Functional relationships of FANCC to homologous recombination, translesion synthesis, and BLM
2004; Springer Nature; Volume: 24; Issue: 2 Linguagem: Inglês
10.1038/sj.emboj.7600534
ISSN1460-2075
AutoresSeiki Hirano, Kazuhiko Yamamoto, Masamichi Ishiai, Mitsuyoshi Yamazoe, Masayuki Seki, Nobuko Matsushita, Mioko Ohzeki, Yukiko Yamashita, Hiroshi Arakawa, Jean-Marie Buerstedde, Takemi Enomoto, Shunichi Takeda, Larry H. Thompson, Minoru Takata,
Tópico(s)PARP inhibition in cancer therapy
ResumoArticle23 December 2004free access Functional relationships of FANCC to homologous recombination, translesion synthesis, and BLM Seiki Hirano Seiki Hirano Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Kazuhiko Yamamoto Kazuhiko Yamamoto Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, JapanPresent address: Division of Hematology/Oncology, Mount Sinai School of Medicine, New York, NY 10029, USA Search for more papers by this author Masamichi Ishiai Masamichi Ishiai Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Mitsuyoshi Yamazoe Mitsuyoshi Yamazoe Department of Radiation Genetics, Graduate School of Medicine, Kyoto University, Yoshida-konoe, Sakyo-ku, Kyoto, Japan Search for more papers by this author Masayuki Seki Masayuki Seki Molecular Cell Biology Laboratory, Graduate School for Pharmaceutical Sciences, Tohoku University, Aoba, Aoba-ku, Sendai, Miyagi, Japan Search for more papers by this author Nobuko Matsushita Nobuko Matsushita Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Mioko Ohzeki Mioko Ohzeki Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Yukiko M Yamashita Yukiko M Yamashita Department of Radiation Genetics, Graduate School of Medicine, Kyoto University, Yoshida-konoe, Sakyo-ku, Kyoto, Japan Search for more papers by this author Hiroshi Arakawa Hiroshi Arakawa GSF, Institute for Molecular Radiobiology, Neuherberg Munich, Germany Search for more papers by this author Jean-Marie Buerstedde Jean-Marie Buerstedde GSF, Institute for Molecular Radiobiology, Neuherberg Munich, Germany Search for more papers by this author Takemi Enomoto Takemi Enomoto Molecular Cell Biology Laboratory, Graduate School for Pharmaceutical Sciences, Tohoku University, Aoba, Aoba-ku, Sendai, Miyagi, Japan Search for more papers by this author Shunichi Takeda Shunichi Takeda Department of Radiation Genetics, Graduate School of Medicine, Kyoto University, Yoshida-konoe, Sakyo-ku, Kyoto, Japan Search for more papers by this author Larry H Thompson Larry H Thompson Biology and Biotechnology Research Program, Lawrence Livermore National Laboratory, Livermore, CA, USA Search for more papers by this author Minoru Takata Corresponding Author Minoru Takata Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Seiki Hirano Seiki Hirano Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Kazuhiko Yamamoto Kazuhiko Yamamoto Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, JapanPresent address: Division of Hematology/Oncology, Mount Sinai School of Medicine, New York, NY 10029, USA Search for more papers by this author Masamichi Ishiai Masamichi Ishiai Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Mitsuyoshi Yamazoe Mitsuyoshi Yamazoe Department of Radiation Genetics, Graduate School of Medicine, Kyoto University, Yoshida-konoe, Sakyo-ku, Kyoto, Japan Search for more papers by this author Masayuki Seki Masayuki Seki Molecular Cell Biology Laboratory, Graduate School for Pharmaceutical Sciences, Tohoku University, Aoba, Aoba-ku, Sendai, Miyagi, Japan Search for more papers by this author Nobuko Matsushita Nobuko Matsushita Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Mioko Ohzeki Mioko Ohzeki Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Yukiko M Yamashita Yukiko M Yamashita Department of Radiation Genetics, Graduate School of Medicine, Kyoto University, Yoshida-konoe, Sakyo-ku, Kyoto, Japan Search for more papers by this author Hiroshi Arakawa Hiroshi Arakawa GSF, Institute for Molecular Radiobiology, Neuherberg Munich, Germany Search for more papers by this author Jean-Marie Buerstedde Jean-Marie Buerstedde GSF, Institute for Molecular Radiobiology, Neuherberg Munich, Germany