Portal de Periódicos da CAPES

Targeted Disruption of FANCC and FANCG in Human Cancer Provides a Preclinical Model for Specific Therapeutic Options

2006; Elsevier BV; Volume: 130; Issue: 7 Linguagem: Inglês

10.1053/j.gastro.2006.03.016

1528-0012

Eike Gallmeier, Eric S. Calhoun, Carlo Rago, Jonathan R. Brody, Steven C. Cunningham, Tomáš Hucl, Myriam Gorospe, Manu Kohli, Christoph Lengauer, Scott E. Kern,

DNA and Nucleic Acid Chemistry

Background & Aims: How specifically to treat pancreatic and other cancers harboring Fanconi anemia gene mutations has raised great interest recently, yet preclinical studies have been hampered by the lack of well-controlled human cancer models. Methods: We endogenously disrupted FANCC and FANCG in a human adenocarcinoma cell line and determined the impact of these genes on drug sensitivity, irradiation sensitivity, and genome maintenance. Results: FANCC and FANCG disruption abrogated FANCD2 monoubiquitination, confirming an impaired Fanconi anemia pathway function. On treatment with DNA interstrand–cross-linking agents, FANCC and FANCG disruption caused increased clastogenic damage, G2/M arrest, and decreased proliferation. The extent of hypersensitivity varied among agents, with ratios of inhibitory concentration 50% ranging from 2-fold for oxaliplatin to 14-fold for melphalan, a drug infrequently used in solid tumors. No hypersensitivity was observed on gemcitabine, etoposide, 3-aminobenzamide, NU1025, or hydrogen peroxide. FANCC and FANCG disruption also resulted in increased clastogenic damage on irradiation, but only FANCG disruption caused a subsequent decrease in relative survival. Finally, FANCC and FANCG disruption increased spontaneous chromosomal breakage, supporting the role of these genes in genome maintenance and likely explaining why they are mutated in sporadic cancer. Conclusions: Our human cancer cell model provides optimal controls to elucidate fundamental biologic features of individual Fanconi anemia gene defects and facilitates preclinical studies of therapeutic options. The impact of Fanconi gene defects on drug and irradiation sensitivity renders these genes promising targets for a specific, genotype-based therapy for individual cancer patients, providing a strong rationale for clinical trials. Background & Aims: How specifically to treat pancreatic and other cancers harboring Fanconi anemia gene mutations has raised great interest recently, yet preclinical studies have been hampered by the lack of well-controlled human cancer models. Methods: We endogenously disrupted FANCC and FANCG in a human adenocarcinoma cell line and determined the impact of these genes on drug sensitivity, irradiation sensitivity, and genome maintenance. Results: FANCC and FANCG disruption abrogated FANCD2 monoubiquitination, confirming an impaired Fanconi anemia pathway function. On treatment with DNA interstrand–cross-linking agents, FANCC and FANCG disruption caused increased clastogenic damage, G2/M arrest, and decreased proliferation. The extent of hypersensitivity varied among agents, with ratios of inhibitory concentration 50% ranging from 2-fold for oxaliplatin to 14-fold for melphalan, a drug infrequently used in solid tumors. No hypersensitivity was observed on gemcitabine, etoposide, 3-aminobenzamide, NU1025, or hydrogen peroxide. FANCC and FANCG disruption also resulted in increased clastogenic damage on irradiation, but only FANCG disruption caused a subsequent decrease in relative survival. Finally, FANCC and FANCG disruption increased spontaneous chromosomal breakage, supporting the role of these genes in genome maintenance and likely explaining why they are mutated in sporadic cancer. Conclusions: Our human cancer cell model provides optimal controls to elucidate fundamental biologic features of individual Fanconi anemia gene defects and facilitates preclinical studies of therapeutic options. The impact of Fanconi gene defects on drug and irradiation sensitivity renders these genes promising targets for a specific, genotype-based therapy for individual cancer patients, providing a strong rationale for clinical trials. The rare familial cancer susceptibility syndrome Fanconi anemia (FA) is caused by mutations in 1 of at least 12 FA genes.1Levitus M. Rooimans M.A. Steltenpool J. Cool N.F. Oostra A.B. Mathew C.G. Hoatlin M.E. Waisfisz Q. Arwert F. de Winter J.P. Joenje H. Heterogeneity in Fanconi anemia evidence for 2 new genetic subtypes.Blood. 2004; 103: 2498-2503Crossref PubMed Scopus (199) Google Scholar, 2Meetei A.R. Medhurst A.L. Ling C. Xue Y. Singh T.R. Bier P. Steltenpool J. Stone S. Dokal I. Mathew C.G. Hoatlin M. Joenje H. de Winter J.P. Wang W. A human ortholog of archaeal DNA repair protein Hef is defective in Fanconi anemia complementation group.Nat Genet. 