Ikuo Kashiwakura, PhD, Department of Radiological Technology, Hirosaki University School of Health Sciences, 66–1 Hon-cho, Hirosaki, Aomori 036–8564, Japan. E-mail: email@example.com
Summary. We investigated whether basic fibroblast growth factor (bFGF) is effective in inducing ex vivo expansion of CD34+ haematopoietic progenitor cells derived from human placental and umbilical cord blood. bFGF significantly promoted the clonal growth of various haematopoietic progenitor cells, including granulocyte–macrophage colony-forming units (CFU-GM), mixed colony-forming units (CFU-Mix) and megakaryocyte colony-forming units (CFU-Meg) under semisolid culture conditions, with an optimal bFGF concentration of 30 ng/ml. CD34+ cells were then cultured in serum-free liquid medium containing various combinations of early-acting cytokines, including thrombopoietin (TPO), stem cell factor (SCF), interleukin 3 (IL-3) and flt3-ligand (FL), with or without bFGF, for 6 and 12 d. Without bFGF, TPO + IL-3, TPO + SCF + FL and TPO +SCF + IL-3 + FL dramatically increased the total numbers of erythroid progenitors, CFU-GM and CFU-Mix by d 12 of culture respectively. However, the addition of bFGF did not promote further proliferation of these progenitors, except for the erythroid progenitors, by d 6 when stimulated with all four cytokines. In contrast, total CFU-Meg numbers were approximately doubled when these cultures were supplemented with bFGF, producing 100- to 120-fold increases compared with the baseline control cultures. These results suggest that bFGF is effective in supporting the generation of megakaryocytic progenitor cells during ex vivo expansion.
Long periods of severe thrombocytopenia have been a problem in many patients who have undergone human bone marrow (BM) or placental and umbilical cord blood (CB) transplantation, as a result of the large proportion of primitive haematopoietic stem/progenitor cells in these sources (Conrad & Emerson, 1998; Huang et al, 1998; Gluckman, 2000). Recently, several studies have used CD34+ cells and/or mononuclear cells mobilized from peripheral blood (PB) rather than marrow cells as a means of shortening the period of cytopenia following high-dose chemotherapy (Cutler & Antin, 2001). This shortening is probably due to the presence of large numbers of differentiated haematopoietic progenitor cells within the mobilized PB autografts. A new approach to the treatment of thrombocytopenia may, therefore, be to transplant differentiated haematopoietic progenitor cells, such as ex-vivo-expanded committed progenitor cells from CB, PB and BM. Indeed, the transplantation of differentiated haematopoietic progenitor cells has been shown to shorten the thrombocytopenic period (Bachier et al, 1999; McNiece et al, 2000; Paquette et al, 2002).
Basic fibroblast growth factor (bFGF) is a member of the heparin-binding growth factor family, and is a multifunctional growth factor known to be involved in angiogenesis as well as the differentiation of endothelial cells and fibroblasts (Allouche, 1995). It also stimulates the formation of an adherent stromal cell layer in human long-term bone marrow cultures, while simultaneously promoting haematopoietic cell development (Oliver et al, 1990; Wilson et al, 1991). Moreover, bFGF has been shown to synergize with other haematopoietic growth factors to enhance colony formation by several classes of haematopoietic progenitor cells in vitro (Gabbianelli et al, 1990; Bikfalvi et al, 1992; Avraham et al, 1994). Although the possibility of using bFGF to induce ex vivo expansion is attractive because of its multifunctional characteristics, few studies have investigated this possibility to date.
In this study, the potential of bFGF to act upon freshly prepared CD34+ cells from human CB and induce the ex vivo expansion of haematopoietic progenitor cells was assessed in serum-free cultures containing various combinations of early-acting cytokines, including thrombopoietin (TPO), stem cell factor (SCF), interleukin 3 (IL-3) and flt3-ligand (FL). We focused particularly on the expansion of megakaryocytes and megakaryocytic progenitor cells.