Search for more papers by this author Takemi Enomoto Takemi Enomoto Molecular Cell Biology Laboratory, Graduate School for Pharmaceutical Sciences, Tohoku University, Aoba, Aoba-ku, Sendai, Miyagi, Japan Search for more papers by this author Shunichi Takeda Shunichi Takeda Department of Radiation Genetics, Graduate School of Medicine, Kyoto University, Yoshida-konoe, Sakyo-ku, Kyoto, Japan Search for more papers by this author Larry H Thompson Larry H Thompson Biology and Biotechnology Research Program, Lawrence Livermore National Laboratory, Livermore, CA, USA Search for more papers by this author Minoru Takata Corresponding Author Minoru Takata Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan Search for more papers by this author Author Information Seiki Hirano1, Kazuhiko Yamamoto1, Masamichi Ishiai1, Mitsuyoshi Yamazoe2, Masayuki Seki3, Nobuko Matsushita1, Mioko Ohzeki1, Yukiko M Yamashita2, Hiroshi Arakawa4, Jean-Marie Buerstedde4, Takemi Enomoto3, Shunichi Takeda2, Larry H Thompson5 and Minoru Takata 1 1Department of Immunology and Molecular Genetics, Kawasaki Medical School, Kurashiki, Okayama, Japan 2Department of Radiation Genetics, Graduate School of Medicine, Kyoto University, Yoshida-konoe, Sakyo-ku, Kyoto, Japan 3Molecular Cell Biology Laboratory, Graduate School for Pharmaceutical Sciences, Tohoku University, Aoba, Aoba-ku, Sendai, Miyagi, Japan 4GSF, Institute for Molecular Radiobiology, Neuherberg Munich, Germany 5Biology and Biotechnology Research Program, Lawrence Livermore National Laboratory, Livermore, CA, USA *Corresponding author. Department of Immunology and Molecular Genetics, Kawasaki Medical School, 577 Matsushima, Kurashiki, Okayama 701-0192, Japan. Tel.: +81 86 462 1111; Fax: +81 86 464 1187; E-mail: [email protected] The EMBO Journal (2005)24:418-427https://doi.org/10.1038/sj.emboj.7600534 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Some of the restarting events of stalled replication forks lead to sister chromatid exchange (SCE) as a result of homologous recombination (HR) repair with crossing over. The rate of SCE is elevated by the loss of BLM helicase or by a defect in translesion synthesis (TLS). We found that spontaneous SCE levels were elevated ∼2-fold in chicken DT40 cells deficient in Fanconi anemia (FA) gene FANCC. To investigate the mechanism of the elevated SCE, we deleted FANCC in cells lacking Rad51 paralog XRCC3, TLS factor RAD18, or BLM. The increased SCE in fancc cells required Xrcc3, whereas the fancc/rad18 double mutant exhibited higher SCE than either single mutant. Unexpectedly, SCE in the fancc/blm mutant was similar to that in blm cells, indicating functional linkage between FANCC and BLM. Furthermore, MMC-induced formation of GFP-BLM nuclear foci was severely compromised in both human and chicken fancc or fancd2 cells. Our cell survival data suggest that the FA proteins serve to facilitate HR, but not global TLS, during crosslink repair. Introduction The integrity of the mammalian genome is essential to suppress oncogenesis, and is maintained by an intricate network of molecules that guard against alterations from DNA damage (Rouse and Jackson, 2002). Fanconi anemia (FA) and Bloom syndrome (BS) are well-known examples of chromosomal instability disorders with an increased incidence of cancer. However, the clinical and cellular features of these diseases are quite distinct. FA is a hereditary disorder characterized by bone marrow failure, skeletal anomalies, and development of leukemia or squamous cell cancer (Grompe and D'Andrea, 2001; Joenje and Patel, 2001; D'Andrea and Grompe, 2003). FA cells are highly sensitive to killing by interstrand crosslinking (ICL) agents such as mitomycin C (MMC) and cisplatin, and display excessive chromosomal aberrations following MMC treatment (Sasaki and Tonomura, 1973; Sasaki, 1975). Complementation analysis indicates that at least 11 causative genes are involved (Levitus et al, 2003), and nine of them have been cloned (FANCA/B/C/D1/D2/E/F/G/L) (Joenje and Patel, 2001; D'Andrea and Grompe, 2003; Meetei et al, 2003a, 2004). Most of the encoded proteins (FANCA/B/C/E/F/G/L) form a nuclear