2005; 37: 958-963Crossref PubMed Scopus (362) Google Scholar Biallelic mutations in some of these genes also occur in cancers of non-FA patients, implicating these genes in tumor suppression or genome maintenance among the general population. In pancreatic cancer, inactivating mutations have been identified in FANCD1/BRCA2, FANCC, and FANCG.3van der Heijden M.S. Yeo C.J. Hruban R.H. Kern S.E. Fanconi anemia gene mutations in young-onset pancreatic cancer.Cancer Res. 2003; 63: 2585-2588PubMed Google Scholar, 4van der Heijden M.S. Brody J.R. Gallmeier E. Cunningham S.C. Dezentje D.A. Shen D. Hruban R.H. Kern S.E. Functional defects in the Fanconi anemia pathway in pancreatic cancer cells.Am J Pathol. 2004; 165: 651-657Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar, 5Goggins M. Schutte M. Lu J. Moskaluk C.A. Weinstein C.L. Petersen G.M. Yeo C.J. Jackson C.E. Lynch H.T. Hruban R.H. Kern S.E. Germline BRCA2 gene mutations in patients with apparently sporadic pancreatic carcinomas.Cancer Res. 1996; 56: 5360-5364PubMed Google Scholar, 6Murphy K.M. Brune K.A. Griffin C. Sollenberger J.E. Petersen G.M. Bansal R. Hruban R.H. Kern S.E. Evaluation of candidate genes MAP2K4, MADH4, ACVR1B, and BRCA2 in familial pancreatic cancer deleterious BRCA2 mutations in 17%.Cancer Res. 2002; 62: 3789-3793PubMed Google Scholar, 7Couch F.J. Johnson M.R. Rabe K. Boardman L. McWilliams R. de Andrade M. Petersen G. Germ line Fanconi anemia complementation group C mutations and pancreatic cancer.Cancer Res. 2005; 65: 383-386PubMed Google Scholar These genes act in a common pathway, which appears to modulate DNA repair, especially the repair of DNA interstrand–cross-links (ICLs) and double-strand breaks.8D'Andrea A.D. Grompe M. The Fanconi anaemia/BRCA pathway.Nat Rev Cancer. 2003; 3: 23-34Crossref PubMed Scopus (684) Google Scholar Defects in this pathway could provide a therapeutic target in a subset of human tumors9Kennedy R.D. D'Andrea A.D. The Fanconi anemia/BRCA pathway new faces in the crowd.Genes Dev. 2005; 19: 2925-2940Crossref PubMed Scopus (354) Google Scholar (eg, by using ICL-forming agents or ionizing radiation [IR]), and should be explored specifically. Preclinical studies to evaluate specific therapeutic options thoroughly before clinical trials have been hampered by the lack of appropriate human cancer FA model systems. The cellular FA phenotype, comprising hypersensitivity to ICL-forming agents, IR, and oxidative stress, increased prevalence of chromosomal breaks, cell-cycle abnormalities, and defects in DNA repair,8D'Andrea A.D. Grompe M. The Fanconi anaemia/BRCA pathway.Nat Rev Cancer. 2003; 3: 23-34Crossref PubMed Scopus (684) Google Scholar has been explored mainly in nonmalignant or nonhuman cells, and, in several instances, remains controversial.10Saadatzadeh M.R. Bijangi-Vishehsaraei K. Hong P. Bergmann H. Haneline L.S. Oxidant hypersensitivity of Fanconi anemia type C-deficient cells is dependent on a redox-regulated apoptotic pathway.J Biol Chem. 2004; 279: 16805-16812Crossref PubMed Scopus (64) Google Scholar, 11Dallapiccola B. Porfirio B. Mokini V. Alimena G. Isacchi G. Gandini E. Effect of oxidants and antioxidants on chromosomal breakage in Fanconi anemia lymphocytes.Hum Genet. 1985; 69: 62-65Crossref PubMed Scopus (63) Google Scholar, 12Zdzienicka M.Z. Arwert F. Neuteboom I. Rooimans M. Simons J.W. The Chinese hamster V79 cell mutant V-H4 is phenotypically like Fanconi anemia cells.Somat Cell Mol Genet. 1990; 16: 575-581Crossref PubMed Scopus (31) Google Scholar, 13Notaro R. Montuori N. di Grazia C. Formisano S. Rotoli B. Selleri C. Fanconi's anemia cells are relatively resistant to H2O2-induced damage.Haematologica. 1998; 83: 868-874PubMed Google Scholar, 14Kalb R. Duerr M. Wagner M. Herterich S. Gross M. Digweed M. Joenje H. Hoehn H. Schindler D. Lack of sensitivity of primary Fanconi's anemia fibroblasts to UV and ionizing radiation.Radiat Res. 2004; 161: 318-325Crossref PubMed Scopus (42) Google Scholar, 15Tebbs R.S. Hinz J.M. Yamada N.A. Wilson J.B. Salazar E.P. Thomas C.B. Jones I.M. Jones N.J. Thompson L.H. New insights into the Fanconi anemia pathway from an isogenic FancG hamster CHO mutant.DNA Repair (Amst). 