Involvement of genes of genome maintenance in the regulation of phase variation frequencies in Neisseria meningitidis

Patricia Martin, Li Sun, Derek W. Hood and E. Richard Moxon

Molecular Infectious Diseases Group, University of Oxford, Department of Paediatrics, Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, Oxford OX3 9DS, UK

Correspondence
Patricia Martin
patricia.martin{at}paediatrics.ox.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In Neisseria meningitidis, the reversible expression of surface antigens, i.e. phase variation, results from changes within repeated simple sequence motifs located in coding or promoter regions of the genes involved in their biosynthesis. The mutation rates of these simple sequences, which have a major influence on the generation of phenotypic diversity, can affect the fitness of the population. The aim of the present study was to investigate the involvement of genetic factors involved (mutS and dam) and not yet analysed (drg and dinB) in the regulation of phase variation frequencies of genes associated with a variety of repeat tracts. The frequency of frameshifts occurring in the polycytidine (polyC) tracts associated with siaD, spr and lgtG and in the tetranucleotide (TAAA) repeat tract associated with nadA was determined by colony immunoblotting or using the lacZ gene as a reporter. Inactivation of mutS increased the frequency of phase variation of genes presenting homopolymeric tracts of diverse length. Overexpression of dinB enhanced the instability of the homopolymeric tract associated with siaD. Investigation of the dam locus in a population of genetically distinct N. meningitidis strains revealed that 27 % of strains associated with invasive disease contained the dam gene. In all strains where a Dam function was absent, the drg gene had been inserted into the dam locus. Disruption of dam and drg in strains representative of each genotype, i.e. dam+/drg and dam/drg+, did not modify phase variation frequencies. In contrast to the effects of certain genes on homopolymeric tracts, none of the genetic factors investigated affected the stability of tetranucleotide repeat tracts.


Abbreviations: MMR, methyl-directed mismatch repair; polyC, polycytidine; polyG, polyguanine


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Neisseria meningitidis, an obligate commensal bacterium which colonizes the human nasopharynx, is also a major cause of septicaemia and meningitis worldwide. The mechanisms by which N. meningitidis is able to transmit between hosts, persist in the upper respiratory tract and evade destruction by the immune system are central to its survival. Its adaptation to changing environments is thought to be achieved, in part, by the generation of genetic variants presenting modified surface structures. This reversible switching of antigens, i.e. phase variation, can result from changes within repeated simple sequence motifs located in coding or promoter regions of genes involved in the biosynthesis of the surface structures. These genes are often referred to as contingency loci (Moxon et al., 1994). Phase variable expression has been experimentally demonstrated for surface antigens such as opacity proteins, lipopolysaccharides, capsular polysaccharides, pili, haemoglobin receptors, PorA and Opc outer-membrane proteins, a putative protease and the NadA adhesin (de Vries et al., 1996; Jennings et al., 1998, 1999; Hammerschmidt et al., 1996; Jonsson et al., 1991; Lewis et al., 1999; Richardson & Stojiljkovic,1999; van der Ende et al., 2000; Sarkari et al., 1994; Martin et al., 2003). We recently showed that 47 genes were likely to be regulated by such a mechanism in N. meningitidis strain MC58 (Martin et al., 2003).

The frequency of phase variation is a major factor controlling the genetic diversity and consequently the fitness of bacterial populations. Therefore, the study of cis- and trans-acting factors regulating the instability of simple sequence repeats is essential to understand the relevance of simple sequence tracts in the biology of the bacterium. The sequence and the length of a repeat tract are major cis-acting factors known to regulate repeat instability in human (Rolfsmeier et al., 2001), yeast (Tran et al., 1997; Wierdl et al., 1997) and bacterial organisms (de Bolle et al., 2000). In N. meningitidis, a linear relationship between the length of the hmbR polyguanine (polyG) tract and phase variation frequency was reported in serogroup A strains (Richardson et al., 2002).

Replication and post-replicative repair systems are also involved in the control of the plasticity of repeat tracts. In yeast, DNA polymerase proofreading and methyl-directed mismatch repair (MMR) correct frameshifts preferentially in short and long homonucleotide runs, respectively (Tran et al., 1997). In Escherichia coli, the overexpression of dinB, which encodes the error-prone DNA polymerase IV, results in an increased frequency of frameshifts in short homopolymeric tracts (Kim et al., 1997). In Haemophilus influenzae, inactivation of mutS, a key component of MMR, and polI destabilize dinucleotide and tetranucleotide repeat tracts, respectively (Bayliss et al., 2002). In N. meningitidis, investigation of two contingency loci revealed that the mismatch repair genes mutS and mutL have a role in controlling phase variation frequencies (Richardson & Stojiljkovic, 2001; Richardson et al., 2002).

Finally, the involvement of Dam (DNA adenine methylase) methylation in the regulation of phase variation is unclear. Dam methylation was shown to have no effect upon phase variation of tetranucleotides in H. influenzae (Bayliss et al., 2002) and of polyG tracts in N. meningitidis (Richardson & Stojiljkovic, 2001). In contrast, Bucci et al. (1999) proposed that the absence of Dam activity was responsible for the high frequency of N. meningitidis capsule phase variation and for the heightened virulence of some strains.

The effect of trans-acting factors on phase variation rates cannot be assumed to be similar for different genes. Indeed, in H. influenzae, mutations in mutS were not found to alter switching rates of dinucleotide repeats of the pilus locus but did alter switching rates of a similar dinucleotide repeat tract in another location in the chromosome (Bayliss et al., 2004). Therefore, the aim of the present study was to investigate the role of genes likely to influence phase variation in N. meningitidis in the regulation of switching frequencies for different types of repeat tracts. The mutS and dinB but not the dam or drg genes influenced the frequency of phase variation of genes associated with homopolymeric tracts. Contrary to this observation, the stability of tetranucleotide repeat tracts did not depend on the genetic factors investigated in the study.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and culture.
N. meningitidis strains used in this study (Table 1) were primarily from a collection of 107 strains representative of all major serogroups (A, B, C, W135, X, Y and Z; Maiden et al., 1998). The serogroup B strain MC58, for which the complete genome sequence was obtained (Tettelin et al., 2000) and which was previously used in related studies, was included. All strains were grown overnight at 37 °C on brain heart infusion (BHI) medium base (Oxoid) with 1·5 % (w/v) added agar supplemented with Levinthal's base (Alexander, 1965), in an atmosphere of 5 % CO2. For selection of transformants, kanamycin (100 µg ml–1) or erythromycin (5 µg ml–1) was included in the growth medium.