core complex, which is responsible for monoubiquitination of FANCD2, a key event in the FA 'pathway' (Garcia-Higuera et al, 2001). The exact biochemical function of FANCD2 remains unknown, but recent evidence indicates a role of FANCD2 in promoting homologous recombination (HR) (Yamamoto et al, 2005), which is crucial for maintaining genome integrity (Venkitaraman, 2004). Indeed, another FA protein, FANCD1/BRCA2 (Howlett et al, 2002), a tumor suppressor for familial breast cancer, is an important regulator of Rad51, the central molecule of HR (Venkitaraman, 2002; West, 2003). BS is also a hereditary disorder characterized by immunodeficiency, small stature, and development of a variety of cancers (German, 1993). In contrast to FA, BS is caused by a defect in a single gene, BLM helicase (Ellis et al, 1995). The hallmark and a diagnostic feature of BS cells is ∼10-fold increase in sister chromatid exchange (SCE) levels (Chaganti et al, 1974). BLM is a member of the eukaryotic RecQ helicase family, which is thought to play multiple roles in DNA replication, recombination, or repair (Bachrati and Hickson, 2003; Hickson, 2003). For example, BLM is a part of BRCA1-containing mega-protein complex (Wang et al, 2000b), and colocalizes in subnuclear foci with Rad51(Wu et al, 2001). The RecQ family contains four more members (RecQ1, WRN, RecQ4, and RecQ5), and a defect in WRN or RecQ4 is known to cause the genome instability disorders, Werner syndrome (Yu et al, 1996) or Rothmund-Thomson syndrome (Kitao et al, 1999), respectively. Unexpectedly, BLM has been recently detected in the FA core complex (Meetei et al, 2003b), suggesting a potential functional link between FA and BLM proteins. It is well accepted that replication forks often stall at various DNA lesions including endogenous base damage (Cox et al, 2000; Cox, 2001). At least two basic processes, translesion synthesis (TLS) and HR, function to ensure that replication can be resumed at stalled or broken forks. When single-strand breaks or blocking lesions are converted to double-strandbreaks or gaps, respectively, during passage of replication forks, these chromatid discontinuities may be repaired by HR occurring between daughter chromatids (Sonoda et al, 1999). In classical models, HR involves formation of double Holliday junctions, and their resolution may or may not result in crossing over. However, HR repair events in somatic cells primarily proceed without double Holliday junctions, hence without crossing over, through synthesis-dependent strand annealing pathway (Paques and Haber, 1999; Johnson and Jasin, 2000; Cromie et al, 2001). Thus, it is likely that only a fraction of restarting events lead to SCE, which is known to depend on Rad51, Rad54 (Sonoda et al, 1999; Dronkert et al, 2000), five Rad51 paralogs (Takata et al, 2000, 2001), and Nbs1 (Tauchi et al, 2002). Since TLS mediates lesion bypass by a group of specialized polymerases, dysfunctional TLS might lead to more SCE if the stalled fork breaks and requires HR to restart. Consistently, at least some of the TLS-deficient cells such as rad18, rev3, or dinB mutants display increased levels of SCE (Okada et al, 2002; Yamashita et al, 2002; Sonoda et al, 2003). In contrast, BLM may directly suppress SCEs by catalyzing the 'dissolution' of the double Holliday junctions (Wu and Hickson, 2003). In this study, we generated FANCC-deficient DT40 cells (hereafter referred to as fancc), and found that levels of spontaneous SCEs increased ∼2-fold compared to wild-type cells, similar to our fancd2 cells (Yamamoto et al, 2005). To investigate the basis of the elevated SCE, we made double mutants of FANCC with mutations in HR (XRCC3), TLS (RAD18), and BLM helicase. Our genetic analyses indicate functional linkage of FANCC with Xrcc3 or BLM but not with Rad18. Furthermore, crosslink damage-induced relocalization of BLM was defective in both human and chicken fancc or fancd2 cells. We propose that BLM, regulated by the FA pathway, functions in restarting stalled replication forks blocked by spontaneous lesions and ICLs. Results Disruption of chicken FANCC gene in DT40 cells We obtained a cDNA