2005; 4: 11-22Crossref PubMed Scopus (58) Google Scholar We endogenously disrupted FANCC and FANCG in the human adenocarcinoma cell line RKO. The generation of naturally immortal, isogenic human cancer cells with specific FA gene defects, combined with the physiologic gene expression in the control cells, allowed the unambiguous evaluation of the impact of FA gene inactivation on human tumors, especially in regard to drug sensitivity, IR sensitivity, and genome maintenance. At least 2 distinct knockout (KO) clones were generated and characterized for each gene, minimizing the effects of clonal variability and allowing well-controlled comparisons between these genes. We disrupted the FANCC and FANCG genes according to the technique of Kohli et al16Kohli M. Rago C. Lengauer C. Kinzler K.W. Vogelstein B. Facile methods for generating human somatic cell gene knockouts using recombinant adeno-associated viruses.Nucleic Acids Res. 2004; 32: e3Crossref PubMed Scopus (143) Google Scholar (Figure 1A), deleting 1 exon of each gene to create a frameshift and a premature stop codon (Table 1). The targeting constructs, containing the selection cassette flanked on either side by loxP sites and approximately 1 kb human sequences adjoining the targeted exon, were ligated into pAAV (Stratagene, La Jolla, CA). pAAV, pRC, and pHelper were cotransfected into RKO cells using Lipofectamine (Invitrogen, Carlsbad, CA). Hygromycin-resistant clones were selected and clones having homologous integration events were identified by polymerase chain reaction (PCR). After Cre-mediated recombination, clones having loss of the selection cassette were used for second allele targeting. Hygromycin expression was removed in the final clones by Cre-mediated excision of the selection cassette.Table 1Design of the Targeting ConstructsDeleted exonDeleted sequencePredicted effectLHA forward 5′ → 3′/ LHA reverse 3′ → 5′RHA forward 5′ → 3′/ RHA reverse 3′ → 5′FANCCExon 10 (100 bp)468 bpPremature stop codon 106 bp downstream exon 9/11 junctionGTTTCTGATTCCCTCCTGTTC/ CTTCGTCTCCACTTGTGAAAGCGCTGTTGAGAGTATTTTGGAC/ CTCAGGTTTTAGCCGATATTTCFANCGExon 8 (152 bp)281 bpPremature stop codon 51 bp downstream exon 7/9 junctionTGAGGCCTGTGTCCTCAAG/ GGAAGTATGGCTCCCAAAGAGAGGCAGTGGCAGGACATAG/ CCTGAGTTGGTGAGCTCCAT Open table in a new tab Genomic DNA and total RNA were extracted using the DNA Blood Mini kit (Qiagen, Valencia, CA) and RNeasy (Qiagen), respectively. Complementary DNA was made from RNA using Superscript (Invitrogen). PCR products were resolved on agarose gels using sodium boric acid electrophoresis (Faster Better Media LLC, Hunt Valley, MD).17Brody J.R. Calhoun E.S. Gallmeier E. Creavalle T.D. Kern S.E. Ultra-fast high-resolution agarose electrophoresis of DNA and RNA using low-molarity conductive media.Biotechniques. 2004; 37 (600, 602): 598PubMed Google Scholar All FA antibodies kindly were provided by Maureen Hoatlin, except for the FANCD2 antibody (Santa Cruz, Santa Cruz, CA). Proteins were separated on 3%–8% Tris-acetate gels for 165 minutes at 150 V (FANCD2) or on 10% Bis-Tris gels for 120 minutes at 150 V (FANCC, FANCG). The primary antibodies were used at 1:500, the corresponding secondary antibodies (Pierce, Rockford, IL) at 1:10,000. Chemiluminescent detection was performed using the SuperSignal West Pico kit (Pierce). Lysates from 20 million cells were precleared with Protein A/G agarose beads (Santa Cruz). After centrifugation, the supernatant was incubated overnight with a polyclonal rabbit anti-human FANCC antibody (#7234) at a 1:200 dilution. After incubation with Protein A/G beads, the immune complexes were isolated by centrifugation, resuspended in Laemmli buffer and heated at 95°C. After sodium dodecyl sulfate–polyacrylamide gel electrophoresis and Western blotting, FANCC protein was detected using a monoclonal mouse anti-human FANCC antibody (clone 8F3). A total of 1300–1800 cells were plated, allowed to adhere, and exposed to various drugs (all from Sigma [St Louis, MO], except SJG-136 from the Drug Synthesis and Chemistry Branch, National Cancer Institute, Bethesda, MD). After 6 days, the cells were washed, lysed in 100 μL H2O, and 0.5% Picogreen (Molecular Probes, Eugene, OR) was added. Fluorescence was measured and the proliferation index was calculated, defining the untreated samples as 1. At least 4 independent experiments were performed per drug, with each data point reflecting triplicate wells. Cells were exposed to mitomycin C (MMC) for 48 hours, fixed in phosphate-buffered saline/3.7% formaldehyde/0.5% Nonidet P40 (US Biochemical, Cleveland, OH), stained with bisbenzimide (Hoechst 33258, Sigma), and analyzed by flow cytometry. A total of 10,000 events were acquired per sample. The data were processed using Cell Quest software (Becton Dickinson, Franklin Lakes, NJ). Cells were plated, allowed to adhere, and exposed to 137Cs gamma rays (dose rate, 1.3 Gy/min, Gammacell 40 Exactor; MDS Nordion, Inc., Ontario, Canada). Cells subsequently were incubated for 14 days, fixed, and stained (10% neutral buffered formalin, 1:500 crystal violet). All macroscopically visible colonies were counted. Two to 7 experiments were performed per sample. Each sample consisted