Growth Defect and Mutator Phenotypes of RecQ-Deficient Neurospora crassa Mutants Separately Result From Homologous Recombination and Nonhomo
http://www.100md.com
遗传学杂志 2006年第1期
ABSTRACT RecQ helicases function in the maintenance of genome stability in many organisms. The filamentous fungus Neurospora crassa has two RecQ homologs, QDE3 and RECQ2. We found that the qde-3 recQ2 double mutant showed a severe growth defect. The growth defect was alleviated by mutation in mei-3, the homolog of yeast RAD51, which is required for homologous recombination (HR), suggesting that HR is responsible for this phenotype. We also found that the qde-3 recQ2 double mutant showed a mutator phenotype, yielding mostly deletions. This phenotype was completely suppressed by mutation of mus-52, a homolog of the human KU80 gene that is required for nonhomologous end joining (NHEJ), but was unaffected by mutation of mei-3. The high spontaneous mutation frequency in the double mutant is thus likely to be due to NHEJ acting on an elevated frequency of double-strand breaks (DSBs) and we therefore suggest that QDE3 and RECQ2 maintain chromosome stability by suppressing the formation of spontaneous DSBs.
THE RecQ family of DNA helicases has been highly conserved during evolution and is present in organisms ranging from bacteria to humans. Members of the RecQ family have been shown to be important for the maintenance of genomic stability. Mutation in three of the five human RecQ homologs, BLM, WRN, and RecQL4, results in Bloom syndrome (BS), Werner syndrome (WS), and Rothmund-Thomson syndrome (RTS), respectively (ELLIS et al. 1995; YU et al. 1996; KITAO et al. 1998, 1999). BS is characterized by growth retardation, a markedly increased incidence of several types of cancer and genomic instability, including chromatid gaps, breaks, and rearrangements (GERMAN 1993). In addition, BS cells exhibit elevated sister-chromatid exchange (CHAGANTI et al. 1974). WS is characterized by genomic instability and the premature appearance of aging phenotypes in young adults, including cancers (EPSTEIN et al. 1966). Cells derived from WS patients show an increased rate of somatic mutations, chromosome loss, and deletions (SALK et al. 1981; FUKUCHI et al. 1989). RTS is also characterized by premature aging and cancer predisposition (VENNOS and JAMES 1995). These three diseases are clinically distinct and the three causative genes are thought to play distinct roles. The genomic instability found in these diseases correlates with cancer predisposition but there is no direct evidence that the genome instability is responsible for the cancer susceptibility. The remaining two human RecQ homologs, RecQL1 (PURANAM and BLACKSHEAR 1994; SEKI et al. 1994) and RecQL5 (KITAO et al. 1998), have not been implicated in any disease. However, in chicken DT40 cells, in which the BLM function is impaired, RecQL1 and RecQL5 appear to be required for cell viability (WANG et al. 2003).
In Saccharomyces cerevisiae, there is only one RecQ homolog, Sgs1. sgs1 mutants show an increase in various types of recombination (WATT et al. 1996; YAMAGATA et al. 1998; ONODA et al. 2000), accumulate gross chromosomal rearrangements (MYUNG et al. 2001), and show impaired sporulation (WATT et al. 1995) and premature aging of mother cells (SINCLAIR et al. 1997).
Much evidence supports the involvement of RecQ helicases in DNA replication. The expression of Sgs1, BLM, WRN, and RecQL4 is cell cycle regulated and peaks during S-phase (KITAO et al. 1998; DUTERTRE et al. 2000; FREI and GASSER 2000; KAWABE et al. 2000). Correspondingly, WRN interacts physically with components required for replication, including DNA polymerase , topoisomerase I, PCNA, FEN-1, and RPA (BROSH et al. 1999, 2001; LEBEL et al. 1999; KAMATH-LOEB et al. 2000). In addition, BLM and WRN localize to stalled replication sites, resulting from cellular exposure to hydroxyurea (HU) (CONSTANTINOU et al. 2000; JIAO et al. 2004). In S. cerevisiae, some Sgs1 foci also colocalize with replication foci (FREI and GASSER 2000). Human cells lacking functional BLM or WRN accumulate aberrant replication intermediates (LONN et al. 1990; POOT et al. 1992). Recent studies indicate that RecQ helicases are required for processing of DNA structures induced by stalled replication forks. It is thought that lesions on the leading strand can cause replication forks to regress, resulting in Holliday-junction-like four-way structures. Holliday junctions are the preferred substrate of BLM, WRN, and Sgs1 (BENNETT et al. 1999; MOHAGHEGH et al. 2001). BLM and WRN can also catalyze branch migration of Holliday junctions (CONSTANTINOU et al. 2000; KAROW et al. 2000). Moreover, defects in Schizosaccharomyces pombe rqh1, the only RecQ homolog in S. pombe, are suppressed by overexpression of a bacterial Holliday junction resolvase, RusA (DOE et al. 2000). However, how replication regression is promoted and how RecQ helicases maintain genomic stability remains to be explained.
In Neurospora crassa, there are two RecQ homologs, QDE3 and RECQ2 (COGONI and MACINO 1999; PICKFORD et al. 2003; KATO et al. 2004). QDE3 is a long-type RecQ homolog, like BLM, WRN, and RTS, while RECQ2 is a short-type homolog, like RecQL1 and RecQL5. It has been reported that QDE3 and RECQ2 are required for DNA repair (PICKFORD et al. 2003; KATO et al. 2004) but it was not previously known whether they are required for genome stability. In this report, we describe the effects of mutations in QDE3 and RECQ2 and their interaction with mutations in homologous recombination (HR) and nonhomologous end joining (NHEJ).
MATERIALS AND METHODS
Strains:
The N. crassa strains used in this study are listed in Table 1. C1-T10-37A and C1-T10-28a (TAMARU and INOUE 1989) are wild-type strains closely related to the standard Oak Ridge wild-type strain. Production of the recQ2 mutant strains KTO-Q2R-5A and KTO-Q2-2A is described in the next section. The qde-3 recQ2 double-mutant strains KTO-Qd-32a and KTO-Qd-35A were derived from a cross between the qde-3 mutant strain KTO-r-20a (KATO et al. 2004) and KTO-Q2-2A. The mus-38 recQ2 double-mutant strain KTO-2m38-1a was derived from a cross between the mus-38 mutant strain CZ272-5a (ISHII et al. 1998) and KTO-Q2-2A. The qde-3 recQ2 mus-38 triple-mutant strain KTO-dm38-26A was derived from a cross between the qde-3 mus-38 double-mutant strain QMU38-2A (KATO et al. 2004) and KTO-2m38-1a. The recQ2 mei-3 double-mutant strain KTO-2mi3-1a was derived from a cross between the mei-3 mutant strain CY-10-9a (ISHII and INOUE 1994) and KTO-Q2-2A. The qde-3 recQ2 mei-3 triple-mutant strain KTO-dmi3-1A was derived from a cross between KTO-Qd-35A and KTO-2mi3-1a. The qde-3 recQ2 mus-52 triple-mutant strain KTO-dk8-33a was derived from a cross between KTO-Qd-32a and the mus-52 mutant strain 54yo-828-3 (NINOMIYA et al. 2004).
Construction of the recQ2 mutant:
A 1908-bp fragment upstream of the recQ2 ORF, beginning 5 bp upstream from the ORF, and a 1836-bp fragment downstream from the recQ2 ORF, beginning 165 bp downstream from the ORF, were amplified by PCR using the following primer sets:
Q2U-1: TTGGGATGATCGAAGAG and Q2U-2: GCAGCGTCGATAGCA
Q2D-1: TTCGGTGACAGGTAGGT and Q2D-2: GTCGTTCTCTGCCTTAG.
These two fragments were inserted on each side of the hygromycin-resistance marker gene (hygr) derived from the plasmid pCSN43 (STABEN et al. 1989) to construct the plasmid pQ2::HYG. This plasmid was used as a PCR template to produce the targeting construct hygr flanked by sequences upstream and downstream of recQ2. The amplified fragment was introduced into the C1-T10-37A wild-type strain by electroporation to replace the endogenous recQ2 with the hygr gene. Hygromycin-B-resistant transformants were isolated and the replacement was confirmed by PCR using the primer Q2I: AACAACAGGCGCGACCAA, which is located in the upstream fragment, and Q2O: AAGCCCGTAGAGTGCAGACAAA, which is located downstream of the downstream fragment. In the wild-type case, a 4-kb fragment was amplified. Since 1.8 kb, including the whole recQ2 ORF, is replaced by the 2.4-kb hygr gene in the recQ2-replaced mutant, a 4.5-kb fragment was amplified. Extra ectopic copies were ruled out and gene replacement was confirmed by Southern analysis. This mutant strain, designated as KTO-Q2R-5A, was backcrossed to the C1-T10-28a wild-type strain, and it was confirmed that the hygr marker gene segregates normally through meiosis. Strain KTO-Q2-2A is one of the progeny of this cross and was used as a standard recQ2 mutant strain.
Growth media and genetic methods:
Growth media and genetic procedures for N. crassa were as described by DAVIS and DE SERRES (1970). Transformation of Neurospora by electroporation was performed as previously described (NINOMIYA et al. 2004).
Qualitative assay of mutagen sensitivity by spot tests:
Sensitivity to chemical mutagens and other chemicals was analyzed by spot tests, as described by WATANABE et al. (1997). Each conidial suspension was diluted and adjusted to densities of 10,000, 1000, 100, and 10 viable conidia per spot. Methyl methanesulfonate (MMS), N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), camptothecin (CPT), tert-butyl hydroperoxide (TBHP), ethyl methanesulfonate (EMS), 4-nitroquinoline-1-oxide (4NQO), and HU were added to agar medium at final concentrations of 0.005 or 0.02%, 0.15 or 0.45 μg/ml, 0.0025 or 0.25 μg/ml, 0.0014 or 0.0028%, 0.1 or 0.2%, 0.0075 or 0.045 μg/ml, and 0.38 or 1.9 mg/ml, respectively. Conidial suspensions were spotted onto these plates and grown at 30° for 2 days. UV sensitivity was investigated by spotting a conidial suspension onto an agar plate and irradiating at 150 or 300 J/m2.
Assay of UV sensitivity:
The UV dose dependency of the survival of N. crassa was measured as described by INOUE and ISHII (1984). Conidial suspensions at a final concentration of 1 x 106 cells/ml were irradiated at various doses of UV, and aliquots were sampled and plated after appropriate dilution. The yeast on all the plates were allowed to grow at 30° for 3 days and the number of colonies on each plate was counted.
Measurement of apical growth and life span:
Apical growth of hyphae was measured in race tubes of 30 cm length. So that the age of each strain would be consistent, mutant strains were backcrossed to the wild-type strain and the resultant ascospores were randomly isolated and grown at 30° for 1 week. Conidia of each were transferred to new media and grown at 30° for another week while the genotype of each progeny was investigated by the spot test or PCR. A silica gel stock was made from the conidia of each strain and stored at 4°. To start each experiment, a few grains of the silica gel were placed on Vogel's minimal agar medium and incubated at 30° for 1 week. The resulting conidia were inoculated at one end of the race tube and incubated at 25°. The position of the growing front was marked once or twice a day. When the mycelia reached the other end of the race tube, a piece of medium bearing mycelia was cut from the growing front and transferred to a fresh tube.
Microscopy:
Mycelia were grown on Vogel's minimal agar medium containing 1.2% sucrose and 2% agar at 30° for 1 day. A piece of mycelium was cut from the growing front, fixed in 70% ethanol, and transferred to a glass slide. SYBR Gold (Molecular Probes, Eugene, OR) dye was used to stain nuclei. Mycelia were observed by fluorescence microscopy (Nikon) and images were taken with a cooled CCD color camera (KEYENCE).
Forward mutation assay:
The direct method described by DE SERRES and KOLMARK (1958) was used for the detection of ad-3 mutations. To eliminate preexisting ad-3 mutation, cultures started by single-colony isolation were used for each experiment. A conidial suspension from each isolate was inoculated into each 5-liter Florence flask at a concentration of 5 x 105 viable conidia per flask. The flask was cultured in the dark at 30° with low aeration. Under these conditions, each conidium makes a bead-like colony after 1 week of culture. As ad-3 mutants accumulated purple pigments in mycelia, purple-colored colonies were isolated and subcultured. To distinguish ad-3A from ad-3B mutants, complementation tests were carried out using ad-3A and ad-3B tester strains (Table 1). Spontaneous mutation rates per nucleus per division were calculated by the method of the median (LEA and COULSON 1949) using the following formula: mutation rate = m/N, where m is the calculated number of mutation events and N is the number of nuclear divisions. To estimate the number of nuclear divisions, the number of nuclei in 50 conidia was counted and the average number of nuclei in a conidium was multiplied by the estimated number of colonies in the culture. The formula r/m – ln(m) – 1.24 = 0, where r is the median number of ad-3 mutants, was solved for m.