clone containing full-length chicken FANCC by searching the chicken EST database (http://swallow.gsf.de/DT40/dt40Est.html). Chicken FANCC encodes a putative 559-amino-acid protein (DDBJ accession number AB176529) compared to 557 amino acids of human FANCC. The identity and similarity between two proteins are 45 and 59%, respectively. There are no domains or motifs suggestive of biochemical function in either protein. Based on the sequence of the cDNA, we PCR-amplified a genomic sequence of chicken FANCC and designed a targeting vector (Figure 1A). A single transfection with the vector abrogated the band in Southern blot analysis (Figure 1B). This is not surprising, given the localization of FANCC on human chromosome 9 and the extensive synteny between human chromosome 9 and chicken Z sex chromosome (Nanda et al, 1999). DT40 was derived from a female chicken, which is hemizygous in terms of the Z chromosome. Another FA gene, FANCG, also lies on human chromosome 9, and we were able to disrupt FANCG in DT40 by single transfection (Yamamoto et al, 2003). RT–PCR analysis confirmed the FANCC gene disruption (Figure 1C) that is expected to delete one exon. Nucleotide sequencing revealed that the faint, shorter transcript was owing to anomalous splicing, which is expected to create a truncated protein (residues 1–55 and six additional amino acids) because of a frame shift. We also examined induction of the long, monoubiquitinated form of FANCD2 protein (FANCD2-L) (Gregory et al, 2003) by MMC treatment using Western blotting with anti-chicken FANCD2 antiserum. As shown in Figure 1D, the FANCD2-L form existed in untreated wild-type cells, and its proportion was clearly increased after MMC treatment. In contrast, FANCD2-L was not detected in fancc cells before or after MMC treatment (Figure 1D), similar to human fancc cells (Garcia-Higuera et al, 2001). Figure 1.Targeted disruption of chicken FANCC loci in DT40 cells. (A) Schematic representation of partial chicken FANCC locus, the gene disruption construct, and the configuration of targeted allele. S, SacI; B, BamHI; H, HindIII. Only a single exon was deleted by the gene disruption. Arrowheads indicate the positions of primers used in RT–PCR. (B) Southern blot analysis of SacI-digested genomic DNA from cells with indicated genotypes using flanking probe as shown in (A). WT, wild type. (C) RT–PCR analysis of total RNA from wild-type and fancc cells. Primers were designed from FANCC sequences of upstream and downstream exons as shown in (A). As a control, the entire coding region of chicken Rad51 was amplified from each RT product. (D) Western blot analysis of whole-cell lysate prepared from wild-type and fancc cells probed with anti-chicken FANCD2 serum. Cells were treated with MMC (500 ng/ml) for 1 h, washed, and lysed 6 h later. L or S, long or short form of FANCD2, respectively. Download figure Download PowerPoint fancc cells have increased levels of spontaneous SCE The FANCC-deficient cells grew more slowly than wild-type cells with an increased proportion of dead cells in the culture (data not shown), and the cell cycle distribution examined by BrdU pulse labelling was not altered (data not shown). As expected, the cells were highly sensitive to the DNA crosslinkers MMC and cisplatin, while sensitivity to X-rays or UV irradiation was only mildly increased (Figure 2A). The cisplatin sensitivity could be complemented to wild-type levels by expression of chicken FANCC cDNA, indicating that this defect was indeed caused by FANCC disruption (Figure 2B). Although spontaneous chromosomal breaks were not elevated, MMC-induced aberrations occurred much more frequently in fancc cells (Figure 2C). Figure 2.Characterization of fancc cells. (A) Sensitivity curves of cells to various DNA-damaging agents. The fraction of surviving colonies in methylcellulose plates is shown for each agent. Mean and standard deviation (s.d.) of at least three independent experiments are shown. WT, wild-type cells. (B) Complementation of fancc cells with FANCC expression. Three fancc clones expressing chicken FANCC cDNA (+chFANCC#1 to #3) were compared with wild-type and