Interferon-γ–induced apoptotic responses of Fanconi anemia group C hematopoietic progenitor cells involve caspase 8–dependent activation of caspase 3 family members

Fanconi anemia (FA) is an autosomal recessive disorder characterized by cellular hypersensitivity to chemical cross-linking agents, bone marrow failure, diverse congenital anomalies, and a marked increase in the incidence of acute myelogenous leukemia.1-4 Phenotypically, the sine qua non of this disorder is hypersensitivity of FA cells to DNA cross-linking agents such as diepoxybutane (DEB) and mitomycin C (MMC).5,6 The disease is genetically heterogeneous, with at least 7 different complementation groups having been identified by somatic cell hybrid analysis.7-10 The genes encoding the A, C, D, F, and G groups have been cloned and have been mapped to chromosomes 16q24.3, 9q22.3, 3p25.3, 11p15, and 9p13, respectively.9 11-14 

The FANCC gene is constitutively expressed in most cells15 and encodes a 63-kd protein16,17 that has no strong amino acid sequence homology with any known gene family and is of unknown function. Although it is hypothesized that the protein plays some role in either facilitating repair of cross-linked DNA or resisting the effects of cross-linking agents on nuclear DNA, and although some of the FANCC protein is found in the nucleus,18,19 much of the protein is cytoplasmic.16,17 At least some critical function of the gene product appears to require cytoplasmic localization.20 

The product of the FANCC gene clearly plays a supportive role in growth or differentiation (or both) of hematopoietic progenitor cells (HPC),21,22 and progenitor cells from FA-C mice are suboptimally responsive to erythropoietin and Steel factor.23 FA-C progenitor cells are also hypersensitive to the apoptosis-inducing effects of interferon-γ (IFN-γ),24,25 in part by priming of the fassignaling pathway.24 Naturally, fas expression alone is insufficient to account for an apoptotic cellular response. Fas-ligand and other factors are required to link thefas-ligation event to distal caspase activation, includingFas-associating protein with death domain (FADD)26,27 and procaspase 8.28 29 

The active forms of the caspases are multimers the subunits of which are cleaved from the same proenzyme. The caspase family has been categorized into 3 subfamilies as determined by their substrate specificity and function. The ICE subfamily consists of caspases 1, 4, and 5.30 Caspase 1 has recently been reported to play a significant role in inflammation,31,32 and its role infas-induced apoptosis in most cells has now been called into question.33,34 The caspase 3 subfamily consists of caspases 3, 6, 7, 8, 9, and 10. Of these, those with large prodomains (ie, caspases 8, 9, and 10) are considered initiator caspases and those with a small prodomain (ie, caspases 3, 6, and 7) are considered effector or executioner caspases.35 A recent model proposes a branched pathway of caspase activation. In this model, caspase 8 activates caspases 3 and 7. Caspase 3 then activates caspase 6, which in turn may feed back on procaspase 3, resulting in a protease amplification cycle.36,37 The third caspase subfamily, ICH-1/Nedd2, consists of caspase 2 and its murine counterpart. It is predicted to act either as an apoptosis effector protein based on its substrate specificity,38 or as an initiator caspase, due to its large prodomain.39 

Because fas-mediated apoptosis in a wide variety of cells involves the ordered activation of the caspases30,40-44followed by cleavage of poly (adenosine diphosphate-ribose) polymerase (PARP),45 lamin,46,47 GATA-1,48VAV1,49 and other critical substrates for hematopoietic cells,30,50-52 we sought to test the notion that the caspases are also involved in the excessive apoptotic activity of FA cells. The ordered activation of caspases and their linkage with thefas pathway are not fully defined in hematopoietic progenitor cells exposed to IFN-γ and are not at all defined in cells bearing inactivating mutations of the FANCC gene, which renders them exquisitely sensitive to IFN-γ24,25 and tumor necrosis factor-α (TNF-α).53 We describe below results of in vitro experiments that demonstrate involvement of caspases 8 and 3 in murine and human FA-C cells primed with IFN-γ and treated with fas agonists.

The FANCC-deficient mice, homozygous for the targeted deletion of exon 9 of the murine Fanconi anemia complementation group C gene on a mixed genetic background of C57BL and 129Sv, were generated as described.25 

Murine bone marrow samples and human CD34⁺ cells were isolated and cultured as described previously.24 In each case signed informed consent was obtained from a parent. Tetrapeptide inhibitors of caspases 1 and 3 (ac-YVAD-cho and ac-DEVD-cho, respectively; BIOMOL, Plymouth Meeting, PA) were used in colony assays at a final concentration of 50 μmol/L. A fluoromethyl ketone inhibitor of caspase 8 (Z-IETD-FMK; 50 μmol/L; Enzyme Systems, Livermore, CA) was used in some experiments.

