Experimental Hematology 32 (2004) 665–672
Quiescent (5-fluorouracil-resistant) aplastic anemia hematopoietic cells in vitro Siaˆn Rizzo, John Scopes, Gwen S. Draycott, Christopher Pocock, Theodora Foukaneli, Timothy R. Rutherford, Edward C. Gordon-Smith, and Frances M. Gibson Department of Cellular and Molecular Medicine (Haematology), St. George’s Hospital Medical School, London, United Kingdom (Received 3 July 2003; revised 5 April 2004; accepted 14 April 2004)
Objective. Bone marrow from aplastic anemia (AA) patients shows reduced numbers in longterm culture (LTC)-initiating cell (LTC-IC) assays. The LTC-IC assay is based on assumptions of the culture kinetics of normal hematopoietic stem cells (HSC), which are not necessarily justified in a disease state. We therefore undertook a detailed examination of the kinetics of quiescent HSC from AA patients in LTC. Methods. Colony formation by quiescent HSC in LTC was tested by pretreating control (n ⫽ 6) and AA bone marrow (n ⫽ 7) with 5-fluorouracil. Secondly, we manipulated normal samples to inoculate cultures with proportions of CD34⫹ cells similar to those from AA samples. We obtained enough CD34⫹ cells to reconstitute one AA sample to “normal” levels. Results. Patient cells showed altered kinetics with rapid proliferation and premature termination of LTC. In vivo, decreased numbers of HSC may induce rapid proliferation and differentiation; a similar phenomenon could explain the observations in culture. We therefore manipulated normal samples to contain a proportion of CD34⫹ HSC similar to that in AA samples. Although absolute numbers of secondary colonies in LTC were reduced, the kinetics of culture were not altered. However, when AA CD34⫹ HSC were reconstituted to “normal” levels, the cultures still demonstrated early termination. Conclusions. The kinetics of LTC are not affected by CD34⫹ HSC number. However, quiescent HSC derived from patients with AA have qualitative differences from normal cells, as reflected by distinct kinetics in long-term culture. This has implications for the interpretation of the LTC-IC assay with AA samples. 쑖 2004 International Society for Experimental Hematology. Published by Elsevier Inc.
Aplastic anemia (AA) is a syndrome of bone marrow (BM) failure characterized by BM hypoplasia and peripheral blood (PB) pancytopenia. AA BM mononuclear cells (BMMC) show a significantly lower proportion of CD34⫹ cells than normal [1,2]. Various phenotypically-defined subpopulations of the CD34⫹ cell fraction, believed to be enriched for primitive hematopoietic stem and progenitor cells (HSC), are also significantly reduced [1–3]. When the hypocellularity of AA BM is taken into account, it is clear that the disease is characterized by a major reduction in phenotypically defined primitive HSC. Evidence has accrued for dysfunction in this reduced HSC compartment: AA BM gives rise to significantly lower numbers of committed progenitor cells
Offprint requests to: Frances M. Gibson, Ph.D., Department of Cellular and Molecular Medicine (Haematology), St. George’s Hospital Medical School, London, UK, SW17 ORE.; E-mail:
[email protected] Dr. Rizzo and Dr. Scopes contributed equally to this work.
0301-472X/04 $–see front matter. Copyright doi: 1 0. 10 1 6 / j .e x p he m.2 0 04 .0 4 .0 0 3
than normal in clonogenic culture and long-term bone marrow culture (LTBMC) [4–7]. Furthermore, isolated AA CD34⫹ cells are dysfunctional in their clonogenic response to a wide range of cytokine stimuli [8]. The quantification of early progenitor cells relies on in vitro systems where the presence of such cells is indirectly measured by their ability to generate progeny. Cobblestone area–forming cells (CAFC) are counted at week 5 of LTBMC [9–11]; long-term culture-initiating cells (LTC-IC) generate progeny that produce colonies after 5 weeks in LTBMC [12,13]. In mice, LTC-IC have the same phenotype and kinetic characteristics as cells capable of providing longterm repopulation to lethally-irradiated animals [14,15]. Several studies described significant reduction in the frequency of CAFC [16] and LTC-IC [17,18] in AA BM. We have demonstrated that this reduction in LTC-IC is a consequence of reduced numbers of LTC-IC within the CD34⫹ cell subpopulation from AA BM rather than due simply to a
쑖 2004 International Society for Experimental Hematology. Published by Elsevier Inc.