Analysis of mutations:
To amplify the ad-3A gene from genomic DNA, PCR was performed with two primers: Pst, AACTGCAGTTCTGATCCGTCCAGTTC, and M2, TTCGTTACAGTGCCGAGTCC. To amplify the ad-3B gene from genomic DNA, PCR was performed with two primers: X1, TTGGGTCGTATGCTCTGTGA, and Y1, CTTGGGCCACTGTCCGT.
To determine the nucleotide sequence of the ad-3A gene, six sequencing primers were used (BigDye Terminator Cycle Sequencing Ready Reaction kit v3.1, Applied Biosystems, Foster City, CA): 2F, GTCCAGTTCCGTCCCGTCCA; 3F, TTTACTACCGACCGCATTTCGG; 5F, CGAGCACCAAGGCAGAGCTG; 4R, GCAGTAAGCTGTGTG AGGAT; 6R, CAGTGGAGTAGACCTTGAGC; and 8R, CCAGCAAACTATCAG ATCAC.
To amplify regions outside of the ad-3A ORF, six primers were used: L1, CAAATGGAAGACGGTGGACT; R1, TCGCAGACTGTGAGGAATTG; L2, GATTGACGCCCTTTTGATGT; R2, GAACTGGAGGCTGGAGTGAG; L4, CGGCGGTCGTTGAGGTAG; and R4, CAGCGAGTATTTGGTGGAAGG.
Analysis of chromosome deletion around the ad-3A gene was performed by Southern hybridization using cosmid G4:E3 as a probe.
RESULTS
The qde-3 recQ2 double mutant showed increased sensitivity to all tested mutagens:
The recQ2 mutant was generated by replacing the whole recQ2 ORF by hygr, as described in MATERIALS AND METHODS. The mutant did not show any defects during vegetative growth and sporulation. The sensitivity of the recQ2 mutant to a variety of mutagens and other chemicals was tested by spot tests (Figure 1). Conidial suspensions adjusted to the same concentrations were spotted on plates containing MMS, MNNG, EMS, TBHP, 4NQO, CPT, or HU, and plates were incubated at 30°. In the UV-sensitivity test, conidia were irradiated after spotting.
While the qde-3 mutant showed high sensitivity to mutagens, as described previously (KATO et al. 2004; Figure 1A), the recQ2 mutant was slightly more sensitive to MMS, MNNG, and CPT than the wild-type strain, the increased sensitivity being detectable only at high concentrations of the DNA-damaging agents (Figure 1B). The qde-3 recQ2 double mutant showed higher sensitivity to the type I topoisomerase inhibitor CPT, the replication inhibitor HU, and to all mutagens tested than the qde-3 single mutant did (Figure 1B). Unlike the sgs1 mutant in S. cerevisiae and the rqh1 mutant in S. pombe, which are sensitive to UV, the UV sensitivity of the qde-3 recQ2 double mutant was similar to that of the wild-type strain (Figure 1, B and C).
To investigate whether QDE3 and RECQ2 function in the resistance to UV, a qde-3 mus-38 double mutant and a qde-3 recQ2 mus-38 triple mutant were made and their UV sensitivity was tested. mus-38 mutants are deficient in nucleotide excision repair (NER) and sensitive to UV. The qde-3 mus-38 double mutant showed a higher sensitivity to UV than the parental mus-38 mutant, and the qde-3 recQ2 mus-38 triple mutant was still more sensitive (Figure 1D). These data indicate that the qde-3 and recQ2 genes are required for resistance to UV in the case of dysfunction of MUS38.
The qde-3 recQ2 double mutant showed a growth defect:
Growth of the qde-3 recQ2 double mutant was poor on glycerol complete medium as well as on minimal medium. About one-fourth of the random progeny of a backcross of the double mutant to the wild-type strain showed slow growth, and all were qde-3 recQ2 double mutants (data not shown). Both qde-3 and recQ2 single mutants grew as fast as the wild-type strain for several consecutive transfers (Figure 2A and data not shown). The double mutant grew as fast as the wild-type strain in the first and second race tubes (Figure 2A) but its growth slowed in the third tube and was dramatically reduced in the fourth tube (Figure 2A). The appearance of mycelia was also altered in the fourth tube, with the accumulation of black and dense orange pigments and spherically formed mycelia (Figure 2B).
Microscopic observation showed that the size and morphology of hyphal tips in the wild-type strain, the qde-3 mutant, and the recQ2 mutant were in all cases similar (Figure 2C). As N. crassa is a coenocytic organism, many nuclei were found in mycelia of all those strains. Hyphal tips of the qde-3 recQ2 double mutant appeared similar when the cells were young and growing vigorously, but in the aged double mutant, whose growth was slow, mycelial tips were thin and the number of nuclei decreased (Figure 2C). Along with the deterioration in linear growth, the colony morphology of the double mutant became flat and spreading, whereas that of the wild-type strain was thick and dense (Figure 2D).
Conidia in the wild-type, the qde-3 mutant, and the recQ2 mutant strains had similar viability (77, 71, and 74%, respectively) but viability was much reduced (to 33%) in the double mutant.
The growth defect of the qde-3 recQ2 double mutant was suppressed by mutation of a gene required for HR:
In S. cerevisiae, the role of another DNA helicase, Srs2, appears to overlap with that of Sgs1. Both Sgs1 and Srs2 have 3'-5' DNA helicase activity (RONG and KLEIN 1993; BENNETT et al. 1998) and both sgs1 and srs2 mutants have a hyperrecombination phenotype (AGUILERA and KLEIN 1988; WATT et al. 1996). Simultaneous deletion of SGS1 and SRS2 results in extremely poor growth or synthetic lethality (LEE et al. 1999; GANGLOFF et al. 2000), with either phenotype alleviated by mutations in the HR genes RAD51, RAD55, and RAD57 (GANGLOFF et al. 2000). A similar result was reported in S. pombe (MAFTAHI et al. 2002). Moreover, overexpression of Sgs1 partially suppresses the MMS and HU sensitivity of the srs2 mutant (MANKOURI et al. 2002).
Similarly, qde-3 and recQ2 are putative DNA helicases with redundant roles in DNA repair in N. crassa. Therefore, the growth defect in the qde-3 recQ2 mutant might be suppressed by mutations in HR genes, for example, in the RAD51 homolog mei-3. The qde-3 recQ2 mei-3 triple mutant grew as fast as the wild-type strain during the entire course of the transfer experiment (Figure 2A). Unlike in the qde-3 recQ2 double mutant, mycelial appearance and growth rate were normal, even after six transfers (data not shown). Colony morphology of the triple mutant was similar to that of the mei-3 mutant (Figure 2D). Conidial viability of the triple mutant and the mei-3 mutant was the same (64 and 61%, respectively).
qde-3 and qde-3 recQ2 double mutants show increased mutability and extensive deletion:
To test the genomic instability of RecQ-homolog mutants of N. crassa, spontaneous mutability was investigated. The spontaneous mutation rate at the ad-3 locus was 0.4 x 10–7 in the recQ2 mutant, which was similar to that in the wild-type strain (Table 2). However, in the qde-3 mutant, the spontaneous mutation rate was 2.4 x 10–7, 11 times higher than that in the wild-type strain. The spontaneous mutation rate of the qde-3 recQ2 double mutant was 4.0 x 10–7, 18 times higher than that of the wild-type strain.
The ad-3 locus consists of two genes, ad-3A and ad-3B, and which gene is mutated in a particular ad-3 mutant can be determined by complementation tests. We wanted to examine the spectrum of mutations generated in qde-3, recQ2, and qde-3 recQ2 double mutants. As the ad-3A gene is smaller than the ad-3B gene, we first determined the sequence of the ad-3A gene.
To determine the sequence, the ad-3A ORF was amplified by PCR using primers Pst and M2 (Figure 3A). The expected 1-kb fragments were amplified using genomic DNA of the wild type, the qde-3 mutant, and the qde-3 recQ2 double mutant as a template, but no fragments were amplified from most of the ad-3A mutants obtained from the qde-3 mutant and the qde-3 recQ2 double mutant (Table 3). Since the ad-3B gene was successfully amplified using primers X1 and Y1 (Figure 3B) in all of these mutants, it was thought that these mutants had experienced deletion in ad-3A. Primers L1 and R1 yield a 3-kb fragment that includes the ad-3A ORF and 1 kb of flanking sequence on each side of the ORF, while primers L2 and R2 give a similar 5-kb fragment that includes 2 kb of each flanking region, and primers L4 and R4 give a 7-kb fragment that includes 3 kb of each flanking region (Figure 3A).
Genomic DNA of the wild-type, the qde-3 mutant, or the qde-3 recQ2 mutant strains yielded the expected fragments with each of these primer pairs (Table 3, "controls"). On the other hand, genomic DNAs of the ad-3A mutants derived from the qde-3 or qde-3 recQ2 strains yielded no fragments in these PCR analyses, suggesting that they have large deletions. In 8 of 11 (72%) ad-3A mutants obtained from the qde-3 mutant strain and 43 of 51 (84%) ad-3A mutants obtained from the qde-3 recQ2 double-mutant strain, no fragments were amplified, even when primers L4 and R4 were used (Table 3). Southern analysis revealed that these mutants all have deletions at the ad-3A locus (data not shown).
The resulting band patterns were the same in all the ad-3A mutants obtained from the same subculture (Jug 9) in the qde-3 strain, but band patterns different from the qde-3 recQ2 strain appeared in the ad-3A mutants (data not shown). Then, the rest of the ad-3A mutations were analyzed by PCR and sequencing. One of 11 ad-3A mutants from the qde-3 strain and 2 of 51 ad-3A mutants from the qde-3 recQ2 strain gave a shorter fragment than expected using primers L1 and R1. In 2 other mutants, shorter fragments were amplified only in the PCR using primers L4 and R4 (Table 3). In the remaining 2 ad-3A mutants derived from the qde-3 strain and 4 ad-3A mutants derived from the qde-3 recQ2 strain, 1-kb fragments were amplified by PCR using the primers Pst and M2 (Figure 3A). From the qde-3 strain, sequence analysis revealed that one ad-3A mutant had a single-base substitution and the other a 10-bp deletion (Figure 4). From the qde-3 recQ2 strain, one ad-3A mutant had a single-base insertion and the other three had deletions of between 20 and 123 bp (Figure 4). No characteristic sequence was found at the junctions.
PCR analysis was also done on the ad-3B mutants. In the PCR using the primers X1 and Y1, more than half of the ad-3B mutants derived from the qde-3 strain or the qde-3 recQ2 strain did not yield any fragments (data not shown). In the ad-3B mutant from the recQ2 strain, 2-kb fragments were amplified in the PCR using the primers X1 and Y1 (data not shown).
The mutator phenotype of the qde-3 recQ2 double mutant was suppressed by mutation in NHEJ, but not in HR:
Plasmid-rejoining experiments in BS cells and WS cells showed that rejoining is error prone and mainly causes deletions (RUNGER and KRAEMER 1989; RUNGER et al. 1994; GAYMES et al. 2002; OSHIMA et al. 2002), suggesting that error-prone NHEJ causes deletion in these RecQ homolog mutants and thus may be responsible for the mutator phenotype of our qde-3 recQ2 double mutant.
In N. crassa, mus-51 and mus-52 genes, homologs of mammalian KU70 and KU80, respectively, are required for NHEJ (NINOMIYA et al. 2004). The spontaneous mutation rate of the mus-52 mutant at the ad-3 loci was 0.4 x 10–7, similar to that of the wild-type strain (Table 2). The qde-3 recQ2 mus-52 triple mutant also showed a rate similar to that of the wild-type strain, 0.2 x 10–7 (Table 2). The mutation rate of the qde-3 recQ2 mei-3 triple mutant was 9.1 x 10–7, indicating that mutation in mei-3 did not suppress the high spontaneous mutation rate and confirming that this suppression effect is due only to the mutation in the NHEJ gene. In addition, analysis of the genomic DNA of the ad-3A mutations derived in the qde-3 recQ2 mei-3 strain indicated that they carry large deletions (data not shown).
