Incubation of murine bone marrow cells in ... - Wiley Online Library

British Journal of Haematology, 2000, 108, 424±429

Incubation of murine bone marrow cells in hypoxia ensures the maintenance of marrow-repopulating ability together with the expansion of committed progenitors Z O RA N I VA NOV I CÂ , 1 B ENE D ETTA B A RTO L O Z Z I , 2 P IE T RO A N TO N I O B E R NA B E I , 2 M AR IA G R AZ I A C I P OL L E S C H I , 1 E L I SAB ETTA R OV I DA , 1 PAV L E M I L E N KOVICÂ , 3 V I N C E N T P R A L O R A N 4 A ND P E R S I O D E L L O S BAR BA 1 1 Department of Experimental Pathology and Oncology, and 2Haematology Unit, A. O. Careggi, University of Florence, Florence, Italy, 3Institute for Medical Research, Belgrade, Yugoslavia, and 4Laboratory of Experimental Haematology, University of Limoges, Limoges, France Received 28 July 1999; accepted for publication 18 October 1999

Summary. We developed previously a hypoxic culture system in which progenitors endowed with marrow-repopulating ability (MRA), unlike committed progenitors, were selected and maintained better than in air. We report here an improvement to this system targeted at combining the maintenance of progenitors sustaining MRA with the numerical expansion of multipotent and committed progenitors. Murine bone marrow cells were incubated at 1% oxygen in liquid medium supplemented with stem cell factor, granulocyte colony-stimulating factor, interleukin-6 and interleukin-3. In day 8 hypoxic cultures, the numbers of high proliferative potential and granulocyte/macrophage colony-forming cells (HPP-CFC and CFU-GM) were increased with respect to time zero. Colonies generated by HPP-CFC derived from hypoxic cultures exhibited a high replating

ability, whereas colonies generated by HPP-CFC derived from control cultures exhibited a low replating ability. MRA was fully maintained in hypoxia and markedly reduced in air. Thus, severe hypoxia is able to ensure a full maintenance of progenitors sustaining MRA, together with a signi®cant expansion of in vitro-detectable clonogenic progenitors, including those endowed with replating ability. This system could contribute to the improvement of current techniques for the in vitro treatment of human haematopoietic cell populations before transplantation.

The in vitro maintenance and expansion of haematopoietic stem and progenitor cells is of great interest for the development of `cell therapy' techniques targeted to reducing the length of cytopenia after bone marrow (BM) transplantation, purging of malignant cells from explants to be used as autografts or transferring genes into haematopoietic cells. The ex vivo ampli®cation of committed progenitors without impairment of stem cell function is a major problem (Williams, 1993). Contradictory data have been published on this issue. A number of reports have shown a relative decrease in the engraftment potential of both murine and human haematopoietic cells expanded in vitro (Traycoff et al, 1988; Peters et al, 1995, 1996; Sekhar et al, 1997; Varas

et al, 1998; GuÈenechea et al, 1999). Later, cultures containing FLT-3-ligand (FLT3L) in combination with early acting cytokines enabled severalfold expansion of human long-term culture-initiating cells (LTC-IC), closely related to cells with short-term repopulating ability, together with a modest expansion of cells capable of repopulating severe combined immunode®cient mice (Connealy et al, 1997). However, in a limiting dilution analysis, these cells produced a three- to severalfold lower number of progenitors in recipient mice than unmanipulated cells, demonstrating a rapid loss of their self-renewal and commitment ability (Connealy et al, 1997). Furthermore, a recent study performed in lethally irradiated cats showed that BM cells (BMCs) expanded in vitro in the presence of FLT3L combined with stem cell factor (SCF), granulocyte colony-stimulating factor (G-CSF), interleukin (IL)-1 and erythropoietin were unable to engraft transplanted animals (Abkowitz et al, 1998), indicating that the

Correspondence: Persio Dello Sbarba, Dipartimento di Patologia e Oncologia Sperimentali, UniversitaÁ di Firenze, viale G.B. Morgagni 50, I-50134 Firenze, Italy.