666
S. Rizzo et al. / Experimental Hematology 32 (2004) 665–672
quantitative reduction of CD34⫹ cell number [19]. However, despite this deficit of hematopoietic cells, autologous hematopoietic reconstitution of aplastic anemia patient marrow occurs frequently following immunosuppressive therapy; a primitive cell population must be present in these cases and further investigation of HSC in AA BM is warranted. This study investigated the in vitro proliferation kinetics of AA BM in comparison to normal donor BM, by assessing the ability of quiescent progenitor cells to generate committed progenitor cells following treatment with 5-fluorouracil (5-FU). This cell cycle–specific toxin allows the study of a quiescent fraction that is enriched for very primitive HSC that are capable of long-term hematopoietic repopulation in vivo [20]. We show here that 5-FU-resistant hematopoietic cells from AA patients have distinctly different kinetics from normal cells in LTBMC, and that this is likely to be intrinsic to the AA cells rather than a consequence of low CD34⫹ cell numbers. Materials and methods Bone marrow collection and preparation Permission to seek informed consent for the use of bone marrow samples was obtained from the St. George’s Hospital Research Ethics Committee with approval from Wandsworth Health Authority for both AA patients and normal subjects. Bone marrow (BM) was obtained from posterior iliac crest marrow aspirates of hematologically normal donors (n ⫽ 15). Bone marrow was collected from patients (n ⫽ 8) with a diagnosis of aplastic anemia [21,22] for clinical reasons at times of routine clinic visits or following admission for treatment. Clinical details [21–23] for aplastic patients are shown in Table 1. Mononuclear cells (MC) were obtained after diluting BM 1:1 with Iscove’s modified Dulbecco’s medium (IMDM; GibcoBRL, Life Technologies, Paisley, Scotland) supplemented with 100 IU/ mL penicillin-streptomycin (PS; GibcoBRL) and 10 U/mL of preservative-free heparin (Leo Pharma, Princes Risborough, Bucks, UK). Samples were centrifuged on Ficoll-Paque (Amersham Pharmacia Biotech AB, Uppsala, Sweden) at 400g for 30 minutes and
the BMMC obtained were washed twice in medium. Viability and cell concentrations were determined from hemocytometer counts after trypan blue staining. Committed progenitor cell assay (colony-forming unit: CFU) Cells were plated up to a maximum of 105 cells/mL of IMDM containing 30% v/v fetal calf serum (FCS; PAA Laboratories GmbH), 10 µg/mL deionized bovine serum albumin (BSA, SigmaAldrich Co. Ltd., Poole, Dorset, UK), 10⫺4 M β-mercaptoethanol (Sigma-Aldrich), and 0.9% w/v methylcellulose (Stem Cell Technologies Inc., Vancouver, Canada), 2 IU/mL erythropoietin (EPO; kindly provided by Janssen-Cilag Ltd., High Wycombe, Bucks, UK), 5 ng/mL granulocyte-macrophage colony-stimulating factor (GM-CSF; 1.39 × 108 U/mg, kindly supplied by Novartis Pharmaceuticals UK Ltd, Camberely, Surrey, UK), and 50 ng/mL of interleukin-3 (IL-3; 4.9 × 106 U/mg, provided by Novartis Pharmaceuticals UK Ltd). Duplicate 1-mL cultures in 35-mm culture dishes were incubated for 14 days at 37⬚C in 5% CO2 in air. Colonies containing at least 50 cells were enumerated by light microscopy and categorized by morphology, as granulocyte-macrophage colony-forming units (CFU-GM), erythroid burst-forming units (BFU-E), or mixed multilineage colonies (CFU-GEM), the sum of which gave the total production of committed progenitors (CFU). Murine stromal cell line: MS5 The murine preadipocyte cell line MS5 forms stromal layers capable of supporting in vitro hematopoiesis [24,25]. Cultures were maintained in RPMI1640 medium (Sigma-Aldrich) supplemented with 100 IU PS/mL, 10% v/v FCS, and 10% v/v horse serum (HS; Sigma) in 25-cm2 flasks that had previously been coated with 5 mL of 0.5% w/v gelatin solution to enhance attachment. Confluent MS5 cells were detached with 0.25% trypsin at 37⬚C for 5 minutes. New cultures were initiated by seeding cells diluted 1:3 into new flasks. Irradiation (1000 rads) of confluent layers 24 hours prior to culture with primary cells extended the duration of cell attachment to the flask base (unpublished observations). Removal of plastic-adherent cells to prevent generation of autologous stroma BMMC were incubated at 33⬚C in 5% CO2 in air at 1 to 2 × 106 cells/ mL of IMDM containing 100 IU PS/mL and 20% v/v FCS in 75cm2 flasks for 2 hours. The flasks were gently agitated to release
Table 1. Aplastic anemia patient clinical characteristics
UPN
Age
Sex
1 2 3 4 5 6 7 8
28 19 19 29 47 16 23 20
M M F F F F M F
Disease duration (mo) at time of marrow
Max disease severity
Severity at time of marrow
Therapy at time of marrow
Previous therapy
ANC × 109/L
Plts × 109/L
Hb g/dL
Transfusion requirements
76 54 39 91 13 6 6 49
NSAA SAA NSAA NSAA NSAA VSAA SAA VSAA
NSAA NSAA NSAA NSAA NSAA NSAA NSAA NSAA
CSA OXY None None None CSA, G-CSF CSA None
ALG,OXY,ALG ALG,OXY CSA,ALG OXY None ALG ALG,OXY ALG,CSA,G-CSF
1.9 1.1 1.9 3.2 2.1 5.2 2.1 1.3
69 93 157 74 44 19 40 118
13.5 13.8 13.5 12.9 7.9 8.4 10.3 11.8
None None None None RBC RBC, Plts RBC, Plts None
Abbreviations: UPN: unique patient number, M: male, F: female, mo: months, NSAA: nonsevere aplastic anemia, SAA: severe AA, VSAA, very severe AA, ALG: horse anti-lymphocyte globulin, G-CSF: granulocyte colony-stimulating factor, CSA: cyclosporin A, OXY: oxymetholone, ANC: absolute neutrophil count, Plts: platelets, Hb: hemoglobin, RBC: red blood cells.