View this table:
[in this window]
[in a new window]
 
Table 1. N. meningitidis Dam+ and Dam strains investigated in this study

NIBSC: National Institute for Biological Standards and Control.

 
E. coli strains DH5{alpha} and GM2163 (dam dcm) were grown on Luria–Bertani (LB) broth or plates at 37 °C and were used to propagate plasmids. When necessary, E. coli was grown in the presence of kanamycin (50 µg ml–1) or ampicillin (100 µg ml–1).

Recombinant DNA techniques and nucleotide sequence analysis.
Restriction endonucleases were obtained from Roche or New England Biolabs and used according to the manufacturers' instructions.

PCR amplification was performed using Taq polymerase from Gibco-BRL. Reactions consisted of 30 cycles including 1 min of denaturation at 94 °C, 1 min of annealing at the appropriate temperature and 1 min of extension at 72 °C. The primers used in the study were purchased from Genosys and are listed in Table 2. Pairs of primers siaDF/siaDR, lacZB1/sprhind, porA1/porA2, lgtGC/lgtGD and nadAF/nadAR were used to PCR amplify DNA fragments containing the repeat tract associated with genes siaD, spr, porA, lgtG and nadA, respectively. The DNA sequences of both strands of these amplicons were obtained in a sequencing reaction with fluorescent dye-labelled dideoxynucleotide terminators using Ampli Taq DNA Polymerase FS (Perkin Elmer), according to the instructions supplied by Applied Biosystems. Primers used in sequencing reactions were siaDF and siaDR for siaD, sprF2, sprF3 and lacZB1 for spr, porA1 and porA2 for porA, and nadAF and nadAR for nadA (Table 2). Clear sequence of the repeat tract could not be obtained from lgtG PCR products, an inherent trait of long homopolymeric tracts. lgtG PCR products were cloned into pT7 Blue vector (Novagen) and sequenced using primers lgtGC, lgtGD and the plasmid-specific primers M13 For and M13 Rev. The sequences were analysed on an automatic sequencer (model 377, Applied Biosystems). Nucleotide sequence data were analysed using the GCG program (Wisconsin package, version 10.2-Unix, Genetics Corporation Group).


View this table:
[in this window]
[in a new window]
 
Table 2. Oligonucleotide primers used in the study

 
Pairs of primers DamF1/DamB2 and DrgF1/DrgB2 (Table 2) were designed using the sequence published by Bucci et al. (1999) and used to PCR amplify internal fragments of dam and drg, respectively (Fig. 1). The upstream and downstream regions of the drg gene and adjacent DNA were amplified by PCR with pairs of primers Drg5F/DrgB3 and Drg3F/DamB5, respectively (Table 2, Fig. 1).



View larger version (8K):
[in this window]
[in a new window]
 
Fig. 1. Location of primers used to amplify internal fragments of dam and drg genes and the regions flanking the drg gene. Gene names are indicated in the boxes representing the genes. Half-arrows symbolize the primers. The names of the primers are also indicated.

 
Southern blots and hybridizations were performed under stringent conditions using [32P]dCTP-labelled probes, essentially as described by Sambrook et al. (1989).

Construction of MC58 strain harbouring the dinB gene under the control of the opa promoter.
Plasmid pCR2-Popa-dinB was derived from pCR2-Popa-gfp, which presents a gfp gene under the control of the N. meningitidis opa promoter (unpublished data). The gfp gene was excised as a NdeI–NsiI fragment from pCR2-Popa-gfp and replaced by the 5' portion of the N. meningitidis strain MC58 dinB gene amplified by PCR with primers Din-NdeI and Din2-Nsi (Table 2), which respectively contain a NdeI and a NsiI restriction site. The NdeI site of the primer Din-NdeI contains the translational start codon of the gene. A 610 bp DNA fragment resulting from the PCR amplification of the 5' region of the dinB gene of strain MC58 with primers Din-down and Din-down-Bam (Table 2) was ligated with the 1·25 kb kanamycin resistance cassette, obtained by digestion of pUC4K vector (Pharmacia), in pT7 Blue vector (Novagen). The 1·86 kb fragment was released by digesting the resulting plasmid with EcoRI and SpeI and cloned upstream of the opa promoter in the plasmid digested with the same enzymes. This construct was transformed into strain MC58. By PCR amplification using primers Din-down2/Din2-Nsi and Din2-Nsi/Kan 5' (Table 2), the transformant MC58PopadinB was shown to present a copy of the wild-type dinB gene and a copy of the gene under the control of the opa promoter. Although additional secondary DNA rearrangements also occurred, the resulting construct allowed the overexpression of the dinB gene.

Construction of mutS, dinB, dam, drg and galE knock-out mutants.
The plasmid pT7-mutS contains a 670 bp DNA fragment amplified by PCR from the 5' region of the mutS gene of strain MC58 with primers MutS-up-Kpn and MutS-up-Bam (Table 2), then the 1·25 kb kanamycin resistance cassette followed by a 500 bp DNA fragment PCR amplified from the 3' region of the mutS gene with primers MutS-down-Bam and MutS-down (Table 2). This plasmid was linearized with ScaI and transformed into strain MC58. The transformant MC58mutS presented a 800 bp deletion of the mutS gene and an insertion of the kanamycin resistance cassette, confirmed by PCR using primers MutS-up2/MutS-down2, MutS-up-Kpn/Kan 5' and MutS-down/Kan 3' (Table 2).

The plasmid pT7-dinB contains a 610 bp DNA fragment amplified by PCR from the 5' region of the dinB gene of strain MC58 with primers Din-down/Din-down-Bam, then the 1·25 kb kanamycin resistance cassette followed by a 630 bp DNA fragment PCR amplified from the 3' region of the dinB gene with primers Din-up-Kpn and Din-up-Bam. This plasmid was linearized by ScaI and transformed into strain MC58. The transformant MC58dinB presented a 800 bp deletion of the dinB gene and an insertion of the kanamycin resistance cassette, confirmed by PCR using primers Din-down2/Din-up2, Din-down/Kan 5' and Din-up-Kpn/Kan 3'.