424

Keywords: hypoxia, stem cell maintenance, marrowrepopulating ability, high proliferative potential colonyforming cells.

q 2000 Blackwell Science Ltd

MRA Maintenance and CFU Expansion in Hypoxia current ex vivo expansion procedures must be improved before their use for long-term grafting in human therapy. All the above studies were performed by incubating the cultures in 20% oxygen and focused on the effects of various growth factor combinations. However, in mouse BM at least, the rapidly proliferating committed progenitors and precursors are located close to the blood vessels, i.e. in better oxygenated areas, whereas more immature progenitors reside preferentially in areas of lower oxygen tension (Lord, 1992). This suggests a role for oxygen tension as a physiological component of the haematopoietic microenvironment and, in particular, as a regulator of the balance between quiescence and proliferation. We have shown previously that mouse BM progenitors endowed with marrow-repopulating ability (MRA) were, unlike committed progenitors, better preserved in vitro at 1% than at 20% oxygen (Cipolleschi et al, 1993), showing that incubation in hypoxia is critical for the maintenance in vitro of very immature progenitors and stem cells. Oxygen tensions signi®cantly higher than 1%, on the other hand, are necessary for maximal clonal expansion and, therefore, colony formation from committed progenitors (Bradley et al, 1978; Ishikawa & Ito, 1988). The study we report here was undertaken to improve our hypoxic culture system in order to combine the maintenance of primitive, marrow-repopulating progenitors with the numerical expansion of multipotent and committed progenitors. We incubated murine BMCs at oxygen levels of 1% or 20% in liquid cultures supplemented with SCF, G-CSF, IL-6 and IL-3. The effects of hypoxia were determined on progenitors sustaining MRA, high proliferative potential and granulocyte/macrophage colony-forming cells (HPP-CFC and CFU-GM respectively). Incubation in hypoxia, but not in air, supported the numerical expansion of CFU-GM and HPP-CFC combined with the maintenance of the replating ability of the latter and the full preservation of progenitors sustaining MRA. MATERIALS AND METHODS BMC recovery and culture. Femoral cells from 10- to 14-week-old, male or female CBA mice were ¯ushed and pooled in RPMI-1640 medium (Gibco, Paisley, UK). Viable cells were then counted in a haemocytometer and plated in RPMI-1640 supplemented with 20% heat-inactivated horse serum (HS; HyClone Europe, Cramlington, UK) in 25 cm2 ¯asks (Nunc-Intermed) at 2 ´ 105 cells/ml and 5 ml/¯ask. Cultures were also supplemented with IL-3 (murine recombinant; PeproTech EC, London, UK; 10 ng/ml), IL-6 (human recombinant; PeproTech; 10 ng/ml), G-CSF (Neupogen, DompeÂ Biotec, Milan, Italy; 20 ng/ml) and SCF (human recombinant; PeproTech; 15 ng/ml). Human SCF, less active on murine than on human cells (Martin et al, 1990), but capable of sustaining the survival of murine HPP-CFC, was used to avoid its effect on differentiation (data not shown). In each experiment, duplicate samples were incubated for 5 or 8 days in a water-saturated atmosphere containing 5% CO2 and 20% O2 or 1% O2. Incubation in hypoxia was carried out in a Heto Cell-House 150 incubator. Viable cells were