S. Rizzo et al. / Experimental Hematology 32 (2004) 665–672
all nonadherent cells (Ad⫺ BMMC); viability and cell concentration were determined as above. Assay for regeneration of CFU in LTBMC following treatment with 5-FU Ad⫺ BMMC were incubated overnight at 37⬚C/5% CO2 in air at 106 cells/mL of IMDM supplemented with 10% FCS, 10% HS, 100 IU PS/mL, either with no 5-FU (control) or with 50 µg/mL 5-FU (normal n ⫽ 6; AA n ⫽ 7, Table 1, UPN 1–7). This dose was selected to cause a 95% reduction of CFU production following overnight incubation (n ⫽ 6, data not shown) and was in line with previous reports [26–28]. Cells were washed in Hank’s buffered saline solution (HBSS; Gibco) three times and resuspended in IMDM containing 100 IU PS/mL. Committed progenitor cell assays were performed as described above, with 105 cells/mL from each of the untreated cell aliquots and with an equal volume of 5-FUtreated cells. Long-term culture (LTC) was performed with the remaining cells (5 × 106 from untreated and an equivalent volume from 5-FU-treated cell populations), which were washed and resuspended in 10 mL IMDM (adjusted to 340 mOsm/Kg), supplemented with 10% v/v FCS, 10% v/v HS, and 10⫺6M hydrocortisone succinate, added to 25-cm2 flasks of irradiated MS5 stromal layers and then incubated at 33⬚C in 5% CO2 in air. At weekly intervals, half of the culture medium was replaced with fresh supplemented medium. The nonadherent cells recovered each week were counted and assayed for colony production as above. A schematic illustrating the process is shown in Figure 1A.
667
Assay for regeneration of CFU in LTBMC with modified CD34⫹ cell content following treatment with 5-FU A schematic illustrating the process of regeneration following modification of CD34⫹ content of Ad⫺ BMMC is shown in Figure 1B and described below. Immunomagnetic separation. CD34⫹ cells were isolated from Ad⫺ BMMC (normal n ⫽ 9; AA n ⫽ 1, Table 1, UPN 8) using the MiniMACS CD34⫹ cell isolation kit (Miltenyi Biotech, Bergisch, Gladbach, Germany). Ad⫺ BMMC were washed twice in 0.6% w/ v acid citrate dextrose in phosphate-buffered saline (PBS), supplemented with 0.5% w/v deionized BSA (Sigma) and adjusted to pH 7.2–7.4 with sodium bicarbonate. Cells were incubated with FcR blocking agent and labeled with monoclonal hapten-conjugated anti-human CD34 antibody (QBEND10, Beckman Coulter, High Wycombe, Bucks, UK), then conjugated to magnetic microbeads, according to the manufacturer’s instructions. The cells were passed through a positive selection column in the presence of a strong magnetic field. CD34⫹-depleted Ad⫺BMMC were collected after they passed unbound through the column. CD34⫹-enriched cells that bound to the column were eluted in wash buffer after removing the magnetic field. Flow cytometry. CD34⫹ cell frequency was assessed by flow cytometry (FACScan, Becton-Dickinson, San Jose, CA, USA) using anti-human CD34 antibody conjugated to phycoerythrin (PE) (HPCA-2, Becton-Dickinson) and compared to an appropriate isotype control. A total of 106 cells (or 105 CD34⫹ selected cells) were washed twice at 4⬚C in PBS supplemented with 0.5% w/v sodium azide and 0.1% v/v FCS. Cells were then incubated for 10 minutes on ice with 40 µL human γ-globulin (20 µg/mL, Sigma), followed by a further 30 minutes on ice with anti-CD34-PE antibody (10 µL), or an isotype-matched irrelevant IgG1-PE control. After labeling, the cells were washed again at 4⬚C and fixed in 2% w/v paraformaldehyde in PBS. Forward light scatter, side-angle light scatter, and fluorescence were acquired for 50,000 events (FACScan): data were analyzed using CellQuest software. Culture preparation. For each normal donor, pairs of cultures were prepared with low (0.50–0.65%) and normal (1.04–2.17%) percentages of CD34⫹ cells by recombining the CD34⫹-enriched and CD34⫹-depleted cell fractions. Specifically, control cultures were generated to mimic the original composition of the Ad⫺ BMMC while those prepared with low numbers of CD34⫹ cells approximated the range of quantitative reduction of CD34⫹ cells described in AA BM. Cell mixtures were incubated overnight in the presence or absence of 5-FU as described above, using 5 × 104 cells/mL in the colony assay and 5 × 106 cells/mL in MS5-supported LTBMC, or the equivalent volume from the treated cultures. For a single AA BM experiment (Table 1, UPN 8), adherent cells were removed and the remaining cell fraction was separated into CD34⫹ and CD34⫺ populations as described above. Pairs of cultures were produced at the original concentration of 0.66% CD34⫹ cells, and with CD34⫹ cell frequency increased to within the normal range (1.27%).
Figure 1. Schematic of experimental design for regeneration of CFU by 5-FU-treated Ad⫺ BMMC in LTBMC (A: dashed line) or following modification of CD34⫹ cell content of Ad⫺ BMMC (B: dotted line); common steps are shown in boxes.