Mutation in NHEJ increases the severity of the growth defect of the qde-3 recQ2 double mutant:
Although the mus-52 mutation suppressed the rate of spontaneous deletions, the growth defect of the qde-3 recQ2 double mutant was more extreme in the qde-3 recQ2 mus-52 triple mutant. The colonies of the triple mutant were very thin and the morphology of each colony was irregular. Conidial viability was also reduced and varied from 17 to 34%. Moreover, linear growth of the triple mutant slowed earlier than that of the qde-3 recQ2 double mutant (Figure 2A). Although the triple mutant grew faster than the double mutant in the fourth tube, the growth rate, once it slowed, was not constant, alternating between a "slow" and "fast" growth pattern (data not shown). This phenotype is like the "stop-and-start" phenotype reported for some DNA repair mutants in N. crassa (NEWMEYER 1984). Although the triple mutant showed a severe growth defect, the abnormal appearance observed for the qde-3 recQ2 double mutant—accumulation of orange and black pigments and spherically formed mycelia—was not seen.
DISCUSSION
RecQ homologs in N. crassa have roles in the maintenance of genome stability as well as DNA repair:
In this report, we demonstrated that QDE3 and RECQ2 are required for maintenance of genomic stability as well as repair of DNA damage induced by mutagens. The qde-3 mutant and the qde-3 recQ2 double mutant show sensitivity to several mutagens, as reported previously (PICKFORD et al. 2003; KATO et al 2004), but are not sensitive to UV. However, the qde-3 mus-38 double mutant is more sensitive to UV than is the mus-38 single mutant. Furthermore, the qde-3 recQ2 mus-38 triple mutant is even more sensitive than either double mutant. As mus-38 is a homolog of S. cerevisiae RAD1, which is required for NER, the role of the Neurospora RecQ homologs in the repair of UV damage must involve pathways separate from that of NER, such as recombination repair, postreplication repair, and DNA damage checkpoint response.
The qde-3 mutant exhibited an increased spontaneous mutation rate, indicating that it is a mutator. The spontaneous mutation rate was much higher in the qde-3 recQ2 double mutant, suggesting that the qde-3 and recQ2 genes are required for the maintenance of genome stability. However, the spontaneous mutation rate in the recQ2 single mutant was similar to that in the wild-type strain. Moreover, the recQ2 mutant did not show significant mutagen sensitivity, although it had a slightly higher sensitivity to high concentrations of MMS, MNNG, and CPT than the wild-type strain did. The qde-3 recQ2 double mutant was more mutagen sensitive than the qde-3 single mutant. Thus, it seems that RECQ2 has a relatively minor role in cellular functions such as DNA repair and the maintenance of genome stability, but the function becomes conspicuous only when the qde-3 function is impaired. These two RecQ homologs have redundant functions in the repair of damaged DNA and also in the maintenance of genome stability, since the double mutant is more sensitive to mutagens and more mutable than each single mutant.
The growth defect of the qde-3 recQ2 mutant is caused by HR, not by NHEJ:
The qde-3 recQ2 double mutant shows low conidial viability, slow growth, and a decreased number of nuclei, indicating a decline of cellular proliferation. The growth defect of the qde-3 recQ2 double mutant is suppressed by a mutation in mei-3, which is required for HR, suggesting that the growth defect in the qde-3 recQ2 double mutant is a result of the HR pathway. In contrast, the qde-3 recQ2 mus-52 triple mutant showed a greater growth defect than the qde-3 recQ2 double mutant did, showing that NHEJ was not responsible for the defect.
HR and NHEJ are two major mechanisms of double-strand break (DSB) repair. HR repair, which depends on homologous sequences, rarely introduces mutations during DSB repair. NHEJ is an error-prone pathway that involves exonucleolytic processing and subsequent rejoining of the DNA ends without dependence on sequence homology. HR is initiated by single-stranded invasion of a homologous DNA sequence and is mediated by Rad51. Their suppression of the growth defect caused by the mei-3 mutation suggests that QDE3 and RECQ2 function after strand invasion. RecQ homologs in human and yeast can unwind Holliday-junction-like DNA structures (BENNETT et al. 1999; MOHAGHEGH et al. 2001). BLM also can melt D-loops, which are formed during HR by RAD51 (VAN BRABANT et al. 2000). Because RecQ proteins are highly conserved, it is likely that N. crassa QDE3 and RECQ2 behave in a similar manner. We speculate that RecQ proteins are involved in the resolution of recombination intermediates and that a lack of such resolution hinders the progress of DNA replication, resulting in the observed growth defect.
The mutator phenotype of the qde-3 recQ2 mutant is caused by NHEJ-dependent deletion formation and not by HR:
Mutation spectrum analysis of the ad-3A locus revealed that almost all the mutations that arose spontaneously in the qde-3 mutant and the qde-3 recQ2 double mutant were deletions. PCR analysis of the ad-3B locus also indicated that the ad-3B mutations obtained from the qde-3 and qde-3 recQ2 mutants were predominantly deletions, especially large deletions. The mutator phenotype of the double mutant was completely suppressed by mutation of mus-52, suggesting that NHEJ is responsible for this phenotype. In contrast, the mutator phenotype was not suppressed by mei-3 mutation, suggesting that HR is not involved. In addition, many large deletions were formed at the ad-3A locus in the qde-3 recQ2 mei-3 triple mutant, suggesting that HR is not required for the formation of deletions in the RecQ homolog mutants.
BS cells and WS cells also show a mutator phenotype and an increased number of deletion events (WARREN et al. 1981; FUKUCHI et al. 1989; TACHIBANA et al. 1996). RUNGER and KRAEMER (1989) and RUNGER et al. (1994) report that joining of linear plasmid DNA is error prone in these cells, and both of these findings offer indirect evidence that NHEJ causes deletion in these cells. Our report provides direct evidence that NHEJ-dependent deletion formation is responsible for the mutator phenotype in RecQ homolog mutants. A number of studies also indicate a relationship between RecQ helicases and NHEJ, which our findings support for the N. crassa RecQ homologs QDE3 and RECQ2. However, the nature of this relationship is unknown. RecQ homologs may control the rejoining efficiency of DSBs and/or the length of the exonucleolytic processing of the DNA ends. The former possibility is supported by the observation that overexpression of Ku70 in Drosophila melanogaster partially suppresses defects conferred by mutations in the RecQ homolog Dmblm (KUSANO et al. 2001). The latter hypothesis is supported by the observation of error-prone joining of linear plasmid DNA in human cells deficient in BLM or WRN (RUNGER and KRAEMER 1989; RUNGER et al. 1994). The characteristic large deletions may result from unregulated end processing by NHEJ machinery.
Abnormal appearance of the qde-3 recQ2 double mutant is dependent on both HR and NHEJ:
The abnormal appearance of the qde-3 recQ2 double mutant is suppressed by mutation in either mei-3 or mus-52, suggesting that this phenotype arises from a pathway involving both HR and NHEJ. Accumulation of orange and black pigments accompanied by the formation of spherical mycelia is a result of the uncontrolled expression of genes involved in conidiation, carotenoid synthesis, hyphal formation, and melanin production. As cancer cells proliferate in an uncontrolled manner, this susceptibility to abnormal appearance is reminiscent of the predisposition to cancer in BS patients. The mean age at which cancer is diagnosed in the BS-inherent population is 24.4 years (GERMAN 1993) and cancer does not occur in children in this population. The abnormal appearance of the qde-3 recQ2 double mutant is found only after growth has occurred for several days. It might be true in all organisms that genomic dysregulation proceeds at a slow tempo, mediated by HR and NHEJ functions, when RecQ homolog genes are mutated.
RecQ homologs may have a role in the suppression of spontaneous DSBs:
The characteristic deletions seen in RecQ mutants of N. crassa might be caused by unregulated NHEJ, as discussed above. However, the increase in spontaneous mutation rate may suggest an elevated frequency of DSBs in RecQ mutants. NHEJ is known to function in DSB repair in the wild-type strain, so it cannot be that loss of QDE3 and RECQ2 activates NHEJ to repair DSBs, resulting in the observed high mutation rate. Since NHEJ is error prone even in the wild-type strain, it is unlikely that loss of QDE3 and RECQ2 increases the tendency of NHEJ to be error prone. We speculate that QDE3 and RECQ2 play roles in the suppression of DSBs. If this is the case, what mechanism increases DSBs in the RecQ-deficient mutant?
The production of DSBs may be stimulated by HR dysfunction. If recombination intermediates accumulate in the RecQ-deficient mutant, physical tension or processing during mitosis may yield additional DSBs. However, since the HR-deficient qde-3 recQ2 mei-3 triple mutant retained the mutator phenotype, it is likely that DSBs arise by another mechanism. DSBs are generated during endonuclease-mediated resolution of Holliday junctions. BLM is thought to disrupt Holliday junctions formed at sites of blocked replication forks by its potential reverse branch migration activity (KAROW et al. 2000; WANG et al. 2000). Therefore, the increase in DSBs in the qde-3 recQ2 double mutant may be due to a decrease in the unraveling of Holliday junctions formed at sites of stalled replication forks and a resultant increase in endonuclease-mediated resolution of these junctions.
The high sensitivity of the qde-3 recQ2 double mutant to HU and CPT supports the idea that QDE3 and RECQ2 function in DNA replication, because replication is hindered by these agents. We propose a model (Figure 5) to explain the mechanism of the growth defect and genomic instability in the qde-3 recQ2 double mutant. If QDE3 and RECQ2 are nonfunctional, additional DSBs are formed by at the sites where replication forks are stalled. Then, proteins involved in HR or NHEJ will be recruited to the DSB sites for the repair of DSBs. HR events may not be completed normally without QDE3 and RECQ2 and thus recombination intermediates accumulate, leading to the inhibition of cell proliferation. In the absence of mei-3, proliferation-inhibiting DNA structures derived from HR will not be made and cells can grow normally, but DSBs must then be repaired by NHEJ and the extensive processing and subsequent ligation of the nonhomologous ends yield deletions. Conversely, in the qde-3 recQ2 mus-52 triple mutant in which NHEJ is impaired, all DSBs will be repaired by HR, so deletions will be less frequent but proliferation-inhibiting DNA structures made by HR will increase, thus amplifying the growth defect. At present, there is no direct evidence that DSBs are more frequent in the qde-3 recQ2 double mutant. Our current studies, focusing on DSB formation during DNA replication, are expected to shed light on the important question of how RecQ helicases function in the maintenance of genomic stability.
ACKNOWLEDGEMENTS
We thank Niji Ota for help in sequencing and Jane Yeadon for reviewing this manuscript. This work was supported by Rational Evolutionary Design of Advanced Biomolecules, Saitama Prefecture Collaboration of Regional Entities for the Advancement of Technological Excellence, Japan Science and Technology Agency.
LITERATURE CITED
AGUILERA, A., and H. L. KLEIN, 1988 Genetic control of intrachromosomal recombination in Saccharomyces cerevisiae. I. Isolation and genetic characterization of hyper-recombination mutations. Genetics 119: 779–790.
BENNETT, R. J., J. A. SHARP and J. C. WANG, 1998 Purification and characterization of the Sgs1 DNA helicase activity of Saccharomyces cerevisiae. J. Biol. Chem. 273: 9644–9650.
BENNETT, R. J., J. L. KECK and J. C. WANG, 1999 Binding specificity determines polarity of DNA unwinding by the Sgs1 protein of S. cerevisiae. J. Mol. Biol. 289: 235–248.
BROSH, R. M., JR., D. K. ORREN, J. O. NEHLIN, P. H. RAVN, M. K. KENNY et al., 1999 Functional and physical interaction between WRN helicase and human replication protein A. J. Biol. Chem. 274: 18341–18350.
BROSH, R. M., JR., C. VON KOBBE, J. A. SOMMERS, P. KARMAKAR, P. L. OPRESKO et al., 2001 Werner syndrome protein interacts with human flap endonuclease 1 and stimulates its cleavage activity. EMBO J. 20: 5791–5801.
CHAGANTI, R. S., S. SCHONBERG and J. GERMAN, 1974 A manyfold increase in sister chromatid exchanges in Bloom's syndrome lymphocytes. Proc. Natl. Acad. Sci. USA 71: 4508–4512.
COGONI, C., and G. MACINO, 1999 Posttranscriptional gene silencing in Neurospora by a RecQ DNA helicase. Science 286: 2342–2344.
CONSTANTINOU, A., M. TARSOUNAS, J. K. KAROW, R. M. BROSH, V. A. BOHR et al., 2000 Werner's syndrome protein (WRN) migrates Holliday junctions and co-localizes with RPA upon replication arrest. EMBO Rep. 1: 80–84.
DAVIS, R. H., and F. J. DE SERRES, 1970 Genetic and microbiological research techniques for Neurospora crassa. Methods Enzymol. 17: 79–143.
