Multiple Inhibitory Cytokines Induce Deregulated Progenitor Growth and Apoptosis in Hematopoietic Cells From Fac^−/− Mice

FANCONI ANEMIA (FA) is a complex genetic disorder characterized by progressive acquisition of bone marrow (BM) aplasia, chromosomal instability, predisposition to malignancies, and hypersensitivity to bifunctional alkylating agents.1-4 At least eight complementation groups have been identified on the basis of somatic cell fusion studies that result in a generally similar phenotype.5,6 Three of the eight complementation groups (A, C, and D) have been mapped to different chromosomal loci,5,7,8 and the cDNAs of the A and C genes have been identified.9-11 Unfortunately, the cloning of the A and C genes have not provided insight into the function of FA genes in cellular homeostasis.

Maintenance of normal hematopoiesis involves complex interactions between the stem/progenitor cells and multiple stimulatory and inhibitory molecules.12,13 The observation that Fanconi anemia complementation type C (FAC) patients acquire a progressive BM failure suggests that FAC plays a role in proliferation and/or survival of hematopoietic stem and progenitor cells. BM aplasia may occur by multiple mechanisms including alterations in the production or intracellular signaling of stimulatory and inhibitory cytokines. Two murine models containing a disruption of the murine homologue of FAC (Fac) have recently been developed to facilitate functional studies in primary cells in vitro and in vivo.14,15 Chen et al14 created a disruption in exon 8 of Fac, while Whitney et al15 used homologous recombination to create a disruption in exon 9. In both models, spontaneous chromosomal aberrations were observed, as well as an increase in chromosome breaks in splenic lymphocytes in response to bifunctional alkylating agents analogous to lymphocytes in FA patients.14 15 

Whitney showed that hematopoietic progenitors from mice containing a disruption of Fac (Fac^−/−) were hypersensitive to IFN-γ and, recently, Rathbun et al, using the model developed by Whitney,15 determined that low doses of IFN-γ affected programmed cell death inFac^−/− hematopoietic progenitors by inducing Fas expression. Similar results were obtained in studies using primary cells from a patient with Fanconi anemia type C.16 

Other inhibitory cytokines, such as tumor necrosis factor-α (TNF-α), have also been implicated in the pathogenesis of aplastic anemia.17 In addition, TNF-α is frequently elevated in patients with Fanconi anemia.18,19 TNF-α is known to induce Fas-mediated apoptosis,17 as well as an alternative apoptosis pathway involving TNF-α receptor-associated death domain (TRADD).20 On the basis of these observations, we hypothesized that Fac-deficient cells may be hypersensitive to TNF-α and potentially other inhibitory cytokines.

In this report, using a genetically distinct model14 from that previously used by Whitney,15 we confirm that disruption of both alleles of Fac(Fac^−/−) results in deregulated colony formation in response to IFN-γ. We also show that a disruption in exon 8 of Fac results in a hypersensitive reduction in progenitor growth in cultures containing TNF-α as well as another inhibitory cytokine macrophage inflammatory protein-1α (MIP-1α). Further, we show that Fac is crucial for prevention of TNF-α and MIP-1α–mediated apoptosis in primary immature myeloid hematopoietic cells. Finally, we observed altered proliferation kinetics in pluripotent and lineage restrictedFac^−/− hematopoietic progenitors in vivo. These results implicate deregulated apoptosis in response to inhibitory cytokines in the pathogenesis of BM aplasia in Fanconi anemia type C.

Heterozygote Fac mice14 were housed in our animal facility. DNA was extracted from mouse tails using a standard phenol-chloroform method. Genotyping was conducted by amplifying sequences unique to exon 8 and neomycin resistance gene by polymerase chain reaction (PCR). The following primers were used: 5′CCTGCCATCTTCAGAATTGT3′ (exon 8 primer), 5′GAGCAACACAAATGGTAAGG3′ (intron 8 primer), and 5′TTGAATGGAAGGATTGGAGC3′ (neomycin resistance gene primer).