425

then counted, washed and processed for in vitro or in vivo assays. In vitro clonal assays. These were performed basically as has been described previously (IvanovicÂ et al, 1999). Brie¯y, cells recovered directly from donor animals or from day 5 or day 8 liquid cultures were plated in 35-mm Petri dishes (Falcon, Becton-Dickinson, UK; 1 ml/dish) in a premixed culture medium containing 0´9% methylcellulose (Methocult H42230; StemCell Technologies, Vancouver, Canada), 30% fetal bovine serum (FBS; Gibco), 10 ng/ml recombinant murine IL-3 and 500 ng/ml recombinant human G-CSF. Colonies ($ 50 cells) were counted at day 7, and cultures were returned to the incubator until day 14, when colonies were counted again. The difference between the number of colonies scored at day 7 and that scored at day 14 represented the number of colonies derived from mature committed progenitors (CFU-GM) that disappeared before the second scoring. Day 14 colonies were identi®ed as generated by HPP-CFC3 or HPP-CFC2 according to their diameter (0´6±1´8 mm or > 1´8 mm respectively) (Ivanovic et al, 1999). To determine the progenitors' secondary clonogenic ability, colonies recovered from day 14 primary methylcellulose cultures were resuspended in 200 ml of RPMI-1640 supplemented with 20% HS. The number of cells per colony was counted, and cells were replated in secondary methylcellulose cultures established as described above. Secondary colonies were scored at day 7. MRA assay. This was performed basically as described by Hodgson et al (1982). Brie¯y, 12- to 14-week-old CBA mice were lethally irradiated (10 Gy, 0´5 Gy/min, with a 1´2 MeV 60 Co source; Theratron 780, Atomic Energy of Canada, Ottawa, Canada). Cells were transplanted at 5 ´ 104 to 2 ´ 105/0´2 ml into four or ®ve mice of the same sex per condition. At day 14 after transplantation, femoral nucleated cells were counted individually, pooled and processed for day 7 in vitro clonal assays established as described above. The MRA of liquid cultures was given by the product between the total number of cells (MRAcell) or the number of day 7 CFU (MRACFU) per recipient femur and the ratio of the total cell number in (time zero or day 8, hypoxic or control) cultures to the number of transplanted cells. MRA was expressed in arbitrary units, 1 U corresponding to the number of cells to be transplanted to repopulate each recipient's femur with 5 ´ 105 BMCs (MRAcell) or 103 day 7 CFU (MRACFU). RESULTS Incubation of BMCs in 20% oxygen resulted in a rapid increase (threefold, and maximal at day 5) in the number of viable cells in culture (Fig 1). In contrast, when BMCs were incubated in 1% oxygen, the number of viable cells was initially reduced (approximately by half at day 5), but later increased signi®cantly to reach or pass the time zero level (at day 8). The numbers of CFU-GM, HPP-CFC3 and HPP-CFC2 also increased, with respect to time zero, after either 5 or 8 days of culture in 20% oxygen (Table I). In day 5 hypoxic cultures, only the number of CFU-GM increased, whereas that of HPP-CFC3 was unchanged and that of HPP-CFC2

q 2000 Blackwell Science Ltd, British Journal of Haematology 108: 424±429

426

Z. IvanovicÂ et al Table I. Effects of hypoxia on the number of in vitro-detectable progenitors. Number of progenitors per ml of culture Day 5

Day 8

Progenitor type

Time zero

20% O2

1% O2

20% O2

1% O2

CFU-GM HPP-CFC3 (0´6±1´8 mm) HPP-CFC2 (> 1´8 mm)

220 6 43 166 6 18 113 6 11

1493 6 711 947 6 393 730 6 338

389 6 104 157 6 31 68 6 13

2224 6 637 1017 6 233 488 6 140

657 6 150 382 6 99 171 6 44

Cells were recovered from cultures at time zero, or after a 5- or 8-day incubation in 20% or 1% oxygen, in liquid medium containing SCF, G-CSF, IL-6 and IL-3. Cells were then transferred to clonal assays in semi-solid medium containing G-CSF and IL-3, always incubated in 20% oxygen. The number of CFU-GM was calculated by subtracting the number of colonies counted at day 14 from the number counted at day 7, as described in Materials and methods. The number of HPP-CFC3- or HPP-CFC2-derived colonies was counted at day 14. The values are the means 6 SEM of six independent experiments. The differences between hypoxic day 8 and time zero cultures, as determined by the Student's t-test for paired samples, were: CFU-GM, P 0´025 (signi®cant); HPP-CFC3, P 0´026 (signi®cant); HPP-CFC2, P 0´073 (marginally signi®cant).