Plasmid pADK1 contains a 540 bp internal fragment of the N. meningitidis strain BZ10 dam gene, which was amplified by PCR with primers DamF1/DamB2 (Table 2, Fig. 1), then interrupted by the insertion of a kanamycin resistance cassette. This plasmid was linearized and transformed into the dam+ strains AK22, 94/155 and NGH38. The disruption of the dam gene in strains AK22dam1, AK22dam2, 94/155dam1 and NGH38dam1 was confirmed by PCR using primers DamF1/DamB2 (Table 2, Fig. 1). This was corroborated by Southern hybridization using the kanamycin resistance gene as a hybridization probe.

The plasmid pT7-drg contains a 640 bp DNA fragment amplified by PCR from the 5' region of the drg gene (or dpnC gene) of strain MC58 with primers Drg-up-Bam/Drg-up, then the 1·25 kb kanamycin resistance cassette followed by a 660 bp DNA fragment PCR amplified from the 3' region of the drg gene with primers Drg-down-Bam and Drg-down-Kpn. This plasmid was linearized by ScaI and transformed into strain MC58. The transformant MC58drg presented a 650 bp deletion of the drg gene and an insertion of the kanamycin resistance cassette, confirmed by PCR using primers Drg-down and Drg-up2.

galE knock-out mutants of strains MC58, MC58mutS, MC58dinB, MC58PopadinB, AK22, AK22dam1 and AK22dam2 were constructed using plasmid containing the galE gene disrupted by the erythromycin cassette (unpublished data). The erythromycin cassette was originally from the Bacillus subtilis plasmid pIM13 (Monod et al., 1986).

In each case, no polar effect was expected as a consequence of the disruption of each of the genes.

Detection of siaD, porA, lgtG and nadA phase variation by colony immunoblots.
Bacteria were grown overnight on solid media, transferred to a nitrocellulose filter (45 µm, pore size, Schleicher and Schuell) and allowed to air-dry. Non-specific binding sites were blocked for 1 h with 5 % (w/v) milk/0·1 % (w/v) azide/PBS-T [phosphate buffered saline–Tween 20, 0·05 % (v/v)]. Filters were washed three times with PBS-T, then incubated for 2 h 30 min with an appropriate dilution of antibody in PBS-T-bovine serum albumin [BSA, 2 % (w/v)]. Monoclonal antibodies (mAb) 735 (Frosch et al., 1985), P1.7 and B5 were used to detect siaD, porA and lgtG phase variation, respectively, and the antiserum containing antibodies raised against NadA (Pizza et al., 2000) was used to detect the switching of nadA expression. The dilutions of antibodies used were 1/10 000, 1/500, 1/3500 and 1/10 000, respectively. Filters were then washed three times in PBS-T. Primary antibody was detected with anti-mouse IgG-Alkaline Phosphatase antibody (Sigma and Cedarlane Laboratories) by incubation for 1 h in PBS-T-BSA. Filters were washed three times in PBS-T and antibody was detected with 5-bromo-4-chloro-3-indolyl phosphate–nitro blue tetrazolium (Perkin Elmer Life Sciences). The colour reaction was stopped after 30 min by several washes with water and the blots were air-dried.

Construction of translational fusion between the lacZ gene and the spr gene in N. meningitidis and detection of spr phase variation.
Chromosomal DNA of strain H44/76spr/lacZ, which contains a promoterless lacZ gene inserted in-frame downstream of the spr gene (Martin et al., 2003), was used to transform strains MC58, MC58mutS, MC58PopadinB, AK22, AK22dam1, AK22dam2 and MC58drg. The same chromosomal DNA was used to transform all the strains. Any additional mutations potentially introduced beside the desired one are most likely present in all the strains transformed. Moreover, there remains a very small possibility of co-transformation events. Transformants were streaked on BHI plates containing X-Gal (80 µg ml–1). Blue transformants expressing {beta}-galactosidase were analysed by PCR amplification and DNA sequencing to confirm the presence of the fusion.

To detect phase variation of the spr gene, the strains were grown overnight on BHI plates. Single colonies were then picked, plated on X-Gal-containing BHI plates and grown overnight. The coexistence of LacZ+ and LacZ variants, i.e. of blue colonies expressing {beta}-galactosidase and white colonies not expressing the enzyme, reflected the switching of the gene.

Determination of phase variation frequencies.
The frequencies of phase variation given are frequencies of mutants and were determined by dividing variant c.f.u. ml–1 by the total viable c.f.u. ml–1, all c.f.u. deriving from the same colony, as defined by Richardson & Stojiljkovic (2001). Our method to estimate mutation frequencies follows the same procedure (median frequency) as Richardson & Stojiljkovic (2001), thus allowing some valid comparison with their results. The relative fitness of the ON and OFF variants in the wild-type and mutant strains is assumed to be identical, since there is no reason to assume that the ON and OFF variants of the wild-type strain when compared to those of the mutant strains have different relative fitnesses when they are being investigated under similar in vitro conditions.

Determination of frequencies of spontaneous mutants resistant to rifampicin.
Frequencies of spontaneous mutants resistant to rifampicin were measured by two different methods. Cells from an overnight culture on plate were harvested in 1 ml PBS; 1010 cells were plated on BHI agar containing 30 µg rifampicin ml–1. Alternatively, 1000 colonies deriving from a single dispersed colony were grown overnight and harvested in 1 ml PBS; 1010 bacteria were plated from this suspension on BHI agar containing 3 µg rifampicin ml–1. Frequencies were determined by dividing resistant c.f.u. ml–1 by the total viable c.f.u. ml–1. Both methods gave similar results for a given strain.