Materials and methods
Reagents. Recombinant human bFGF was provided by Kaken Pharmaceuticals (Tokyo, Japan). Recombinant human TPO, human SCF and human IL-3 were kindly provided by Kirin Brewery, (Tokyo, Japan). FL was purchased from Gibco BRL (Grand Island, USA). Recombinant human granulocyte colony-stimulating factor (G-CSF) and erythropoietin (Epo) were purchased from Sankyo (Tokyo, Japan). Recombinant human granulocyte–macrophage colony-stimulating factor (GM-CSF) was kindly provided by Shering-Plough (Ireland). The amounts of these factors added per millilitre of medium were: TPO, 50 ng; SCF, 100 ng; IL-3, 50 ng; FL, 100 ng; G-CSF, 10 ng; GM-CSF, 10 ng; and Epo, 4 units. Fluorescence-labelled monoclonal antibodies (mAb) were obtained as follows: fluorescein isothiocyanate (FITC)-conjugated anti-human CD34 (FITC-CD34), FITC-conjugated anti-human CD41 (FITC-CD41), phycoerythrin (PE)-conjugated anti-human CD41 (PE-CD41), PE-cyanin 5·1 (PC5)-conjugated anti-human CD45 (PC5-CD45) and allophycocyanin (APC)-conjugated anti-human CD56 (APC-CD56) were purchased from Beckman-Coulter-Immunotech (Marseille, France). FITC-conjugated anti-human CD3 (FITC-CD3), PE-conjugated anti-human CD3 (PE-CD3) and APC-conjugated anti-human CD8 (APC-CD8) were purchased from Pharmingen (Becton Dickinson Biosciences, Franklin Lakes, USA). PE-conjugated anti-human CD4 (PE-CD4) and FITC-conjugated anti-human CD19 (FITC-CD19) were purchased from Becton Dickinson Biosciences. Mouse IgG1-FITC, mouse IgG1-PE, mouse IgG1-PC5 and mouse IgG1-APC (Becton Dickinson Biosciences) were used as isotype controls.
Collection of CB and CD34+ cell purification. After obtaining informed consent from the mothers, CB was collected at the end of full-term deliveries using a sterile collection bag containing citrate–phosphate dextrose as anticoagulant, according to the guidelines of the Tokyo Cord Blood Bank. Light-density mononuclear CB cells were separated by centrifugation on Ficoll–Paque (1·077 g/ml; Amersham Pharmacia Biotech AB, Uppsala, Sweden) for 30 min at 300 g and washed three times with phosphate-buffered saline (PBS) containing 5 mmol/l EDTA (EDTA-PBS). They were then processed for CD34+ cell enrichment according to the manufacturer's instructions, and magnetic cell sorting (Miltenyi Biotec, Germany) was used for the positive selection of CD34+ cells. At the end of the procedure, the CD34+ cell recovery from the light-density mononuclear cells was approximately 0·13–0·66%, and the purity, measured using a fluorescence cell analyser (EPICS-XL; Beckman-Coulter, Tokyo, Japan), was 90–95%.
Ex vivo expansion. CD34+ cells were placed into 24-well plates (Falcon; Becton Dickinson Biosciences) at a concentration of 1–2 × 103 cells per well. Each well contained 0·5 ml liquid medium plus the growth factors of interest. The culture medium consisted of serum-free Iscove's modified Dulbecco's medium (IMDM; Gibco BRL) supplemented with BIT9500 (StemCell Technologies, Vancouver, Canada), a serum substitute for serum-free culture. The cytokines, used in various combinations, included recombinant human TPO, IL-3, SCF and FL. The cultures were incubated at 37°C in a humidified atmosphere containing 5% CO2. On d 6 and 12, cells were harvested from the culture and the number of viable cells was determined by trypan blue exclusion. Cell counts were also performed using a haemocytometer. The cells from these aliquots were assayed for the number of haematopoietic progenitor cells. Erythroid burst-forming units (BFU-E), granulocyte–macrophage colony-forming units (CFU-GM) and mixed colony-forming units (CFU-Mix) were assayed using the methylcellulose culture technique. The number of megakaryocyte colony-forming units (CFU-Meg) was measured using a plasma clot culture technique. The total number of each type of progenitor cell was calculated from the total number of cells harvested and the number of each type of colony per well.
