Mismatch-induced lethality due to a defect in Escherichia coli RecQ helicase in exonuclease-deficient background: Dependence on MutS and UvrD functions

Plasmid 63 (2010) 119–127

Contents lists available at ScienceDirect

Plasmid journal homepage: www.elsevier.com/locate/yplas

Mismatch-induced lethality due to a defect in Escherichia coli RecQ helicase in exonuclease-deficient background: Dependence on MutS and UvrD functions Yoshimasa Yamana a,b, Shuji Sonezaki b, Hiroaki I. Ogawa a, Kohji Kusano a,c,* a

Department of Biological Functions and Engineering, Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology, 2-4 Hibikino, Wakamatsu-ku, Kita-Kyushu 808-0196, Japan b Research Institute, TOTO LTD, 2-1-1 Nakashima, Kokurakita-ku, Kita-kyushu 802-8601, Japan c Genetic Resource Education & Development Center, Drosophila Genetic Resource Center, Kyoto Institute of Technology, Saga-Ippongi-cho, Ukyo-ku, Kyoto 616-8354, Japan

a r t i c l e

i n f o

Article history: Received 3 September 2009 Revised 1 December 2009 Available online 14 December 2009 Communicated by Ichizo Kobayashi Keywords: RecQ Exonuclease Mismatch repair SOS response

a b s t r a c t Escherichia coli DNA-unwinding protein RecQ has roles in the regulation of general recombination and the processing of stalled replication forks. In this study, we found that knockout of the recQ gene in combination with xonA xseA recJ mutations, which inhibit methyldirected mismatch repair (MMR), caused about 100-fold increase in sensitivity to a purine analog 2-aminopurine (2AP). Intriguingly, inactivation of a MMR initiator due to the either mutation mutS or uvrD completely suppressed the 2AP sensitivity caused by recQ xonA xseA recJ mutations, suggesting that RecQ helicase might act on the DNA structures that are generated by the processing of DNA by the MutSLH complex and UvrD helicase. Moreover, the recQ gene knockout in combination with xonA xseA recJ mutations enhanced 2AP-induced filament formation, and increased by twofold the rate of spontaneous forward mutations in the thyA locus but did not increase the rate of rifampicin-resistant mutations. We discuss about the possible interplay between E. coli RecQ helicase and mismatch recognition factors. Ó 2010 Published by Elsevier Inc.

1. Introduction The human recessive disorders Werner syndrome and Bloom syndrome are caused by mutations in RECQ helicase family genes, WRN and BLM, respectively (Ellis et al., 1995; Yu et al., 1996). These genetic diseases are associated with cancer predisposition, which results from chromosome instability including deletions and insertions. Escherichia coli RecQ helicase is well known to be involved in the maintenance and rearrangement of genome structure through conjugational recombination (Nakayama et al.,

* Corresponding author. Address: Genetic Resource Education & Development Center, Drosophila Genetic Resource Center, Kyoto Institute of Technology, Saga-Ippongi-cho, Ukyo-ku, Kyoto 616-8354, Japan. Fax: +81 75 861 0881. E-mail address: [email protected] (K. Kusano). 0147-619X/$ - see front matter Ó 2010 Published by Elsevier Inc. doi:10.1016/j.plasmid.2009.12.001

1984; Lovett and Sutera Jr., 1995), regulation of gene conversion with crossover (Kusano et al., 1994), and suppression of illegitimate recombination (Hanada et al., 1997). When DNA replication forks are stalled, E. coli RecQ processes the regressed replication forks (Courcelle and Hanawalt, 1999; Courcelle et al., 2003) or recruits RecA protein (Hishida et al., 2004). However, there have been no reports implicating interplay between E. coli RecQ helicase and the methyl-directed mismatch repair pathway, which the MutSLH complex and UvrD helicase initiate. When errors in DNA replication escape the proof-reading function of DNA polymerases, they are corrected by methyldirected MMR to ensure the fidelity of E. coli chromosome replication (Modrich and Lahue, 1996). In the initial steps of the MMR process, MutS, MutL, MutH, and UvrD function cooperatively. First, the MutS–MutL sliding clamp complex recognizes mismatches resulting from errors missed by

120

Y. Yamana et al. / Plasmid 63 (2010) 119–127

DNA polymerases (Su et al., 1988; Grilley et al., 1989; Acharya et al., 2003). Subsequently, this complex activates the endonuclease MutH to cleave an unmethylated strand of a hemimethylated GATC sequence that should be located on either side of the misincorporation site (Au et al., 1992). Finally, UvrD helicase invades into the resultant single-strand break and unwinds the strand containing a misincorporated base (Dao and Modrich, 1998; Mechanic et al., 2000). During later steps of the MMR process, the strand displaced by UvrD is subject to degradation by one of several single-strand specific exonucleases including Exonuclease I (xonA gene product), Exonuclease VII (xseA and xseB gene products) and the exonuclease RecJ (Cooper et al., 1993). This exonucleotic degradation assists repair synthesis by the holoenzyme DNA polymerase III, followed by restoration of the covalent bond to the repaired strand by DNA ligase (Lahue et al., 1989). Consistently, DxonA DxseA recJ mutant cells display an increased rate of spontaneous mutations (Viswanathan et al., 2000) and increased lethality after treatment with 2aminopurine (2AP), which is incorporated instead of guanine or adenine and induces mismatches (Ronen, 1979); however, the removal of any of MutS, MutL, MutH and UvrD from DxonA DxseA recJ mutant cells substantially increase the mutation rate (Viswanathan et al., 2000), and suppresses 2AP-induced cell death (Burdett et al., 2001). These biochemical and genetic observations indicate that the three exonucleases operate after the actions of MutSLH and UvrD in E. coli MMR. In this study, we show that the mutation of recQ increases the sensitivity of DxonA DxseA recJ mutant cells to 2AP. Inactivation of MutS or UvrD completely suppressed the 2AP sensitivity of these cells. The recQ mutation on a DxonA DxseA recJ mutant background modestly, but significantly, increased the rate of spontaneous mutations conferring trimethoprim resistance and enhanced a characteristic SOS response; however, did not affect that of spontaneous mutations conferring rifampicin resistance. We discuss the possible functions of RecQ helicase in the processing of mismatch-induced DNA lesions generated by MutSLH and UvrD.