Statistical analysis Wilcoxon’s signed-rank test was used to assess the significance of treatment with 5-FU on the generation of CFU compared with untreated controls. The Mann-Whitney U-test was used to assess
668
S. Rizzo et al. / Experimental Hematology 32 (2004) 665–672
the significance of differences between normal and AA cultures before and after treatment with 5-FU. The Student’s t-test was applied to determine the significance of differences between pairs of normal donor cultures with normal or modified CD34⫹ cell content.
Results LTBMC of unmodified adherent cell–depleted (Ad⫺) BMMC: normal vs aplastic anemia Treatment of normal and AA Ad⫺ BMMC with 5-FU resulted in the elimination of approximately 95% of CFU at the initiation of culture (week 0). The regeneration of CFU production in LTBMC (Table 2) began by week 2. Normal donor CFU increased to a maximum between weeks 2 and 4. In AA cultures, maximum CFU production occurred earlier, between weeks 1 and 3; CFU were produced until week 3 in 3/7 cases and until week 4 in 2/7 cases. No 5FU-treated AA culture produced CFU at week 5 or beyond. In untreated cultures, AA BM produced significantly lower numbers of CFU compared to normal from week 0 through week 7 (Table 2). Following 5-FU treatment, AA BM produced significantly lower numbers of CFU compared to normal from week 2 to week 5 (Table 2). Therefore, to determine whether the 5-FU-treated cultures simply reflected this overall deficit, the data were expressed as the mean values of the proportion of CFU produced following treatment with 5-FU compared with untreated cultures for individual normal and AA cases respectively (Fig. 2). Results were similar in AA and normal cultures on weeks 1 and 2, but were significantly lower in AA cultures compared to normal cultures at weeks 3, 4, and 5. Most notably, in normal cultures, CFU production expressed in this way reached a maximum mean value of 0.247 at week 5. For AA cultures, the maximum mean value was 0.072 at week 2. The difference between CFU output each week by pairs of untreated and 5-FU-treated flasks became less marked as the duration of the experiment increased, but treated cultures consistently produced fewer colonies.
Figure 2. Mean CFU production following treatment with 5-FU expressed as proportions of CFU produced in control (untreated) cultures. The means of normal (mean ⫾ SEM, n ⫽ 6, solid squares) and AA (mean ⫾ SEM, n ⫽ 7, open squares) cultures are shown and significant differences between normal and AA cultures were determined by the Wilcoxon’s signed-rank test (∗∗p ⬍ 0.01).
Control cultures vs cultures with modified CD34⫹ cell content The apparent difference in cell culture kinetics from AA patient cells could be simply due to the low numbers of CD34⫹ cells in these samples. In order to test whether CD34⫹ cell concentration directly affects the kinetics of LTBMC, normal donor Ad⫺BMMC were separated into CD34⫹ and CD34⫺ cells and then recombined to either the original CD34⫹ cell content or to a lower concentration, as found in AA patient BM. 5-FU treatment resulted in elimination of 95.5% ⫾ 1.9% CFU in the control group reconstituted to original levels of CD34⫹ cells and of 94.2% ⫾ 2.4% in the low CD34⫹ cell group (n ⫽ 9, p n.s.). All cultures had commenced colony production by week 1 following 5-FU treatment (Table 3). In the absence of 5-FU treatment, the mean number of colonies produced during LTBMC containing low numbers
Table 2. Absolute numbers of CFU produced in LTBMC by normal and AA Ad⫺ BMMC before and after treatment with 5-FU Week
Untreated normal (n ⫽ 6)
Untreated AA (n ⫽ 7)
p
5-FU treated normal (n ⫽ 6)
0 1 2 3 4 5 6 7 8
4217 ⫾ 745 (1650–7050) 1686 ⫾ 191 (543–3528) 908 ⫾ 63 (396–1336) 373 ⫾ 58 (0–714) 230 ⫾ 48 (42–542) 97 ⫾ 41 (12–404) 63 ⫾ 24 (6–166) 12 ⫾ 3 (0–24) 4 ⫾ 2 (0–12)
1007 ⫾ 258 (500–2050) 444 ⫾ 126 (92–1248) 176 ⫾ 40 (42–387) 74 ⫾ 17 (24–162) 32 ⫾ 8 (12–60) 11 ⫾ 3 (0–24) 1 ⫾ 1 (0–6) 0 0
⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.05 ⬎0.05
117 ⫾ 71 (0–450) 57 ⫾ 14 (0–190) 75 ⫾ 19 (12–276) 59 ⫾ 13 (7–180) 38 ⫾ 7 (18–78) 18 ⫾ 4 (0–42) 9 ⫾ 4 (0–24) 1 ⫾ 1 (0–6) 0
Mean ⫾ SEM (range), with results of Mann-Whitney U-test (p) of normal vs AA.