Methylcellulose cultures. Colony-forming cells, including BFU-E, CFU-GM and CFU-Mix, were assayed by methylcellulose culturing using MethoCult7 (StemCell Technologies). Freshly prepared CD34+ cells (d 0) or expanded cells (d 6 or 12) were plated into 35-mm plastic Petri-dishes (Falcon) in culture medium containing SCF, IL-3, G-CSF, GM-CSF and Epo as colony-stimulating factors. Each dish was incubated at 37°C in a humidified atmosphere of 5% CO2 for 14 d. Colonies consisting of more than 50 cells were counted under an inverted microscope.
Plasma clot cultures. CFU-Meg was assayed by the plasma clot technique using platelet-poor human plasma. The culture medium contained freshly prepared CD34+ cells (d 0) or expanded cells (d 6 or 12) plus 15–20% human platelet-poor AB plasma and growth factor(s) in IMDM with the following additives: 100 U/ml penicillin (Gibco BRL), 100 μg/ml streptomycin (Gibco BRL), 1 mmol/l sodium pyruvate (Gibco BRL), 1% minimal essential medium (MEM) vitamins (Gibco BRL), 0·1 mmol/l MEM non-essential amino acids (Gibco BRL), 1 × 10−5 mol/l thioglycerol (Sigma, St Louis, USA), 2 μg/ml l-asparagine (Wako Pure Chemicals, Tokyo, Japan), 74 μg/ml CaCl2 (Wako Pure Chemicals) and 0·2% bovine serum albumin (BSA; Boehringer Mannheim, Germany). The medium was plated into the wells of 24-well culture plates (0·3 ml per well) and incubated at 37°C in a humidified atmosphere of 5% CO2 for 11–12 d.
Immunofluorescence staining to identify megakaryocyte colonies. Each well was fixed twice with a 2:1 mixture of acetone and methanol for 15 min. The plates were air dried overnight and then kept at −20°C until required for staining. For the staining, the plates were removed from the freezer and returned to room temperature. PBS containing 0·5% BSA (PBS-B) was then added to soften the clot. After discarding the solution, FITC-CD41 mAb diluted 1:100 in PBS-B was added, then the plates were incubated for 1 h at room temperature and washed once with PBS-B. The nuclei were counter-stained with propidium iodine (0·3 ng/ml; Sigma). The colonies were washed again and counted using a fluorescence microscope (Olympus, Tokyo, Japan) at a magnification of ×100. The megakaryocyte colonies were classified into two types: large colonies containing more than 50 cells (immature CFU-Meg), and small colonies containing 3–50 cells (mature CFU-Meg) (Siena et al, 1993; Hagiwara et al, 1998). The total number of CFU-Meg (total CFU-Meg) was obtained by summing the immature and mature CFU-Meg.
Immunological marker analysis. The expression of cell surface antigens was analysed by direct immunofluorescence flow cytometry using triple-staining combinations of mAb, including PC5-CD45, FITC-CD34 and PE-CD41. Briefly, the cells were incubated with saturated concentrations of the relevant mAb for 20 min at room temperature, washed and analysed with a flow cytometer. For each experiment, a negative control experiment was performed on isotype-matched irrelevant control mAb.
Statistical analysis. The significance of differences between the control and experimental groups was determined by the Student's t-test. Data for multiple groups were analysed using one-way layout analysis of variance (anova) and Fisher's least significant difference test.