2. Materials and methods 2.1. E. coli strains and media Strains used in this work are presented in Table 1. Strains were grown in LB medium composed of 1% Bactotryptone (Difco), 0.5% yeast extract (Difco), 0.5% NaCl and 1.5% Bacto-agar (Difco) (for plate media) or minimal media consisting of M9 salts (50 mM Na2PO4, 50 mM KHPO4, 1 mM MgSO4, 0.1 mM CaCl2), supplemented with 0.4% glucose, 50 mg/ml of the appropriate required amino acids, 1 mg/ml thiamine and 2% agar (for plate media). Thymine (50 mg/ml) and guanosine (25 mg/ml) were added to the LB and minimal media used for measurement of thyA mutation rates. Trimethoprim (Tmp; 10 mg/ml) was also added to the minimal medium used for measurement of thyA mutation rates. Rifampicin (Rif; 100 lg/ml) was added to the LB medium used for measurement of Rifresistant mutation rates. All of the strains that were made in this work were constructed by P1vir-mediated transduction from either the strain BT199 or the strain STL2701 (Table 1). Antibiotics (ampicillin, 50 mg/ml; chloramphenicol, 30 mg/ml; kanamycin, 50 mg/ml; and tetracycline, 10 mg/ ml) were added to media for selecting strains possessing the appropriate antibiotic-resistance markers. 2.2. Measurement of 2AP sensitivity Cultures were grown in LB with appropriate antibiotics to 2  108 cells per ml. After appropriate dilution with M9 salts, cells were spread onto LB plates containing various concentrations of 2AP. After growth for 24 h at 37 °C, colonies were counted. 2.3. Measurement of cell-length distribution A fresh single colony was inoculated into 2 ml of M9 medium with appropriate antibiotics in the absence or presence of 0.2 mM 2AP and was grown at 37 °C with

Table 1 Description of the E. coli K12 strains used. Strain

Genotype

Source or derivation

BT199 STL2701b BMH71–18mutS 21336 KUW64 KTYY11 KTYY33 KTYY13 KTYY23 KTYY25 KTYY41

AB1157a but his+ arg+ BT199 but DxonA300::cat D(xseA-guaB)zff3139::Tn10Kan recJ2052::Tn10Kan F (traD36 proAB lacIq (lacZ)D(M15)) D(lac-proAB) thi supE mutS215::Tn10 AB1157 but uvrD260::Tn5 AB1157 but recQ1803::Tn3 KUW64 but uvrD260::Tn5 BT199 but recQ1803::Tn3 STL2701 but recQ1803::Tn3 STL2701 but mutS215::Tn10 KTYY13 but mutS215::Tn10 STL2701 but recQ1803::Tn3 uvrD260::Tn5

W. Wackernagel (Thoms and Wackernagel, 1998) S.T. Lovett (Viswanathan et al., 1999) National Institute of Genetics (Ogawa et al., 1997) National Institute of Genetics K. Kusano collection KUW64  P1(21336)c BT199  P1(KUW64)d STL2701  P1(KUW64)d STL2701  P1(BMH71–18mutS)e KTYY13  P1(BMH71–18mutS)e STL2701  P1(KTYY11)f

a The genotype of AB1157 strain is F-, thr-1 ara-14 leuB6 D(gpt-proA)62 lacY1 tsx-33 supE44 galK2k- rac- hisG4(Oc) rfbD1 mgl-51 rpsL31 kdgK51 xyl-5 mtl-1 argE3(Oc) thi-1 qsr- (Bachmann, 1996). b The allele DxonA300::cat lacks the gene coding Exonuclease I. The allele D(xseA-guaB)zff3139::Tn10Kan lacks the gene coding Exonuclease VII. c The kanamycin-resistant transductants were selected. d The ampicillin-resistant transductants were selected. e The tetracycline-resistant transductants were selected. f The ampicillin-resistant transductants were first selected and then more UV-sensitive transductants out of them were screened.

Y. Yamana et al. / Plasmid 63 (2010) 119–127

shaking up to late-log phase. Each 2AP-free and 2AP-containing culture was spun down and 25-fold concentrated in the same fresh M9 medium. About 10 ll from each was put on a microscope slide and mixed with 10 ll of fixation solution (M9 medium/1.2% GTG agar) in the absence or presence of 0.2 mM 2AP. After 1 min at 25 °C, a glass coverslip was placed onto each fixed mixture. Images were visualized by bright field microscopy using BX51 (OLYMPUS, JAPAN), captured with a cooled charge-coupled device (CCD) camera VB6010/6000 (KEYENCE, JAPAN), and then transferred to Photoshop (Adobe). The images were overlapped with a picture of the ruled grid of the counting chamber (50 lm  50 lm) as internal standard to determine cell lengths.