fancc cells in colony survival with cisplatin. (C) Chromosome aberrations after MMC treatment in wild-type and fancc cells. A total of 200 metaphases were scored. Download figure Download PowerPoint To test whether fancc cells have HR defects, we examined gene targeting at three genomic loci. Wild-type and fancc cells were transfected with linearized targeting vectors and selected in media containing appropriate drugs. After expansion, each colony was examined for targeting events by Southern blot analysis of genomic DNA. Fancc cells had dramatically reduced gene-targeting efficiency compared to wild-type cells (Table I). In contrast, we also found that the frequency of spontaneous SCE in fancc cells was elevated ∼2-fold compared to wild-type cells (Figures 3C and 4B), similar to our fancd2 cells (Yamamoto et al, 2005). Figure 3.Genetic analysis of fancc mutation combined with xrcc3. (A) Generation of conditional xrcc3 cells and disruption of FANCC in those cells. OH-TAM treatment activates MerCreMer recombinase that removes the human Xrcc3 (hXrcc3)-IRES-EGFP expression cassette. WT (cond), xrcc3 cells that express hXrcc3, GFP, and MerCreMer. (B) CDDP sensitivity curves of cells with indicated genotypes. Fancc cells used were derived from wild-type DT40 cells, not from the conditional xrcc3 mutant. Mean and s.d. from three independent experiments are shown. (C) Spontaneous SCE levels. Numbers represent mean and s.d. of scores from 50 metaphases. In this analysis, we used fancc cells derived from the conditional xrcc3 mutant, which express hXrcc3. Statistical significance was detected between WT (cond) versus fancc, WT (cond) versus xrcc3, and fancc versus xrcc3 (Bonferroni/Dunn test, P<0.0001 in all comparisons). The difference was insignificant between xrcc3 versus fancc/xrcc3 double-mutant clones. Download figure Download PowerPoint Figure 4.Genetic analysis of fancc mutation combined with rad18 or blm. (A) CDDP or MMS survival curves of cells with indicated genotypes. Mean and s.d. from three independent experiments are shown. WT, wild type. (B) Spontaneous SCE levels. Numbers represent mean and s.d. of scores from 50 metaphases. Statistical significance was detected between WT versus fancc, WT versus rad18, WT versus fancc/rad18, fancc versus facc/rad18, and rad18 versus fancc/rad18 (Bonferroni/Dunn test, P<0.0001 in all comparisons). (C) Survival curves of cells with indicated genotypes. Mean and s.d. from three independent experiments are shown. (D) Spontaneous SCE levels. Numbers represent mean and s.d. of scores from 50 metaphases. Statistical significance was detected between fancc versus blm, fancc versus fancc/blm#1, and fancc versus fancc/blm#2 (Bonferroni/Dunn test, P<0.0001 in all comparisons). There was no statistically significant difference between blm and fancc/blm double mutants. (E) MMC (20 ng/ml)-induced SCE levels. Numbers represent mean and s.d. of scores from 50 metaphases. Statistical significance was detected between fancc versus blm, fancc versus fancc/blm#1, and fancc versus fancc/blm#2 (Bonferroni/Dunn test, P<0.0001 in all comparisons). There was no statistically significant difference between blm and fancc/blm double mutants. Download figure Download PowerPoint Table 1. Targeted integration efficiencies in fancc cells Targeting vectors Genotypes Ku70-His Xrcc2-puro Ovalubumin-puro Wild type 14/34 (41%) 12/24 (50%) 10/24 (42%) fancc 3/20 (15%) 1/11 (9%) 0/35 (0%) Data are numbers of targeted clones per number of clones analyzed by Southern blotting. The percentage of targeted integration events is given in parentheses. Elevated SCEs in fancc cells depend on Rad51 paralog Xrcc3 To investigate the mechanism of SCE elevation in fancc cells, and to better define the FA pathway, we performed genetic analysis by disrupting FANCC in cells that are deficient in HR (xrcc3), TLS (rad18), or BLM helicase (blm). We disrupted FANCC in conditional xrcc3 background (Ishiai et al, 2004), and exposed the cells, as well as the parental cells, to 4-hydroxy tamoxifen (OH-TAM) (Figure 3A) to excise the Xrcc3-EGFP expression cassette by activating the MerCreMer recombinase (Zhang et al, 1998). The removal was ensured by subcloning, and was further verified by the loss of GFP fluorescence and by Southern blotting using human XRCC3 probe (data not shown). Thus, two independent clones of fancc/xrcc3 cells were established. We compared the fancc/xrcc3 cells with fancc cells in terms of cisplatin sensitivity. Fancc cells were much more cisplatin sensitive to killing compared to xrcc3 cells. However, fancc/xrcc3 cells displayed about the same cisplatin sensitivity as the fancc single mutant, indicating functional overlap between FANCC and XRCC3 (Figure 3B). In addition, spontaneous SCE was clearly decreased in xrcc3, but elevated in fancc cells compared to the parental conditional cells (Figure 3C). Not surprisingly, the SCE frequency in two clones of fancc/xrcc3 cells was similar to that of xrcc3 cells (Figure 3C), indicating that spontaneous SCE in fancc cells is partially Xrcc3-dependent as in wild-type cells (Takata et al, 2001). Genetic analysis of FANCC and TLS factor Rad18 Since several TLS-deficient vertebrate cells including Rad18-deficient DT40 (Yamashita et al, 2002) have increased SCE (Okada et al, 2002; Sonoda et al, 2003), it is quite possible that elevated SCE in fancc cells might be related to TLS defects. In yeast Saccharomyces cerevisiae, most TLS activity is dependent on Rad6, and its heterodimeric partner Rad18. To examine the relationship between the FA pathway and TLS, the FANCC gene was targeted in rad18 cells (Yamashita et al, 2002). Although both rad18 and fancc cells had high sensitivity to cisplatin or methylmethanesulfonate (MMS), the fancc/rad18 double mutant displayed even higher sensitivity to both agents, indicating their distinct functions (Figure 4A). Moreover, fancc/rad18 cells had even greater increased SCE than either rad18 or fancc cells (Figure 4B). Thus, we conclude that spontaneous SCEs in fancc cells are elevated by a mechanism separate from Rad18 function. It is still possible that FANCC plays a role in Rad18-independent TLS, since not all TLS events require Rad18 in vertebrate cells (Okada et al, 2002). A functional link between FA pathway and BLM in suppressing spontaneous SCE Next, we deleted FANCC in BLM-deficient cells (Wang et al, 2000a). In contrast to fancc/rad18 cells, we found that the blm and fancc/blm mutants displayed about the same high levels of spontaneous and MMC-induced SCE (Figure 4D and E). Given the functional link suggested by these findings, we measured cell survival of blm, fancc, and fancc/blm cells following treatment with several mutagens. While MMS-treated fancc/blm cells survived more poorly than either single mutant (additive phenotype), survival following cisplatin or MMC treatment was similar between fancc and fancc/blm cells (Figure 4C). These results suggest that FANCC and BLM act in a common pathway in ICL repair as well as in controlling SCE levels. However, in repairing DNA lesions created by MMS, they likely have more distinct functions. In addition, fancc/blm cells grew more slowly than either fancc or blm cells (data not shown), indicating that the requirement of both genes extends to cell viability. To further verify the potential link between FA proteins and BLM, we created double mutant lacking both FANCD2 and BLM by targeting FANCD2 gene in blm cells. FANCD2 disruption was confirmed by Southern (data not shown) and Western blotting (Supplementary Figure 1A). We found that spontaneous SCE levels and cisplatin sensitivity of the double mutant were essentially the same as those of blm mutant or fancd2 mutant, respectively (Supplementary Figure 1B and C). This result is consistent with the essential role of FANCC in activation and/or monoubiquitination of FANCD2 (Figure 1D). Defective BLM focus formation in the absence of FANCC and FANCD2 While blm cells have extremely elevated SCEs, the elevation is only ∼2-fold both in fancc and fancd2 cells. As BLM helicase was shown to mediate the 'dissolution' of double Holliday junctions without crossing over (Wu and Hickson, 2003), it likely suppresses SCE in a direct manner. We hypothesized that increased SCE in fancc and fancd2 cells is