Dose–response relationship for bFGF in terms of the clonal growth of CB haematopoietic progenitor cells
The effect of bFGF concentration on the in vitro clonogenic potentials of haematopoietic progenitor cells isolated from freshly purified CB CD34+ cells was assessed using methylcellulose culture and plasma clot culture techniques. In the absence of exogenous cytokines, no colonies were observed in any of the cultures, regardless of whether or not bFGF was present (data not shown). In the methylcellulose cultures, a combination of SCF, IL-3, G-CSF, GM-CSF and Epo, which yields maximal colony formation, supported the growth of 194 colonies (including CFU-GM-, BFU-E- and CFU-Mix-derived colonies) per 1 × 103 cells when no bFGF was added (Fig 1). The addition of bFGF (30 ng/ml) significantly enhanced the growth of CFU-GM and CFU-Mix. The maximal effect of bFGF was, therefore, investigated by comparing several different bFGF doses. In contrast, no significant increase in BFU-E-derived colony formation was seen at any concentration of bFGF. In the plasma clot cultures used for assessing CFU-Meg-derived colony formation, TPO alone supported the growth of approximately 63 colonies per 1 × 103 cells (Fig 2), of which 66·8% were large megakaryocyte colonies (i.e. immature CFU-Meg-derived colonies). The addition of bFGF (10–100 ng/ml) induced a 41–60% increase in the total number of colonies (P < 0·05). bFGF was also added to a culture supplemented with a combination of TPO with SCF, which yields maximal megakaryocyte colony formation. Though no significant enhancement was observed, bFGF produced a tendency towards an increase in megakaryocyte colony formation (Fig 3). Based on the results obtained above, all other experiments in this study were performed with bFGF 30 ng/ml.
Effects of various combinations of cytokines plus bFGF on the expansion of total mononuclear cells
To avoid problems arising from the inclusion of animal serum, a serum-free medium, BIT9500-IMDM, was used for ex vivo expansion in this study. Various combinations of the early-acting cytokines TPO, SCF, IL-3 and FL were used as haematopoietic growth factors to expand the haematopoietic stem/progenitor cells. Isolated CD34+ cells were incubated in the presence of TPO alone, TPO + SCF, TPO + IL-3, TPO + SCF + FL or TPO + SCF + IL-3 + FL, combined with bFGF (30 ng/ml). On d 6 and 12 of culture, cells were harvested and the number of viable cells was counted (Fig 4). Regardless of whether or not bFGF had been added, TPO alone stimulated a three to fourfold increase in total cell numbers by d 6, and a 27- to 28-fold increase by d 12 of culture, compared with the baseline cell concentration (Fig 4). Stimulation with the other four combinations induced 21- to 112-fold increases in the total number of cells by d 6, and 140- to 480-fold increases by d 12 of culture. However, bFGF did not affect the total number of cells in these cultures. The expression of CD8+CD4+, CD56+CD19+, CD34+CD45+ and CD41+CD45+ cells among those harvested from a 12-d culture was analysed using flow cytometry (Fig 5). The proportions of each cell type in the culture stimulated with TPO + IL-3 were 0·019%, 0·0%, 9·74% and 15·3% respectively. The cells from the culture stimulated with TPO + SCF + FL (0·01%, 0·028%, 9·63% and 26·6% respectively) and TPO + SCF + IL-3 + FL (0·038%, 0·00%, 8·21% and 22·8% respectively) showed almost the same values. No significant differences in the expression of these antigens were observed in cells harvested from cultures supplemented with, or lacking, bFGF. These results indicated that, with this cell expansion system, there was little lymphocyte production and a decrease in the proportion of CD34+ cells (i.e. the relatively immature cells in haematopoiesis), with megakaryocytes becoming the major component of the generated cells.