of 2 different cell concentrations, each performed in duplicate. The amount of cells plated was 20 and 50 cells at 0 Gy, 50 and 100 at 2 Gy, 100 and 200 at 4 Gy, 1000 and 2000 at 6 Gy, 10,000 and 20,000 at 8 Gy, and 100,000 and 200,000 at 10 Gy. Chromosome breakage analysis and karyotyping were performed by the cytogenetics facility at Johns Hopkins according to standard protocols. For spontaneous chromosome breakage analysis, untreated cells were harvested at 3–5 time points during cell culture. For IR experiments, the cells were treated with 2 Gy for 30 minutes. For diepoxybutane (DEB) testing,18Auerbach A.D. Diagnosis of Fanconi anemia by diepoxybutane analysis.in: Boyle A. Current protocols in human genetics. John Wiley & Sons, Inc, Hoboken, NJ2003: 8.7.1-8.7.15Google Scholar the cells were treated with DEB at 3–1.0 μg/mL for 48 hours. For chromosome breakage analysis, 50 or 500 metaphases of each sample were analyzed in a blinded fashion for breakage and other structural abnormalities (ie, rings or radials). For karyotyping, 20 cells were analyzed per sample. Centromeric probes for chromosomes 7, 17, and 18 were used to determine the frequency of variance from the modal chromosome number and the frequency of tetraploid cells, as described.19Lengauer C. Kinzler K.W. Vogelstein B. Genetic instability in colorectal cancers.Nature. 1997; 386: 623-627Crossref PubMed Scopus (1672) Google Scholar Two hundred interphase nuclei were evaluated for each cell population and chromosome. We disrupted FANCG in RKO cells by serially deleting the 2 alleles of exon 8 (Figure 1A). Two FANCG+/− and 3 FANCG−/− clones were obtained. RKO cells contain an unbalanced translocation (der20)t(9;20)(q22;p13), resulting in segmental trisomy at the FANCC locus. Consequently, we disrupted FANCC by serially deleting 1, 2, or 3 alleles of exon 10. Five FANCC+/+/−, 6 FANCC+/−/− and 2 FANCC−/−/− clones were obtained. The control heterozygote FANCC and FANCG clones were derived from cells that had nonhomologously integrated the targeting construct in the second or third targeting round, respectively, and thus underwent the same number of selection steps as did the FANCC−/−/− and FANCG−/− clones. We performed PCR, reverse-transcription PCR, and Northern and Western blotting to confirm FANCC and FANCG gene disruption. By using PCR, we detected a band the size of the gene wild-type sequence in parental and heterozygote cells, and a smaller band corresponding to the deleted Cre-recombinant sequence, in heterozygote and homozygote KO cells (Figure 1B). By using reverse-transcription PCR, we detected truncated gene transcripts in heterozygote and homozygote KO cells (Figure 1C). By using Northern blotting, we did not detect stable messenger RNA of the respective gene in the homozygote KO clones (data not shown). By using Western blotting, we detected FANCG or FANCC protein, respectively, in parental and heterozygote, but not in homozygote KO cells (Figure 1D). Because of its low endogenous expression levels, FANCC required enrichment by immunoprecipitation before detection. We confirmed an impaired FA pathway function in the FANCC−/−/− and FANCG−/− clones by assessing FANCD2 monoubiquitination. Whereas normal cells express a monoubiquitinated form of FANCD2, which becomes pronounced on DNA damage, FA cells with defects in the proximal FA pathway, including FANCC and FANCG, lack FANCD2 monoubiquitination.20Shimamura A. de Oca R.M. Svenson J.L. Haining N. Moreau L.A. Nathan D.G. D'Andrea A.D. A novel diagnostic screen for defects in the Fanconi anemia pathway.Blood. 2002; 100: 4649-4654Crossref PubMed Scopus (94) Google Scholar Consistently, the monoubiquitinated form of FANCD2 was detected in the control cells, especially after treatment with MMC, but was undetectable in FANCC−/−/− and FANCG−/− clones (Figure 1E). We further confirmed an impaired FA pathway function using the DEB test, which serves as a diagnostic assay for the FA syndrome.18Auerbach A.D. Diagnosis of Fanconi anemia by diepoxybutane analysis.in: Boyle A. Current protocols in human genetics. John Wiley & Sons, Inc, Hoboken, NJ2003: 8.7.1-8.7.15Google Scholar, 21Auerbach A.D. Fanconi anemia diagnosis and the diepoxybutane (DEB) test.Exp Hematol. 