DE SERRES, F. J., and H. G. KLMARK, 1958 A direct method for determination of forward-mutation rates in Neurospora crassa. Nature 182: 1249–1250.
DOE, C. L., J. DIXON, F. OSMAN and M. C. WHITBY, 2000 Partial suppression of the fission yeast rqh1(-) phenotype by expression of a bacterial Holliday junction resolvase. EMBO J. 19: 2751–2762.
DUTERTRE, S., M. ABABOU, R. ONCLERCQ, J. DELIC, B. CHATTON et al., 2000 Cell cycle regulation of the endogenous wild-type Bloom's syndrome DNA helicase. Oncogene 19: 2731–2738.
ELLIS, N. A., J. GRODEN, T. Z. YE, J. STRAUGHEN, D. J. LENNON et al., 1995 The Bloom's syndrome gene product is homologous to RecQ helicases. Cell 83: 655–666.
EPSTEIN, C. J., G. M. MARTIN, A. L. SCHULTZ and A. G. MOTULSKY, 1966 Werner's syndrome: a review of its symptomatology, natural history, pathologic features, genetics and relationship to the natural aging process. Medicine 45: 177–221.
FREI, C., and S. M. GASSER, 2000 The yeast Sgs1p helicase acts upstream of Rad53p in the DNA replication checkpoint and colocalizes with Rad53p in S-phase-specific foci. Genes Dev. 14: 81–96.
FUKUCHI, K., G. M. MARTIN and R. J. MONNAT, JR., 1989 Mutator phenotype of Werner syndrome is characterized by extensive deletions. Proc. Natl. Acad. Sci. USA 86: 5893–5897.
GANGLOFF, S., C. SOUSTELLE and F. FABRE, 2000 Homologous recombination is responsible for cell death in the absence of the Sgs1 and Srs2 helicases. Nat. Genet. 25: 192–194.
GAYMES, T. J., P. S. NORTH, N. BRADY, I. D. HICKSON, G. J. MUFTI et al., 2002 Increased error-prone non homologous DNA end-joining—a proposed mechanism of chromosomal instability in Bloom's syndrome. Oncogene 21: 2525–2533.
GERMAN, J., 1993 Bloom syndrome: a Mendelian prototype of somatic mutational disease. Medicine 72: 393–406.
INOUE, H., and C. ISHII, 1984 Isolation and characterization of MMS-sensitive mutants of Neurospora crassa. Mutat. Res. 125: 185–194.
ISHII, C., and H. INOUE, 1994 Mutagenesis and epistatic grouping of the Neurospora meiotic mutants, mei-2 and mei-3, which are sensitive to mutagens. Mutat. Res. 315: 249–259.
ISHII, C., K. NAKAMURA and H. INOUE, 1998 A new UV-sensitive mutant that suggests a second excision repair pathway in Neurospora crassa. Mutat. Res. 408: 171–182.
JIAO, R., C. Z. BACHRATI, G. PEDRAZZI, P. KUSTER, M. PETKOVIC et al., 2004 Physical and functional interaction between the Bloom's syndrome gene product and the largest subunit of chromatin assembly factor 1. Mol. Cell. Biol. 24: 4710–4719.
KAMATH-LOEB, A. S., E. JOHANSSON, P. M. BURGERS and L. A. LOEB, 2000 Functional interaction between the Werner Syndrome protein and DNA polymerase delta. Proc. Natl. Acad. Sci. USA 97: 4603–4608.
KAROW, J. K., A. CONSTANTINOU, J. L. LI, S. C. WEST and I. D. HICKSON, 2000 The Bloom's syndrome gene product promotes branch migration of Holliday junctions. Proc. Natl. Acad. Sci. USA 97: 6504–6508.
KATO, A., Y. AKAMATSU, Y. SAKURABA and H. INOUE, 2004 The Neurospora crassa mus-19 gene is identical to the qde-3 gene, which encodes a RecQ homologue and is involved in recombination repair and postreplication repair. Curr. Genet. 45: 37–44.
KAWABE, T., N. TSUYAMA, S. KITAO, K. NISHIKAWA, A. SHIMAMOTO et al., 2000 Differential regulation of human RecQ family helicases in cell transformation and cell cycle. Oncogene 19: 4764–4772.
KITAO, S., I. OHSUGI, K. ICHIKAWA, M. GOTO, Y. FURUICHI et al., 1998 Cloning of two new human helicase genes of the RecQ family: biological significance of multiple species in higher eukaryotes. Genomics 54: 443–452.
KITAO, S., A. SHIMAMOTO, M. GOTO, R. W. MILLER, W. A. SMITHSON et al., 1999 Mutations in RECQL4 cause a subset of cases of Rothmund-Thomson syndrome. Nat. Genet. 22: 82–84.
KUSANO, K., D. M. JOHNSON-SCHLITZ and W. R. ENGELS, 2001 Sterility of Drosophila with mutations in the Bloom syndrome gene: complementation by Ku70. Science 291: 2600–2602.
LEA, D. E., and C. A. COULSON, 1949 The distribution of the numbers of mutants in bacterial populations. J. Genet. 49: 264–285.
LEBEL, M., E. A. SPILLARE, C. C. HARRIS and P. LEDER, 1999 The Werner syndrome gene product co-purifies with the DNA replication complex and interacts with PCNA and topoisomerase I. J. Biol. Chem. 274: 37795–37799.
LEE, S. K., R. E. JOHNSON, S. L. YU, L. PRAKASH and S. PRAKASH, 1999 Requirement of yeast SGS1 and SRS2 genes for replication and transcription. Science 286: 2339–2342.
LONN, U., S. LONN, U. NYLEN, G. WINBLAD and J. GERMAN, 1990 An abnormal profile of DNA replication intermediates in Bloom's syndrome. Cancer Res. 50: 3141–3145.
MAFTAHI, M., J. C. HOPE, L. DELGADO-CRUZATA, C. S. HAN and G. A. FREYER, 2002 The severe slow growth of srs2 rqh1 in Schizosaccharomyces pombe is suppressed by loss of recombination and checkpoint genes. Nucleic Acids Res. 30: 4781–4792.
MANKOURI, H. W., T. J. CRAIG and A. MORGAN, 2002 SGS1 is a multicopy suppressor of srs2: functional overlap between DNA helicases. Nucleic Acids Res. 30: 1103–1113.
MOHAGHEGH, P., J. K. KAROW, R. M. BROSH, JR, V. A. BOHR and I. D. HICKSON, 2001 The Bloom's and Werner's syndrome proteins are DNA structure-specific helicases. Nucleic Acids Res. 29: 2843–2849.
MYUNG, K., A. DATTA, C. CHEN and R. D. KOLODNER, 2001 SGS1, the Saccharomyces cerevisiae homologue of BLM and WRN, suppresses genome instability and homeologous recombination. Nat. Genet. 27: 113–116.
NEWMEYER, D., 1984 Neurospora mutants sensitive both to mutagens and to histidine. Curr. Genet. 9: 65–74.
NINOMIYA, Y., K. SUZUKI, C. ISHII and H. INOUE, 2004 Highly efficient gene replacements in Neurospora strains deficient for nonhomologous end-joining. Proc. Natl. Acad. Sci. USA 101: 12248–12253.
ONODA, F., M. SEKI, A. MIYAJIMA and T. ENOMOTO, 2000 Elevation of sister chromatid exchange in Saccharomyces cerevisiae sgs1 disruptants and the relevance of the disruptants as a system to evaluate mutations in Bloom's syndrome gene. Mutat. Res. 459: 203–209.
OSHIMA, J., S. HUANG, C. PAE, J. CAMPISI and R. H. SCHIESTL, 2002 Lack of WRN results in extensive deletion at nonhomologous joining ends. Cancer Res. 62: 547–551.
PICKFORD, A., L. BRACCINI, G. MACINO and C. COGONI, 2003 The QDE-3 homologue RecQ-2 co-operates with QDE-3 in DNA repair in Neurospora crassa. Curr. Genet. 42: 220–227.
POOT, M., H. HOEHN, T. M. RUNGER and G. M. MARTIN, 1992 Impaired S-phase transit of Werner syndrome cells expressed in lymphoblastoid cell lines. Exp. Cell Res. 202: 267–273.
PURANAM, K. L., and P. J. BLACKSHEAR, 1994 Cloning and characterization of RECQL, a potential human homologue of the Escherichia coli DNA helicase RecQ. J. Biol. Chem. 269: 29838–29845.
RONG, L., and H. L. KLEIN, 1993 Purification and characterization of the SRS2 DNA helicase of the yeast Saccharomyces cerevisiae. J. Biol. Chem. 268: 1252–1259.
RUNGER, T. M., and K. H. KRAEMER, 1989 Joining of linear plasmid DNA is reduced and error-prone in Bloom's syndrome cells. EMBO J. 8: 1419–1425.
RUNGER, T. M., C. BAUER, B. DEKANT, K. MOLLER, P. SOBOTTA et al., 1994 Hypermutable ligation of plasmid DNA ends in cells from patients with Werner syndrome. J. Invest. Dermatol. 102: 45–48.
SALK, D., K. AU, H. HOEHN and G. M. MARTIN, 1981 Cytogenetics of Werner's syndrome cultured skin fibroblasts: variegated translocation mosaicism. Cytogenet. Cell Genet. 30: 92–107.
SEKI, M., H. MIYAZAWA, S. TADA, J. YANAGISAWA, T. YAMAOKA et al., 1994 Molecular cloning of cDNA encoding human DNA helicase Q1 which has homology to Escherichia coli Rec Q helicase and localization of the gene at chromosome 12p12. Nucleic Acids Res. 22: 4566–4573.
SINCLAIR, D. A., K. MILLS and L. GUARENTE, 1997 Accelerated aging and nucleolar fragmentation in yeast sgs1 mutants. Science 277: 1313–1316.
STABEN, C., B. JENSEN, M. SINGER, J. POLLOCK, M. SCHECHTMAN et al., 1989 Use of a bacterial hygromycin B resistance gene as a dominant selectable marker in Neurospora crassa transformation. Fungal Genet. Newsl. 36: 79–81.
TACHIBANA, A., K. TATSUMI, T. MASUI and T. KATO, 1996 Large deletions at the HPRT locus associated with the mutator phenotype in a Bloom's syndrome lymphoblastoid cell line. Mol. Carcinogen. 17: 41–47.
TAMARU, H., and H. INOUE, 1989 Isolation and characterization of a laccase-derepressed mutant of Neurospora crassa. J. Bacteriol. 171: 6288–6293.
VAN BRABANT, A. J., T. YE, M. SANZ, I. J. GERMAN, N. A. ELLIS et al., 2000 Binding and melting of D-loops by the Bloom syndrome helicase. Biochemistry 39: 14617–14625.
VENNOS, E. M., and W. D. JAMES, 1995 Rothmund-Thomson syndrome. Dermatol. Clin. 13: 143–150.
WANG, W., M. SEKI, Y. NARITA, E. SONODA, S. TAKEDA et al., 2000 Possible association of BLM in decreasing DNA double strand breaks during DNA replication. EMBO J. 19: 3428–3435.
WANG, W., M. SEKI, Y. NARITA, T. NAKAGAWA, A. YOSHIMURA et al., 2003 Functional relation among RecQ family helicases RecQL1, RecQL5, and BLM in cell growth and sister chromatid exchange formation. Mol. Cell. Biol. 23: 3527–3535.
WARREN, S. T., R. A. SCHULTZ, C. C. CHANG, M. H. WADE and J. E. TROSKO, 1981 Elevated spontaneous mutation rate in Bloom syndrome fibroblasts. Proc. Natl. Acad. Sci. USA 78: 3133–3137.
WATANABE, K., Y. SAKURABA and H. INOUE, 1997 Genetic and molecular characterization of Neurospora crassa mus-23: a gene involved in recombinational repair. Mol. Gen. Genet. 256: 436–445.
WATT, P. M., E. J. LOUIS, R. H. BORTS and I. D. HICKSON, 1995 Sgs1: a eukaryotic homolog of E. coli RecQ that interacts with topoisomerase II in vivo and is required for faithful chromosome segregation. Cell 81: 253–260.
WATT, P. M., I. D. HICKSON, R. H. BORTS and E. J. LOUIS, 1996 SGS1, a homologue of the Bloom's and Werner's syndrome genes, is required for maintenance of genome stability in Saccharomyces cerevisiae. Genetics 144: 935–945.