RT-PCR.
Following 12 h growth, total RNA was prepared using the SV total RNA isolation system (Promega) from N. meningitidis strains MC58 and MC58PopadinB, the strain presenting the dinB gene under the control of the opa promoter. Five hundred and fifty nanograms of total RNA was reverse transcribed to produce cDNA using Moloney murine leukaemia virus (MMLV) reverse transcriptase (RT) and random primers (Promega). Co-amplification of the dinB transcript with an internal control allowed comparison of the levels of dinB gene expression between the RNA samples. PCR amplification of the dinB cDNA was performed with oligonucleotides Din-up2 and Din-NdeI. Primers specific for the constitutively expressed housekeeping gene gdh (glucose-6-phosphate dehydrogenase), gdh-P1 and gdh-P2 (Table 2), were also included in each PCR amplification to provide an internal control. To control for chromosomal DNA contamination, RNA was directly used for PCR amplification. The relative intensities of the dinB-specific and gdh-specific products were measured by densitometry (ImageQuant, Amersham Biosciences) for three independent experiments.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
mutS influences the stability of homopolymeric tracts but not of tetranucleotide repeat tracts
Previously, by investigating two phase variable genes associated with polyG tracts, i.e. hmbR and hpuAB, Richardson and colleagues showed that mutS was involved in the regulation of phase variation rates (Richardson & Stojiljkovic, 2001; Richardson et al., 2002). However, the effect of trans-acting factors on phase variation rates cannot be assumed to be similar for different genes and different repeat tracts. Therefore, we constructed a mutS knock-out mutant of strain MC58 in order to analyse rates of phase variation of several genes, associated with a variety of repeat tracts, in a mutator background.

An 11-fold increase of the frequency of spontaneous acquisition of rifampicin resistance was observed in the MC58mutS background (3·0x10–8) compared to wild-type (2·7x10–9) (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 3. Spontaneous mutation frequency and frequency of siaD, spr, lgtG and nadA phase variation in different wild-type (WT) and mutant strains of N. meningitidis

The sequence of the repeat tract associated with each phase variable gene is indicated. The frequency range determined is indicated in parentheses.

 
Five phase-variable genes were investigated to test whether there was any difference in phase variation frequency between strain MC58 and the isogenic MC58mutS mutant. siaD (NMB0067) encodes a polysialyltransferase involved in the biosynthesis of the capsule, spr (NMB1969) a putative protease, lgtG (NMB2032) a lipopolysaccharide (LPS) glucosyltransferase, porA (NMB1429) an outer-membrane protein and nadA (NMB1994) an adhesin. In the published genome sequence of strain MC58, the reading frames of siaD, spr and lgtG contain homopolymeric tracts of 7, 10 and 12 cytosine residues, respectively, and the promoter regions of porA and nadA present homopolymeric tracts of 11 guanines and 9 repeated motifs of the tetranucleotide (TAAA), respectively (Saunders et al., 2000). The phase variable expression of these genes was previously demonstrated (Hammerschmidt et al., 1996; van der Ende et al., 2000; Mackinnon et al., 2002; Martin et al., 2003). We determined the frequencies of phase variation of siaD, porA, lgtG and nadA by performing colony immunoblotting experiments using antibodies recognizing the relevant surface structures. Frequencies of spr phase variation were determined after constructing a translational fusion between the lacZ gene and the chromosomal spr gene and plating the fusion strains on BHI plates containing X-Gal.

To estimate the frequency of siaD phase variation (‘on’ to ‘off’ switching), 105 colonies of the wild-type strain MC58 were plated. However, we could not detect any non-reactive colonies using mAb 735. This finding was in agreement with the low frequency of capsule switching already reported in N. meningitidis (Hammerschmidt et al., 1996). It was therefore not possible to determine the frequency of capsule phase variation in the wild-type strain MC58 (Table 3). In contrast, in MC58mutS, ‘on to off’ switching was observed. Amplification by PCR and sequencing of the siaD polyC tract in 22 capsule-negative variants revealed that five displayed eight cytosine residues, three exhibited six residues and 14 had seven residues in the repeat tract. mAb 735 reactive colonies are typically associated with seven cytosines. Interestingly, capsule-negative variants presenting a frameshift in the polyC tract as well as variants displaying seven cytosine residues in the repeat tract were obtained in distinct experiments. Although the quoted frequencies may be underestimated due to the possibility of overlooking non-reactive colonies, in strain MC58mutS, capsule-negative variants resulting from a frameshift in the repeat tract located in siaD arose at a frequency of 3·7x10–4 (Table 3) and capsule-negative mutants resulting from a different mutational event were obtained at a similar frequency (2·5x10–4).

We measured the frequency of ‘off to on’ spr phase variation by enumerating blue colonies expressing (‘on’) and white colonies not expressing (‘off’) {beta}-galactosidase. The frequency was increased at least 700-fold in the MC58mutS background (8·3x10–3) when compared to the wild-type frequency (1·2x10–5) (Table 3).

To detect lgtG phase variation, we used mAb B5. Part of the epitope recognized by mAb B5 comprises the phosphoethanolamine (PEA) at the 3 position of heptose II in the LPS. In the ‘on’ state, lgtG encodes a glucosyltransferase which adds a glucose to the 3 position of heptose II, precluding the addition of PEA (Mackinnon et al., 2002). This LPS structure therefore impedes the recognition and the binding of mAb B5 (Plested et al., 1999). In the lgtG ‘off’ state, glucose is not added; therefore PEA can be added and consequently mAb B5 can recognize the relevant epitope. Thus, a mAb B5-positive phenotype implies the existence of an ‘off’ state and a mAb B5-negative phenotype an ‘on’ state of lgtG gene expression. We showed that strains MC58 and MC58mutS both presented 12 cytosine residues in the repeat tract located in lgtG. This number of bases makes the ORF out of frame, preventing the synthesis of a mature protein. The frequency of lgtG phase variation measured was therefore the ‘off to on’ frequency. lgtG phase variation was increased by about 120-fold in the MC58mutS background (2·0x10–3) as compared to the wild-type (2·6x10–5) (Table 3).

Preliminary data obtained from a single experiment revealed an increase in the frequency of frameshifts in the porA gene in the MC58mutS background (3·8x10–4) compared to the wild-type frequency (3·3x10–5). Sequencing of the repeat tract associated with porA in 13 colonies that reacted with mAb P1.7 and in five colonies that did not react with mAb P1.7 revealed that the polyG tract contained 11 and 10 guanine residues, respectively.

The tetranucleotide (TAAA) repeat tract associated with nadA is located in the promoter region of the gene. In strain MC58, nine repeated motifs constitute the repeat tract, a number of repeats associated with a low level of nadA transcription and a weak reactivity with the specific polyclonal antibody (Martin et al., 2003). Loss or gain of one tetranucleotide repeat, leading to eight or ten reiterated motifs, was shown to result in increased expression of nadA, as evidenced by strong antibody reactivity. No differences in switching between strong and weak antibody reactivities were detected between strains MC58 (4·4x10–4) and MC58mutS (3·0x10–4) (Table 3).