decreased. On the other hand, at day 8, all the progenitors were signi®cantly increased in hypoxia, although to a lesser extent than in air. The replating ability of colonies derived from HPP-CFC2 was assessed in experiments reported in Table II. When colonies were generated by HPP-CFC2 derived from hypoxic cultures, their replating ability was very high, similar to that of colonies grown from BMCs recovered directly from mice of the same strain (IvanovicÂ et al, 1999). In contrast, the replating ability of colonies generated by HPP-CFC2 derived from control cultures was close to zero. The MRA of cultures incubated for 8 days is reported in Table III. MRAcell was

maintained in cultures incubated in 20% oxygen, whereas it was reduced to two-thirds in hypoxia. MRACFU, however, was maintained in hypoxia and reduced to one-third in 20% oxygen. The results in Tables I and III are summarized together in Fig 2, in which the content of progenitors in day 8 cultures is expressed as a percentage of the time zero value. The ®gure makes it evident that all in vitro-detected progenitors increased in both 1% and 20% oxygen, MRAcell was substantially unaffected by the oxygen percentage, while MRACFU was fully maintained in hypoxia and strongly reduced in control cultures. Table II. Effects of hypoxia on the secondary clonogenic ability of HPP-CFC2. Number of day 7 secondary colonies per HPP-CFC2 recovered from Day 5

Day 8

liquid cultures at

Fig 1. Effects of hypoxia on the number of viable cells in BMC cultures. Pooled cells from three mice were incubated in liquid medium containing SCF, G-CSF, IL-6 and IL-3 for 5 or 8 days in an atmosphere of 20% or 1% oxygen. Histograms represent the total numbers of viable cells per culture and are means 6 SEM of six independent experiments. The horizontal line represents the time zero value. The difference between day 5 and day 8 hypoxic cultures was statistically signi®cant according to the Student's t-test for paired samples (P 0´043).

Expt no.

20% O2

1% O2

20% O2

1% O2

1 2 3

0 0´5 1´0

3´0 69´5 25´0

0 0 0´5

35´0 8´0 113´0

The values represent the number of secondary colonies per primary colony generated from HPP-CFC2 recovered from day 5 or day 8, hypoxic or control, liquid cultures in three independent experiments. Cells from four primary colonies were pooled, counted and replated at identical numbers across individual experiments and experimental conditions. The count of secondary colonies (at day 7 only, to determine the total replating potential) enabled the estimation of the average number of progenitors per primary colony.


MRA Maintenance and CFU Expansion in Hypoxia

427

Table III. Effects of hypoxia on marrow-repopulating ability. Number of arbitrary units per ml of culture Progenitor type MRAcell Mean 6 SEM MRACFU Mean 6 SEM

Day 8 Expt no.

Time zero

20% O2

20%O2/time zero

1% O2

1% O2/time zero

1 2 3

24´3 17´6 11´4

17´6 9´4 16´6

14´2 8´9 9´1 (A)

1 2 3

2´6 14´9 6´7

0´6 8´1 2´6

0´72 0´53 1´46 0´90 6 0´28 (C) 0´22 0´54 0´38 0´38 6 0´09

0´58 0´51 0´80 0´63 6 0´09 (D) 1´45 1´08 0´82 1´12 6 0´18

3´7 16´1 5´5 (B)

Cells were recovered from cultures at time zero or after an 8-day incubation in 20% or 1% oxygen, in liquid medium containing SCF, G-CSF, IL-6 and IL-3. Cells were then transplanted into lethally irradiated syngeneic mice of the same sex, and the numbers of cells or day 7 CFU contained in the recipient's femur at day 14 after transplantation were determined. MRA is given by the product between these numbers and the ratio of the total cell number in culture to the number of transplanted cells. MRA is expressed in arbitrary units, 1 U corresponding to the number of cells to be transplanted to repopulate each recipient's femur with 5 ´ 105 cells (MRAcell) or 103 day 7 CFU (MRACFU). Results from three independent experiments and the relative means 6 SEM are reported. The differences were, according to the Student's t-test: line (A), not signi®cant; line (B), signi®cant (P 0´023); column (C), not signi®cant; column (D), marginally signi®cant (P 0´074).