BM cells were flushed from femurs and tibias ofFac^−/− and Fac^+/+littermate controls with Iscove's modified Dulbecco's medium (IMDM) (GIBCO-BRL, Gaithersburg, MD) supplemented with 5% fetal calf serum (FCS) (Hyclone, Logan, UT). Spleen cell suspensions were prepared by flushing cells from spleens and passing cells through a 23-gauge needle. Total nucleated cell number of BM and spleens was determined before manipulation of cells for further experimentation. Low-density mononuclear cells (LDMNC) were prepared by centrifugation on Ficoll-Hypaque (density 1.119; Sigma Chemical Co, St Louis, MO). Unfractionated BM and spleen cells were used in all experiments except for mitomycin C (MMC) dose-response experiments, which used LDMNC.

Clonogenic methylcellulose assays were plated in triplicate at 1 × 10⁴ to 5 × 10⁴ BM cells/mL and 5 × 10⁵ spleen cells/mL. Cultures were established in 1% IMDM methylcellulose (Stem Cell Technology, Vancouver, BC, Canada), with 30% FCS, 50 ng/mL recombinant murine Steel factor (a generous gift from Immunex, Seattle, WA) or SCF (Peprotech, Rocky Hill, NJ), 4 U/mL recombinant human erythropoietin (Amgen), 5% vol/vol pokeweed mitogen spleen conditioned media (PWMSCM) and 0.1 mmol hemin (Sigma). Cells were incubated at 37°C, 5% CO₂, and lowered (5%) O₂. BM cells were cultured in methylcellulose progenitor assays with increasing concentrations of MMC (Sigma), recombinant murine IFN-γ (R&D Systems, Minneapolis, MN), recombinant murine TNF-α (R&D Systems), recombinant murine MIP-1α (R&D Systems), and thrombopoietin (Genzyme, Cambridge, MA) to generate dose-response curves for Fac^−/− andFac^+/+ hematopoietic progenitor cells. Cultures to establish the responsiveness of Fac^−/−cells to MMC were conducted exactly as above, except recombinant murine GM-CSF (Peprotech) was substituted for PWMSCM, and cultures were incubated at 21% O₂. CFU-GM, BFU-E, and CFU-GEMM colonies were scored on day 7 of culture. Levels of significance were determined using Student's t-distribution.

The proportion of hematopoietic progenitor cells in S phase was estimated by means of suicide assays that used either³H-thymidine or hydroxyurea as previously described.21-23 BM and spleen cells were pulse-treated for 30 minutes at 37°C with either high specific activity³H-thymidine (50 mCi/mL, specific activity = 20 Ci/mmol; New England Nuclear, Boston, MA) or control medium. Cells were washed twice with control medium before plating in methylcellulose cultures as described above. The percentage suicide was determined by the following calculation: (progenitors in control − progenitors in³H-thymidine) divided by progenitors in control. The level of significance was determined using Student's t-test. These results were compared with the percentage suicide determined by treatment of cells with hydroxyurea (HU) as described elsewhere.23 

Pooled samples of BM LDMNC from fourFac^−/− and fourFac^+/+ animals were stained with Sca1-PE, B220-FITC, and CD3-FITC (PharMingen, San Diego, CA) for 15 minutes at 4°C, washed twice, and sorted by a Becton Dickinson fluorescence activated cell sorter for Sca1⁺B220⁻CD3⁻ and Sca1⁻B220⁻CD3⁻. Sca1⁺B220⁻CD3⁻ cells were incubated at a density of 2 × 10⁵ cells/mL in a 96-well tissue culture plate in IMDM 20% FCS and were cultured in the following combinations of cytokines: (1) GM-CSF (200 ng/mL) and SCF (100 ng/mL) only, (2) GM-CSF and SCF together with 1 ng/mL TNF-α or 100 ng/mL MIP-1α, and (3) TNF-α (1 ng/mL) or MIP-1α (100 ng/mL). Cultures containing Sca1⁻B220⁻CD3⁻cells were established at a density of 2 × 10⁶cells/mL using the same concentrations of GM-CSF, TNF-α, and MIP-1α as above described for Sca1⁺B220⁻CD3⁻. After 24 hours, cytospins were made for each condition.