1993; 21: 731-733PubMed Google Scholar We found a significant increase in chromosomal aberrations, including breaks, radials, and rings, 48 hours after treatment with DEB at 0.3 μg/mL, in FANCC−/−/− and FANCG−/− clones. The control cells had baseline levels of chromosomal breaks but no radials or rings (Figure 1F). Similar results were obtained 72 hours after DEB treatment. Higher DEB concentrations (1.0 μg/mL) caused an increased frequency of breaks and radials in the FANCC−/−/− and FANCG−/−, but not in the control cells (data not shown). By using flow cytometry, we determined the fractions of cells at the different stages of the cell cycle 48 hours after treatment with MMC (Figure 2). A pronounced G2/M arrest (>40% of the cells) was observed in the FANCC−/−/− and FANCG−/− clones at 4-fold lower MMC concentrations than in the control cells (25 vs 100 nmol/L). The amount of cells arresting in G2/M was dose-dependent and increased with incremental doses of MMC. At very high concentrations (400 nmol/L), MMC uniformly caused a near-complete G2/M arrest in all cell lines, independent of FANCC or FANCG genotype. In contrast, the quotient of the G1/S ratio of the treated samples divided by the G1/S ratio of the untreated samples did not exceed 2.4 at any dose, indicating only a modest (or no) G1 arrest. We surveyed a broad panel of ICL-forming drugs for their effects on proliferation among our clones, including the natural product MMC, platinum drugs (cisplatin, carboplatin, oxaliplatin), nitrogen mustards with either an aromatic substituent (melphalan) or a phosphoramide group (cyclophosphamide), a chloroethylnitrosurea (carmustin), and SJG-136, a rationally designed, sequence-specific ICL-forming agent in preclinical development.22Gregson S.J. Howard P.W. Hartley J.A. Brooks N.A. Adams L.J. Jenkins T.C. Kelland L.R. Thurston D.E. Design, synthesis, and evaluation of a novel pyrrolobenzodiazepine DNA-interactive agent with highly efficient cross-linking ability and potent cytotoxicity.J Med Chem. 2001; 44: 737-748Crossref PubMed Scopus (186) Google Scholar FANCC−/−/− and FANCG−/− clones were more sensitive than control cells to all of these agents. However, one group of drugs caused prominent proliferation differences between the FA-proficient and FA-deficient clones over a broad concentration range (from 16-fold to >64-fold), with inhibitory concentration 50% (IC50) ratios between 7 and 14 (melphalan > MMC > carboplatin > cisplatin > SJG-136). A second group caused only slight to intermediate differences with IC50 ratios between 2 and 3.5 (carmustin > cyclophosphamide > oxaliplatin) (Figure 3A). In contrast to the ICL-forming agents, no proliferation differences were found between FA-proficient and FA-deficient clones on treatment with gemcitabine, H2O2, the poly(ADP-ribose) polymerase (PARP) inhibitors 3-aminobenzamide and NU1025, or etoposide (Figure 3B and data not shown). We assessed the impact of FANCC and FANCG disruption on the frequency of chromosomal aberrations 30 minutes after treatment with IR at 2 Gy. FANCC−/−/− and FANCG−/− cells developed more chromosomal aberrations than did control cells. Accordingly, the fraction of cells lacking chromosomal aberrations was decreased (Figure 4A). We then assessed the impact of FANCC and FANCG disruption on survival after treatment with IR. FANCG−/− but not FANCC−/−/− clones had a decreased relative survival as compared with the control cells. This effect was pronounced especially at higher doses: at 6–10 Gy the relative survival in the FANCG−/− clones was decreased by approximately 10-fold as compared with control or FANCC−/−/− clones (Figure 4B). We analyzed chromosome breakage to quantify spontaneous chromosomal aberrations. The cells were harvested at several time points during cell culture and 500 metaphases analyzed per sample in each experiment (Table 2). We found significantly more chromosomal breaks in the FANCC−/−/− and FANCG−/− clones than in the parental or heterozygote cells (both comparisons, P < .001, χ2 test). We next performed karyotyping to detect heritable chromosomal changes. Before analysis, the cells were passed for approximately 200 cell doublings after the generation of the gene defect to allow heritable chromosomal changes to occur and subclones to develop. Sidelines with newly acquired structural and numeric clonal abnormalities were observed in 2 of 3 FANCG−/− clones, as compared with the karyotype of the parental cells, but not in 2 FANCC−/−/− or 2 heterozygote clones. One FANCG−/− line harbored a new translocation t(1;10)(q25;q22.3) in 5 of 20 cells. A second FANCG−/− line had an addition of unknown material on the short arm of chromosome 15 [add(15)(p11.2)] in 6 of 20 cells and loss of chromosome 9 in 3 of 20 cells. Random, single-cell, nonclonal abnormalities were observed in all lines. Finally, we performed fluorescence in situ hybridization analysis to assess chromosomal copy number changes. Before analysis, the cells were passed for approximately 100 cell doublings after the generation of the gene defect. The frequency of cells having chromosomal copy number changes ranged from 2.0% to 6.5%, without significant differences among the clones (Table 3). These numbers reflect methodologic background levels previously reported for RKO cells (≤10%).19Lengauer C. Kinzler K.W. Vogelstein B. Genetic instability in colorectal cancers.Nature. 