Fig 2. Effects of hypoxia on the expansion or maintenance of progenitors in culture. The ®gure summarizes the data reported in Tables I and III. Histograms represent the total numbers of progenitors in culture expressed as percentages of the corresponding time zero values (horizontal lines) and are means 6 SEM of the numbers of independent experiments indicated in the legends to Tables I and III. The statistical signi®cance of the differences are given in the legends to Tables I and III.

DISCUSSION The main ®nding of this work was that BMCs incubated for 8 days in 1% oxygen completely maintained their time zero stem cell potential, as detected by MRACFU assay, yet allowed a signi®cant expansion in vitro of committed clonogenic progenitors with respect to time zero. In 1% oxygen, the replating ability of HPP-CFC2-derived colonies was also fully maintained, indicating that hypoxia preserved in vitro the detectable progenitors capable of generating secondary colonies. In contrast, in 20% oxygen, the expansion of the total number of viable cells and of the number of clonogenic progenitors was higher than in hypoxia, but the secondary colony-forming ability of committed progenitors was completely lost, and only one-third of MRACFU was maintained (38% vs. 112% in hypoxia). These results seem to indicate

that hypoxia was necessary to ensure the combination of stem cell maintenance with a signi®cant production of committed progenitors. As this combination is believed to be the distinguishing feature of `haematopoietic niches' (Scho®eld, 1978), the results lend support to our hypothesis that hypoxia contributes to establishing the typical `niche' environment in vivo (Cipolleschi et al, 1993). The mechanisms underlying the observed effects of hypoxia are unclear. Hypoxia might act directly on haematopoietic cells by slowing down the cycling rate (Loe¯er, 1992; Webster et al, 1998) or shifting the balance between self-renewal and commitment towards self-renewal. This would be a consequence of the activation of hypoxiainducible genes in stem cells (Pugh et al, 1997), including those controlling the expression of receptors for stem cellactive cytokines. Hypoxia might also act on haematopoietic


428

Z. IvanovicÂ et al

cells indirectly, in¯uencing tissue growth factor production (Derevianko et al, 1996). The acquisition by stem cells of a particular type of cytokine responsiveness could explain the results reported here (La Iuppa et al, 1998). Indeed, a subset of human stem cells that does not respond to cytokine stimulation and therefore cannot be expanded in vitro has been identi®ed (Berardi et al, 1995; Bertolini et al, 1997). These cells do not divide, are 5-¯uorouracil (5-FU) resistant and are believed to be responsible for long-term marrowrepopulating ability. On the other hand, we recently demonstrated an excellent correlation between progenitors sustaining MRA and culture-repopulating progenitors, on the basis of their maintenance under prolonged incubation in hypoxia (unpublished observations). When transferred to air, culture-repopulating cells were capable of generating in vitro committed progenitors in the presence of growth factors. Thus, MRA most probably detects cytokineresponsive progenitors endowed with short-term marrowrepopulating ability, which are less primitive than cytokineunresponsive, non-dividing stem cells (Ladd et al, 1997). On this basis, one can hypothesize that the latter, in hypoxia but not in air, become cytokine responsive and maintain the pool of MRACFU progenitors, while some of them are used to generate committed progenitors. A mechanism related to the response of haematopoietic cells to cytokines may also explain the signi®cant advancement represented by the results reported here with respect to the study in which, in 1% oxygen, we obtained a less ef®cient maintenance of MRACFU, together with a marked reduction in committed progenitors (Cipolleschi et al, 1993). As, in that study, we did not use stem cell-active factors, such as SCF, GCSF and IL6 (Bodine et al, 1991), it is possible that it is only when these factors are present together with IL-3 that hypoxia ensures a full maintenance of MRACFU progenitors, as well as a sizeable expansion of committed progenitors. In keeping with this hypothesis, neither IL-3 alone, nor the other three factors in the absence of IL-3, prevented GM-CFC and HPP-CFC from decreasing in hypoxia (not shown). In this respect, it would be interesting to test in our system the effects of FLT3L, which has been shown so far to sustain a modest expansion of human-repopulating cells, but not their capacity to generate committed progenitors in vivo (Connealy et al, 1997). Whatever the mechanism underlying the results presented here, they are of practical interest for the improvement of techniques for the in vitro expansion of BMCs. These results, indeed, represent the basis for the de®nition of growth factor combinations and oxygen concentrations, as well as of procedures (such as the change in oxygen concentration at critical points of incubation) targeted to optimize the ratio of stem cells to committed progenitors and to best meet the clinical requirements for BMC transplantation. ACKNOWLEDGMENTS The authors thank Professor Massimo Olivotto, Department of Experimental Pathology and Oncology, University of Florence, for moral and material support for this work. The