In conclusion, we showed that mutS was involved in the control of the instability of homopolymeric tracts (C7 in siaD, C10 in spr, G11 in porA and C12 in lgtG) but not of tetranucleotide repeat tracts [(TAAA)9 in nadA].

dinB influences the stability of the siaD homopolymeric tract
To investigate the involvement of dinB in the regulation of phase variation, we constructed a dinB knock-out mutant of strain MC58 (MC58dinB) and a MC58 mutant strain in which the dinB gene was placed under the control of the opa promoter (MC58PopadinB). Using RT-PCR we found that, under the growth conditions used, expression of dinB was increased by approximately 70 %, relative to strain MC58 (Fig. 2). Disruption or overexpression of dinB did not modify the global mutation frequency, as detected by spontaneous resistance to rifampicin, compared to the frequency determined in the wild-type background (1·8x10–9 and 2·9x10–9 vs 2·7x10–9, respectively) (Table 3).



View larger version (102K):
[in this window]
[in a new window]
 
Fig. 2. RT-PCR analysis of the expression of the dinB gene. Oligonucleotides used were Din-up2/din-NdeI (dinB) and gdh-P1/gdh-P2 (gdh, used as an internal control to semi quantify the level of dinB gene expression). Total RNA was extracted from N. meningitidis strain MC58 (lane 2) and from MC58PopadinB (lane 3), a derivative of strain MC58 presenting the dinB gene under the control of the opa promoter. The relative intensities of the dinB-specific and gdh-specific products were measured by densitometry (ImageQuant, Amersham Biosciences). Lane 1 contains 1 kb ladder.

 
The five phase variable genes siaD, spr, lgtG, porA and nadA were investigated to determine whether there was any difference in phase variation frequency between the strains MC58, MC58dinB and MC58PopadinB. In the MC58dinB background, no increase or decrease in the switching frequencies of genes associated with homopolymeric tracts tested could be detected (Table 3). Disruption of dinB did not alter the frequency of frameshifts (3·5x10–4) occurring in a tetranucleotide repeat tract when compared to wild-type (4·4x10–4) (Table 3).

We detected capsule-negative variants in strain MC58PopadinB. The siaD repeat tract was sequenced in seven capsule-negative colonies. Two variants displayed eight cytosines and five had six cytosines in the repeat tract. Thus, the siaD ‘on to off’ phase variation frequency was augmented in the strain MC58PopadinB (1·4x10–4), when compared with the wild-type frequency (<1·0x10–5) (Table 3). Interestingly, overexpression of dinB did not affect the frequency of switching of the spr gene when compared to the wild-type (2·0x10–5 vs 1·2x10–5, respectively). Likewise, we showed that the frequency of lgtG phase variation was unchanged (Table 3). Preliminary data also suggested that the porA switching frequency was not affected by the overexpression of the dinB gene (3·3x10–5 vs 1·4x10–5). Finally, overexpression of dinB did not alter the frequency of nadA phase variation when compared to wild-type (3·7x10–4 vs 4·4x10–4, respectively) (Table 3).

Conservation of two distinct dam loci in Dam+ and Dam N. meningitidis strains
To investigate the consequences of dam inactivation on the bacterial phenotype, we constructed mutants in which the dam gene was disrupted. Contrary to what Richardson & Stojiljkovic (2001) reported, Bucci et al. (1999) reported that the pathogenic strains they investigated presented a Dam phenotype. Therefore, we investigated the presence or absence of the dam gene in a global collection of 84 disease isolates representative of the genetic diversity of N. meningitidis. Phylogenetic profiles of these geographically and genetically distinct N. meningitidis strains were previously characterized (Maiden et al., 1998). The chromosomal DNA of each of the 84 strains was extracted and digested with the restriction enzymes DpnI, DpnII and Sau3AI. Fig. 3(a) shows an example of patterns obtained for a subset of N. meningitidis strains. DpnI recognizes and cuts its cognate targets only when GATC sites are methylated, DpnII cuts only unmethylated sites, and Sau3AI cuts both methylated and unmethylated target sequences. Since Dam directs the methylation of the deoxyadenosine site at the GATC sequence, hence rendering DNA resistant to cleavage by DpnII, but not by DpnI or Sau3AI, the functional integrity of the dam gene can be ascertained via a combination of these three enzymes.



View larger version (119K):
[in this window]
[in a new window]
 
Fig. 3. Restriction endonuclease digestion of chromosomal DNA (a) isolated from representative dam+ and dam N. meningitidis strains and (b) of dam knock-out mutant strains. (a) Strains in lanes 1–6 were digested by DpnI, those in lanes 8–13 were digested by DpnII, and those in lanes 16–21 were digested by Sau3AI. Lanes 1, 8 and 16, AK22 (dam+); 2, 9 and 17, 94/155 (dam+); 3, 10 and 18, NGH38 (dam+); 4, 11 and 19, H44/76 (dam); 5, 12 and 20, BZ169 (dam); 6, 13 and 21, 2059001 (dam); 7, 14 and 15, 1 kb DNA ladder. (b) Strains in lanes 2–7 were digested with DpnI, those in lanes 10–15 were digested by DpnII, and those in lane numbers 17–22 were digested by Sau3AI. Lanes 2, 10 and 17, AK22 (dam+); 3, 11 and 18, 94/155 (dam+); 4, 12 and 19, NGH38 (dam+); 5, 13 and 20, AK22dam1 (dam); 6, 14 and 21, 94/155dam1 (dam); 7, 15 and 22, NGH38dam1 (dam); 1, 8, 9 and 16, 1 kb DNA ladder.

 
The presence of the dam gene was investigated by PCR using primers DamF1 and DamB2 (Fig. 1). A 540 bp fragment was amplified from the dam gene for each of the Dam+ strains, but not from any of the Dam strains. We subsequently used this 540 bp fragment as a probe for Southern analysis, and detected positive hybridization signals only from the Dam+ strains (data not shown). This analysis revealed that 23 strains out of 84 (27 %) possessed a dam gene and an active Dam function (Table 1). In contrast to Bucci et al. (1999), who reported an absence of Dam activity in pathogenic isolates, here we demonstrate that there is no relationship between the dam phenotypes and their association with commensal or virulence behaviour.