YAMAGATA, K., J. KATO, A. SHIMAMOTO, M. GOTO, Y. FURUICHI et al., 1998 Bloom's and Werner's syndrome genes suppress hyperrecombination in yeast sgs1 mutant: implication for genomic instability in human diseases. Proc. Natl. Acad. Sci. USA 95: 8733–8738.
YU, C. E., J. OSHIMA, Y. H. FU, E. M. WIJSMAN, F. HISAMA et al., 1996 Positional cloning of the Werner's syndrome gene. Science 272: 258–262.
Laboratory of Genetics, Department of Regulation Biology, Faculty of Science, Saitama University, 338-8570 Saitama, Japan(Akihiro Kato and Hirokazu)
THE RecQ family of DNA helicases has been highly conserved during evolution and is present in organisms ranging from bacteria to humans. Members of the RecQ family have been shown to be important for the maintenance of genomic stability. Mutation in three of the five human RecQ homologs, BLM, WRN, and RecQL4, results in Bloom syndrome (BS), Werner syndrome (WS), and Rothmund-Thomson syndrome (RTS), respectively (ELLIS et al. 1995; YU et al. 1996; KITAO et al. 1998, 1999). BS is characterized by growth retardation, a markedly increased incidence of several types of cancer and genomic instability, including chromatid gaps, breaks, and rearrangements (GERMAN 1993). In addition, BS cells exhibit elevated sister-chromatid exchange (CHAGANTI et al. 1974). WS is characterized by genomic instability and the premature appearance of aging phenotypes in young adults, including cancers (EPSTEIN et al. 1966). Cells derived from WS patients show an increased rate of somatic mutations, chromosome loss, and deletions (SALK et al. 1981; FUKUCHI et al. 1989). RTS is also characterized by premature aging and cancer predisposition (VENNOS and JAMES 1995). These three diseases are clinically distinct and the three causative genes are thought to play distinct roles. The genomic instability found in these diseases correlates with cancer predisposition but there is no direct evidence that the genome instability is responsible for the cancer susceptibility. The remaining two human RecQ homologs, RecQL1 (PURANAM and BLACKSHEAR 1994; SEKI et al. 1994) and RecQL5 (KITAO et al. 1998), have not been implicated in any disease. However, in chicken DT40 cells, in which the BLM function is impaired, RecQL1 and RecQL5 appear to be required for cell viability (WANG et al. 2003).
In Saccharomyces cerevisiae, there is only one RecQ homolog, Sgs1. sgs1 mutants show an increase in various types of recombination (WATT et al. 1996; YAMAGATA et al. 1998; ONODA et al. 2000), accumulate gross chromosomal rearrangements (MYUNG et al. 2001), and show impaired sporulation (WATT et al. 1995) and premature aging of mother cells (SINCLAIR et al. 1997).
Much evidence supports the involvement of RecQ helicases in DNA replication. The expression of Sgs1, BLM, WRN, and RecQL4 is cell cycle regulated and peaks during S-phase (KITAO et al. 1998; DUTERTRE et al. 2000; FREI and GASSER 2000; KAWABE et al. 2000). Correspondingly, WRN interacts physically with components required for replication, including DNA polymerase , topoisomerase I, PCNA, FEN-1, and RPA (BROSH et al. 1999, 2001; LEBEL et al. 1999; KAMATH-LOEB et al. 2000). In addition, BLM and WRN localize to stalled replication sites, resulting from cellular exposure to hydroxyurea (HU) (CONSTANTINOU et al. 2000; JIAO et al. 2004). In S. cerevisiae, some Sgs1 foci also colocalize with replication foci (FREI and GASSER 2000). Human cells lacking functional BLM or WRN accumulate aberrant replication intermediates (LONN et al. 1990; POOT et al. 1992). Recent studies indicate that RecQ helicases are required for processing of DNA structures induced by stalled replication forks. It is thought that lesions on the leading strand can cause replication forks to regress, resulting in Holliday-junction-like four-way structures. Holliday junctions are the preferred substrate of BLM, WRN, and Sgs1 (BENNETT et al. 1999; MOHAGHEGH et al. 2001). BLM and WRN can also catalyze branch migration of Holliday junctions (CONSTANTINOU et al. 2000; KAROW et al. 2000). Moreover, defects in Schizosaccharomyces pombe rqh1, the only RecQ homolog in S. pombe, are suppressed by overexpression of a bacterial Holliday junction resolvase, RusA (DOE et al. 2000). However, how replication regression is promoted and how RecQ helicases maintain genomic stability remains to be explained.
In Neurospora crassa, there are two RecQ homologs, QDE3 and RECQ2 (COGONI and MACINO 1999; PICKFORD et al. 2003; KATO et al. 2004). QDE3 is a long-type RecQ homolog, like BLM, WRN, and RTS, while RECQ2 is a short-type homolog, like RecQL1 and RecQL5. It has been reported that QDE3 and RECQ2 are required for DNA repair (PICKFORD et al. 2003; KATO et al. 2004) but it was not previously known whether they are required for genome stability. In this report, we describe the effects of mutations in QDE3 and RECQ2 and their interaction with mutations in homologous recombination (HR) and nonhomologous end joining (NHEJ).
MATERIALS AND METHODS
Strains:
The N. crassa strains used in this study are listed in Table 1. C1-T10-37A and C1-T10-28a (TAMARU and INOUE 1989) are wild-type strains closely related to the standard Oak Ridge wild-type strain. Production of the recQ2 mutant strains KTO-Q2R-5A and KTO-Q2-2A is described in the next section. The qde-3 recQ2 double-mutant strains KTO-Qd-32a and KTO-Qd-35A were derived from a cross between the qde-3 mutant strain KTO-r-20a (KATO et al. 2004) and KTO-Q2-2A. The mus-38 recQ2 double-mutant strain KTO-2m38-1a was derived from a cross between the mus-38 mutant strain CZ272-5a (ISHII et al. 1998) and KTO-Q2-2A. The qde-3 recQ2 mus-38 triple-mutant strain KTO-dm38-26A was derived from a cross between the qde-3 mus-38 double-mutant strain QMU38-2A (KATO et al. 2004) and KTO-2m38-1a. The recQ2 mei-3 double-mutant strain KTO-2mi3-1a was derived from a cross between the mei-3 mutant strain CY-10-9a (ISHII and INOUE 1994) and KTO-Q2-2A. The qde-3 recQ2 mei-3 triple-mutant strain KTO-dmi3-1A was derived from a cross between KTO-Qd-35A and KTO-2mi3-1a. The qde-3 recQ2 mus-52 triple-mutant strain KTO-dk8-33a was derived from a cross between KTO-Qd-32a and the mus-52 mutant strain 54yo-828-3 (NINOMIYA et al. 2004).
Construction of the recQ2 mutant:
A 1908-bp fragment upstream of the recQ2 ORF, beginning 5 bp upstream from the ORF, and a 1836-bp fragment downstream from the recQ2 ORF, beginning 165 bp downstream from the ORF, were amplified by PCR using the following primer sets:
Q2U-1: TTGGGATGATCGAAGAG and Q2U-2: GCAGCGTCGATAGCA
Q2D-1: TTCGGTGACAGGTAGGT and Q2D-2: GTCGTTCTCTGCCTTAG.
These two fragments were inserted on each side of the hygromycin-resistance marker gene (hygr) derived from the plasmid pCSN43 (STABEN et al. 1989) to construct the plasmid pQ2::HYG. This plasmid was used as a PCR template to produce the targeting construct hygr flanked by sequences upstream and downstream of recQ2. The amplified fragment was introduced into the C1-T10-37A wild-type strain by electroporation to replace the endogenous recQ2 with the hygr gene. Hygromycin-B-resistant transformants were isolated and the replacement was confirmed by PCR using the primer Q2I: AACAACAGGCGCGACCAA, which is located in the upstream fragment, and Q2O: AAGCCCGTAGAGTGCAGACAAA, which is located downstream of the downstream fragment. In the wild-type case, a 4-kb fragment was amplified. Since 1.8 kb, including the whole recQ2 ORF, is replaced by the 2.4-kb hygr gene in the recQ2-replaced mutant, a 4.5-kb fragment was amplified. Extra ectopic copies were ruled out and gene replacement was confirmed by Southern analysis. This mutant strain, designated as KTO-Q2R-5A, was backcrossed to the C1-T10-28a wild-type strain, and it was confirmed that the hygr marker gene segregates normally through meiosis. Strain KTO-Q2-2A is one of the progeny of this cross and was used as a standard recQ2 mutant strain.
Growth media and genetic methods:
Growth media and genetic procedures for N. crassa were as described by DAVIS and DE SERRES (1970). Transformation of Neurospora by electroporation was performed as previously described (NINOMIYA et al. 2004).
Qualitative assay of mutagen sensitivity by spot tests:
Sensitivity to chemical mutagens and other chemicals was analyzed by spot tests, as described by WATANABE et al. (1997). Each conidial suspension was diluted and adjusted to densities of 10,000, 1000, 100, and 10 viable conidia per spot. Methyl methanesulfonate (MMS), N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), camptothecin (CPT), tert-butyl hydroperoxide (TBHP), ethyl methanesulfonate (EMS), 4-nitroquinoline-1-oxide (4NQO), and HU were added to agar medium at final concentrations of 0.005 or 0.02%, 0.15 or 0.45 μg/ml, 0.0025 or 0.25 μg/ml, 0.0014 or 0.0028%, 0.1 or 0.2%, 0.0075 or 0.045 μg/ml, and 0.38 or 1.9 mg/ml, respectively. Conidial suspensions were spotted onto these plates and grown at 30° for 2 days. UV sensitivity was investigated by spotting a conidial suspension onto an agar plate and irradiating at 150 or 300 J/m2.
Assay of UV sensitivity:
The UV dose dependency of the survival of N. crassa was measured as described by INOUE and ISHII (1984). Conidial suspensions at a final concentration of 1 x 106 cells/ml were irradiated at various doses of UV, and aliquots were sampled and plated after appropriate dilution. The yeast on all the plates were allowed to grow at 30° for 3 days and the number of colonies on each plate was counted.
Measurement of apical growth and life span:
Apical growth of hyphae was measured in race tubes of 30 cm length. So that the age of each strain would be consistent, mutant strains were backcrossed to the wild-type strain and the resultant ascospores were randomly isolated and grown at 30° for 1 week. Conidia of each were transferred to new media and grown at 30° for another week while the genotype of each progeny was investigated by the spot test or PCR. A silica gel stock was made from the conidia of each strain and stored at 4°. To start each experiment, a few grains of the silica gel were placed on Vogel's minimal agar medium and incubated at 30° for 1 week. The resulting conidia were inoculated at one end of the race tube and incubated at 25°. The position of the growing front was marked once or twice a day. When the mycelia reached the other end of the race tube, a piece of medium bearing mycelia was cut from the growing front and transferred to a fresh tube.
Microscopy:
Mycelia were grown on Vogel's minimal agar medium containing 1.2% sucrose and 2% agar at 30° for 1 day. A piece of mycelium was cut from the growing front, fixed in 70% ethanol, and transferred to a glass slide. SYBR Gold (Molecular Probes, Eugene, OR) dye was used to stain nuclei. Mycelia were observed by fluorescence microscopy (Nikon) and images were taken with a cooled CCD color camera (KEYENCE).
Forward mutation assay:
The direct method described by DE SERRES and KOLMARK (1958) was used for the detection of ad-3 mutations. To eliminate preexisting ad-3 mutation, cultures started by single-colony isolation were used for each experiment. A conidial suspension from each isolate was inoculated into each 5-liter Florence flask at a concentration of 5 x 105 viable conidia per flask. The flask was cultured in the dark at 30° with low aeration. Under these conditions, each conidium makes a bead-like colony after 1 week of culture. As ad-3 mutants accumulated purple pigments in mycelia, purple-colored colonies were isolated and subcultured. To distinguish ad-3A from ad-3B mutants, complementation tests were carried out using ad-3A and ad-3B tester strains (Table 1). Spontaneous mutation rates per nucleus per division were calculated by the method of the median (LEA and COULSON 1949) using the following formula: mutation rate = m/N, where m is the calculated number of mutation events and N is the number of nuclear divisions. To estimate the number of nuclear divisions, the number of nuclei in 50 conidia was counted and the average number of nuclei in a conidium was multiplied by the estimated number of colonies in the culture. The formula r/m – ln(m) – 1.24 = 0, where r is the median number of ad-3 mutants, was solved for m.
Analysis of mutations:
To amplify the ad-3A gene from genomic DNA, PCR was performed with two primers: Pst, AACTGCAGTTCTGATCCGTCCAGTTC, and M2, TTCGTTACAGTGCCGAGTCC. To amplify the ad-3B gene from genomic DNA, PCR was performed with two primers: X1, TTGGGTCGTATGCTCTGTGA, and Y1, CTTGGGCCACTGTCCGT.