Apoptosis was evaluated using the terminal deoxynucleotidyl transferase (tDt)-mediated dUTP nick end-labeling (TUNEL) assay24 as specified by the manufacturer (Boehringer Mannheim, Indianapolis, IN). Briefly, cells were fixed with 4% formaldehyde (Sigma) for 30 minutes at room temperature and permeabilized with 0.1% sodium citrate (Fisher Scientific, Fair Lawn, NJ) 0.1% Triton-X100 (Boehringer Mannheim) for 2 minutes at 4°C. Cells were incubated with tDt and dUTP-FITC in a humidified environment at 37°C, 5% CO₂ for 1 hour. After incubation, cytospins were evaluated by fluorescence microscopy. Photographs were obtained of each condition and scoring of apoptotic cells was conducted. As an independent control to verify the function of the assay, BM cells were irradiated (700 rads) and then maintained in liquid culture for 24 hours. The TUNEL assay was performed on these cells with and without the addition of tDt as positive and negative controls respectively. Three or four independent experiments were conducted with each inhibitory cytokine. Statistical significance was determined using the Student's t-test.

A characteristic feature of Fanconi anemia is the hypersensitivity of cells to clastogenic agents such as mitomycin C. Because FA patients have defects in hematopoietic progenitor cell function, it was critical to determine whether hematopoietic progenitors fromFac^−/− mice were hypersensitive to MMC as observed in FA patients. As shown in Fig1, a 50% reduction in maximal colony formation was observed at a 10-fold lower concentration of MMC inFac^−/− progenitors as compared withFac^+/+ littermates. We conclude that the MMC hypersensitivity observed in Fac^−/−progenitors is remarkably similar to that seen in FA patients.

IFN-γ, TNF-α, and MIP-1α are cytokines known to inhibit hematopoietic cell growth.25-29 We examined the effect of these inhibitory cytokines on growth of multipotential and lineage restricted progenitor cell growth of Fac mice in vitro (Fig 2). Multipotent (CFU-GEMM), erythroid (BFU-E), and granulocyte-macrophage (CFU-GM) progenitors cultured fromFac^−/− BM cells showed a 50% reduction in maximal colony formation at a 50- to 100-fold lower concentration of each of the respective cytokines as compared with progenitors cultured from WT mice. No differences in colony formation were observed betweenFac^−/− and Fac^+/+hematopoietic progenitors when increasing concentrations of a noninhibitory cytokine (thrombopoietin) were added to the cultures (data not shown).

To test whether Fac^−/− cells are predisposed to TNF-α– or MIP-1α–mediated apoptosis, two populations of hematopoietic cells were cultured in the presence of TNF-α or MIP-1α, then analyzed by the TUNEL assay. One population was highly enriched for myeloid hematopoietic progenitors (Sca1⁺B220⁻CD3⁻), and a second population was highly enriched for myeloid differentiated cells (Sca1⁻B220⁻CD3⁻). More than 90% of Sca1⁻B220⁻CD3⁻cells had macrophage (Mac1) or granulocyte (GR-1) surface markers (data not shown). Figure 3 shows representative examples of Sca1⁺B220⁻CD3⁻ cells after liquid culture with TNF-α and TUNEL assay. The summary of experiments evaluating apoptosis ofFac^−/− cells in response to TNF-α are shown in Table 1. Low equivalent levels of apoptosis were observed in both Fac^−/−and Fac^+/+ cells when cultured with GM-CSF and SCF; cytokines that protect hematopoietic cells from apoptosis.30 31 The addition of TNF-α to cultures containing GM-CSF and SCF resulted in increased apoptosis inFac^−/− primitive myeloid cells, but not in the Fac^+/+ cells. When hematopoietic cells were cultured with TNF-α alone, a dramatic increase in apoptosis was observed in immature (Sca1⁺B220⁻CD3⁻) and differentiated (Sca1⁻B220⁻CD3⁻) myeloid cells from Fac^−/− mice.

Experiments evaluating apoptosis inFac^−/− cells after culture with MIP-1α are summarized in Table 2. Again, low levels of apoptosis were detected inFac^−/− and Fac^+/+primitive and differentiated cells when cultured with GM-CSF and SCF. Similar levels of apoptosis were observed when MIP-1α was added to the cultures. However, culture of Sca1⁺B220⁻CD3⁻ cells from Fac^−/− mice with MIP-1α alone resulted in a dramatically increased level of apoptosis as compared with Fac^+/+ mice whose level of apoptosis was unchanged from conditions with protective cytokines. A trend (though not statistically significant) toward increased apoptosis was observed in Sca1⁻B220⁻CD3⁻cells from Fac^−/− mice. Together these data are consistent with the enhanced growth inhibition ofFac^−/− progenitors in response to TNF-α and MIP-1α observed in the clonogenic assays (Fig 2).