5-FU treated AA (n ⫽ 7) 0 19 ⫾ 5 9⫾3 4⫾2 2⫾1 0 0 0 0
(0–52) (0–30) (0–18) (0–6)
p ⬎0.05 ⬎0.05 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬎0.05 ⬎0.05 ⬎0.05
S. Rizzo et al. / Experimental Hematology 32 (2004) 665–672
669
Table 3. Absolute numbers of CFU produced by control (reconstituted to original percentage of CD34⫹ cells) and low CD34⫹ cell normal donor cultures Week 0 1 2 3 4 5 6∗
Untreated control
Untreated low
p
5-FU control
5-FU low
p
6217 ⫾ 2133 (850–21,550) 3570 ⫾ 604 (1240–6888) 1926 ⫾ 517 (607–5730) 1077 ⫾ 265 (0–2097) 1002 ⫾ 307 (117–2762) 701 ⫾ 211 (39–1787) 521 ⫾ 175 (36–975)
4044 ⫾ 1115 (650–11,600) 1464 ⫾ 283 (530–2535) 788 ⫾ 189 (214–1966) 491 ⫾ 136 (57–1164) 443 ⫾ 169 (9–1248) 223 ⫾ 86 (0–579) 134 ⫾ 74 (3–387)
⬎0.05 ⬍0.01 ⬍0.05 ⬍0.01 ⬍0.05 ⬍0.05 ⬍0.05
494 ⫾ 285 (0–2550) 501 ⫾ 204 (25–2007) 319 ⫾ 98 (6–942) 338 ⫾ 118 (0–1104) 339 ⫾ 136 (0–1221) 300 ⫾ 138 (6–1347) 161 ⫾ 57 (6–372)
339 ⫾ 188 (0–1650) 135 ⫾ 64 (6–595) 81 ⫾ 34 (0–309) 96 ⫾ 48 (0–453) 99 ⫾ 59 (0–516) 99 ⫾ 69 (0–615) 22 ⫾ 19 (0–117)
⬎0.05 ⬍0.05 ⬍0.01 ⬍0.05 ⬍0.05 ⬍0.05 ⬎0.05
Mean ⫾ SEM (range) of 9 cultures (∗n ⫽ 6), and results of paired Student’s t-test (p).
of CD34⫹ cells in comparison to control cultures was significantly reduced from week 1 to week 6 of culture, but colony production was maintained in the majority of these cultures in both the control group and the low CD34⫹ cell group over the 6 weeks of the experiment (Table 3). A similar pattern was found in 5-FU-treated cultures (Table 3) where colony production was maintained, but at a significantly lower output, from week 1 to week 5 in cultures with reduced CD34⫹ cell content. The mean proportion of colonies produced by 5-FU-resistant cells was comparable for control and low CD34⫹ cell cultures (Fig. 3), except at week 1 where control cultures contained significantly greater proportions of 5-FU-resistant colony-forming cells than found in modified cultures (p ⬍ 0.05). Examples of individual experiments are illustrated in Figure 4 (raw data shown in Table 4); Panel A shows a typical normal donor experiment. The proportion of 5-FUresistant CFU production in the low CD34⫹ cell cultures is reduced in comparison to the cultures with normal CD34⫹ cell content from week 2 to week 4, but by week 5 they are
Figure 3. The proportion of colonies produced by 5-FU-resistant cells during control and modified LTBMC. The mean ⫾ SEM (n ⫽ 9) number of colonies produced by 5-FU-resistant cells expressed as a proportion of CFU from untreated cultures in control cultures (1.04–2.17% CD34⫹) and by cultures reconstituted to low levels of CD34⫹ cells (0.50–0.65% CD34⫹) are depicted by solid circles and open circles, respectively. Significant differences were determined using the paired Student’s t-test (∗p ⬍ 0.05).
of a similar value. Over several years of collection of samples from a large group of AA patients, only one BM sample (Table 1, UPN 8) contained sufficient cells to produce cultures where CD34⫹ cells could be increased to normal levels (Fig. 4B; raw data shown in Table 4). However, in these corrected cultures, the duration and proportion of colonies produced by 5-FU-resistant cells remained consistent with AA culture dynamics. Therefore, these results suggest that correction of the quantitative defect did not alter the kinetics of 5-FU-resistant cells in AA LTBMC.
Discussion The hematopoietic cell defect in aplastic anemia has been defined by quantitative and qualitative measurement in vitro. Maciejewski et al. [17] found that LTC-IC were severely reduced in AA compared to normal, but postulated that this was primarily a result simply of the quantitative reduction of CD34⫹ cells. However, in our recent publication reporting LTC-IC measurements in CD34⫹ cells purified from AA patient BM [19], 8/13 patients had LTC-IC frequencies below the normal range, suggesting that the composition of the CD34⫹ population differs significantly from normal. Furthermore, Schrezenmeier et al. [16] found lower proportions of CAFC in AA BM compared to normal, although, as they state, such an assay cannot discriminate the nature of the stem cell defect. 5-FU is toxic to cells in S-phase of the cell cycle [29– 31]. Experiments with mouse [32–36] and early human hematopoietic progenitors have shown similar resistance to 5-FU [27,37–39], indicating that these cells are quiescent in the G0 phase of cell cycle. Treatment with 5-FU, therefore, provides a system for enriching early hematopoietic progenitors, including the proportion of LTC-IC that are quiescent. In order to investigate differences in the proliferation of early progenitors from normal and AA BM, our first series of experiments used BMMC populations treated with 5-FU at a concentration known to eliminate cycling cells [26– 28] and recharged onto confluent MS5 feeder layers in a LTBMC coculture system. Untreated AA cultures produced significantly lower numbers of CFU compared to normal cultures, throughout the culture period, and AA CFU production ceased prematurely.
