone marrow than for cord blood (77.4 ± 13.1versus 57.1 ± 9.9%, p < 0.01). SCF increased the sensitivity of cord blood CFU-C to all suboptimal doses of G-CSF, whereas it increased the sensitivity of bone marrow CFU-C to G-CSF only at 0.1 ng/mL.

Figure 2
figure 2

G-CSF dose-response curves in the absence or presence of SCF for GM colony formation from isolated CD34+ progenitors from cord blood (CB) and bone marrow (BM). CD34+ cells(103/mL) were cultured in the presence of various concentrations of G-CSF (0-500 ng/mL) alone or combined with SCF (100 ng/mL). Data are from Table 1 and are expressed as means of the percentage of maximal colony growth. *p < 0.05, G-CSF alone vs G-CSF + SCF for CB or BM as indicated; †p < 0.05, CB vs BM with SCF alone; ‡p < 0.01, CB vs BM with G-CSF alone.

Full size image

In the presence of G-CSF at a concentration of 1 ng/mL, the average increases in GM colony number were 2.0- versus 1.4-fold (not significant) in response to SCF at 1 ng/mL and 6.6- versus 3.2-fold(p < 0.01) in response to SCF at 10 ng/mL for six cord blood and six bone marrow samples, respectively.

In combination with G-CSF at concentrations of 1 or 100 ng/mL, SCF (100 ng/mL) significantly increased the size of colonies generated from cord blood CFU-C (Fig. 3A). For bone marrow, SCF increased the size of colonies in the presence of G-CSF at 100 ng/mL but not at 1 ng/mL (Fig. 3B).

Figure 3
figure 3

Effects of G-CSF and SCF on the size of GM colonies generated from cord blood (A) and bone marrow (B). Progenitors were cultured with G-CSF (1 or 100 ng/mL) in the absence or presence of SCF (100 ng/mL), and the number of cells per colony was determined. Statistical significance was assessed by Mann-Whitney's U test. NS, not significant.

Full size image

In three samples of cord blood, c-Kit was expressed on 38.1, 40.7, and 46.0% of cells, values that did not appear to differ from those obtained with bone marrow samples (27.9, 36.7, and 50.7%). The frequency of c-Kit+ cells among HPCA-2+ cells in a cord blood sample (54.8%) was also similar to that for a bone marrow sample (59.4%). Differences in the intensity of c-Kit expression also were not apparent between the two cell sources.

DISCUSSION

We have shown that GM colony formation by cord blood CFU-C is markedly more sensitive to SCF in combination with G-CSF, especially at suboptimal concentrations of G-CSF, than is GM colony formation by bone marrow CFU-C.

Previous studies have shown that the responsiveness of cord blood to CSFs differs from that of bone marrow hematopoietic progenitors(7, 8, 17–19). Such differences may be attributable to differences in cell composition. However, the purity of our CD34+ cell preparations was virtually identical for cord blood and bone marrow. The percentage of mature hematopoietic cells has been shown to be 0-5% for CD34+ cells purified by the same method as that used in the present study(13). It is therefore unlikely that mature cells affect colony formation by cord blood progenitors.

Cairo et al.(7) showed that the combination of SCF, IL-3, and IL-6 induced the highest increase in GM colony formation from cord blood CD34+ cells, whereas the most effective combination of CSFs for adult bone marrow was SCF, IL-3, and GM-CSF(7). Because IL-6 induces colony formation from cells that are more immature than those affected by GM-CSF, the researchers suggested that cord blood contains a higher percentage of earlier progenitor cells than does adult bone marrow. If the greater responsiveness of cord blood to SCF in our study was attributable to differences in composition of the two cell preparations, the percentage of c-Kit+ CD34+ cells would be expected to be greater for cord blood than for adult bone marrow. However, we failed to detect such a difference in the number of cells expressing c-Kit. Characterization of c-Kit+ CD34+ cells according to the intensity of c-Kit expression showed the c-Kitlow cell fraction to be richer in primitive progenitor cells than the c-Kithigh fraction(20, 21). The c-Kit mRNA is present in primitive hematopoietic progenitor cells, with the c-Kit protein being abundantly expressed in more differentiated CD34+ progenitor cells(22). Further investigation is required to clarify whether cord blood contains a higher percentage of more primitive progenitor cells, such as c-Kitlow CD34+ cells or c-Kit mRNA+ CD34+ cells, than does bone marrow.

Rather than being attributable to a difference in cell composition, the difference in responsiveness between progenitors from cord blood and bone marrow may be due to qualitative changes in the cells. Lansdorp et al.(19) detected ontogeny-related differences in CD34+ cells from fetal liver, cord blood, and adult bone marrow with regard to the ability to generate CD34+ daughter cells. Furthermore, Kubo and Nakahata(23) showed that erythroid progenitors in cord blood were more responsive to SCF than those in adult bone marrow or peripheral blood. The latter researchers suggested that erythroid progenitors expressing c-Kit may be anchored to bone marrow stromal cells by interaction of c-Kit with SCF and that the immaturity of this system in newborns may allow the progenitors to circulate in peripheral blood. In contrast, progenitors in the well developed hematopoietic environment of adult bone marrow may be affected by stromal cells directly or indirectly, resulting in a greater responsiveness to lineage-specific CSFs. Our observation that myeloid progenitors in cord blood showed a greater responsiveness to SCF than those in adult bone marrow may therefore also be related to the ontogenetic development of the c-Kit-SCF system and its immaturity in cord blood myeloid progenitors. Ontogeny-related differences in the affinity of c-Kit for SCF or in intracellular signal transduction may also explain the quantitative differences in the responsiveness of the two hematopoietic progenitor cell populations to CSFs.

Although G-CSF acts at later stages of granulopoiesis, it also has an effect on multipotential progenitors(24). In combination with SCF, differences in this earlier action between the two cell sources may contribute to the difference in responsiveness to G-CSF observed in the present study. Little is known about the G-CSF receptors on hematopoietic progenitor cells. Shimoda et al.(25) detected no significant difference in the percentage of cells expressing the G-CSF receptor between cord blood and bone marrow progenitors. It is possible that the low responsiveness of cord blood progenitors to G-CSF reflects an immaturity of the G-CSF receptor.

The concentration of SCF in normal human plasma is ≈3 ng/mL(26). In our study, the synergistic effect of SCF was apparent at concentrations of 10 and 100 ng/mL. Whether the administration of SCF as a supplemental therapy for cord blood transplantation would prove beneficial by enhancing the sensitivity of progenitors to G-CSF should be clarified by further investigation.