due to mildly compromised BLM function by the absence of the FA 'pathway'. Since BLM resides in a nuclear complex with several FA proteins (Meetei et al, 2003b), we examined the distribution of BLM inside fancc and fancd2 cells. GFP-human BLM (GFP-hBLM) has been extensively utilized in a number of studies to examine its subnuclear localization (Hu et al, 2001; Suzuki et al, 2001; Yankiwski et al, 2001; Stavropoulos et al, 2002). To visualize chicken BLM localization, we transfected a GFP-chicken BLM (GFP-chBLM) expression construct into wild-type, fancc, and fancd2 cells. Stably expressing transfectants were analyzed by FACSCalibur, and clones with equal expression levels were selected for further analysis. GFP-chBLM was fully functional in suppressing SCE in blm DT40 cells (data not shown). In wild-type cells expressing GFP-chBLM, we observed relatively large foci (5–10 foci per cell) in 2–3% of the untreated cells (Figure 5A and B). These foci might correspond to PML nuclear bodies (Ishov et al, 1999), although we could not confirm this because of the lack of suitable reagents. In addition, we detected a number of less intense, smaller foci in untreated cells (Figure 5A and B). MMC treatment produced brighter and more numerous foci (Figure 5A and B). Some cells had a rather faint patch-like GFP accumulation that seems to be BLM localized in the nucleolus (Figure 5B). Figure 5.Analysis of GFP-chBLM in wild-type, fancc, and fancd2 DT40 cells. (A) Focus formation of GFP-chBLM. Cells expressing GFP-chBLM were observed before and 7 h after MMC treatment (500 ng/ml, 1 h). (B) Fraction of cells having large foci, indicated number of small foci, and patch-like accumulation. Cells were treated as in (A). At least 200 cells were scored in each preparation. (C) Colocalization of GFP-chBLM and FANCD2 in wild-type cells treated with MMC as in (A). (D) Co-immunoprecipitation of GFP-chBLM with FANCD2. Cells with indicated genotypes were treated with MMC (500 ng/ml for 7 h) or left untreated and lysed. A portion of the lysate (2.5%) was saved as an input control, and the rest was subjected to immunoprecipitation with anti-GFP. The lysates and immunoprecipitates were separated and probed with the indicated antibodies. GFP-chBLM (−), wild-type cells that were not transfected with the GFP-chBLM construct serving as negative control. The experiments were repeated three times with similar results. Download figure Download PowerPoint Surprisingly, in fancc or fancd2 mutant cells, the formation of both large and small GFP-chBLM foci was largely suppressed after MMC treatment. In particular, the number of the small foci per cell and the % positive cells were severely decreased in fancc and fancd2 cells (Figure 5A and B). This reduction was not caused by differences in cell cycle, since both wild-type and mutant cells showed progressive accumulation at late S to G2 phase in an essentially similar way after MMC treatment (data not shown). In untreated cells, the reduction was not as clear as in MMC-treated cells (Figure 5B). These results indicate that nuclear relocalization of BLM is influenced by the FA proteins. Next, we looked at whether BLM and FANCD2 can colocalize in DNA damage-induced subnuclear foci. FANCD2 foci were detected by staining with anti-chicken FANCD2. As shown in Figure 5C, GFP-chBLM and FANCD2 foci partially colocalized in MMC-treated cells. Among the 200 cells scored 7 h after MMC treatment, ∼40% were positive for FANCD2/GFP-chBLM colocalizing foci, while ∼40 or 10% were positive only for FANCD2 foci or GFP-chBLM foci, respectively. The rest of the cells (∼10%) had no focus formation. Furthermore, we were able to detect FANCD2-L form in anti-GFP immunoprecipitates from wild-type but not fancc cells expressing GFP-chBLM following MMC treatment (Figure 5D). These data suggest that monoubiquitinated FANCD2 physically interacts with BLM either directly or indirectly, resulting in the colocalizing foci. Of note, we could not detect any reproducible mobility change, which may suggest post-translational modifications, in immunoprecipitated GFP-chBLM treated with λ-phosphatase foll
Referência(s)