To determine the nucleotide sequence of the ad-3A gene, six sequencing primers were used (BigDye Terminator Cycle Sequencing Ready Reaction kit v3.1, Applied Biosystems, Foster City, CA): 2F, GTCCAGTTCCGTCCCGTCCA; 3F, TTTACTACCGACCGCATTTCGG; 5F, CGAGCACCAAGGCAGAGCTG; 4R, GCAGTAAGCTGTGTG AGGAT; 6R, CAGTGGAGTAGACCTTGAGC; and 8R, CCAGCAAACTATCAG ATCAC.
To amplify regions outside of the ad-3A ORF, six primers were used: L1, CAAATGGAAGACGGTGGACT; R1, TCGCAGACTGTGAGGAATTG; L2, GATTGACGCCCTTTTGATGT; R2, GAACTGGAGGCTGGAGTGAG; L4, CGGCGGTCGTTGAGGTAG; and R4, CAGCGAGTATTTGGTGGAAGG.
Analysis of chromosome deletion around the ad-3A gene was performed by Southern hybridization using cosmid G4:E3 as a probe.
RESULTS
The qde-3 recQ2 double mutant showed increased sensitivity to all tested mutagens:
The recQ2 mutant was generated by replacing the whole recQ2 ORF by hygr, as described in MATERIALS AND METHODS. The mutant did not show any defects during vegetative growth and sporulation. The sensitivity of the recQ2 mutant to a variety of mutagens and other chemicals was tested by spot tests (Figure 1). Conidial suspensions adjusted to the same concentrations were spotted on plates containing MMS, MNNG, EMS, TBHP, 4NQO, CPT, or HU, and plates were incubated at 30°. In the UV-sensitivity test, conidia were irradiated after spotting.
While the qde-3 mutant showed high sensitivity to mutagens, as described previously (KATO et al. 2004; Figure 1A), the recQ2 mutant was slightly more sensitive to MMS, MNNG, and CPT than the wild-type strain, the increased sensitivity being detectable only at high concentrations of the DNA-damaging agents (Figure 1B). The qde-3 recQ2 double mutant showed higher sensitivity to the type I topoisomerase inhibitor CPT, the replication inhibitor HU, and to all mutagens tested than the qde-3 single mutant did (Figure 1B). Unlike the sgs1 mutant in S. cerevisiae and the rqh1 mutant in S. pombe, which are sensitive to UV, the UV sensitivity of the qde-3 recQ2 double mutant was similar to that of the wild-type strain (Figure 1, B and C).
To investigate whether QDE3 and RECQ2 function in the resistance to UV, a qde-3 mus-38 double mutant and a qde-3 recQ2 mus-38 triple mutant were made and their UV sensitivity was tested. mus-38 mutants are deficient in nucleotide excision repair (NER) and sensitive to UV. The qde-3 mus-38 double mutant showed a higher sensitivity to UV than the parental mus-38 mutant, and the qde-3 recQ2 mus-38 triple mutant was still more sensitive (Figure 1D). These data indicate that the qde-3 and recQ2 genes are required for resistance to UV in the case of dysfunction of MUS38.
The qde-3 recQ2 double mutant showed a growth defect:
Growth of the qde-3 recQ2 double mutant was poor on glycerol complete medium as well as on minimal medium. About one-fourth of the random progeny of a backcross of the double mutant to the wild-type strain showed slow growth, and all were qde-3 recQ2 double mutants (data not shown). Both qde-3 and recQ2 single mutants grew as fast as the wild-type strain for several consecutive transfers (Figure 2A and data not shown). The double mutant grew as fast as the wild-type strain in the first and second race tubes (Figure 2A) but its growth slowed in the third tube and was dramatically reduced in the fourth tube (Figure 2A). The appearance of mycelia was also altered in the fourth tube, with the accumulation of black and dense orange pigments and spherically formed mycelia (Figure 2B).
Microscopic observation showed that the size and morphology of hyphal tips in the wild-type strain, the qde-3 mutant, and the recQ2 mutant were in all cases similar (Figure 2C). As N. crassa is a coenocytic organism, many nuclei were found in mycelia of all those strains. Hyphal tips of the qde-3 recQ2 double mutant appeared similar when the cells were young and growing vigorously, but in the aged double mutant, whose growth was slow, mycelial tips were thin and the number of nuclei decreased (Figure 2C). Along with the deterioration in linear growth, the colony morphology of the double mutant became flat and spreading, whereas that of the wild-type strain was thick and dense (Figure 2D).
Conidia in the wild-type, the qde-3 mutant, and the recQ2 mutant strains had similar viability (77, 71, and 74%, respectively) but viability was much reduced (to 33%) in the double mutant.
The growth defect of the qde-3 recQ2 double mutant was suppressed by mutation of a gene required for HR:
In S. cerevisiae, the role of another DNA helicase, Srs2, appears to overlap with that of Sgs1. Both Sgs1 and Srs2 have 3'-5' DNA helicase activity (RONG and KLEIN 1993; BENNETT et al. 1998) and both sgs1 and srs2 mutants have a hyperrecombination phenotype (AGUILERA and KLEIN 1988; WATT et al. 1996). Simultaneous deletion of SGS1 and SRS2 results in extremely poor growth or synthetic lethality (LEE et al. 1999; GANGLOFF et al. 2000), with either phenotype alleviated by mutations in the HR genes RAD51, RAD55, and RAD57 (GANGLOFF et al. 2000). A similar result was reported in S. pombe (MAFTAHI et al. 2002). Moreover, overexpression of Sgs1 partially suppresses the MMS and HU sensitivity of the srs2 mutant (MANKOURI et al. 2002).
Similarly, qde-3 and recQ2 are putative DNA helicases with redundant roles in DNA repair in N. crassa. Therefore, the growth defect in the qde-3 recQ2 mutant might be suppressed by mutations in HR genes, for example, in the RAD51 homolog mei-3. The qde-3 recQ2 mei-3 triple mutant grew as fast as the wild-type strain during the entire course of the transfer experiment (Figure 2A). Unlike in the qde-3 recQ2 double mutant, mycelial appearance and growth rate were normal, even after six transfers (data not shown). Colony morphology of the triple mutant was similar to that of the mei-3 mutant (Figure 2D). Conidial viability of the triple mutant and the mei-3 mutant was the same (64 and 61%, respectively).
qde-3 and qde-3 recQ2 double mutants show increased mutability and extensive deletion:
To test the genomic instability of RecQ-homolog mutants of N. crassa, spontaneous mutability was investigated. The spontaneous mutation rate at the ad-3 locus was 0.4 x 10–7 in the recQ2 mutant, which was similar to that in the wild-type strain (Table 2). However, in the qde-3 mutant, the spontaneous mutation rate was 2.4 x 10–7, 11 times higher than that in the wild-type strain. The spontaneous mutation rate of the qde-3 recQ2 double mutant was 4.0 x 10–7, 18 times higher than that of the wild-type strain.
The ad-3 locus consists of two genes, ad-3A and ad-3B, and which gene is mutated in a particular ad-3 mutant can be determined by complementation tests. We wanted to examine the spectrum of mutations generated in qde-3, recQ2, and qde-3 recQ2 double mutants. As the ad-3A gene is smaller than the ad-3B gene, we first determined the sequence of the ad-3A gene.
To determine the sequence, the ad-3A ORF was amplified by PCR using primers Pst and M2 (Figure 3A). The expected 1-kb fragments were amplified using genomic DNA of the wild type, the qde-3 mutant, and the qde-3 recQ2 double mutant as a template, but no fragments were amplified from most of the ad-3A mutants obtained from the qde-3 mutant and the qde-3 recQ2 double mutant (Table 3). Since the ad-3B gene was successfully amplified using primers X1 and Y1 (Figure 3B) in all of these mutants, it was thought that these mutants had experienced deletion in ad-3A. Primers L1 and R1 yield a 3-kb fragment that includes the ad-3A ORF and 1 kb of flanking sequence on each side of the ORF, while primers L2 and R2 give a similar 5-kb fragment that includes 2 kb of each flanking region, and primers L4 and R4 give a 7-kb fragment that includes 3 kb of each flanking region (Figure 3A).
Genomic DNA of the wild-type, the qde-3 mutant, or the qde-3 recQ2 mutant strains yielded the expected fragments with each of these primer pairs (Table 3, "controls"). On the other hand, genomic DNAs of the ad-3A mutants derived from the qde-3 or qde-3 recQ2 strains yielded no fragments in these PCR analyses, suggesting that they have large deletions. In 8 of 11 (72%) ad-3A mutants obtained from the qde-3 mutant strain and 43 of 51 (84%) ad-3A mutants obtained from the qde-3 recQ2 double-mutant strain, no fragments were amplified, even when primers L4 and R4 were used (Table 3). Southern analysis revealed that these mutants all have deletions at the ad-3A locus (data not shown).
The resulting band patterns were the same in all the ad-3A mutants obtained from the same subculture (Jug 9) in the qde-3 strain, but band patterns different from the qde-3 recQ2 strain appeared in the ad-3A mutants (data not shown). Then, the rest of the ad-3A mutations were analyzed by PCR and sequencing. One of 11 ad-3A mutants from the qde-3 strain and 2 of 51 ad-3A mutants from the qde-3 recQ2 strain gave a shorter fragment than expected using primers L1 and R1. In 2 other mutants, shorter fragments were amplified only in the PCR using primers L4 and R4 (Table 3). In the remaining 2 ad-3A mutants derived from the qde-3 strain and 4 ad-3A mutants derived from the qde-3 recQ2 strain, 1-kb fragments were amplified by PCR using the primers Pst and M2 (Figure 3A). From the qde-3 strain, sequence analysis revealed that one ad-3A mutant had a single-base substitution and the other a 10-bp deletion (Figure 4). From the qde-3 recQ2 strain, one ad-3A mutant had a single-base insertion and the other three had deletions of between 20 and 123 bp (Figure 4). No characteristic sequence was found at the junctions.
PCR analysis was also done on the ad-3B mutants. In the PCR using the primers X1 and Y1, more than half of the ad-3B mutants derived from the qde-3 strain or the qde-3 recQ2 strain did not yield any fragments (data not shown). In the ad-3B mutant from the recQ2 strain, 2-kb fragments were amplified in the PCR using the primers X1 and Y1 (data not shown).
The mutator phenotype of the qde-3 recQ2 double mutant was suppressed by mutation in NHEJ, but not in HR:
Plasmid-rejoining experiments in BS cells and WS cells showed that rejoining is error prone and mainly causes deletions (RUNGER and KRAEMER 1989; RUNGER et al. 1994; GAYMES et al. 2002; OSHIMA et al. 2002), suggesting that error-prone NHEJ causes deletion in these RecQ homolog mutants and thus may be responsible for the mutator phenotype of our qde-3 recQ2 double mutant.
In N. crassa, mus-51 and mus-52 genes, homologs of mammalian KU70 and KU80, respectively, are required for NHEJ (NINOMIYA et al. 2004). The spontaneous mutation rate of the mus-52 mutant at the ad-3 loci was 0.4 x 10–7, similar to that of the wild-type strain (Table 2). The qde-3 recQ2 mus-52 triple mutant also showed a rate similar to that of the wild-type strain, 0.2 x 10–7 (Table 2). The mutation rate of the qde-3 recQ2 mei-3 triple mutant was 9.1 x 10–7, indicating that mutation in mei-3 did not suppress the high spontaneous mutation rate and confirming that this suppression effect is due only to the mutation in the NHEJ gene. In addition, analysis of the genomic DNA of the ad-3A mutations derived in the qde-3 recQ2 mei-3 strain indicated that they carry large deletions (data not shown).
Mutation in NHEJ increases the severity of the growth defect of the qde-3 recQ2 double mutant:
Although the mus-52 mutation suppressed the rate of spontaneous deletions, the growth defect of the qde-3 recQ2 double mutant was more extreme in the qde-3 recQ2 mus-52 triple mutant. The colonies of the triple mutant were very thin and the morphology of each colony was irregular. Conidial viability was also reduced and varied from 17 to 34%. Moreover, linear growth of the triple mutant slowed earlier than that of the qde-3 recQ2 double mutant (Figure 2A). Although the triple mutant grew faster than the double mutant in the fourth tube, the growth rate, once it slowed, was not constant, alternating between a "slow" and "fast" growth pattern (data not shown). This phenotype is like the "stop-and-start" phenotype reported for some DNA repair mutants in N. crassa (NEWMEYER 1984). Although the triple mutant showed a severe growth defect, the abnormal appearance observed for the qde-3 recQ2 double mutant—accumulation of orange and black pigments and spherically formed mycelia—was not seen.