2.4. Measurement of mutation rates Rif resistance: A frozen stock (70 °C) of each strain tested was streaked on LB plates containing appropriate antibiotics needed for the selection of the strains and allowed to grow for 18 h at 37 °C. Whole single colonies were transferred one by one into 5.0 ml of LB with antibiotics and were grown for 24 h at 37 °C in the dark. After appropriate dilution by M9 salts, cells were spread on LB plates with or without Rif and grown for 24 h at 37 °C. The number of viable cells in the culture and the number of Rif-resistant mutant cells in the same culture were determined by plating without and with Rif, respectively. Mutations rates were calculated as the total number of Rif-resistant cells/the total number of viable cells in each of 4–12 independent cultures for each strains and averaged in each strain. Tmp resistance: A frozen stock (70 °C) of each strain tested was streaked on LB plates containing thymine, guanosine and appropriate antibiotics needed for the selection of the strains and allowed to grow for 18 h at 37 °C. Whole single colonies were transferred one by one into 2.5 ml of LB with thymine, guanosine and antibiotics and were grown for 20 h at 37 °C in the dark. Cells from each culture were trapped on sterile 0.45-lm filter membranes, washed with 5 ml of M9 salts, and resuspended in 2.5 ml of M9 salts. After appropriate dilutions, cells were spread on M9 plates with or without Tmp and grown for 48 h at 37 °C. The number of viable cells in the culture (n) and the number of thyA mutant cells in the same culture (r0) were determined by plating with and without Tmp, respectively. Mutation rates were calculated by the method of the median (Lea and Coulson, 1949) using the formula: mutation rate = m/n, where m was estimated based on the formula: r0/m – log m = 1.24 and Table 3 from Lea and Coulson (1949).

2.5. Sequencing of thyA mutations One Tmp-resistant clone from each culture was used to prepare a genomic DNA sample for PCR template. A PCR product including the entire thyA gene was amplified by the two primers TCCCCGCGGCCAATGGGAGCTGTCTCAGG and ATCGATTTCTTCGGCGCATCTTCCGG, and was directly sequenced with the same primer pair.

121

3. Results 3.1. Effect of the recQ mutation on 2AP sensitivity in the ExoI ExoVII RecJ deficient background 2AP, a purine analog, causes cell lethality of triple-exonuclease-deficient mutants (Burdett et al., 2001) and Dam methylase-deficient mutants (Glickman and Radman, 1980), which fail to drive methyl-directed MMR. We quantified the 2AP sensitivities of the wild-type strain, the recQ mutant and various DxonA DxseA recJ mutants. The wildtype strain showed no sensitivity to 2AP at high concentrations (Fig. 1A). We confirmed that the DxonA DxseA recJ mutant was sensitive to 2AP, because its survival decreased to around 0.03 at a concentration of P0.8 mM 2AP (Fig. 1B). Next, we tested whether a recQ defect caused sensitivity to 2AP. The recQ null mutation alone did not affect resistance to 2AP (Fig. 1A); however, on the DxonA DxseA recJ mutant background, the recQ mutation caused 100-fold increase in 2AP sensitivity, as the survival fractions of the recQ DxonA DxseA recJ mutant decreased to 0.00025 at P0.8 mM 2AP (Fig. 1B). The 100-fold increase in 2AP sensitivity was observed even at 0.2 mM 2AP (Fig. 1B). 3.2. Suppression of 2AP-induced lethality due to the recQ defect by mutation of mutS or uvrD The removal of any of the MMR initiator proteins MutS, MutL, MutH and UvrD from DxonA DxseA recJ mutant cells suppresses 2AP-induced cell death (Burdett et al., 2001). We tested the effect of inactivation of MutS or UvrD on 2AP-induced lethality due to the recQ defect. The addition of the mutS insertion mutation or the uvrD insertion mutation completely suppressed 2AP sensitivity in the recQ DxonA DxseA recJ mutant (Fig. 1C), because these two strains were as resistant to 2AP as the wild-type strain (Fig. 1A). These results show that the functions of MutS and UvrD lead to increased sensitivity to 2AP due to RecQ inactivation. 3.3. Effect of the recQ mutation on filamentation in the ExoI ExoVII RecJ deficient background 2AP induces SOS responses in a MutSLH-dependent manner (Bebenek and Janion, 1985). DxonA DxseA recJ mutant cells display filament formations characteristic of SOS responses in the presence of 2AP at lower concentrations (Burdett et al., 2001). We tested whether a recQ defect caused 2AP-induced filamentation. The recQ null mutation alone did not affect cell-length distribution in the presence of 2AP (data not shown). Next, in order to test whether the addition of the recQ mutation to DxonA DxseA recJ mutant cells affects filament formation, we compared cell-length distributions of the DxonA DxseA recJ mutant and its recQ derivative. In the absence of 2AP, the recQ mutation did not affect cell-length distribution (Fig. 2). We confirmed that 2AP significantly increased the fraction of DxonA DxseA recJ mutant cells over 5 lm in length (Fig. 2A and C; (Burdett et al., 2001)). We found that the fraction of recQ DxonA DxseA recJ mutant cells over 5 lm in length was significantly