670
S. Rizzo et al. / Experimental Hematology 32 (2004) 665–672
Figure 4. The proportion of colonies produced by 5-FU-resistant cells during normal donor and aplastic anemia patient LTBMC. Both panels show mean colony production per 5-FU-treated flask expressed as a fraction of a paired untreated flask. (A): An example of normal donor cell culture. The solid circles represent data from the pair of flasks reconstituted to the original CD34⫹ cell percentage of 1.92%. A second pair of flasks contained cell mixtures with CD34⫹ cells reduced to 0.5% (open circles, lower dashed line), as would typically be found in an AA culture. (B): Data of cultures from an aplastic anemia patient. The solid circles, connected by a solid line, represent the pair of Ad⫺ BMMC flasks at the original 0.66% CD34⫹. Column-separated cells were recombined to produce a pair of cultures with increased CD34⫹ cell content (1.27%; open circles, dotted line).
Since MS5 feeder layers support human hematopoiesis as effectively as human stroma [24,25,40], our results agree with other coculture experiments that indicate an HSC defect in AA [5,41], since the aplastic pattern of growth is not corrected by the support of a normal microenvironment. Treatment with 5-FU eliminated the majority of CFU at the start of culture, indicating the cycling status of these cells. AA cultures produced CFU for a shorter period than normal and the numbers of CFU produced by treated AA cultures were significantly lower than normal. Since AA BM has a lower proportion of CD34⫹ cells than normal [1,2,8], this might account for the lower numbers of CFU both before and after 5-FU treatment. For this reason, CFU production in each 5-FU-treated culture was expressed as a proportion of CFU production in its untreated control. Generally, AA
cultures recovered at a similar rate to normal cells in the 2 weeks following 5-FU exposure, but subsequently the proportion of 5-FU resistant CFU was significantly reduced. AA BM is therefore dysfunctional in its ability to regenerate CFU production following 5-FU treatment, and this dysfunction is not simply a reflection of a reduced proportion of CD34⫹ cells in AA bone marrow. The dysfunction indicates an abnormality in the proliferation of early progenitor cells that give rise to CFU later in LTBMC, which does not prove or exclude a defect at the level of a primitive CD34⫺ HSC population. Maciejewski et al. [2] have shown that the cycling status of normal CD34⫹ cells is associated with the expression of the c-kit antigen. Since they found that significantly fewer AA CD34⫹ cells express c-kit, they proposed that AA bone
Table 4. Examples of absolute numbers of CFU produced per flask from LTBMC generated from a single normal donor and a single AA patient (shown in Fig. 4A and B, respectively) following modification of CD34⫹ cell content and treatment with 5-fluorouracil Normal donor LTBMC 1.92% (original) Week 0 1 2 3 4 5 6
Aplastic anemia LTBMC ⫹
0.50% CD34 (reduced)
0.66% (original)
1.27% CD34⫹ (increased)
no 5-FU
⫹ 5-FU
no 5-FU
⫹ 5-FU
no 5-FU
⫹ 5-FU
no 5-FU
⫹ 5-FU
3350 2428 1335 552 270 192 132
0 25 186 174 63 81 75
1500 530 801 315 84 18 21
0 6 9 13 6 6 3
5550 3044 630 285 54 0 0
100 23 6 0 0 0 0
8200 2980 1089 342 42 0 0
150 58 3 0 0 0 0
S. Rizzo et al. / Experimental Hematology 32 (2004) 665–672
marrow contains a lower proportion of cycling CD34⫹ cells than normal bone marrow. Our results appear to contradict this, since 5-FU has a relatively greater effect on CFU production by AA CD34⫹ cells in LTBMC. We suggest that the quantitative reduction of CD34⫹ cells in AA BM may lead to an increase in cell cycling to try and maintain steady-state hematopoiesis. The abnormal in vitro kinetics of 5-FU-resistant CD34⫹ cells from AA BM could then simply reflect the same phenomenon as proposed in vivo, so that decreased proportions of CD34⫹ cells result in feedback mechanisms within the LTBMC, resulting in increased cell cycling and more rapid differentiation, thus explaining the premature termination of AA LTBMC. We tested this hypothesis by culturing normal CD34⫹ cells in LTBMC at concentrations equivalent to those in the AA samples. We demonstrated that reduced CD34⫹ cell content led to a proportionate decrease in the total number of colonies generated, but duration of colony production was not significantly reduced. Furthermore, the proportion of 5-FU-resistant cells was largely unaffected in those cultures containing a reduced number of CD34⫹ cells. Therefore, it is unlikely that the primitive hematopoietic population within the CD34⫹ cell compartment is merely driven to cycle as a direct consequence of low cell numbers, whereas an intrinsic defect in the 5-FU-resistant population in aplastic anemia BM would be supported by our results. This was exemplified in the experiment where increasing CD34⫹ cell content of aplastic anemia cultures to within the normal range did not extend the duration of colony production or increase the proportion of 5-FU-resistant cells capable of producing CFU during LTBMC. It should be noted that we have only manipulated CD34⫹ HSC in this culture system and our results need to be interpreted in that context. However, it would be interesting to examine further the role of CD34⫺ HSC [42] in the human LTC-IC assay, which may shed light on the differences between normal and aplastic hematopoiesis. The CAFC and LTC-IC assays are based on the assumption that HSC proliferate and differentiate at a uniform rate, given the same culture conditions, since the number of LTCIC present in a population of cells is taken to be proportional to the number of CFU these cells give rise to at week 5 in culture. As shown here, AA HSC do not proliferate normally in vitro. What is particularly striking about the 5-FU experiments presented here is that, whereas all normal cultures produced CFU at week 5, in AA BM no CFU were produced at week 5 or beyond. Therefore, our results would suggest that the LTC-IC assay is an unreliable method of assessing early AA progenitor cells. Our data provide further evidence for defective stem cell function in AA, and also highlight the need for detailed analysis of the kinetics of proliferation in this disease. Understanding cell proliferation kinetics is essential if current systems of stem cell quantification are to be reliably used and may also provide a key to the pathogenesis of BM failure.