The genetic basis of the dam gene inactivation was investigated in the Dam strains. This revealed that all the Dam strains were the result of an ancestral inactivation event. We surveyed the presence of drg (dam replacing gene) in the Dam strains (Bucci et al., 1999). All the Dam strains yielded a PCR product of 705 bp with primers DrgF1 and DrgB2 (Fig. 1), whereas all the Dam+ strains did not (data not shown). Southern analyses using the 705 bp PCR amplicon as a probe confirmed that only Dam strains possessed the drg gene (data not shown). The chromosomal location of drg was determined by analysing the immediate up- and down-stream sequences which flanked the drg gene. Each pair of primers (Drg5F/DrgB3 and Drg3F/DamB5, Fig. 1) generated a PCR product of similar size from all the Dam strains (data not shown), indicating that drg gene insertion indeed occurred at the dam locus, at the same place. This suggests that the acquisition of drg most likely occurred once during the evolution of the bacterium and indicates that the two types of dam loci have been conserved in the dam+/drg and dam/drg+ strains, respectively.

dam and drg do not influence the stability of homopolymeric and tetranucleotide repeat tracts
We constructed dam knock-out mutants in strains AK22, 94/155 and NGH38, arbitrarily chosen to represent the dam+/drg genotype. The inactivation of the Dam function in these mutant strains was subsequently verified by using the restriction enzymes DpnI, DpnII and Sau3A1. All the DNAs were susceptible to complete digestion by DpnII and Sau3A1, but resistant to DpnI (Fig. 3b), demonstrating the absence of Dam function in the strains.

We determined the spontaneous mutation frequency in the above three pairs of isogenic strains, nine natural dam+ strains and 12 natural dam strains randomly selected from the same meningococcal strain collection. In general, no marked difference was detected between dam+ and dam strains, regardless of the mechanism by which the Dam activity had been abolished (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4. Spontaneous mutation frequency of dam+ and dam strains

The values given are the mean of two experiments, except for AK22 and AK22dam1.

 
We tested whether there was any difference in phase variation frequency between wild-type strain AK22 and two independent mutant strains AK22dam1 and AK22dam2. For each strain, 105 colonies were screened to determine the frequencies of siaD phase variation, and no instance of capsule-negative variants was detected (Table 3). Measurement of frequencies of spr phase variation showed that disruption of dam had no effect on the frequency of spr switching (Table 3). Sequencing of the lgtG repeat tract in strains AK22galE, AK22dam1galE and AK22dam2galE revealed 11 cytosines in the homopolymeric tract. This number of bases makes the ORF in-frame, permitting the synthesis of a mature protein. The phase variation frequency determined for these strains was therefore the ‘on to off’ frequency. We observed a 3·6-fold increase in lgtG phase variation frequency in the AK22dam1galE background (1·6x10–4) compared to that of strain AK22galE (4·4x10–5). In contrast, strain AK22dam2galE (1·6x10–5) switched from ‘on’ to ‘off’ at a frequency similar to that of the wild-type strain (Table 3). The difference in lgtG phase variation frequencies observed between strains AK22galE and AK22dam1galE was most likely due to factors other than the activity of Dam, since none of the switching frequencies of the other genes investigated were affected in both AK22dam1galE and AK22dam2galE mutant strains. We showed that the frequency of nadA phase variation was similar in the wild-type AK22 (3·0x10–4) and the AK22dam1 (3·3x10–4) backgrounds (Table 3). Altogether these data suggested that dam gene inactivation alone had no effect on the frequency of phase variation.

We constructed a drg knock-out mutant in MC58, a strain representative of the dam/drg+ genotype. The spontaneous mutation frequency determined for the MC58drg mutant (0·9x10–9) was not significantly different from the frequency determined for the wild-type strain (2·7x10–9, Table 3). The frequency of phase variation was determined for the siaD, spr and nadA genes in the MC58drg background. This revealed that disruption of the drg gene did not affect the frequency of frameshifts occurring in homopolymeric and tetranucleotide repeat tracts (Table 3). This therefore complements and extends the findings reported by Bucci et al. (1999).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In this study we have identified trans-acting factors that influence the stability of mononucleotide but not of tetranucleotide repeat tracts in N. meningitidis. There are several approaches to estimating rates of phenotypic or genotypic switching. A theoretical paper that deals with some complexities and interpretative issues when measuring switching rates was recently published (Saunders et al., 2003). However, the key issue in our study is whether the observed rates of switching reflect differences, even if these estimates would need to be modified to better reflect an estimate of mutation rates. It is not critical that the measurement of switching rates may be an over- or underestimate of mutation rates, providing that there is consistency. Moreover, there is no reason to assume that the ON and OFF variants of the wild-type strain when compared to those of the mutant strains have different relative fitnesses when they are being investigated under similar in vitro conditions. In addition, to circumvent any jackpot problem, replicate experiments were carried out a sufficient number of times, from at least four and up to eight times, to be able to exclude outlying results with confidence. Therefore, we followed the same procedure as Richardson & Stojiljkovic (2001) used to study hmbR and hpuAB phase variation rates, thus allowing some valid comparison with their results, and showed that mutS influences both the frequency of frameshifts in homopolymeric tracts differing in length and sequence and the spontaneous mutation frequency, but neither mutS, dinB, dam nor drg influence the rate of changes occurring in the tetranucleotide repeat tract studied.

According to the classification established by Richardson & Stojiljkovic (2001), strain MC58 has a low overall mutation rate, as defined by spontaneous acquisition of rifampicin resistance. We have shown that the mutS mutant of N. meningitidis strain MC58 presented a spontaneous mutation frequency 11-fold higher than the wild-type rate. Thus, we confirm the weak mutator phenotype resulting from the inactivation of the mutS gene previously reported in N. meningitidis (Richardson & Stojiljkovic, 2001). Likewise, in H. influenzae, mutS, mutL and mutH mutants exhibit a significant three- to fourfold higher global mutation frequency than wild-type strain Rd (Bayliss et al., 2002). In addition, inactivation of dam did not result in a global mutator phenotype in N. meningitidis, in agreement with previous findings in this organism (Richardson & Stojiljkovic, 2001) and in H. influenzae strain Rd (Bayliss et al., 2002). Strain MC58 does not appear to have the mutH gene (Tettelin et al., 2000), suggesting that mismatch repair in N. meningitidis may be different from the system characterized in E. coli.