1997; 386: 623-627Crossref PubMed Scopus (1672) Google Scholar Likewise, the fraction of tetraploid cells ranged from 1.9% to 3.8% and did not significantly differ among the clones.Table 2Spontaneous Chromosomal Breaks by Cytogenetic AnalysisCellsExperiment 1aThe cells were harvested 3–5 times during cell culture.Experiment 2Experiment 3Experiment 4Experiment 5Experiment 6Total breaks/cellsParentalWild type3 (3)b500 metaphases were analyzed per sample: the first value indicates the total number of breaks in 500 cells, the second value (in parenthesis) indicates the total number of cells with breaks in 500 cells.10 (10)—2 (2)6 (6)2 (2)23/2500 (0.9%)HeterozygoteFANCC+/−/−——4 (4)4 (4)4 (3)—12/1500 (0.8%)FANCG+/−2 (2)15 (15)——8 (8)—25/1500 (1.7%)Homozygote KOFANCC−/−/−——25 (25)11 (9)13 (13)—49/1500 (3.3%)FANCG−/−10 (10)36 (35)——30 (29)—76/1500 (5.1%)a The cells were harvested 3–5 times during cell culture.b 500 metaphases were analyzed per sample: the first value indicates the total number of breaks in 500 cells, the second value (in parenthesis) indicates the total number of cells with breaks in 500 cells. Open table in a new tab Table 3Spontaneous Chromosomal Copy Number Changes by Fluorescence In Situ Hybridization AnalysisCellsPercentage of cells having tetraploidyPercentage of cells having copy number changesMeanaTetraploid cells were excluded from the totals of changes in copy number.Chromosome 7Chromosome 17Chromosome 18ParentalbTwo hundred interphase nuclei were analyzed per sample.Wild type3.92.53.06.54.0HeterozygoteFANCC+/−/−2.53.55.03.03.8FANCG+/−2.92.04.05.53.8Homozygote KOFANCC−/−/−1.82.56.54.54.5FANCG−/−3.52.53.53.53.2a Tetraploid cells were excluded from the totals of changes in copy number.b Two hundred interphase nuclei were analyzed per sample. Open table in a new tab A variety of phenotypes have been described in human nonmalignant FA cells or in nonhuman (ie, mouse, chicken, hamster) FA KO cells. The unequivocal evaluation of whether cancer cells show a similar phenotype as do nonmalignant FA cells, especially in regard to drug and IR sensitivity, is a prerequisite before clinical trials. Therefore, we disrupted the endogenous gene loci of FANCC and FANCG in a human adenocarcinoma cell line. The acute disruption of FANCC or FANCG in RKO yielded viable clones. In contrast, no viable clones were obtained on sequential gene targeting of both alleles of the downstream FA genes FANCD2 and BRCA2 using the same cell line (E.G., T.H., S.E.K., unpublished observations). FANCC−/−/− and FANCG−/− clones immediately displayed the characteristic FA phenotype (ie, defective FANCD2 monoubiquitination and hypersensitivity to ICL-forming agents). On treatment with the ICL-forming agent DEB, FANCC−/−/− and FANCG−/− clones developed more chromosomal aberrations than control cells, including the characteristic formation of rings and radials, a phenotype observed so stringently in nonmalignant FA cells that this test serves as a diagnostic assay for the FA syndrome.18Auerbach A.D. Diagnosis of Fanconi anemia by diepoxybutane analysis.in: Boyle A. Current protocols in human genetics. John Wiley & Sons, Inc, Hoboken, NJ2003: 8.7.1-8.7.15Google Scholar, 21Auerbach A.D. Fanconi anemia diagnosis and the diepoxybutane (DEB) test.Exp Hematol. 1993; 21: 731-733PubMed Google Scholar Furthermore, FANCC−/−/− and FANCG−/− clones arrested in G2/M on lower doses of the ICL-forming agent MMC than did control cells, again resembling nonmalignant FA cells.23Kaiser T.N. Lojewski A. Dougherty C. Juergens L. Sahar E. Latt S.A. Flow cytometric characterization of the response of Fanconi's anemia cells to mitomycin C treatment.Cytometry. 1982; 2: 291-297Crossref PubMed Scopus (71) Google Scholar Consistently, proliferation on MMC treatment was decreased in the FANCC−/−/− and FANCG−/− clones as compared with control cells. To explore potential treatment options for patients having FA-deficient tumors, we directly compared a broad panel of ICL-forming drugs, including all major classes of these agents. We found that not all drugs performed similarly in regard to the in vitro therapeutic window defined by differing FANCC and FANCG gene status. One group of drugs that caused pronounced proliferation differences between FA-proficient and FA-deficient clones (melphalan > MMC > carboplatin > cisplatin > SJG-136) was contrasted by a second group of drugs causing only slight to intermediate differences (carmustin > cyclophosphamide > oxaliplatin). These differences likely reflect the contribution of ICLs to the overall cytotoxicity of the respective drug. Notably, FANCC and FANCG gene status had a different impact on treatment with the structurally and functionally similar platinum-based drugs. In contrast to cisplatin and especially carboplatin, oxaliplatin caused only slight proliferation differences between FA-proficient and FA-deficient cells. This is in agreement with studies by Woynarowski et al,24Woynarowski J.M. Faivre S. Herzig M.C. Arnett B. Chapman W.G. Trevino A.V. Raymond E. Chaney S.G. Vaisman A. Varchenko M. Juniewicz P.E. Oxaliplatin-induced damage of cellular DNA.Mol Pharmacol. 