122

Y. Yamana et al. / Plasmid 63 (2010) 119–127

A Survival fraction

1 0.1 Wild-type 0.01

recQ

0.001 0.0001

B

xonA xseA recJ

Survival fraction

1

xonA xseA recJ recQ

0.1

0.01

0.001 0.0001

C

Survival fraction

1 0.1

xonA xseA recJ recQ mutS 0.01

xonA xseA recJ recQ uvrD

0.001 0.0001

0

0.4

0.8 2AP, mM

1.2

1.6

Fig. 1. Effect of 2AP treatment on the growth of E. coli recQ and exonuclease mutants. (A) 2AP sensitivities of the wild-type strain and its recQ derivative. (B) 2AP sensitivities of the DxonA DxseA recJ deficient mutant and its recQ derivative. (C) 2AP sensitivities of the DxonA DxseA recJ recQ uvrD mutant and the DxonA DxseA recJ recQ mutS mutant. Survival fractions for the wild-type strain and its derivatives, with the relevant genotypes indicated in each paenl, were determined by dividing the viable cell titers on 2AP-containing plates by those grown on 2AP-free plates. The average values (the central points in symbols) with two measured values (the ends of vertical bars), which were obtained from two cultures for each strain, were shown as survival fractions.

higher than that of DxonA DxseA recJ mutant cells in the presence of 2AP only (Fig. 2C). Thus, these results show that the recQ defect causes enhancement of filamentation at lower doses of 2AP that give a higher survival fraction. We will discuss the relationship between the enhancement of the 2AP-induced filamentation and the increase of the 2AP sensitivity due to the recQ mutation. 3.4. Effect of the recQ mutation on spontaneous mutation rate We measured the frequency of Rif-resistant clones generated in normal cell growth to estimate the rate of

spontaneously generated mutations, which occur exclusively in the rpoB gene (Jin and Gross, 1988). The wild-type strain BT199 showed a mutation rate of 1.3  108 mutations/cell generation (Table 2). The DxonA DxseA recJ mutant lacking three single-strand specific exonucleases showed a mutation rate of 1.5  107 mutations/cell generation, which was higher several fold than that of a strain with the same genotype in a previous report (Viswanathan and Lovett, 1998). This mutation rate is 12-fold higher than that of the wild-type strain (Table 2). We tested whether a recQ defect increased the Rif-resistant mutation rate. The recQ null mutation alone did not increase the mutation

123

Y. Yamana et al. / Plasmid 63 (2010) 119–127

A Number of cells

25

STL2701 ( xonA xseA recJ )

20 None

15

2AP

10 5 0 30

Number of cells

B

KTYY13 ( xonA xseA recJ recQ )

25 20 15 10 5 0

0

1

2

3

4

5

6

7

C

8 9 10 11 12 13 14 15 16 17 18 19 Cell length, µm

none

Strain

STL2701

2AP 2

Cells (%) 0-5 µm

5-10 µm

> 10 µm

80

20

0

significance (P ) STL2701vs KTYY13

0-5 µm

5-10 µm

> 10 µm

44

40

16

81

18

1

significance (P ) STL2701vs KTYY13

significance (P ) none vs 2AP

1.1E-05* 0.00021*

0.56 KTYY13

2 2

Cells (%)

14

50

36

3.9E-15*

Fig. 2. Effect of 2AP treatment on cell-length distributions of recQ and exonuclease mutants. (A) A graphical presentation of cell-length distributions of the DxonA DxseA recJ deficient mutant (STL2701 strain). (B) A graphical presentation of cell-length distributions of the DxonA DxseA recJ recQ deficient mutant (KTYY13 strain). Cultures of two strains, with the relevant genotypes indicated, were grown with or without 0.2 mM 2AP and cell lengths were determined as described in Section 2. For each culture 70 cells were classified based on their length. (C) Statistical analysis of cell-length distributions using the v2 test. The probability (P) of the difference being due to chance is shown, with *indicating a high level of significant difference.

Table 2 Rates of spontaneous mutations in the rpoB locus of recQ and exonuclease mutants. Strain

Relevant genotype

Rifr mutation rate  108 [SD  108]

Rate relative to BT199

Rate relative to STL2701

n

BT199 KTYY33 STL2701 KTYY13 KTYY23 KTYY25

Wild-type recQ DxonA DxseA DxonA DxseA DxonA DxseA DxonA DxseA

1.28 [0.442] 1.63 [0.504] 14.9 [5.35] 8.53 [8.14] 188 [98.6] 234 [83.7]

1 1.3 12 7 147 183

— — 1 0.57 13 16

12 11 4 4 4 4

recJ recJ recQ recJ mutS recJ recQ mutS

rate, and also did not increase the mutation rates on a DxonA DxseA recJ background and on a mutS DxonA DxseA recJ background (Table 2). These results indicate that RecQ knockout does not cause any effects on mutagenesis conferring Rif resistance. Since a mutation in thyA encoding thymidilate synthetase confers resistance to Tmp, a potent inhibitor of dihydrofolate reductase, we measured the frequency of Tmp-resistant clones generated in normal cell growth to estimate the rate of spontaneously generated mutations in the thyA gene (Viswanathan et al., 2000). The wild-type strain BT199 showed a mutation rate of 2.1  107 mutations/cell genera-

tion (Table 3). The DxonA DxseA recJ mutant lacking three single-strand specific exonucleases showed about a 13-fold increase in the mutation rate, relative to the wild-type strain (Table 3). These two results were similar to those reported previously (Viswanathan et al., 2000). We first tested whether a recQ defect increased the thyA mutation rate. The recQ null mutation alone did not increase the mutation rate; however, on a DxonA DxseA recJ background, the recQ null mutation increased the mutation rate 2.4-fold relative to the DxonA DxseA recJ mutant (Table 3). This increase was small but statistically significant. We next examined the effect of the recQ mutation in a mutS DxonA DxseA recJ genetic

124

Y. Yamana et al. / Plasmid 63 (2010) 119–127

Table 3 Rates of spontaneous forward mutations in the thyA locus of recQ and exonuclease mutants. Strain