project was funded by grants from Ministero della UniversitaÁ e della Ricerca Scienti®ca e Tecnologica, Consiglio Nazionale delle Ricerche, Associazione Italiana per la Ricerca sul Cancro (AIRC) and Regione Toscana (Progetto QualitaÁ). Z.I. was the recipient of a fellowship from AIRC. P.M. was supported by a grant from the Ministry of Science and Technology of Serbia. V.P. was supported by a grant from Conseil Regional du Limousin. REFERENCES Abkowitz, J.L., Taboada, M.R., Sabo, K.M. & Shelton, G.H. (1998) The ex vivo expansion of feline marrow cells leads to increased numbers of BFU-E and CFU-GM but a loss of reconstituting ability. Stem Cells, 16, 288±293. Berardi, A.C., Wang, A., Levine, J.D., Lopez, P. & Scadden, D.T. (1995) Functional isolation and characterisation of human haematopoietic stem cells. Science, 267, 104±108. Bertolini, F., Battaglia, M., Soligo, D., Corsini, C., Curioni, C., Lazzari, L., Pedrazzoli, P. & Thalmeier, K. (1997) `Stem cell candidates' puri®ed by liquid culture in the presence of steel factor, IL-3, and 5FU are strictly stroma-dependent and have myeloid, lymphoid, and megakaryocytic potential. Experimental Haematology, 25, 350±356. Bodine, D.M., Crosier, P.S. & Clark, S.C. (1991) Effects of haematopoietic growth factors on the survival of primitive stem cells in liquid suspension culture. Blood, 78, 914±920. Bradley, T.R., Hodgson, G.S. & Rosendaal, M. (1978) The effect of oxygen tension on haematopoietic and ®broblast cell proliferation in vitro. Journal of Cellular Physiology, 97, 517±522. Cipolleschi, M.G., Dello Sbarba, P. & Olivotto, M. (1993) The role of hypoxia in the maintenance of haematopoietic stem cells. Blood, 82, 2031±2037. Connealy, E., Cashman, J., Petzer, A. & Eaves, C. (1997) Expansion in vitro of transplantable human cord blood stem cells demonstrated using a quantitative assay of their lympho-myeloid repopulating activity in nonobese diabetic scid/scid mice. Proceedings National Academy of Sciences of the USA, 94, 9836±9841. Derevianko, A., D'Amico, R. & Simms, H. (1996) Polymorphonuclear leucocyte (PMN)-derived in¯ammatory cytokines ± regulation by oxygen tension and extracellular matrix. Clinical and Experimental Immunology, 103, 560±567. GuÈenechea, G., Segovia, J.C., Albella, B., Lamana, M., Ramirez, M., Regidor, C., Fernandez, M.N. & Bueren, J.A. (1999) Delayed engraftment of nonobese diabetic/severe combined immunode®cient mice transplanted with ex vivo-expanded human CD34 cord blood cells. Blood, 93, 1097±1105. Hodgson, G.S., Bradley, T.R. & Radley, J.M. (1982) The organization of haematopoietic tissue as inferred from the effects of 5¯uorouracil. Experimental Haematology, 10, 26±35. Ishikawa, Y. & Ito, T. (1988) Kinetics of haematopoietic stem cells in hypoxic cultures. European Journal of Haematology, 40, 126±129. IvanovicÂ, Z., Bartolozzi, B., Bernabei, P.A., Cipolleschi, M.G., MilencovicÂ, P., Praloran, P. & Dello Sbarba, P. (1999) A simple, one-step clonal assay allows the sequential detection of committed (CFU-GM-like) and of several subsets of primitive (HPP-CFC) murine progenitors. Stem Cells, 17, 219±225. Ladd, A.C., Pyatt, R., Gothot, A., Rice, S., McMahel, J., Traycoff, C.M. & Srour, E.F. (1997) Orderly process sequential cytokine stimulation is required for activation and maximal proliferation of primitive human bone marrow CD34 haematopoietic progenitor cells residing in G0. Blood, 90, 658±668. La Iuppa, J.A., Papoutsakis, E.T. & Miller, W.M. (1998) Oxygen tension alters the effects of cytokines on the megakaryocyte,