The EBV-transformed lymphoblast cell line HSC536N (a gift from Manuel Buchwald) was derived from peripheral blood cells of a child with FA-C. In the cells of this compound heterozygote, oneFANCC allele is deleted and the other carries a L-to-P substitution at amino acid position 554 (single-letter amino acid code). The EBV-transformed cell line JY was derived from EBV-infected normal peripheral blood mononuclear leukocytes.

Plasmid construction, packaging, transduction, selection, and complementation analysis were performed as previously described.24 

Immunoblot analyses were performed using cell lysates from human EBV-transformed lymphoblasts from normal volunteers and FA-C patients. An agonistic antihuman fas antibody (100 ng/mL, Upstate Biotechnology, Inc [UBI], Lake Placid, NY) and recombinant human IFN-γ (1 ng/mL, R&D Systems, Minneapolis, MN) were added for the intervals indicated. In some experiments, inhibitors to caspases 1, 3, and 8 (ac-YVAD-cho, ac-DEVD-cho, and Z-IETD-FMK, respectively) were added to a final concentration of 50 μmol/L for the times indicated.

The cells were harvested, washed 3 times with phosphate-buffered saline (PBS), and the cell pellets were solubilized in RIPA (10 mmol/L Tris-Cl, pH 7.6, 150 mmol/L NaCl, 1% sodium deoxycholate, 1% Triton X-100, 0.1% sodium dodecyl sulfate [SDS], and freshly added 1% aprotinin, 2 mmol/L Na₃VO₄, 1 μg/mL leupeptin, 1 mmol/L pepstatin A, and 1 mmol/L phenylmethylsulfonyl fluoride [PMSF]). Lysates were centrifuged at 16 000g for 15 minutes at 4°C. Protein concentrations were determined on supernatants using a protein microassay of the Bradford method (Bio-Rad, Hercules, CA). Cell lysates were heated at 94°C for 5 minutes in the presence of SDS and β-mercaptoethanol, and subjected to SDS-polyacrylamide gel electrophoresis (PAGE). Separated proteins were electroblotted onto Bio-Blot nitrocellulose (Costar, Cambridge, MA) as previously described.54 Each blot was stained with Ponceau S stain to confirm equal protein loading from lane to lane and photocopies were kept as a record. Nonspecific binding was blocked by incubating the blots for 1 hour in 5% (w/v) nonfat dry milk in water.

Caspase 1 and caspase 10 were detected by incubating blots with a rabbit polyclonal antihuman caspase 1 (ICE) antibody, or a goat polyclonal antihuman caspase 10 (MCH-4) antibody (no. sc-515 and no. sc-6185, respectively, Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:1000 in 5% nonfat dry milk. Caspase 3 (CPP32), caspase 7 (MCH-3), and FADD were detected with monoclonal antibodies (no. C31720, no. M64620, and no. F36620, respectively, Transduction Laboratories, Lexington, KY), each diluted 1:1000 in 5% milk. PARP was detected by incubating blots with a monoclonal anti-PARP antibody (no. SA-250; BIOMOL) diluted 1:3000. Caspase 8 was detected with a monoclonal antihuman caspase 8 (FLICE) antibody (no. 66231A; PharMingen, San Diego, CA) diluted 1:1000 in 5% milk. Caspase 9 was detected with a rabbit polyclonal antihuman caspase 9 antibody (no. AAP-109, Stressgen, Victoria, British Colombia, Canada) diluted 1:2000 in 5% milk. All primary antibody incubations were 1 hour except for anticaspase 9, which was for 2 hours. These were followed by 6 (5 minutes) washes with Tris-buffered saline containing 0.005% Tween-20 (TBS-T). The blots were incubated with secondary antibodies (goat antirabbit IgG-horseradish peroxidase [HRP] conjugate, goat antimouse IgG-HRP conjugate [Bio-Rad], or donkey antigoat IgG-HRP conjugate [Santa Cruz]) for 30 minutes at 1:10 000, 1:5000, or 1:4000 dilutions, respectively, then washed as above with TBS-T. Antibody-reactive proteins were detected using enhanced chemiluminescence (ECL) reagents (Amersham, Piscataway, NJ).