We showed that disruption of mutS in N. meningitidis increased the frequency of frameshifts occurring in homonucleotide tracts of various lengths (C7 in siaD, C10 in spr, G11 in porA and C12 in lgtG). This extends the conclusions drawn previously from the analysis of two contingency loci associated with polyG tracts of similar length, 9 and 10 bases (Richardson & Stojiljkovic, 2001). The effect of mutS on phase variation rates could not be assumed a priori to be similar for different genes (Bayliss et al., 2004). As most of the phase variable genes in N. meningitidis are associated with polyC or polyG tracts (Saunders et al., 2000), bacterial cells contain many genes that appear to be hypermutable targets in mismatch-repair-deficient strains. Consistently, strains exhibiting a general mutator phenotype resulting from defects in mismatch-repair pathways were found amongst clinical isolates of N. meningitidis (Richardson & Stojiljkovic, 2001; Richardson et al., 2002). Pathogenic bacteria can face strong selective pressures when colonizing a new host or a new environment within the same host. Thus, mutator strains of N. meningitidis could be advantageous, at least in the short term, because of their adaptive capacities, resulting from an enhanced variability of their surface-exposed structures.

The present study has identified for the first time a role for dinB in the regulation of phase variation in a bacterial pathogen. DinB is the recently discovered DNA polymerase IV which belongs to the SOS regulon (Kim et al., 1997). Investigations are currently under way to elucidate its role in the bacterial cell (Wagner et al., 2002; Tompkins et al., 2003). It has been shown that N. gonorrhoeae does not possess an SOS-like system that is induced in response to DNA damage (Black et al., 1998). Moreover, the lexA gene is missing in the N. meningitidis strain MC58 genome sequence (Tettelin et al., 2000). One can therefore speculate that in N. meningitidis, dinB is not under the control of a DNA damage-inducible system. Consequently, the context in which the DNA polymerase IV is recruited could be different in E. coli and in N. meningitidis. In our study we showed that, contrary to what was observed in E. coli (Kim et al., 1997), overexpression of dinB in N. meningitidis did not result in a global mutator phenotype. Although this absence of a mutator phenotype could be due to a quantitative difference of DinB molecules in both organisms, the amount of molecules of DNA polymerase IV in N. meningitidis was sufficient to increase the frequency of frameshifts in the siaD homopolymeric tract. We also showed that the frequencies of frameshifts in the other homonucleotide tracts we investigated were not affected by dinB overexpression. In yeast, the impact of proofreading in preventing frameshift mutations was shown to decrease when the length of a homonucleotide run increases (Tran et al., 1997). In addition, in E. coli, the proofreading activity of DNA polymerase III recognizes frameshifts close to the 3' end of the nascent DNA strand (Strauss et al., 1997). Bulged nucleotides resulting from replication slippage within a long run of simple repeats could locate at a distance from the 3' end of the growing DNA strand and thus escape from DNA proofreading. Thus, to explain the lack of effect of overexpression of dinB on the frequency of frameshifts in long homopolymeric tracts, we hypothesize that in N. meningitidis neither DNA polymerase III nor DNA polymerase IV mediates efficient proofreading on long repeat tracts. Construction of mutants containing increasing tract lengths incorporated into the siaD gene, and investigation of the influence of dinB overexpression on the instability of these various repeat tracts, would be required to test this hypothesis.

In this study we have shown that the instability of tetranucleotide repeats did not depend on genetic factors such as mutS, dinB, dam and drg. Similarly, in H. influenzae, where phase variation is almost exclusively mediated by tetranucleotide repeat tracts (Hood et al., 1996), mutation of polI but not of genes involved in MMR or other pathways of DNA repair or recombination was shown to increase phase variation rates mediated by tetranucleotide repeat tracts (Bayliss et al., 2002). Thus, in N. meningitidis as in H. influenzae, tetranucleotide repeat tracts seem to be refractory to most of the mismatch correction systems.

We also demonstrated that 27 % of N. meningitidis strains possessed a dam gene and an active Dam function, 70 % of which belonged to either the A4 cluster or ET-37 lineage, which are classical hypervirulent lineages. Dam is associated with less than a third of virulent strains, findings not consistent with previous data suggesting an absence of Dam activity in pathogenic isolates (Bucci et al., 1999). The selection of strains from Italy and France investigated in this latter study is unlikely to be representative of the global meningococcal population. Bucci and colleagues also showed that dam-deficiency in N. meningitidis created a mutator phenotype, resulting in an increased frequency of phase variation in the siaD gene (Bucci et al., 1999). However, we and others (Richardson & Stojiljkovic, 2001) could not reproduce this finding, showing a lack of correlation between Dam methylation and the frequency of phase variation. In contrast, a multifactorial role of Dam methylation was involved in Salmonella virulence (Garcia-Del Portillo et al., 1999). Mutants of Salmonella typhimurium lacking dam were shown to be attenuated for virulence and to present a distinct pattern of secreted proteins compared to dam+ strains (Garcia-Del Portillo et al., 1999). The strict correlation between the presence of drg and the absence of dam evidenced in our study confirmed that the presence of dam and drg was mutually exclusive (Bucci et al., 1999). We showed that the drg gene was located at an identical locus in all dam/drg+ strains. This suggests that the common ancestor of the N. meningitidis strains investigated was likely to be a dam-proficient strain. The drg gene shares significant homology with the Streptococcus pneumoniae type II restriction endonuclease gene dpnI (Cantalupo et al., 2001), suggesting that drg has been horizontally acquired. Investigation of the involvement of drg in the regulation of the phase variation frequencies suggested that it did not have a role in the control of repeat tract stability. The function of the drg gene, which must be important for the bacteria since it has been widely conserved, remains to be elucidated.