671
Acknowledgments The authors thank the clinical and nursing hematology staff of St. George’s Hospital for their assistance in obtaining bone marrow samples from normal donors and AA patients. In addition, we would like to acknowledge the generous gifts of cytokines provided by Janssen-Cilag Ltd. and by Novartis Pharmaceuticals UK Ltd. Financial support was provided by grants from the Aplastic Anaemia Trust (formerly the Marrow Environment Fund), Action Research, and the Leukaemia Research Fund.
References 1. Scopes J, Bagnara M, Gordon-Smith EC, Ball SE, Gibson FM. Haemopoietic progenitor cells are reduced in aplastic anaemia. Br J Haematol. 1994;86:427–430. 2. Maciejewski JP, Anderson S, Katevas P, Young NS. Phenotypic and functional analysis of bone marrow progenitor cell compartment in bone marrow failure. Br J Haematol. 1994;87:227–234. 3. Manz CY, Nissen C, Wodnar-Filipowicz A. Deficiency of CD34⫹ c-kit⫹ and CD34⫹38⫺ hematopoietic precursors in aplastic anemia after immunosuppressive treatment. Am J Hematol. 1996;52:264–274. 4. Juneja HS, Lee S, Gardner FH. Human long-term bone marrow cultures in aplastic anemia. Int J Cell Cloning. 1989;7:129–135. 5. Marsh JC, Chang J, Testa NG, Hows JM, Dexter TM. The hematopoietic defect in aplastic anemia assessed by long-term marrow culture. Blood. 1990;76:1748–1757. 6. Gibson FM, Gordon-Smith EC. Long-term culture of aplastic anaemia bone marrow. Br J Haematol. 1990;75:421–427. 7. Bacigalupo A, Figari O, Tong J, et al. Long-term marrow culture in patients with aplastic anemia compared with marrow transplant recipients and normal controls. Exp Hematol. 1992;20:425–430. 8. Scopes J, Daly S, Atkinson R, Ball SE, Gordon-Smith EC, Gibson FM. Aplastic anemia: evidence for dysfunctional bone marrow progenitor cells and the corrective effect of granulocyte colony-stimulating factor in vitro. Blood. 1996;87:3179–3185. 9. Ploemacher RE, Van der Sluijs JP, Voerman JS, Brons NH. An in vitro limiting-dilution assay of long-term repopulating hematopoietic stem cells in the mouse. Blood. 1989;74:2755–2763. 10. Ploemacher RE, Van der Sluijs JP, van Beurden CA, Baert MR, Chan PL. Use of limiting-dilution type long-term marrow cultures in frequency analysis of marrow-repopulating and spleen colony-forming hematopoietic stem cells in the mouse. Blood. 1991;78:2527–2533. 11. Neben S, Anklesaria P, Greenberger J, Mauch P. Quantitation of murine hematopoietic stem cells in vitro by limiting dilution analysis of cobblestone area formation on a clonal stromal cell line. Exp Hematol. 1993;21:438–443. 12. Sutherland HJ, Lansdorp PM, Henkelman DH, Eaves AC, Eaves CJ. Functional characterization of individual human hematopoietic stem cells cultured at limiting dilution on supportive marrow stromal layers. Proc Natl Acad Sci U S A. 1990;87:3584–3588. 13. Eaves CJ. Peripheral blood stem cells reach new heights. Blood. 1993; 82:1957–1959. 14. Orlic D, Bodine DM. What defines a pluripotent hematopoietic stem cell (PHSC): will the real PHSC please stand up. Blood. 1994;84: 3991–3994. 15. Lemieux ME, Rebel VI, Lansdorp PM, Eaves CJ. Characterization and purification of a primitive hematopoietic cell type in adult mouse marrow capable of lymphomyeloid differentiation in long-term marrow “switch” cultures. Blood. 1995;86:1339–1347. 16. Schrezenmeier H, Jenal M, Herrmann F, Heimpel H, Raghavachar A. Quantitative analysis of cobblestone area-forming cells in bone marrow of patients with aplastic anemia by limiting dilution assay. Blood. 1996;88:4474–4480. 17. Maciejewski JP, Selleri C, Sato T, Anderson S, Young NS. A severe and consistent deficit in marrow and circulating primitive hematopoietic
672
18.
19.
20.
21.
22.
23. 24.
25.
26.
27.
28.
29.