MRA Maintenance and CFU Expansion in Hypoxia erythrocyte, and granulocyte lineages. Experimental Haematology, 26, 835±843. Loe¯er, M. (1992) A cytokinetic approach to determine the range of O2-dependence of pyrimidine(deoxy)nucleotide biosynthesis relevant for cell proliferation. Cell Proliferation, 25, 169±179. Lord, B.I. (1992) The architecture of bone marrow cell populations. In: Concise Reviews in Clinical and Experimental Haematology (ed. by M. Murphy, Jr), p. 225. AlphaMed Press, Dayton, OH. Martin, F.H., Stuggs, S.V., Langley, K.H., Lu, H.S., Ting, J., Okino, K.H., Morris, F.C., McNiece, I.K., Jacobson, F.W., Mendiaz, E.A., Birkett, N.C., Smith, K.A., Johnson, M.J., Parker, V.P., Flores, J.C., Patel, A.C., Fisher, E.F., Erjavec, H.O., Herrera, C.J., Wypych, J., Sachdev, R.K., Pope, J.A., Leslie, I., Wen, D., Lin, C.-H., Cupples, R.L. & Zsebo, K.M. (1990) Primary structure and functional expression of rat and human stem cell factor DNAs. Cell, 63, 203±211. Peters, S.O., Kittler, E.L.W., Ramshaw, H.S. & Quesenberry, P.J. (1995) Murine marrow cells expanded in culture with IL-3, IL-6, IL-11, and SCF acquire an engraftment defect in normal hosts. Experimental Haematology, 23, 461±469. Peters, S.O., Kittler, E.L.W., Ramshaw, S. & Quesenberry, P.J. (1996) Ex vivo expansion of murine marrow cells with interleukin-3 (IL-3), IL-6, IL-11, and stem cell factor leads to impaired engraftment in irradiated hosts. Blood, 87, 30±37. Pugh, C.W., O'Rourke, J.F., Nagao, M., Gleadle, J.M. & Ratcliffe, P.J.

429

(1997) Activation of hypoxia-inducible factor-1; de®nition and regulatory domains within the a subunit. Journal of Biological Chemistry, 272, 11205±11214. Scho®eld, R. (1978) The relationship between the spleen colonyforming cells and the haematopoietic stem cell ± a hypothesis. Blood Cells, 4, 7±25. Sekhar, M., Yu, J.-M., Soma, T. & Dunbar, C. (1997) Murine long-term repopulating ability is compromised by ex vivo culture in serum-free medium despite preservation of committed progenitors. Journal of Hematotherapy, 6, 543±549. Traycoff, C.M., Orazi, A., Ladd, A.C., Rice, S., McMahel, J. & Srour, E.F. (1988) Proliferation-induced decline of primitive haematopoietic progenitor cell activity is coupled with an increase in apoptosis of ex vivo expanded CD34 cells. Experimental Haematology, 26, 53±62. Varas, F., Bernad, A. & Bueren, J.A. (1998) Restrictions in the stem cell function of murine bone marrow grafts after ex vivo expansion of short-term repopulating progenitors. Experimental Haematology, 26, 100±109. Webster, L., Hodgkiss, R.J. & Wilson, G.D. (1998) Cell cycle distribution of hypoxia and progression of hypoxic tumour cells in vivo. British Journal of Cancer, 77, 227±234. Williams, D.A. (1993) Ex vivo expansion of haematopoietic stem and progenitor cells ± robbing Peter to pay Paul? Blood, 81, 3169±3172.