Effects of various combinations of cytokines with bFGF on the expansion of haematopoietic progenitor cells
After ex vivo expansion, the total numbers of CFU-GM, BFU-E and CFU-Mix were assayed using methylcellulose cultures. Compared with freshly prepared CD34+ cells at baseline, the total CFU-GM count increased dramatically during culture with all cytokine combinations except TPO alone, rising by approximately 20- to 230-fold by d 6, and 125- to 1440-fold by d 12 of culture (Fig 6). However, no further significant increase was observed when bFGF was added to any of the cytokine combinations during any of the culture periods examined. Similarly, considerable proportional increases in the total numbers of both BFU-E and CFU-Mix were observed in cultures without bFGF (Figs 7 and 8). However, only a combination of all four cytokines, TPO + SCF + IL-3 + FL, stimulated a 650-fold increase of BFU-E by d 6 of culture, which was statistically significant compared with culture containing no bFGF. However, bFGF showed little effect on the ex vivo expansion of CFU-GM, BFU-E and CFU-Mix in most combinations. At the same time, the total number of CFU-Meg was also assessed by plasma clot culture (Fig 9). The cytokine combinations without bFGF amplified the number of CFU-Meg by 16- to 44-fold by d 6, and 51- to 56-fold by d 12 of culture. The highest level of expansion of CFU-Meg (102- to 120-fold) was observed on d 12 with TPO + SCF, TPO + IL-3, TPO + SCF + FL and TPO + SCF + IL-3 + FL in the presence of bFGF. Therefore, bFGF resulted in an approximate doubling of total CFU-Meg by d 12 of culture compared with cultures containing no bFGF.
In the present study, various combinations of the early acting cytokines TPO, SCF, IL-3 and FL were combined with bFGF to determine their effects on the ex vivo expansion of haematopoietic stem/progenitor cells. In particular, as a long period of severe thrombocytopenia is a problem in many patients after cord blood transplantation, we focused on the expansion of megakaryocytes and megakaryocytic progenitor cells; for this purpose, we used four combinations that contained TPO, whose effectiveness in inducing ex vivo expansion of these cells has been demonstrated in previous studies (Maurer et al, 2000; Sasayama et al, 2001). Combinations of TPO + IL-3, TPO + SCF + FL or TPO + SCF + IL-3 + FL without bFGF induced 270- to 470-fold increases in the total numbers of cells by d 12 of culture (Fig 4). The total numbers of CFU-GM, BFU-E and CFU-Mix generated in these cultures increased dramatically to 720- to 1440-fold, 170- to 440-fold and 85- to 750-fold the baseline cell concentration respectively. However, bFGF did not promote further proliferation except when added to the combination of TPO + SCF + IL-3 + FL. This cocktail stimulated a significant increase of BFU-E by d 6 of culture, showing promoting activity by bFGF (Fig 7). In contrast, a culture containing bFGF with all cytokine combinations except TPO alone resulted in the doubling of CFU-Meg production compared with the non-bFGF control by d 12 of culture (Fig 9), indicating the usefulness of bFGF in inducing CFU-Meg expansion. Thus, although bFGF significantly promoted clonal growth of haematopoietic progenitor cells under the semisolid culture conditions tested in this study, it appeared to have an effect only on the expansion of CFU-Meg in further liquid assay cultures. One reason for this could be that differences between the semisolid cultures conditions, such as methylcellulose and plasma clot, and the liquid culture conditions may have affected the proliferation of the haematopoietic stem/progenitor cells. Iscove et al (1989), in their investigation of increases in pluripotential haematopoietic precursors in suspension cultures with IL-1 and IL-3, detected different kinds of cells in methylcellulose colony assays and suspension culture assays. With regard to other possibilities, Ratajczak et al (1996) reported that the FGF receptor is either absent, or present at very low levels, on primitive haematopoietic cells, whereas they are expressed more abundantly on relatively more mature cells, i.e. human megakaryoblasts (Brunner et al, 1993). In addition, binding of bFGF to its high-affinity receptors is highly dependent on the presence of cell surface heparan sulphate proteoglycans or free heparin (Brunner et al, 1994; Allouche, 1995). Furthermore, FGF has been reported to prevent haematopoietic cells from undergoing apoptosis (Majka et al, 2001). Therefore, as the mechanisms of bFGF action are likely to be complex under ex vivo culture conditions, more detailed studies are required.