DISCUSSION
RecQ homologs in N. crassa have roles in the maintenance of genome stability as well as DNA repair:
In this report, we demonstrated that QDE3 and RECQ2 are required for maintenance of genomic stability as well as repair of DNA damage induced by mutagens. The qde-3 mutant and the qde-3 recQ2 double mutant show sensitivity to several mutagens, as reported previously (PICKFORD et al. 2003; KATO et al 2004), but are not sensitive to UV. However, the qde-3 mus-38 double mutant is more sensitive to UV than is the mus-38 single mutant. Furthermore, the qde-3 recQ2 mus-38 triple mutant is even more sensitive than either double mutant. As mus-38 is a homolog of S. cerevisiae RAD1, which is required for NER, the role of the Neurospora RecQ homologs in the repair of UV damage must involve pathways separate from that of NER, such as recombination repair, postreplication repair, and DNA damage checkpoint response.
The qde-3 mutant exhibited an increased spontaneous mutation rate, indicating that it is a mutator. The spontaneous mutation rate was much higher in the qde-3 recQ2 double mutant, suggesting that the qde-3 and recQ2 genes are required for the maintenance of genome stability. However, the spontaneous mutation rate in the recQ2 single mutant was similar to that in the wild-type strain. Moreover, the recQ2 mutant did not show significant mutagen sensitivity, although it had a slightly higher sensitivity to high concentrations of MMS, MNNG, and CPT than the wild-type strain did. The qde-3 recQ2 double mutant was more mutagen sensitive than the qde-3 single mutant. Thus, it seems that RECQ2 has a relatively minor role in cellular functions such as DNA repair and the maintenance of genome stability, but the function becomes conspicuous only when the qde-3 function is impaired. These two RecQ homologs have redundant functions in the repair of damaged DNA and also in the maintenance of genome stability, since the double mutant is more sensitive to mutagens and more mutable than each single mutant.
The growth defect of the qde-3 recQ2 mutant is caused by HR, not by NHEJ:
The qde-3 recQ2 double mutant shows low conidial viability, slow growth, and a decreased number of nuclei, indicating a decline of cellular proliferation. The growth defect of the qde-3 recQ2 double mutant is suppressed by a mutation in mei-3, which is required for HR, suggesting that the growth defect in the qde-3 recQ2 double mutant is a result of the HR pathway. In contrast, the qde-3 recQ2 mus-52 triple mutant showed a greater growth defect than the qde-3 recQ2 double mutant did, showing that NHEJ was not responsible for the defect.
HR and NHEJ are two major mechanisms of double-strand break (DSB) repair. HR repair, which depends on homologous sequences, rarely introduces mutations during DSB repair. NHEJ is an error-prone pathway that involves exonucleolytic processing and subsequent rejoining of the DNA ends without dependence on sequence homology. HR is initiated by single-stranded invasion of a homologous DNA sequence and is mediated by Rad51. Their suppression of the growth defect caused by the mei-3 mutation suggests that QDE3 and RECQ2 function after strand invasion. RecQ homologs in human and yeast can unwind Holliday-junction-like DNA structures (BENNETT et al. 1999; MOHAGHEGH et al. 2001). BLM also can melt D-loops, which are formed during HR by RAD51 (VAN BRABANT et al. 2000). Because RecQ proteins are highly conserved, it is likely that N. crassa QDE3 and RECQ2 behave in a similar manner. We speculate that RecQ proteins are involved in the resolution of recombination intermediates and that a lack of such resolution hinders the progress of DNA replication, resulting in the observed growth defect.
The mutator phenotype of the qde-3 recQ2 mutant is caused by NHEJ-dependent deletion formation and not by HR:
Mutation spectrum analysis of the ad-3A locus revealed that almost all the mutations that arose spontaneously in the qde-3 mutant and the qde-3 recQ2 double mutant were deletions. PCR analysis of the ad-3B locus also indicated that the ad-3B mutations obtained from the qde-3 and qde-3 recQ2 mutants were predominantly deletions, especially large deletions. The mutator phenotype of the double mutant was completely suppressed by mutation of mus-52, suggesting that NHEJ is responsible for this phenotype. In contrast, the mutator phenotype was not suppressed by mei-3 mutation, suggesting that HR is not involved. In addition, many large deletions were formed at the ad-3A locus in the qde-3 recQ2 mei-3 triple mutant, suggesting that HR is not required for the formation of deletions in the RecQ homolog mutants.
BS cells and WS cells also show a mutator phenotype and an increased number of deletion events (WARREN et al. 1981; FUKUCHI et al. 1989; TACHIBANA et al. 1996). RUNGER and KRAEMER (1989) and RUNGER et al. (1994) report that joining of linear plasmid DNA is error prone in these cells, and both of these findings offer indirect evidence that NHEJ causes deletion in these cells. Our report provides direct evidence that NHEJ-dependent deletion formation is responsible for the mutator phenotype in RecQ homolog mutants. A number of studies also indicate a relationship between RecQ helicases and NHEJ, which our findings support for the N. crassa RecQ homologs QDE3 and RECQ2. However, the nature of this relationship is unknown. RecQ homologs may control the rejoining efficiency of DSBs and/or the length of the exonucleolytic processing of the DNA ends. The former possibility is supported by the observation that overexpression of Ku70 in Drosophila melanogaster partially suppresses defects conferred by mutations in the RecQ homolog Dmblm (KUSANO et al. 2001). The latter hypothesis is supported by the observation of error-prone joining of linear plasmid DNA in human cells deficient in BLM or WRN (RUNGER and KRAEMER 1989; RUNGER et al. 1994). The characteristic large deletions may result from unregulated end processing by NHEJ machinery.
Abnormal appearance of the qde-3 recQ2 double mutant is dependent on both HR and NHEJ:
The abnormal appearance of the qde-3 recQ2 double mutant is suppressed by mutation in either mei-3 or mus-52, suggesting that this phenotype arises from a pathway involving both HR and NHEJ. Accumulation of orange and black pigments accompanied by the formation of spherical mycelia is a result of the uncontrolled expression of genes involved in conidiation, carotenoid synthesis, hyphal formation, and melanin production. As cancer cells proliferate in an uncontrolled manner, this susceptibility to abnormal appearance is reminiscent of the predisposition to cancer in BS patients. The mean age at which cancer is diagnosed in the BS-inherent population is 24.4 years (GERMAN 1993) and cancer does not occur in children in this population. The abnormal appearance of the qde-3 recQ2 double mutant is found only after growth has occurred for several days. It might be true in all organisms that genomic dysregulation proceeds at a slow tempo, mediated by HR and NHEJ functions, when RecQ homolog genes are mutated.
RecQ homologs may have a role in the suppression of spontaneous DSBs:
The characteristic deletions seen in RecQ mutants of N. crassa might be caused by unregulated NHEJ, as discussed above. However, the increase in spontaneous mutation rate may suggest an elevated frequency of DSBs in RecQ mutants. NHEJ is known to function in DSB repair in the wild-type strain, so it cannot be that loss of QDE3 and RECQ2 activates NHEJ to repair DSBs, resulting in the observed high mutation rate. Since NHEJ is error prone even in the wild-type strain, it is unlikely that loss of QDE3 and RECQ2 increases the tendency of NHEJ to be error prone. We speculate that QDE3 and RECQ2 play roles in the suppression of DSBs. If this is the case, what mechanism increases DSBs in the RecQ-deficient mutant?
The production of DSBs may be stimulated by HR dysfunction. If recombination intermediates accumulate in the RecQ-deficient mutant, physical tension or processing during mitosis may yield additional DSBs. However, since the HR-deficient qde-3 recQ2 mei-3 triple mutant retained the mutator phenotype, it is likely that DSBs arise by another mechanism. DSBs are generated during endonuclease-mediated resolution of Holliday junctions. BLM is thought to disrupt Holliday junctions formed at sites of blocked replication forks by its potential reverse branch migration activity (KAROW et al. 2000; WANG et al. 2000). Therefore, the increase in DSBs in the qde-3 recQ2 double mutant may be due to a decrease in the unraveling of Holliday junctions formed at sites of stalled replication forks and a resultant increase in endonuclease-mediated resolution of these junctions.
The high sensitivity of the qde-3 recQ2 double mutant to HU and CPT supports the idea that QDE3 and RECQ2 function in DNA replication, because replication is hindered by these agents. We propose a model (Figure 5) to explain the mechanism of the growth defect and genomic instability in the qde-3 recQ2 double mutant. If QDE3 and RECQ2 are nonfunctional, additional DSBs are formed by at the sites where replication forks are stalled. Then, proteins involved in HR or NHEJ will be recruited to the DSB sites for the repair of DSBs. HR events may not be completed normally without QDE3 and RECQ2 and thus recombination intermediates accumulate, leading to the inhibition of cell proliferation. In the absence of mei-3, proliferation-inhibiting DNA structures derived from HR will not be made and cells can grow normally, but DSBs must then be repaired by NHEJ and the extensive processing and subsequent ligation of the nonhomologous ends yield deletions. Conversely, in the qde-3 recQ2 mus-52 triple mutant in which NHEJ is impaired, all DSBs will be repaired by HR, so deletions will be less frequent but proliferation-inhibiting DNA structures made by HR will increase, thus amplifying the growth defect. At present, there is no direct evidence that DSBs are more frequent in the qde-3 recQ2 double mutant. Our current studies, focusing on DSB formation during DNA replication, are expected to shed light on the important question of how RecQ helicases function in the maintenance of genomic stability.
ACKNOWLEDGEMENTS
We thank Niji Ota for help in sequencing and Jane Yeadon for reviewing this manuscript. This work was supported by Rational Evolutionary Design of Advanced Biomolecules, Saitama Prefecture Collaboration of Regional Entities for the Advancement of Technological Excellence, Japan Science and Technology Agency.
LITERATURE CITED
AGUILERA, A., and H. L. KLEIN, 1988 Genetic control of intrachromosomal recombination in Saccharomyces cerevisiae. I. Isolation and genetic characterization of hyper-recombination mutations. Genetics 119: 779–790.
BENNETT, R. J., J. A. SHARP and J. C. WANG, 1998 Purification and characterization of the Sgs1 DNA helicase activity of Saccharomyces cerevisiae. J. Biol. Chem. 273: 9644–9650.
BENNETT, R. J., J. L. KECK and J. C. WANG, 1999 Binding specificity determines polarity of DNA unwinding by the Sgs1 protein of S. cerevisiae. J. Mol. Biol. 289: 235–248.
BROSH, R. M., JR., D. K. ORREN, J. O. NEHLIN, P. H. RAVN, M. K. KENNY et al., 1999 Functional and physical interaction between WRN helicase and human replication protein A. J. Biol. Chem. 274: 18341–18350.
BROSH, R. M., JR., C. VON KOBBE, J. A. SOMMERS, P. KARMAKAR, P. L. OPRESKO et al., 2001 Werner syndrome protein interacts with human flap endonuclease 1 and stimulates its cleavage activity. EMBO J. 20: 5791–5801.
CHAGANTI, R. S., S. SCHONBERG and J. GERMAN, 1974 A manyfold increase in sister chromatid exchanges in Bloom's syndrome lymphocytes. Proc. Natl. Acad. Sci. USA 71: 4508–4512.
COGONI, C., and G. MACINO, 1999 Posttranscriptional gene silencing in Neurospora by a RecQ DNA helicase. Science 286: 2342–2344.
CONSTANTINOU, A., M. TARSOUNAS, J. K. KAROW, R. M. BROSH, V. A. BOHR et al., 2000 Werner's syndrome protein (WRN) migrates Holliday junctions and co-localizes with RPA upon replication arrest. EMBO Rep. 1: 80–84.
DAVIS, R. H., and F. J. DE SERRES, 1970 Genetic and microbiological research techniques for Neurospora crassa. Methods Enzymol. 17: 79–143.
DE SERRES, F. J., and H. G. KLMARK, 1958 A direct method for determination of forward-mutation rates in Neurospora crassa. Nature 182: 1249–1250.
DOE, C. L., J. DIXON, F. OSMAN and M. C. WHITBY, 2000 Partial suppression of the fission yeast rqh1(-) phenotype by expression of a bacterial Holliday junction resolvase. EMBO J. 19: 2751–2762.