A microfluorescence assay was adapted from a protocol established by Enari and colleagues42 55; 2 × 10⁶ lymphoblasts were washed 2 times in 4°C PBS after timed exposure to human agonistic anti-fas antibody (100 ng/mL, UBI) and human IFN-γ (1 ng/mL, R&D Systems). In some experiments, inhibitors to caspases 1 (Z-WEHD-FMK), 1 and 4 (ac-YVAD-cho, Cell Permeable [CP]-YVAD-cho, Z-YVAD-FMK), 2 (Z-VDVAD-FMK), 3 (ac-DEVD-cho, CP-DEVD-cho, Z-DEVD-FMK), 6 (Z-VEID-FMK), and 8 (ac-IETD-cho and ac-IETD-FMK) were added for the times and final concentrations indicated (aldehyde inhibitors were obtained from BIOMOL, FMK inhibitors were obtained from Enzyme Systems). Cytosolic extracts were prepared by resuspending cell pellets in 50 μL extraction buffer (50 mmol/L PIPES-NaOH, pH 7.0, 50 mmol/L KCl, 5 mmol/L EGTA, 2 mmol/L MgCl₂, 1 mmol/L DTT, 20 μmol/L cytochalasin B, 1 mmol/L PMSF, 1 μg/mL leupeptin, 1 μg/mL pepstatin A, 50 μg/mL antipain, and 10 μg/mL chymopapain). Cells were disrupted by 5 cycles of freezing and thawing, followed by centrifugation at 10 000g for 12 minutes at 4°C. Protein concentrations of the supernatants were determined by a microassay of the Bradford method (Bio-Rad). A total reaction volume of 50 μL included 6.25 to 25 μg cell lysate, eitherN-acetyl-YVAD-MCA, or N-acetyl-DEVD-MCA fluorogenic substrate (10 μmol/L, BIOMOL), and assay buffer (100 mmol/L HEPES-KOH buffer, pH 7.5, 10% sucrose, 0.1% CHAPS, 10 mmol/L DTT, 0.1 mg/mL ovalbumin). Reactions were incubated at 30°C for 60 minutes in 96-well microtiter plates (Falcon) and enzyme activity detected by a Cytofluor II (PerSeptive Biosystems, Framingham, MA) at an excitation λ of 360 nm and an emission λ of 460 nm. Additional experiments used fluorogenic substrates for caspases 1, 2, 4, 6, and 7 (N-acetyl-WEHD-MCA, MCA-VDVAD, MCA-LEVDGW[K-DNP]-NH₂,N-acetyl-VEID-MCA, and MCA-VDOVDGW[K-DNP]-NH₂, respectively; Enzyme Systems).

Lymphoblasts (1 × 10⁶/mL) were plated in 24-well plates. IFN-γ (1 ng/mL, R&D Systems) and activating anti-fas antibody (100 ng/mL, UBI) were added individually or in combination for 3 or 48 hours, then washed twice with 2 mL staining buffer (PBS, 2% fetal bovine serum [FBS], 0.1% sodium azide). Cells were resuspended in 400 μL staining buffer and 400 μL cytofix/cytoperm (PharMingen) and incubated on ice for 20 minutes, were washed with 2 mL perm/wash buffer (PharMingen) and were resuspended in 100 μL perm/wash buffer for staining. Normal rabbit IgG (20 μL, Caltag) was added for 20 minutes on ice. Twenty microliters phycoerythrin (PE)-conjugated polyclonal rabbit antiactive caspase 3 antibody (no. 67345x, PharMingen) was then added and cells were incubated 30 minutes on ice in the dark. Cells were then washed twice with 2 mL perm/wash buffer and resuspended in 500 μL staining buffer for analysis. Cells were analyzed using a FACSCalibur (Becton Dickinson, Franklin Lakes, NJ) flow cytometer.

A portion of the cells cultured under various conditions were quantified for apoptosis using fluorescence microscopy and the TUNEL assay (ApopTag in situ Apoptosis Detection Kit; Oncor, Gaithersburg, MD). Cells with nuclear fragmentation or green fluorescence or both were scored as apoptotic.