Relevant genotype

Tmpr mutation ratea  107 [95% confidence interval]b

Rate relative to BT199

Rate relative to STL2701

n

BT199 KTYY33 STL2701 KTYY13 KTYY23 KTYY25

Wild-type recQ DxonA DxseA DxonA DxseA DxonA DxseA DxonA DxseA

2.08 [1.52–2.71] 1.43 [0.899–1.97] 26.6 [22.1–30.6] 62.6 [53.9–78.5] 128 [110–156] 105 [57.9–128]

1 0.69 13 30 61 50

— — 1 2.4 4.8 4.0

24 24 49 57 24 24

recJ recJ recQ recJ mutS recJ recQ mutS

a The mutation rate to Tmp resistance for each culture was determined as described in the Section 2. The mutation rate is an estimation of the number of mutation events that occurred in the population based on the observed number of Tmp-resistant colonies for each culture. b The estimated mutation rate for each culture was ranked in a given data set of n cultures and a 95% confidence interval was established for the median of the n samples based on Table A-25a of Dixon and Massey (1969).

background. The mutS DxonA DxseA recJ mutant showed a 4.8-fold increase in the mutation rate relative to the DxonA DxseA recJ mutant; and the recQ mutS DxonA DxseA recJ mutant showed a 4.0-fold increase relative to the DxonA DxseA recJ mutant (Table 3), indicating that the recQ mutation did not cause an increase in the mutation rate of mutS DxonA DxseA recJ mutant cells. These results indicate that RecQ knockout causes inhibitory effects on the methyl-directed mismatch repair process, which is initiated by the MutSLH complex, only in the absence of the three exonucleases. 3.5. Sequence analysis of thyA forward mutations We determined the nucleotide sequences of the thyA mutations generated from the DxonA DxseA recJ mutant

or the recQ DxonA DxseA recJ mutant. As shown in Table 4, these mutations fell into four categories: base substitution, sequence substitution, frameshift, and duplication/deletion. The class of sequence substitution that we observed was identical to that seen at a mutational hotspot generated in a natural quasipalindrome characteristic of the E. coli thyA locus (Viswanathan et al., 2000). This is another important source of mutational hotspots than endogenous DNA lesions and replicational errors (Maki, 2002). We also identified the putative misalignment products from both mutants. Their sequences indicate that the duplication or deletion events of a segment of 40–220 nucleotides occurred between repeated tracts of two to nine nucleotides (last row of Table 4 and Fig. 3). An example (at site 480) is the duplication of a 44-nucleotide segment between

Table 4 Spontaneous forward mutations in the thyA locus of DxonA DxseA recJ and DxonA DxseA recJ recQ mutants. Class of mutation

STL2701 (DxonA DxseA recJ)

Base substitution

KTYY13(DxonA DxseA recJ recQ)

Original nucleotide

Mutant nucleotide

n

Sitea

Original nucleotide

Mutant nucleotide

n

52 289 452

A C A

G A T

1 1 1

59 294 437 616 616 626

A G G A A A

C T T C G T

1 1 1 1 1 1

131

T

A

124 131

TTCC T

A A

204

CA

–

64 679

CTGCCG

ACCGGA –

90

TGGTC(220)TGGT C

TGGTC(220)TGGT C(220)TGGTC

1

480

TTGCCAGCT(44)TT GCCAGCT

482

GCCAGC(47)GCC AGC TGGA(53)TGGA

GCCAGC

1

556

CA(69)CA

TGGA(53)TGGA(5 3)TGGA CCGGTGGC(168)C CGGTGGC(168)CC GGTGGC

1

670

CCGC(86)CCGC

TTGCCAGCT(44)T TGCCAGCT(44)TT GCCAGCT CA(69)CA(69) CA CCGC(86)CCGC(86) CCGC

Site

a

Fractionb Sequence substitution

3/33

Fractionb Frameshift

605

a

1

CCGGTGGC(168)C CGGTGGC

2 21 23/34

1/33

581

Fractionb

6/34

25/33

Fractionb Duplication/ Deletion

25

1 1 2/34 1

1 1

1

4/33

3/34

Numbered on the coordinate for the thyA coding region (1–792). b The number of mutations in each class divided by the total number of mutations. The level of statistical significance (P) between the fractions of STL2701 strain and the fractions of KTYY13 strain is calculated using v2 test; P is 0.67, indicating that they are not significantly different.

125

579

TTGCCAGCT AACGGTCGA

541

533

488

A

480

Y. Yamana et al. / Plasmid 63 (2010) 119–127

TTGCCAGCT AACGGTCGA

CH3 GATC CTAG

B

CH3 GATC

UvrD CTAG 5

C

-Exo

CH3 GATC

D

CT AG

CH3 GATC CTAG

CTAG

E

Fig. 3. A misalignment model for the duplication events. This diagram shows an example at site 480. In the duplex DNA around the coding region 480–541, two identical nine-base tracts (written in bold) are separated by 44 bases (A). The lower strand of the duplex is newly synthesized by DNA replication machinery, but not modified at the closest GATC site (579) yet (A). The unmodified GATC site (579) is incised by the MutSLH complex. UvrD helicase invades from the incision, unwinds the duplex leftward by its 30 –50 helicase activity (B). UvrD continues to unwind beyond the region 480–541(C). A displaced, long single strand remains in the absence of 50 –30 single-strand exonucleases; thus, this strand tends to realign (C). In this case, the complementary strand of the second tract 533–541 is misaligned with the first tract 480–488, associated with the formation of an unpaired 50-base loop (C). DNA polymerase III starts repair synthesis at the GATC site (577) but collapses at the obstruction, which is the misaligned duplex associated with the 50-base loop (D). The unpaired 50 -end (40-base single strand) is cleaved by an endonuclease and the resultant 50 end of the mis-paired tract, 30 -AACGGTCGA-50 , is ligated to the tip of the synthesizing strand (dotted line with arrowhead) (E). The next round of chromosome replication completes this duplication product (E).