The maintenance of normal hematopoietic cell populations results from a complex homeostasis of cell production and apoptosis. To determine whether there were differences in the absolute number of progenitors per organ in mice homozygous for disruption of Fac,methylcellulose cultures of splenic and BM cells were established. The absolute number of progenitors in the BM and spleen as well as total nucleated cells per organ of Fac^−/−mice were similar to those of their Fac^+/+littermates (Table 3). BecauseFac^−/− hematopoietic progenitors are hypersensitive to multiple inhibitory cytokines and both immature and differentiated myeloid cells undergo increased apoptosis in response to TNF-α and MIP-1α, it is curious thatFac^−/− mice have equivalent numbers of myeloid progenitors and differentiated cells asFac^+/+ mice (Table 3). We hypothesized that if deregulated apoptosis were occurring in vivo,Fac^−/− progenitors may exhibit abnormal proliferation kinetics. To investigate this hypothesis, BM and spleen cells from Fac^−/− andFac^+/+ mice were pulse treated with tritiated thymidine or hydroxyurea and cultured in methylcellulose for growth of progenitors. The percentage suicide of multipotential (CFU-GEMM), erythroid (BFU-E), and granulocyte-macrophage (CFU-GM) progenitors in both the spleen and BM was dramatically higher inFac^−/− mice as compared withFac^+/+ controls (Table 3). The percentage suicide was comparable using either tritiated thymidine or hydroxyurea. These data infer that an increased proportion of clonogenic hematopoietic progenitors from Fac^−/− mice are in S phase and that altered proliferation kinetics exist in vivo.

The development of murine models using homologous recombination to delete the murine homologue of a gene known to cause human disease is extremely useful for understanding the normal function of the gene, the pathogenesis of the human disease, and potential treatment strategies. Some murine models replicate the human disease completely, whereas the phenotype of other models is somewhat variant from the human disease.32 A characteristic feature of Fanconi anemia is the hypersensitive reduction in growth of hematopoietic progenitors cultured in methylcellulose in the presence of MMC. Therefore, it was critical to determine whether hematopoietic progenitors cultured fromFac^−/− mice used in these studies were hypersensitive to MMC. The observation that hematopoietic progenitors derived from the murine BM are hypersensitive to MMC is particularly significant because it supports using this model to gain insight into the molecular mechanisms involved in the development of the progressive BM failure in FA patients.

BM failure or acquisition of myeloid leukemia, or both, are the most common causes of mortality in patients with FA.33,34 The precise role of FA genes in maintaining normal hematopoietic cellular homeostasis is the current challenge of many laboratories.35 The role of FA proteins in DNA repair,36 redox status of the cell,37-39 and apoptosis38,40,41 are three active areas of investigation. Cumming et al40 provided evidence that FAC may have a role in apoptosis when they demonstrated that overexpression of FACin a megakaryocytic IL-3–dependent cell line (MO7e cells), resulted in the prevention of growth factor dependent apoptosis. Other studies reported increased spontaneous apoptosis in FA lymphoblasts but decreased apoptosis when stressed with γ-irradiation as compared with lymphoblasts derived from normal donors.41 Although these studies in immortalized cell lines implicate FAC in the modulation of apoptosis, it is difficult to determine how accurately they reproduce the biochemical defect and cellular physiology of primary hematopoietic progenitor cells.