Cells of the FA-C EBV-transformed lymphoblast cell lines HSC536N and the non-FA-C EBV-transformed cell line JY were exposed to an agonistic anti-fas antibody and a suboptimal (for normal cells) “priming” dose of IFN-γ (1 ng/mL) for 0, 10, 20, 30, 60, 120, and 180 minutes. Immunoblots performed on lysates of these cells demonstrated procaspase 3 cleavage at 60 to 120 minutes and PARP cleavage by 180 minutes (Figure 1A-B). Immunoblots also detected the cleaved form of caspase 7 by 180 minutes (not shown). Activation of caspases 1 (Figure 1C), 9, and 10 (not shown) was not detected at any time point. A cell-free fluorogenic assay confirmed activation of caspase 3 by 60 minutes and nearly maximal activity by 120 minutes, whereas no activation of caspase 1 was detected at any time point (Figure 1D). Flow cytometric analysis of lymphoblasts detected constitutive caspase 3 activation in FA-C cells, and a profound increase in this activation in response to IFN-γ and anti-fas antibody treatment (Figure 1E). In both cases, the number of cells containing active caspase 3 was increased from 2- to 4-fold in FA-C cells. A fluorogenic assay for caspase 6 using ac-VEID-MCA as a substrate detected enzyme activity at 180 minutes (not shown). Activation of caspases 2 and 4 was not detected (not shown).

We sought to rule out caspase 1 involvement in these cells by inhibiting caspase 1 with the inhibitor ac-YVAD-cho and quantifying caspase 3 activation. Such treatment prior to a 120-minute exposure to IFN-γ and agonistic anti-fas antibody did not inhibit cleavage of procaspase 3 in an immunoblot assay, whereas the addition of the caspase 3 inhibitor, ac-DEVD-cho, did (Figure2A, lanes 7 and 8). Caspase 1 activation was not affected by the presence of either the caspase 1 or 3 inhibitors. Fluorogenic substrate assays confirmed caspase 3 activation in response to IFN-γ and fas (Figure 2B) and this activation was blocked by ac-DEVD-cho but not ac-YVAD-cho.

Because we detected no caspase 1 activation and observed no effect of caspase 1 inhibitors on caspase 3 activation, we sought to confirm that the inhibitors were functional. The addition of ac-YVAD-cho (5-50 μmol/L) or YVAD-FMK (0.1 μmol/L) to a fluorogenic assay with N-ac-WEHD-MCA as substrate, in the presence of recombinant human caspase 1 (R&D Systems), reduced substrate cleavage to 16.3% or 9.3%, respectively, of caspase 1 alone. A negative control (Z-FA-FMK) had no effect (data not shown).

Apoptosis increased from 4% to 33% in mutant lymphoblasts (as analyzed by TUNEL assay) after exposure to IFN-γ for 24 hours followed by a 24-hour treatment with the agonistic anti-fasantibody (Figure 3). This apoptotic response is blunted by preexposure to an inhibitor of caspase 3 (25 μmol/L Z-DEVD-FMK). Although the increase in apoptosis in FANCC-corrected lymphoblasts is less pronounced, it is also blunted by preexposure to an inhibitor of caspase 3. The design of these studies was initially based on the well-known inductive effects of IFN-γ onfas expression and our previous findings that IFN-γ primed FA-C cells for subsequent responses to fas ligation. However, we later determined that simultaneous treatment with both agonistic anti-fas antibody and IFN-γ for 24 hours revealed the same patterns of apoptosis. Specifically, FANCC-deficient cells had constitutively higher levels of apoptosis and a greater increase in apoptosis as a result of the treatment (data not shown).

Optimal doses for the inhibitors to be used in colony assays were determined in experiments using a range of doses from 0.05 to 300 μmol/L. The dose-response curve shown (Figure4A) demonstrates that the 50-μmol/L concentration of the tetrapeptide aldehyde inhibitors was optimal in the following experiments where the caspase 3 (ac-DEVD-cho) inhibitor blunted the suppressive effect of IFN-γ, whereas the caspase 1 (ac-YVAD-cho) inhibitor did not. We have previously shown that FAC^−/− mice are sensitive to doses of IFN-γ too low to have an effect on FAC^−/+ mice.24 Because we have previously reported that Fas-ligand is expressed on CD34⁺, and also demonstrated that an anti-fasblocking antibody reduced IFN-γ-triggered inhibition of clonal growth in FA-C cells,24 we did not add the agonistic anti-fas antibody to the colony assays performed here.