repeated tracts of TTGCCAGCT (Fig. 3). Another one (at site 482) is the deletion of a 47-nucleotide segment between repeated tracts of GCCAGC. The fraction of each class among the mutations analyzed was almost equal between these two host strains (Table 4). This indicates that the twofold increase of the thyA mutation rate was not attributed to increase of a limited class of mutations. 4. Discussion 4.1. Mismatch-induced lethality due to a recQ defect Previously it was reported that 2AP induces lethality in single-strand specific exonuclease-deficient mutants (xonA xseA recJ mutations) (Burdett et al., 2001). This finding indicates that there may be deleterious structures gener-

ated in the absence of the three exonucleases. We found that a recQ defect increases 2AP-induced lethality in exonucleases-deficient background, but that the recQ defect alone did not have any effect. These results indicate that the specific structures leading to cell death might be made more abundant by the addition of the recQ defect. Inactivation of MutS or UvrD totally suppressed the 2AP hypersensitivity caused by the defects of RecQ and the exonucleases. These suppression phenomena indicate that MutSLH and UvrD successively act on a 2AP:pyrimidine mismatch and that this action might produce the cytotoxic structure. What structure would be cytotoxic? The suppression of the toxicity by uvrD mutation suggests that the twin-nick structure made by MutSLH is not yet toxic. After MutSLH makes two nicks, UvrD invades at the 50 -nick and displaces the complementary strand to generate a

126

Y. Yamana et al. / Plasmid 63 (2010) 119–127

single-strand gap region. Without the exonucleases, the displaced complementary strand might partly realign or misalign (Fig. 3) onto the single-strand gap. DNA polymerase III holoenzyme fills the single-strand gap while displacing the complementary strand. Strand-displacement synthesis (SDS) can lead to rolling circle replication with deleterious consequences when the 50 -end of an Okazaki fragment is incompatible with the synthesizing tip (Garg and Burgers, 2005). If the 30 -nick is closed by ligation before DNA polymerase III enzyme reaches that site, SDS might continue to produce a longer single-stranded DNA, which leads to cytotoxic effects. The realigned or the misaligned structure generated in exonucleases-deficient cells might delay repair synthesis, meanwhile the 30 -nick might be closed. MMR exonucleases, therefore, are thought to be necessary for quick digestion of the displaced strand, leading to efficient repair synthesis. If RecQ contributes to eliminate the realigned or the misaligned structure, the uncontrolled SDS might occur more frequently by adding a recQ mutation to the exonuclease-deficient strain. 4.2. Is RecQ involved in regulation of mismatch-induced SOS response? 2AP induces SOS responses, which are greatly enhanced in dam3 mutants (Bebenek and Janion, 1985). Moreover, it was shown that 2AP-induced lethality of single-strand exonuclease mutants was associated with cell filamentation due to inhibition of septum formation in a MutSLHand UvrD-dependent manner (Viswanathan et al., 2001). As mentioned above, uncontrolled SDS (like rolling circle replication) generates a long single strand. This product could operate as an SOS signal. We found that 2AP-induced filamentation was enhanced by the introduction of the recQ mutation into the exonuclease mutant. RecQ might be needed for quick removal of the realigned strand onto the single-strand gap. If RecQ dissolves the realigned structure to assist MMR synthesis by the DNA polymerase III holoenzyme, the recQ defect, conversely, would delay the MMR synthesis. During this delay, the 30 -nick might be closed and SDS might then be converted to rolling circle replication in association with displacement of a long single-strand DNA. RecQ promotes SOS signaling at stalled replication forks (Hishida et al., 2004); thus, RecQ might be involved in both the production and the attenuation of SOS signals. 4.3. Is RecQ involved in methyl-directed MMR? The recQ defect did not increase at all in the rate of Rifresistant mutations in the exonuclease-deficient genetic background, but slightly increased in the rate of thyA mutations in the same background. The gene of rpoB is mostly causative for generation of Rif-resistant cells and the mutations conferring Rif resistance in the gene are limited to the classes of single base substitution and three- or six-nucleotide insertion within the central portion of 180 out of 1342 sense codons (Jin and Gross, 1988). It might be too narrow to detect the weak effect of the recQ defect on mutation rates. As the three exonucleases are thought to have a major role in the quick digestion of the displaced