Rathbun et al16 recently made the first observation that deregulated apoptosis was evident in primaryFac^−/− hematopoietic progenitor cells These investigators reported data demonstrating decreased clonogenic growth in human and murine Fac^−/−progenitors in response to IFN-γ and/or anti-Fas activating antibody, which infers an increased sensitivity to Fas-mediated apoptosis.16 We chose to extend those observations by evaluating multiple inhibitory cytokines that modulate hematopoietic cell growth by alternative intracellular signaling pathways.17,20,42-44 IFN-γ and TNF-α have been implicated in the pathogenesis of acquired aplastic anemia by upregulating Fas expression on hematopoietic cells.19,42,43TNF-α has additional signaling mechanisms to regulate apoptosis through TNF receptor associated proteins, such as TRADD and receptor-interacting protein (RIP), and activation of NF-κB.20 MIP-1α is a chemokine that appears to induce growth inhibition through the Raf-1 kinase pathway.44 A further distinction between MIP-1α and both TNF-α and IFN-γ is that TNF-α and IFN-γ induce apoptosis in normal hematopoietic cells at high cytokine concentrations,45 whereas MIP-1α had not yet been evaluated for induction of apoptosis.

Our data, using a genetically distinct murine model, confirm previous findings that homozygous disruption of Fac results in hypersensitivity to IFN-γ.15,16 This hypersensitivity is particularly interesting as few data are available regarding the role of inhibitory cytokines in other genomic instability syndromes.46-49 In addition to hypersensitivity ofFac^−/− progenitors to IFN-γ, we have shown decreased clonogenic growth in response to TNF-α and MIP-1α which was not observed previously.15 Furthermore, our data also support the concept of Fac directly or indirectly influencing programmed cell death by showing TNF-α and MIP-1α induced apoptosis using a methodology that permitted direct detection of apoptosis in primary progenitor and differentiated cells. We showed that purified populations of primary cells enriched for either hematopoietic progenitor or myeloid differentiated cells fromFac^−/− mice undergo increased cytokine-mediated apoptosis after culture with TNF-α alone or in combination with cytokines that protect hematopoietic cells from apoptosis. In addition, we showed that primitive myeloid cells (Sca1⁺B220⁻CD3⁻) undergo enhanced apoptosis when cultured with MIP-1α as a single cytokine. This is the first description of a chemokine inducing apoptosis in hematopoietic cells. Together our findings and previous data16 are consistent with the hypothesis that there is a progressive loss of hematopoietic progenitor and stem cells inFac^−/− mice and FA patients resulting directly or indirectly from inhibitory cytokine-mediated apoptosis.

We hypothesized that if apoptosis is involved in the pathogenesis of FA BM failure in vivo in Fac^−/− mice, an increase in the number of proliferating hematopoietic progenitors would be required to sustain normal numbers of differentiated cells. To test this hypothesis we evaluated the suicide of hematopoietic progenitors using two independent agents. The markedly increased suicide observed in multipotent and lineage-restricted clonogenic cells from BM and spleen of Fac^−/− mice supports the hypothesis that a higher proportion ofFac^−/− hematopoietic progenitors in vivo are cycling due to loss of immature and differentiated cells from increased apoptosis. Alternatively, the increased suicide rate observed in Fac^−/− progenitors may reflect an increased sensitivity to tritium and hydroxyurea. The rationale for using hydroxyurea as an agent to assess the cycling status of clonogenic progenitors was because of recent data showing that FAC-deficient lymphoblast cell lines exhibit normal cell-cycle kinetics after 24-hour exposure to the drug.50 The lack of hypersensitivity to hydroxyurea and the similar suicide rates between the two agents support the hypothesis that the increased suicide rate in Fac^−/− progenitors is due to an accelerated proportion of clonogenic cells in S phase, rather than to hypersensitivity to the agents used to conduct these studies. However, to further delineate and confirm that an increased proportion ofFac^−/− progenitor cells are in S phase, future studies evaluating phenotypically defined populations of cells are indicated.

In summary, we have shown that loss of Fac results in deregulated apoptosis in myeloid hematopoietic cells in response to inhibitory cytokines with distinctive intracellular signaling pathways. It will be interesting in future experiments to evaluate how the loss of Fac disturbs the tightly regulated intracellular signaling of IFN-γ, TNF-α, and MIP-1α in mediating hematopoietic cell growth. Fac^−/− mice provide a valuable model system to evaluate the role of Fac in myeloid growth control.

We thank our colleagues at Indiana University and Dr Kevin Shannon (University of California, San Francisco) for reading the manuscript. We also thank Patricia Fox for secretarial support.