In a series of 7 experiments we have found that IFN-γ-treated HPC from the bone marrow of these mice form more burst-forming units-erythroid (BFU-E) in the presence of a tetrapeptide aldehyde inhibitor of caspase 3 (ac-DEVD-cho) (Figure 4B), whereas clonal growth was not affected by the addition of an inhibitor of caspase 1 (ac-YVAD-cho; 50 μmol/L; Figure 4C). We emphasize that the IFN dose used for these studies (50 U/mL) was high enough to suppress clonal growth in both normal and FA progenitor cells.

On 2 separate occasions we obtained bone marrow cells from a child with FA-C. Parallel studies were performed using bone marrow cells obtained from normal volunteers. CD34⁺ cells from the FA-C patient were hypersensitive to the clonal inhibitory effects of IFN-γ on colony-forming units-granulocyte/macrophage (CFU-GM; Figure5A) and BFU-E (Figure 5B) when compared to those of the control. The low IFN dose, 0.1 ng/mL, was selected to maximize the differences between FA and normal progenitor cells. The caspase 3 inhibitor (ac-DEVD-cho) augmented colony formation in the IFN-γ-treated cells of the FA-C patient, but the caspase 1 inhibitor (ac-YVAD-cho) had no effect (Figure 5).

Seeking to clarify the role of caspase 8 in the activation of caspase 3, we first measured caspase 8 activation in cells of the FA-C EBV-transformed lymphoblast lines in response to a 120-minute exposure to IFN-γ and anti-fas antibody. This activation is shown in FA-C cells as well as isogenic cells corrected by introduction of FANCC cDNA (Figure6A). We then treated these cells with an inhibitor of caspase 8 (ac-IETD-FMK) before IFN-γ and anti-fas antibody exposure. The caspase 8 inhibitor prevented cleavage of caspase 3 as detected by fluorogenic (Figure 6B) and immunoblot assays (Figure 6C). No induction or inhibition of activation of caspase 1 was observed as a result of treatment with ac-IETD-FMK. The caspase 8 inhibitor also blocked caspase 8 activation in an immunoblot assay, whereas inhibitors of caspases 1 and 3 did not (data not shown). These findings indicate a role for caspase 8, either directly or through another molecular intermediate (not caspase 1) in caspase 3 activation. Experiments using the aldehyde inhibitor of caspase 8, ac-IETD-cho (BIOMOL), resulted in the inhibition of cleavage of caspase 3 as well (data not shown).

The FADD is required to link fas ligation with caspase 8 activation. Because this is a proximal event in the apoptotic activation pathway, we did not expect the caspase inhibitors to affect FADD protein levels, but needed to rule this out in the cell types we studied. FADD levels were not influenced by inhibitors in anti-fas antibody-stimulated IFN-γ-primed cells (data not shown).

Three types of caspase inhibitors were used in these experiments: tetrapeptide aldehyde inhibitors (cho), cell-permeable tetrapeptide aldehyde inhibitors (CP-cho), and fluoromethyl ketone inhibitors (FMK). Because of the reported overlap of activity for some of these inhibitors,38 particularly those for YVAD (caspases 1 and 4) and DEVD (caspase 3), we carefully determined concentrations that were high enough to inhibit apoptosis but low enough to reveal differential specificities. Cytosolic extracts were made from lymphoblasts that had been treated with IFN-γ and anti-fasantibody for 180 minutes after pretreatment with either a dose from 0.1 to 5 μmol/L of CP-cho inhibitors of caspases 1 and 3 (Figure7A), or from 0.1 to 50 μmol/L range of FMK inhibitors of caspases 1, 3, 1 and 4, or 8 (Figure 7B). Fluorogenic assays demonstrated that 1.0 μmol/L CP-DEVD-cho was sufficient to inhibit caspase 3 activation in these cells, whereas 5 μmol/L CP-YVAD-cho had no effect. YVAD-cho in control experiments did inhibit the activity of caspase 1 in fluorogenic assays (not shown). Similarly, a dose of 0.1 μmol/L Z-DEVD-FMK had the most suppressive effect of all FMK inhibitors tested on IFN/fas-induced activation of caspase 3. These results suggest the inhibition of IFN-γ- and anti-fas antibody-induced apoptosis by caspase 3 inhibitors is indeed selective and specific.