2000; 58: 920-927Crossref PubMed Scopus (268) Google Scholar, 25Woynarowski J.M. Chapman W.G. Napier C. Herzig M.C. Juniewicz P. Sequence- and region-specificity of oxaliplatin adducts in naked and cellular DNA.Mol Pharmacol. 1998; 54: 770-777Crossref PubMed Scopus (205) Google Scholar showing that oxaliplatin, despite higher equimolar cytotoxicity, forms fewer ICLs than cisplatin. Given the preferable side-effect profile of carboplatin as compared with cisplatin26Wang D. Lippard S.J. Cellular processing of platinum anticancer drugs.Nat Rev Drug Discov. 2005; 4: 307-320Crossref PubMed Scopus (3168) Google Scholar along with its stronger effect on FANCC−/−/− and FANCG−/− clones, carboplatin might be a drug of choice among the platinum compounds to exploit FA pathway deficiency in human tumors (together with MMC and melphalan). The high concentrations of cyclophosphamide necessary to observe an effect likely are explained by the inactive prodrug, which requires activation by cytochrome P450 enzymes,27Fleming R.A. An overview of cyclophosphamide and ifosfamide pharmacology.Pharmacotherapy. 1997; 17: 146S-154SPubMed Google Scholar which in turn are expressed only at low levels in cancer cells.28McFadyen M.C. Melvin W.T. Murray G.I. Cytochrome P450 enzymes novel options for cancer therapeutics.Mol Cancer Ther. 2004; 3: 363-371PubMed Google Scholar Taken together, our data exemplify the usefulness of this cancer cell model in the preclinical assessment of candidate drugs for FA pathway-deficient tumors. Furthermore, our model should be applicable to high-throughput drug screening of compound libraries in an effort to identify agents to which FA-deficient cells may be hypersensitive over an even broader range of concentrations. In striking contrast to ICL-forming agents, we observed no significant proliferation differences among our clones on treatment with several other drugs. First, we were not able to confirm our previously reported increased resistance of FA-deficient cells to gemcitabine, which were obtained in a less physiologically relevant gene-overexpression model.29van der Heijden M.S. Brody J.R. Dezentje D.A. Gallmeier E. Cunningham S.C. Swartz M.J. DeMarzo A.M. Offerhaus G.J. Isacoff W.H. Hruban R.H. Kern S.E. In vivo therapeutic responses contingent upon Fanconi/BRCA2 tumor status.Clin Cancer Res. 2005; 11: 7508-7515Crossref PubMed Scopus (140) Google Scholar Second, we did not find decreased proliferation in our model on treatment with the PARP inhibitor NU1025, an agent that has been reported to specifically kill cells with defects in the downstream FA gene BRCA2.30Bryant H.E. Schultz N. Thomas H.D. Parker K.M. Flower D. Lopez E. Kyle S. Meuth M. Curtin N.J. Helleday T. Specific killing of BRCA2-deficient tumours with inhibitors of poly(ADP-ribose) polymerase.Nature. 2005; 434: 913-917Crossref PubMed Scopus (3905) Google Scholar Third, our data do not support an impact of FANCC or FANCG on H2O2 sensitivity in human cancer cells, an issue discussed controversially in previous studies.10Saadatzadeh M.R. Bijangi-Vishehsaraei K. Hong P. Bergmann H. Haneline L.S. Oxidant hypersensitivity of Fanconi anemia type C-deficient cells is dependent on a redox-regulated apoptotic pathway.J Biol Chem. 2004; 279: 16805-16812Crossref PubMed Scopus (64) Google Scholar, 11Dallapiccola B. Porfirio B. Mokini V. Alimena G. Isacchi G. Gandini E. Effect of oxidants and antioxidants on chromosomal breakage in Fanconi anemia lymphocytes.Hum Genet. 1985; 69: 62-65Crossref PubMed Scopus (63) Google Scholar, 12Zdzienicka M.Z. Arwert F. Neuteboom I. Rooimans M. Simons J.W. The Chinese hamster V79 cell mutant V-H4 is phenotypically like Fanconi anemia cells.Somat Cell Mol Genet. 1990; 16: 575-581Crossref PubMed Scopus (31) Google Scholar, 13Notaro R. Montuori N. di Grazia C. Formisano S. Rotoli B. Selleri C. Fanconi's anemia cells are relatively resistant to H2O2-induced damage.Haematologica. 