strand in later steps of methyl-directed MMR, we hypothesized that RecQ prevents the formation of a realigned structure, which tends to be caused by the absence of the exonucleases. If so, recQ mutations would greatly increase mutation rates; however, the recQ mutation did not induce mutagenesis so much. Classes of spontaneous mutations that we observed were base substitution, sequence substitution, frameshift, and duplication/deletion. Classes except for duplication/deletion might be caused by realignment (not misalignment) between a single-strand gap region and a displaced complementary strand with a wrong single nucleotide or an oligonucleotide loop. The fourth class of mutations duplication/deletion might be induced by misalignment as depicted in Fig. 3C. Secondary DNA structures, acting as obstructions, might impede the progression of MMR synthesis when they are generated during methyldirected MMR. These sequence information confirmed that the absence of exonucleases induced the realignment and misalignment reactions, and the addition of the RecQ inactivation weakly caused overall increase of the classes of mutations. The fact that the recQ mutation enhanced a SOS response suggests the possibility that the small increase in thyA mutation rate was caused by SOS mutagenesis involving other DNA polymerases, e.g. DNA polymerases II and V (Pham et al., 2001) rather than by failure of MMR. 4.4. Examples of RecQ reactions independent of RecJ If a double defect of RecQ and RecJ causes more severe phenotypes than either single defect, then that would suggest that RecQ and RecJ apparently work independently. However, if a double defect of RecQ and RecJ causes no more severe phenotypes than either single defect, then that would suggest that RecQ and RecJ apparently work together in a common function. Several genetic results have shown either dependent or independent actions of RecQ and RecJ. In the recBC sbcA mutant background, a double defect of RecQ and RecJ never caused a more severe phenotype than either single defect; the double defect partially rescued the dysfunctions due to single defect of RecJ (Kusano et al., 1994; Lovett and Sutera Jr., 1995). These results indicated that RecQ works with RecJ in RecET-directed homologous recombination. However, on the DxonA DxseA mutant background, a double defect of RecQ and RecJ caused more severe phenotypes in terms of 2AP sensitivity than either single defect. In addition, on the recBC sbcBC mutant background, a double defect of RecQ and RecJ caused a more severe phenotype in terms of UV sensitivity than either single defect (data not shown). These results indicate that RecQ works independently of RecJ in these aspects. We suppose that the interplay of RecQ helicase with a MutSLH-dependent process occurs independently of RecJ that works as a MMR exonuclease. 4.5. Interplay between human RecQ homologues and mismatch repair initiation factors A human RecQ-type protein RECQ1 physically interacts with mismatch recognition complex MSH2/6 (MutS homologues) and MLH1 (a MutL homologue) and its helicase

Y. Yamana et al. / Plasmid 63 (2010) 119–127

activity for a 30 -flap DNA substrate is stimulated by RECQ1MSH2/6 interaction (Doherty et al., 2005). Another human RecQ-type protein WRN physically interacts with mismatch recognition complex MSH2/6, MSH2/3 and MLH1/ PMS2 complexes and its helicase activity for forked DNA substrates is stimulated by WRN-MSH2/6 and WRNMSH2/3 interactions (Saydam et al., 2007). It is thought that the role of the interactions of these human RecQ-type proteins with the mismatch recognition complexes are the suppression of homologous recombination by unwinding the heteroduplex with mismatches, which is a recombination intermediate generated by homologous pairing proteins, rather than human DNA mismatch repair (Doherty et al., 2005; Saydam et al., 2007). Acknowledgments We are grateful to Susan T. Lovett and Wilfried Wackernagel for providing E. coli strains. We thank Gerald Smith, Susan T. Lovett and Ayumi Kusano for helpful comments on the paper. This work has been supported by grant-inaids from Exploratory Research Program and Scientific Research Program of the Ministry of Education Culture, Sports, Science, and Technology of Japan. References Acharya, S., Foster, P.L., Brooks, P., Fishel, R., 2003. The coordinated functions of the E. coli MutS and MutL proteins in mismatch repair. Mol. Cell 12, 233–246. Au, K.G., Welsh, K., Modrich, P., 1992. Initiation of methyl-directed mismatch repair. J. Biol. Chem. 267, 12142–12148. Bachmann, B.J., 1996. Derivation and genotypes of some mutant derivatives of Escherichia coli K12. In: Neidhardt, F.C., Curtiss-III, R., Ingraham, J.L., Lin, E.C.C., Low, K.B., Magasanik, B., Reznikoff, W.S., Riley, M., Schaechter, M., Umbarger, H.E. (Eds.), Escherichia coli and Salmonella. Cellular and Molecular Biology. ASM Press, Washington, D.C., pp. 2460–2488. Bebenek, K.K., Janion, C., 1985. Ability of base analogs to induce the SOS response: effect of a dam mutation and mismatch repair system. Mol. Gen. Genet. 201, 519–524. Burdett, V., Baitinger, C., Viswanathan, M., Lovett, S.T., Modrich, P., 2001. In vivo requirement for RecJ, ExoVII, ExoI, and ExoX in methyldirected mismatch repair. Proc. Natl. Acad. Sci. USA 98, 6765–6770. Cooper, D.L., Lahue, R.S., Modrich, P., 1993. Methyl-directed mismatch repair is bidirectional. J. Biol. Chem. 268, 11823–11829. Courcelle, J., Donaldson, J.R., Chow, K.H., Courcelle, C.T., 2003. DNA damage-induced replication fork regression and processing in Escherichia coli. Science 299, 1064–1067. Courcelle, J., Hanawalt, P.C., 1999. RecQ and RecJ process blocked replication forks prior to the resumption of replication in UVirradiated Escherichia coli. Mol. Gen. Genet. 262, 543–551. Dao, V., Modrich, P., 1998. Mismatch-, MutS-, MutL-, and helicase IIdependent unwinding from the single-strand break of an incised heteroduplex. J. Biol. Chem. 273, 9202–9207. Dixon, W.J., Massey, F.J., 1969. Introduction to Statistical Analysis. McGraw-Hill, New York. Doherty, K.M., Sharma, S., Uzdilla, L.A., Wilson, T.M., Cui, S., Vindigni, A., Brosh-Jr, R.M., 2005. RECQ1 helicase interacts with human mismatch repair factors that regulate genetic recombination. J. Biol. Chem. 280, 28085–28094. Ellis, N.A., Groden, J., Ye, T.Z., Straughen, J., Lennon, D.J., Ciocci, S., Proytcheva, M., German, J., 1995. The Bloom’s syndrome gene product is homologous to RecQ helicases. Cell 83, 655–666. Garg, P., Burgers, P.M.J., 2005. How the cell deals with DNA nicks. Cell Cycle 4, 221–224.