Hematopoietic progenitor cells from children with FA-C are apoptotic.56,57 Cells are hypersensitive to IFN-γ and TNF-α as well as hypersensitive to MCC (these factors induce apoptosis in Fanconi cells at doses too low to influence normal progenitors cells).24,57 Because the faspathway underlies at least some of the apoptotic responses in IFN-stimulated FA cells, and in view of the clearly defined pathway from fas ligation to caspase activation, we suspected thatfas-induced apoptosis in FA cells would involve a caspase pathway. To test this hypothesis, immunoblots were performed with 8 antibodies to caspases to determine which of the caspase subfamilies might be involved in effecting apoptosis in FA-C cells. Additionally, cell-free assays were performed with 7 fluorogenic caspase substrates, as well as 11 aldehyde and fluoromethyl ketone inhibitors of various caspases, to strengthen our observations. Recent work indicates that the caspase 1 family is involved largely in inflammatory and not apoptotic responses.31 32 Accordingly, we expected thatfas-induced caspase activation would involve caspase 8 and 3 but not 1.

Immunoblots (Figure 1) and cell-free fluorescence assays performed on lysates from the isogenic EBV-transformed cell lines (Figure 2) treated with IFN-γ and an agonistic anti-fas antibody, as well as flow cytometry on these cells, demonstrated a functional role for caspase 3 in the apoptosis of these cells, and also established that its activation is independent of caspase 1. IFN-γ inhibited clonal progenitor cell growth of marrow cells from mice nullizygous at theFANCC locus and this inhibitory effect was substantially blocked by ac-DEVD-cho, but not by ac-YVAD-cho, suggesting that caspase 3, but not caspase 1, was involved in the aberrant FA cell responses to these agents (Figure 4). Clonal growth of normal and FA-C CD34⁺ cells (Figure 5) under similar conditions likewise demonstrated a role for caspase 3 but not caspase 1. These results contrast somewhat with the report of Krantz and coworkers58that caspase 1 was activated by IFN-γ alone in erythroid cells. Because this group exposed highly purified late progenitor cells to IFN-γ and we used a mixture of CD34⁺ cells and exposed only early (BFU-E) progenitor cells, we suspect that the observed differences reflect the different cell types studied.

Caspases 6 and 7 (CPP32 family members) were also activated by treatment with IFN-γ and an agonistic anti-fas antibody. Because caspase 3 was shown to be activated by this treatment, it was not surprising that caspases 6 and 7 were also, because they are effector caspases of the caspase 3 subfamily that serve as potential modulators of the apoptotic response.36,37 Furthermore, when treated in this manner, a higher percentage of FA-C lymphoblasts undergo apoptosis than cells that have been FANCC corrected. This induced apoptotic response can be blunted by prior exposure of the cells to an inhibitor of caspase 3 (Figure 3). Additional experiments performed with these cells confirm that fas-induced caspase 3 activation does depend on caspase 8 activation (Figure 6). It is increasingly apparent that bone marrow failure so prevalent in children with FA derives from excessive apoptosis in progenitor cells depending, in part, on activation of the fas pathway.24 56We now demonstrate that the FA phenotype, at least in children with the C complementation group, involves the activation of caspases 8 and 3, but apparently not caspase 1. Although we did not detect higher protein levels of these activated caspases in FA-C cells undergoing apoptosis than in normal controls by immunoblot (Figure 1B), on a single cell basis there was a higher fractional level of constitutive caspase 3 activation detected by flow cytometry and a greater level of activation in response to IFN-γ and anti-fas antibody (Figure 1E). Flow cytometry also detected constitutive levels of caspase 3 activation that immunoblots and fluorogenic assays did not. For these reasons we believe that the flow cytometric method is more sensitive in detecting differential caspase 3 activation in FA-C and normal cells. In addition, the results clearly demonstrate that the antiapoptotic function of the FANCC protein must be positioned, at least in part, “upstream” of caspase 3 activation.

Considering the mutability of somatic cells of children with FA, we speculate that the apoptotic phenotype in hematopoietic progenitor cells and stem cells creates a perfect selective pressure for the emergence of mutant clones fully resistant to one or more mitogenic inhibitors. Such somatic mutants may have taken the first genetic steps in the conversion of a normal stem cell to one set on a course toward myelodysplasia and acute myeloblastic leukemia, clinical disorders for which children with FA are at substantial risk.59 60 We speculate that at least some of these mutations might result in the inactivation of proteins in signaling pathways that ordinarily control caspase activation.

We thank Dr Manuel Buchwald for providing cells and FANCC cDNA for our studies. The authors thank Dr Michael Heinrich and David and Lynn Frohnmayer for helpful discussions. We also thank Tara Koretsky for invaluable technical assistance and Markus Grompe for providing FA-C knockout mice for our use. Markus Grompe and Robb Moses provided valuable advice.