S. Rizzo et al. / Experimental Hematology 32 (2004) 665–672 cells (long-term culture-initiating cells) in acquired aplastic anemia. Blood. 1996;88:1983–1991. Podesta M, Piaggio G, Frassoni F, et al. The assessment of the hematopoietic reservoir after immunosuppressive therapy or bone marrow transplantation in severe aplastic anemia. Blood. 1998;91:1959–1965. Rizzo S, Scopes J, Elebute MO, Papadaki HA, Gordon-Smith EC, Gibson FM. Stem cell defect in aplastic anemia: reduced long-term culture-initiating cells (LTC-IC) in CD34⫹ cells isolated from aplastic anemia patient bone marrow. Hematol J. 2002;3:230–236. Bertolini F, Battaglia M, Lanza A, et al. Multilineage long-term engraftment potential of drug-resistant hematopoietic progenitors. Blood. 1997;90:3027–3036. Camitta BM, Rappeport JM, Parkman R, Nathan DG. Selection of patients for bone marrow transplantation in severe aplastic anemia. Blood. 1975;45:355–363. Bacigalupo A, Hows J, Gluckman E, et al. Bone marrow transplantation (BMT) vs immunosuppression for the treatment of severe aplastic anaemia (SAA): a report of the EBMT SAA working party. Br J Haematol. 1988;70:177–182. Camitta BM. What is the definition of cure for aplastic anemia? Acta Haematol. 2000;103:16–18. Itoh K, Tezuka H, Sakoda H, et al. Reproducible establishment of hemopoietic supportive stromal cell lines from murine bone marrow. Exp Hematol. 1989;17:145–153. Suzuki J, Fujita J, Taniguchi S, Sugimoto K, Mori KJ. Characterization of murine hemopoietic-supportive (MS-1 and MS-5) and non-supportive (MS-K) cell lines. Leukemia. 1992;6:452–458. Dooley DC, Law P. Detection and quantitation of long-term cultureinitiating cells in normal human peripheral blood. Exp Hematol. 1992; 20:156–160. Breems DA, Blokland EA, Neben S, Ploemacher RE. Frequency analysis of human primitive haematopoietic stem cell subsets using a cobblestone area forming cell assay. Leukemia. 1994;8:1095–1104. Rice A, Boiron JM, Barbot C, et al. Cytokine-mediated expansion of 5-FU-resistant peripheral blood stem cells. Exp Hematol. 1995;23: 303–308. Heidelberger C, Danenberg PV, Moran RG. Fluorinated pyrimidines and their nucleosides. Adv Enzymol Relat Areas Mol Biol. 1983;54: 58–119.
30. Lenz HJ, Manno DJ, Danenberg KD, Danenberg PV. Incorporation of 5-fluorouracil into U2 and U6 snRNA inhibits mRNA precursor splicing. J Biol Chem. 1994;269:31962–31968. 31. Mirjolet J-F, Didelot M, Barberi-Hayob M, Merlin J-L. G1/S but not G0/G1 cell fraction is related to 5-fluorouracil cytotoxicity. Cytometry. 2002;48:6–12. 32. Hodgson GS, Bradley TR. Properties of hematopoietic stem cells surviving 5-fluorouracil treatment: evidence for a pre-CFU-S cell? Nature. 1979;281:381–382. 33. Van Zant G. Sudies of hematopoietic stem cells spared by 5-fluorouracil. J Exp Med. 1984;159:679–690. 34. Lerner C, Harrison DE. 5-Fluorouracil spares hemopoietic stem cells responsible for long-term repopulation. Exp Hematol. 1990;18:114– 118. 35. Down JD, Ploemacher RE. Transient and permanent engraftment potential of murine hematopoietic stem cell subsets: differential effects of host conditioning with gamma radiation and cytotoxic drugs. Exp Hematol. 1993;21:913–921. 36. Stewart FM, Crittenden RB, Lowry PA, Pearson-White S, Quesenberry PJ. Long-term engraftment of normal and post-5-fluorouracil murine marrow into normal nonmyeloablated mice. Blood. 1993;81: 2566–2571. 37. Brandt JE, Baird N, Lu L, Srour EF, Hoffman R. Characterization of a human hematopoietic progenitor cell capable of forming blast cell containing colonies in vitro. J Clin Invest. 1988;82:1017–1027. 38. McNiece IK, Stewart FM, Deacon DM, et al. Detection of a human CFC with a high proliferative potential. Blood. 1989;74:609–612. 39. Stewart FM, Temeles D, Lowry P, Thraves T, Grosh WW, Quesenberry PJ. Post-5-fluorouracil human marrow: stem cell characteristics and renewal properties after autologous marrow transplantation. Blood. 1993;81:2283–2289. 40. Issaad C, Croisille L, Katz A, Vainchenker W, Coulombel. A murine stromal cell line allows the proliferation of very primitive human CD34⫹⫹/CD38⫺ progenitor cells in long-term cultures and semisolid assays. Blood. 1993;81:2916–2924. 41. Marsh JC, Chang J, Testa NG, Hows JM, Dexter TM. In vitro assessment of marrow “stem cell” and stromal cell function in aplastic anaemia. Br J Haematol. 1991;78:258–267. 42. Kuci S, Wessels JT, Buhring HJ, et al. Identification of a novel class of human adherent CD34⫺ stem cells that give rise to SCID-repopulating cells. Blood. 2003;101:869–876.