127

Glickman, B.W., Radman, M., 1980. Escherichia coli mutator mutants deficient in methylation-instructed DNA mismatch correction. Proc. Natl. Acad. Sci. 77, 1063–1067. Grilley, M., Welsh, K.M., Su, S.S., Modrich, P., 1989. Isolation and characterization of the Escherichia coli mutL gene product. J. Biol. Chem. 264, 1000–1004. Hanada, K., Ukita, T., Kohno, Y., Saito, K., Kato, J., Ikeda, H., 1997. RecQ DNA helicase is a suppressor of illegitimate recombination in Escherichia coli. Proc. Natl. Acad. Sci. USA 94, 3860–3865. Hishida, T., Han, Y.W., Shibata, T., Kubota, Y., Ishino, Y., Iwasaki, H., Shinagawa, H., 2004. Role of the Escherichia coli RecQ DNA helicase in SOS signaling and genome stabilization at stalled replication forks. Genes Dev. 18, 1886–1897. Jin, D.J., Gross, C.A., 1988. Mapping and sequencing of mutations in Escherichia coli rpoB gene that lead to Rifampicin resistance. J. Mol. Biol. 202, 45–58. Kusano, K., Sunohara, Y., Takahashi, N., Yoshikura, H., Kobayashi, I., 1994. DNA double-strand break repair: genetic determinants of flanking crossing-over. Proc. Natl. Acad. Sci. USA 91, 1173–1177. Lahue, R.S., Au, K.G., Modrich, P., 1989. DNA mismatch correction in a defined system. Science 245, 160–164. Lea, D.E., Coulson, C.A., 1949. The distribution of the numbers of mutants in bacterial populations. J. Genet. 49, 264–285. Lovett, S.T., Sutera Jr., V.A., 1995. Suppression of RecJ exonuclease mutants of Escherichia coli by alterations in DNA helicase II (uvrD) and IV (helD). Genetics 140, 27–45. Maki, H., 2002. Origins of spontaneous mutations: Specificity and directionality of base-substitution, frameshift, and sequencesubstitution mutagenesis. Annu. Rev. Genet. 36, 279–303. Mechanic, L.E., Frankel, B.A., Matson, S.W., 2000. Escherichia coli MutL loads DNA helicase II onto DNA. J. Biol. Chem. 275, 38337–38346. Modrich, P., Lahue, R., 1996. Mismatch repair in replication fidelity, genetic recombination, and cancer biology. Annu. Rev. Biochem. 65, 101–133. Nakayama, H., Nakayama, K., Nakayama, R., Irino, N., Nakayama, Y., Hanawalt, P.C., 1984. Isolation and genetic characterization of a thymineless death-resistant mutant of Escherichia coli K12: identification of a new mutation (recQ1) that blocks the RecF recombination pathway. Mol. Gen. Genet. 195, 474–480. Ogawa, H., Ohyama, T., Katsura, E., Katoh, Y., 1997. Characterization of the pUC19-lacZC141 reversion system for assaying chemical mutagenesis. Mutat. Res. 394, 141–151. Pham, P., Rangarajan, S., Woodgate, R., Goodman, M.F., 2001. Roles of DNA polymerases V and II in SOS-induced error-prone and error-free repair in Escherichia coli. Proc. Natl. Acad. Sci. USA 98, 8350–8354. Ronen, A., 1979. 2-AMINOPURINE. Mutat. Res. 75, 1–47. Saydam, N., Kanagaraj, R., Dietschy, T., Garcia, P.L., PeÒa-Diaz, J., Shevelev, I., Stagljar, I., Janscak, P., 2007. Physical and functional interactions between Werner syndrome helicase and mismatch-repair initiation factors. Nucleic Acids Res. 35, 5706–5716. Su, S.S., Lahue, R.S., Au, K.G., Modrich, P., 1988. Mispair specificity of methyl-directed DNA mismatch correction in vitro. J. Biol. Chem. 263, 6829–6835. Thoms, B., Wackernagel, W., 1998. Interaction of RecBCD enzyme with DNA at double-strand breaks produced in UV-irradiated Escherichia coli: requirement for DNA end processing. J. Bacteriol. 180, 5639– 5645. Viswanathan, M., Burdett, V., Baitinger, C., Modrich, P., Lovett, S.T., 2001. Redundant exonuclease involvement in Escherichia coli methyldirected mismatch repair. J. Biol. Chem. 276, 31053–31058. Viswanathan, M., Lacirignola, J.J., Hurley, R.L., Lovett, S.T., 2000. A novel mutational hotspot in a natural quasipalindrome in Escherichia coli. J. Mol. Biol. 302, 553–564. Viswanathan, M., Lanjuin, A., Lovett, S.T., 1999. Identification of RNase T as a high-copy suppressor of the UV sensitivity associated with single-strand DNA exonuclease deficiency in Escherichia coli. Genetics 151, 929–934. Viswanathan, M., Lovett, S.T., 1998. Single-strand DNA-specific exonucleases in Escherichia coli: roles in repair and mutation avoidance. Genetics 149, 7–16. Yu, C.E., Oshima, J., Fu, Y.H., Wijsman, E.M., Hisama, F., Alisch, R., Matthews, S., Nakura, J., Miki, T., Ouais, S., Martin, G.M., Mulligan, J., Schellenberg, G.D., 1996. Positional cloning of the Werner’s syndrome gene. Science 272, 258–262.