DUTERTRE, S., M. ABABOU, R. ONCLERCQ, J. DELIC, B. CHATTON et al., 2000 Cell cycle regulation of the endogenous wild-type Bloom's syndrome DNA helicase. Oncogene 19: 2731–2738.
ELLIS, N. A., J. GRODEN, T. Z. YE, J. STRAUGHEN, D. J. LENNON et al., 1995 The Bloom's syndrome gene product is homologous to RecQ helicases. Cell 83: 655–666.
EPSTEIN, C. J., G. M. MARTIN, A. L. SCHULTZ and A. G. MOTULSKY, 1966 Werner's syndrome: a review of its symptomatology, natural history, pathologic features, genetics and relationship to the natural aging process. Medicine 45: 177–221.
FREI, C., and S. M. GASSER, 2000 The yeast Sgs1p helicase acts upstream of Rad53p in the DNA replication checkpoint and colocalizes with Rad53p in S-phase-specific foci. Genes Dev. 14: 81–96.
FUKUCHI, K., G. M. MARTIN and R. J. MONNAT, JR., 1989 Mutator phenotype of Werner syndrome is characterized by extensive deletions. Proc. Natl. Acad. Sci. USA 86: 5893–5897.
GANGLOFF, S., C. SOUSTELLE and F. FABRE, 2000 Homologous recombination is responsible for cell death in the absence of the Sgs1 and Srs2 helicases. Nat. Genet. 25: 192–194.
GAYMES, T. J., P. S. NORTH, N. BRADY, I. D. HICKSON, G. J. MUFTI et al., 2002 Increased error-prone non homologous DNA end-joining—a proposed mechanism of chromosomal instability in Bloom's syndrome. Oncogene 21: 2525–2533.
GERMAN, J., 1993 Bloom syndrome: a Mendelian prototype of somatic mutational disease. Medicine 72: 393–406.
INOUE, H., and C. ISHII, 1984 Isolation and characterization of MMS-sensitive mutants of Neurospora crassa. Mutat. Res. 125: 185–194.
ISHII, C., and H. INOUE, 1994 Mutagenesis and epistatic grouping of the Neurospora meiotic mutants, mei-2 and mei-3, which are sensitive to mutagens. Mutat. Res. 315: 249–259.
ISHII, C., K. NAKAMURA and H. INOUE, 1998 A new UV-sensitive mutant that suggests a second excision repair pathway in Neurospora crassa. Mutat. Res. 408: 171–182.
JIAO, R., C. Z. BACHRATI, G. PEDRAZZI, P. KUSTER, M. PETKOVIC et al., 2004 Physical and functional interaction between the Bloom's syndrome gene product and the largest subunit of chromatin assembly factor 1. Mol. Cell. Biol. 24: 4710–4719.
KAMATH-LOEB, A. S., E. JOHANSSON, P. M. BURGERS and L. A. LOEB, 2000 Functional interaction between the Werner Syndrome protein and DNA polymerase delta. Proc. Natl. Acad. Sci. USA 97: 4603–4608.
KAROW, J. K., A. CONSTANTINOU, J. L. LI, S. C. WEST and I. D. HICKSON, 2000 The Bloom's syndrome gene product promotes branch migration of Holliday junctions. Proc. Natl. Acad. Sci. USA 97: 6504–6508.
KATO, A., Y. AKAMATSU, Y. SAKURABA and H. INOUE, 2004 The Neurospora crassa mus-19 gene is identical to the qde-3 gene, which encodes a RecQ homologue and is involved in recombination repair and postreplication repair. Curr. Genet. 45: 37–44.
KAWABE, T., N. TSUYAMA, S. KITAO, K. NISHIKAWA, A. SHIMAMOTO et al., 2000 Differential regulation of human RecQ family helicases in cell transformation and cell cycle. Oncogene 19: 4764–4772.
KITAO, S., I. OHSUGI, K. ICHIKAWA, M. GOTO, Y. FURUICHI et al., 1998 Cloning of two new human helicase genes of the RecQ family: biological significance of multiple species in higher eukaryotes. Genomics 54: 443–452.
KITAO, S., A. SHIMAMOTO, M. GOTO, R. W. MILLER, W. A. SMITHSON et al., 1999 Mutations in RECQL4 cause a subset of cases of Rothmund-Thomson syndrome. Nat. Genet. 22: 82–84.
KUSANO, K., D. M. JOHNSON-SCHLITZ and W. R. ENGELS, 2001 Sterility of Drosophila with mutations in the Bloom syndrome gene: complementation by Ku70. Science 291: 2600–2602.
LEA, D. E., and C. A. COULSON, 1949 The distribution of the numbers of mutants in bacterial populations. J. Genet. 49: 264–285.
LEBEL, M., E. A. SPILLARE, C. C. HARRIS and P. LEDER, 1999 The Werner syndrome gene product co-purifies with the DNA replication complex and interacts with PCNA and topoisomerase I. J. Biol. Chem. 274: 37795–37799.
LEE, S. K., R. E. JOHNSON, S. L. YU, L. PRAKASH and S. PRAKASH, 1999 Requirement of yeast SGS1 and SRS2 genes for replication and transcription. Science 286: 2339–2342.
LONN, U., S. LONN, U. NYLEN, G. WINBLAD and J. GERMAN, 1990 An abnormal profile of DNA replication intermediates in Bloom's syndrome. Cancer Res. 50: 3141–3145.
MAFTAHI, M., J. C. HOPE, L. DELGADO-CRUZATA, C. S. HAN and G. A. FREYER, 2002 The severe slow growth of srs2 rqh1 in Schizosaccharomyces pombe is suppressed by loss of recombination and checkpoint genes. Nucleic Acids Res. 30: 4781–4792.
MANKOURI, H. W., T. J. CRAIG and A. MORGAN, 2002 SGS1 is a multicopy suppressor of srs2: functional overlap between DNA helicases. Nucleic Acids Res. 30: 1103–1113.
MOHAGHEGH, P., J. K. KAROW, R. M. BROSH, JR, V. A. BOHR and I. D. HICKSON, 2001 The Bloom's and Werner's syndrome proteins are DNA structure-specific helicases. Nucleic Acids Res. 29: 2843–2849.
MYUNG, K., A. DATTA, C. CHEN and R. D. KOLODNER, 2001 SGS1, the Saccharomyces cerevisiae homologue of BLM and WRN, suppresses genome instability and homeologous recombination. Nat. Genet. 27: 113–116.
NEWMEYER, D., 1984 Neurospora mutants sensitive both to mutagens and to histidine. Curr. Genet. 9: 65–74.
NINOMIYA, Y., K. SUZUKI, C. ISHII and H. INOUE, 2004 Highly efficient gene replacements in Neurospora strains deficient for nonhomologous end-joining. Proc. Natl. Acad. Sci. USA 101: 12248–12253.
ONODA, F., M. SEKI, A. MIYAJIMA and T. ENOMOTO, 2000 Elevation of sister chromatid exchange in Saccharomyces cerevisiae sgs1 disruptants and the relevance of the disruptants as a system to evaluate mutations in Bloom's syndrome gene. Mutat. Res. 459: 203–209.
OSHIMA, J., S. HUANG, C. PAE, J. CAMPISI and R. H. SCHIESTL, 2002 Lack of WRN results in extensive deletion at nonhomologous joining ends. Cancer Res. 62: 547–551.
PICKFORD, A., L. BRACCINI, G. MACINO and C. COGONI, 2003 The QDE-3 homologue RecQ-2 co-operates with QDE-3 in DNA repair in Neurospora crassa. Curr. Genet. 42: 220–227.
POOT, M., H. HOEHN, T. M. RUNGER and G. M. MARTIN, 1992 Impaired S-phase transit of Werner syndrome cells expressed in lymphoblastoid cell lines. Exp. Cell Res. 202: 267–273.
PURANAM, K. L., and P. J. BLACKSHEAR, 1994 Cloning and characterization of RECQL, a potential human homologue of the Escherichia coli DNA helicase RecQ. J. Biol. Chem. 269: 29838–29845.
RONG, L., and H. L. KLEIN, 1993 Purification and characterization of the SRS2 DNA helicase of the yeast Saccharomyces cerevisiae. J. Biol. Chem. 268: 1252–1259.
RUNGER, T. M., and K. H. KRAEMER, 1989 Joining of linear plasmid DNA is reduced and error-prone in Bloom's syndrome cells. EMBO J. 8: 1419–1425.
RUNGER, T. M., C. BAUER, B. DEKANT, K. MOLLER, P. SOBOTTA et al., 1994 Hypermutable ligation of plasmid DNA ends in cells from patients with Werner syndrome. J. Invest. Dermatol. 102: 45–48.
SALK, D., K. AU, H. HOEHN and G. M. MARTIN, 1981 Cytogenetics of Werner's syndrome cultured skin fibroblasts: variegated translocation mosaicism. Cytogenet. Cell Genet. 30: 92–107.
SEKI, M., H. MIYAZAWA, S. TADA, J. YANAGISAWA, T. YAMAOKA et al., 1994 Molecular cloning of cDNA encoding human DNA helicase Q1 which has homology to Escherichia coli Rec Q helicase and localization of the gene at chromosome 12p12. Nucleic Acids Res. 22: 4566–4573.
SINCLAIR, D. A., K. MILLS and L. GUARENTE, 1997 Accelerated aging and nucleolar fragmentation in yeast sgs1 mutants. Science 277: 1313–1316.
STABEN, C., B. JENSEN, M. SINGER, J. POLLOCK, M. SCHECHTMAN et al., 1989 Use of a bacterial hygromycin B resistance gene as a dominant selectable marker in Neurospora crassa transformation. Fungal Genet. Newsl. 36: 79–81.
TACHIBANA, A., K. TATSUMI, T. MASUI and T. KATO, 1996 Large deletions at the HPRT locus associated with the mutator phenotype in a Bloom's syndrome lymphoblastoid cell line. Mol. Carcinogen. 17: 41–47.
TAMARU, H., and H. INOUE, 1989 Isolation and characterization of a laccase-derepressed mutant of Neurospora crassa. J. Bacteriol. 171: 6288–6293.
VAN BRABANT, A. J., T. YE, M. SANZ, I. J. GERMAN, N. A. ELLIS et al., 2000 Binding and melting of D-loops by the Bloom syndrome helicase. Biochemistry 39: 14617–14625.
VENNOS, E. M., and W. D. JAMES, 1995 Rothmund-Thomson syndrome. Dermatol. Clin. 13: 143–150.
WANG, W., M. SEKI, Y. NARITA, E. SONODA, S. TAKEDA et al., 2000 Possible association of BLM in decreasing DNA double strand breaks during DNA replication. EMBO J. 19: 3428–3435.
WANG, W., M. SEKI, Y. NARITA, T. NAKAGAWA, A. YOSHIMURA et al., 2003 Functional relation among RecQ family helicases RecQL1, RecQL5, and BLM in cell growth and sister chromatid exchange formation. Mol. Cell. Biol. 23: 3527–3535.
WARREN, S. T., R. A. SCHULTZ, C. C. CHANG, M. H. WADE and J. E. TROSKO, 1981 Elevated spontaneous mutation rate in Bloom syndrome fibroblasts. Proc. Natl. Acad. Sci. USA 78: 3133–3137.
WATANABE, K., Y. SAKURABA and H. INOUE, 1997 Genetic and molecular characterization of Neurospora crassa mus-23: a gene involved in recombinational repair. Mol. Gen. Genet. 256: 436–445.
WATT, P. M., E. J. LOUIS, R. H. BORTS and I. D. HICKSON, 1995 Sgs1: a eukaryotic homolog of E. coli RecQ that interacts with topoisomerase II in vivo and is required for faithful chromosome segregation. Cell 81: 253–260.
WATT, P. M., I. D. HICKSON, R. H. BORTS and E. J. LOUIS, 1996 SGS1, a homologue of the Bloom's and Werner's syndrome genes, is required for maintenance of genome stability in Saccharomyces cerevisiae. Genetics 144: 935–945.
YAMAGATA, K., J. KATO, A. SHIMAMOTO, M. GOTO, Y. FURUICHI et al., 1998 Bloom's and Werner's syndrome genes suppress hyperrecombination in yeast sgs1 mutant: implication for genomic instability in human diseases. Proc. Natl. Acad. Sci. USA 95: 8733–8738.
YU, C. E., J. OSHIMA, Y. H. FU, E. M. WIJSMAN, F. HISAMA et al., 1996 Positional cloning of the Werner's syndrome gene. Science 272: 258–262.
Laboratory of Genetics, Department of Regulation Biology, Faculty of Science, Saitama University, 338-8570 Saitama, Japan(Akihiro Kato and Hirokazu)