bFGF is capable of inducing the production of other cytokines from accessory cells, providing another potential mechanism for the modulation of megakaryocytopoiesis by this growth factor (Avraham et al, 1994). Under liquid culture conditions, there was a dramatic increase in the generation of various kinds of mature cells, such as neutrophils and megakaryocytes. Flow cytometric analysis of the cells harvested from the cultures showed no significant differences between the cells generated in cultures with or without bFGF. It has been demonstrated that various haematopoietic growth factors, cytokines and chemokines are secreted by human bone-marrow-derived CD34+ cells, CFU-GM-, CFU-Meg- and BFU-E-derived cells in an autocrine and/or a paracrine manner (Majka et al, 2001). In this study, most combinations induced dramatic increases of haematopoietic progenitor cells, such as CFU-GM and BFU-E (Figs 6 and 7). Furthermore, the results of flow cytometric analysis showed that the number of mature megakaryocytes increased dramatically (Fig 5) and, although there was a decrease in the proportion of CD34+ cells, the total number of CD34+ cells increased by approximately 70- to 120-fold, as the number of cells increased rapidly by approximately 300- to 500-fold by d 12. In a preliminary quantitative determination of IL-6, IL-3 and GM-CSF concentrations in harvested culture media (data not shown), there were no significant differences in the concentrations of these factors in conditioned media harvested from cultures grown with or without bFGF. Bruno et al (1993) reported that bFGF appeared to be capable of acting directly upon human megakaryocytic progenitor cells, while Chen et al (1999) reported that bFGF acts directly on CD34+ cells prepared from CB. Although bFGF might possibly stimulate the release of molecules that regulate the pathways by which CFU-GM, BFU-E and CFU-Mix are formed from mature generated cells, these results suggest that bFGF may have a direct promotional effect on in vitro thrombopoiesis. In addition, although the details of the mechanism by which FGF specifically promoted CFU-Meg production are unclear, it is possible that the affinity of FGF for cytokines, especially TPO, and to CD34+ cells is deeply involved.
The well known Mpl ligand, TPO, has been shown to be the central physiological regulator of megakaryocytopoiesis and platelet production, and clinical trials of its effectiveness in thrombocytopenia have now been carried out in many countries (Miyazaki & Kato, 1999; Vadhan-Raj, 2000). However, investigations of TPO have recently been hampered by the discovery of induced antibodies that neutralize TPO and cause thrombocytopenia (Basser, 2000). Nevertheless, CB is increasingly being used as a source of stem and progenitor cells for transplantation (Barker & Wagner, 2002), and a significant delay in platelet recovery has been observed in patients receiving CB transplantations (Conrad & Emerson, 1998; Gluckman, 2000). One possible approach to solving this problem is to transplant megakaryocytic progenitor cells and megakaryocytes expanded ex vivo from haematopoietic stem cells. Many studies that are now under way worldwide to evaluate the usefulness of these cells have reported a certain degree of effectiveness (Bachier et al, 1999; McNiece et al, 2000; Paquette et al, 2002). Furthermore, it is well known that the bone marrow microenvironment may be an important limiting factor that modulates blood production and recovery after haematopoietic stem cell transplantation or irradiation- and chemotherapy-induced myelosuppression (Novitzky & Mohamed, 1997). bFGF has been found to stimulate the formation of an adherent stromal cell layer in long-term human bone marrow cultures, while promoting haematopoietic cell development (Oliver et al, 1990; Wilson et al, 1991). Novitzky et al (2001) have further shown that bFGF improves the growth of adherent cell layers and increases their support of the proliferation of stroma-adherent CD34+ precursors. The development of more effective amplifying systems for haematopoietic stem/progenitor cells can, therefore, be expected as the multiple functions of bFGF are utilized to a greater extent.