1998; 83: 868-874PubMed Google Scholar The mismatch-repair–deficient background of our cells, however, might have influenced our results because mismatch-repair proteins may contribute to the repair of oxidative damage.31Slupphaug G. Kavli B. Krokan H.E. The interacting pathways for prevention and repair of oxidative DNA damage.Mutat Res. 2003; 531: 231-251Crossref PubMed Scopus (432) Google Scholar Fourth, we did not find proliferation differences on treatment with the DNA double-strand break–inducing drug etoposide. Because FA cells appear to be fully proficient in the initial processing of ICLs to DNA double-strand breaks,32Rothfuss A. Grompe M. Repair kinetics of genomic interstrand DNA cross-links evidence for DNA double-strand break-dependent activation of the Fanconi anemia/BRCA pathway.Mol Cell Biol. 2004; 24: 123-134Crossref PubMed Scopus (200) Google Scholar our findings that the DNA double-strand break–inducing agents H2O2 and etoposide do not cause similar proliferation discrepancies between FA-deficient and FA-proficient cells as do ICL-forming agents, further support the idea that DNA double-strand breaks induced by ICL-forming agents are distinguishable from those induced by other agents33Scharer O.D. DNA interstrand crosslinks natural and drug-induced DNA adducts that induce unique cellular responses.Chembiochem. 2005; 6: 27-32Crossref PubMed Scopus (155) Google Scholar and might be repaired differently. The sensitivity of FA cells to IR remains controversial.14Kalb R. Duerr M. Wagner M. Herterich S. Gross M. Digweed M. Joenje H. Hoehn H. Schindler D. Lack of sensitivity of primary Fanconi's anemia fibroblasts to UV and ionizing radiation.Radiat Res. 2004; 161: 318-325Crossref PubMed Scopus (42) Google Scholar In our model, we found more chromosomal aberrations in FANCC−/−/− and FANCG−/− clones on treatment with IR than in control cells, suggesting that the former may experience higher initial DNA damage and/or impaired DNA repair. Unexpectedly, the relative survival on IR was decreased significantly only in the FANCG−/−, but not in FANCC−/−/− clones. Thus, FANCC−/−/− and FANCG−/− cells differ in their repair (or survival) capabilities on treatment with IR only during the later stages of DNA repair. Finally, we used our model to study whether disruption of FANCC or FANCG would promote gross chromosomal instability (the chromosomal instability [CIN] phenotype) in a karyotypically stable, microsatellite-instable cancer cell line (the microsatellite instability [MSI] phenotype), as has been reported for the CDC4 gene.34Rajagopalan H. Jallepalli P.V. Rago C. Velculescu V.E. Kinzler K.W. Vogelstein B. Lengauer C. Inactivation of hCDC4 can cause chromosomal instability.Nature. 2004; 428: 77-81Crossref PubMed Scopus (484) Google Scholar Nonneoplastic cells from FA patients often show increased spontaneous chromosomal breakage, although the numbers are highly variable even in the same individual,21Auerbach A.D. Fanconi anemia diagnosis and the diepoxybutane (DEB) test.Exp Hematol. 1993; 21: 731-733PubMed Google Scholar, 35Schroeder T.M. Tilgen D. Kruger J. Vogel F. Formal genetics of Fanconi's anemia.Hum Genet. 1976; 32: 257-288Crossref PubMed Scopus (154) Google Scholar indicating that the FA genes are involved in maintaining chromosomal integrity. Our findings of increased spontaneously occurring chromosomal breaks in FANCC−/−/− and FANCG−/− cancer cells might explain why these genes are mutated in sporadic cancer and would classify FANCC and FANCG as genome-maintenance genes. On the other hand, the significance of the observed FANCG−/− subclones harboring newly acquired karyotypic changes remains elusive because the long-term rate of karyotypic changes in the parental RKO cell line is not known and we did not find an increase in chromosomal copy number changes in FANCC−/−/− and FANCG−/− clones. Taken together, FANCC and FANCG mutations likely contribute to chromosomal instability in cancer but are insufficient to cause the full CIN phenotype. In conclusion, our syngenetic human cancer model provides optimal controls to elucidate fundamental biologic features, either shared or gene-specific, of individual FA gene defects and facilitates preclinical studies of therapeutic options. The data obtained support a role of FANCC and FANCG in genome maintenance and highlight the impact of these genes on drug and IR sensitivity. We provide proof of principle that the mere disruption of a single FA gene acutely sensitizes human cancer cells to common chemotherapy and IR, which renders these genes promising targets for a specific genotype-based therapy for individual cancer patients. Therefore, our study strongly should encourage the initiation of FA genotype–specific clinical trials. The authors thank L. A. Morsberger for the cytogenetic analyses, B. Vogelstein for the selection cassette, M. Hoatlin for the FA antibodies, and C. Griffin and T. L. deWeese for helpful discussions.