   ACKNOWLEDGEMENTS
 
This work was supported by a programme grant from the Wellcome Trust (E. R. M.). We would like to acknowledge the gift of anti-NadA antiserum provided by Rino Rappuoli, Mariagrazia Pizza and Maurizio Comanducci. We would like to acknowledge Dr Nigel J. Saunders for supplying plasmid pCR2-Popa-gfp, Dr J. Claire Wright for supplying the plasmid containing the galE gene disrupted by the erythromycin cassette, and Dr Keith A. Jolley, Dr Martin C. Maiden and Dr Christopher D. Bayliss for critical reading of the manuscript.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Alexander, H. E. (1965). The Haemophilus group. In Bacterial and Mycotic Infections of Man, pp. 724–741. Edited by R. J. Dabos & J. G. Horsch. London: Pitman Medical Publishing.

Bayliss, C. D., van de Ven, T. & Moxon, E. R. (2002). Mutations in polI but not mutSLH destabilize Haemophilus influenzae tetranucleotide repeats. EMBO J 21, 1465–1476.[Abstract/Free Full Text]

Bayliss, C. D., Sweetman, W. A. & Moxon, E. R. (2004). Mutations in Haemophilus influenzae mismatch repair genes increase mutation rates of dinucleotide repeat tracts but not dinucleotide repeat-driven pilin phase variation rates. J Bacteriol 186, 2928–2935.[Abstract/Free Full Text]

Black, C. G., Fyfe, J. A. & Davies, J. K. (1998). Absence of an SOS-like system in Neisseria gonorrhoeae. Gene 208, 61–66.[CrossRef][Medline]

Bucci, C., Lavitola, A., Salvatore, P., Del Giudice, L., Massardo, D. R., Bruni, C. B. & Alifano, P. (1999). Hypermutation in pathogenic bacteria: frequent phase variation in meningococci is a phenotypic trait of a specialized mutator biotype. Mol Cell 3, 435–445.[Medline]

Cantalupo, G., Bucci, C., Salvatore, P., Pagliarulo, C., Roberti, V., Lavitola, A., Bruni, C. B. & Alifano, P. (2001). Evolution and function of the neisserial dam-replacing gene. FEBS Lett 495, 178–183.[CrossRef][Medline]

de Bolle, X., Bayliss, C. D., Field, D., van de Ven, T., Saunders, N. J., Hood, D. W. & Moxon, E. R. (2000). The length of a tetranucleotide repeat tract in Haemophilus influenzae determines the phase variation rate of a gene with homology to type III DNA methyltransferases. Mol Microbiol 35, 211–222.[CrossRef][Medline]

de Vries, F. P., van der Ende, A., van Putten, J. P. & Dankert, J. (1996). Invasion of primary nasopharyngeal epithelial cells by Neisseria meningitidis is controlled by phase variation of multiple surface antigens. Infect Immun 64, 2998–3006.[Abstract]

Frosch, M., Gorgen, I., Boulnois, G. J., Timmis, K. N. & Bitter-Suermann, D. (1985). NZB mouse system for production of monoclonal antibodies to weak bacterial antigens: isolation of an IgG antibody to the polysaccharide capsules of Escherichia coli K1 and group B meningococci. Proc Natl Acad Sci U S A 82, 1194–1198.[Abstract]

Garcia-Del Portillo, F., Pucciarelli, M. G. & Casadesus, J. (1999). DNA adenine methylase mutants of Salmonella typhimurium show defects in protein secretion, cell invasion, and M cell cytotoxicity. Proc Natl Acad Sci U S A 96, 11578–11583.[Abstract/Free Full Text]

Hammerschmidt, S., Muller, A., Sillmann, H. & 7 other authors (1996). Capsule phase variation in Neisseria meningitidis serogroup B by slipped-strand mispairing in the polysialyltransferase gene (siaD): correlation with bacterial invasion and the outbreak of meningococcal disease. Mol Microbiol 20, 1211–1220.[Medline]

Hood, D. W., Deadman, M. E., Jennings, M. P., Bisercic, M., Fleischmann, R. D., Venter, J. C. & Moxon, E. R. (1996). DNA repeats identify novel virulence genes in Haemophilus influenzae. Proc Natl Acad Sci U S A 93, 11121–11125.[Abstract/Free Full Text]

Jennings, M. P., Virji, M., Evans, D., Foster, V., Srikhanta, Y. N., Steeghs, L., van der Ley, P. & Moxon, E. R. (1998). Identification of a novel gene involved in pilin glycosylation in Neisseria meningitidis. Mol Microbiol 29, 975–984.[CrossRef][Medline]

Jennings, M. P., Srikhanta, Y. N., Moxon, E. R., Kramer, M., Poolman, J. T., Kuipers, B. & van der Ley, P. (1999). The genetic basis of the phase variation repertoire of lipopolysaccharide immunotypes in Neisseria meningitidis. Microbiology 145, 3013–3021.[Medline]

Jonsson, A. B., Nyberg, G. & Normark, S. (1991). Phase variation of gonococcal pili by frameshift mutation in pilC, a novel gene for pilus assembly. EMBO J 10, 477–488.[Abstract]

Kim, S. R., Maenhaut-Michel, G., Yamada, M., Yamamoto, Y., Matsui, K., Sofuni, T., Nohmi, T. & Ohmori, H. (1997). Multiple pathways for SOS-induced mutagenesis in Escherichia coli: an overexpression of dinB/dinP results in strongly enhancing mutagenesis in the absence of any exogenous treatment to damage DNA. Proc Natl Acad Sci U S A 94, 13792–13797.[Abstract/Free Full Text]

Lewis, L. A., Gipson, M., Hartman, K., Ownbey, T., Vaughn, J. & Dyer, D. W. (1999). Phase variation of HpuAB and HmbR, two distinct haemoglobin receptors of Neisseria meningitidis DNM2. Mol Microbiol 32, 977–989.[CrossRef][Medline]

Mackinnon, F. G., Cox, A., Plested, J. S. & 8 other authors (2002). Identification of a gene (lpt-3) required for the addition of phosphoethanolamine to the lipopolysaccharide inner core of Neisseria meningitidis and its role in mediating susceptibility to bactericidal killing and opsonophagocytosis. Mol Microbiol 43, 931–943.[CrossRef][Medline]

Maiden, M. C., Bygraves, J. A., Feil, E. J. & 10 other authors (1998). Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc Natl Acad Sci U S A 95, 3140–3145.[Abstract/Free Full Text]

Martin, P., van de Ven, T., Mouchel, N., Jeffries, A. C., Hood, D. W. & Moxon, E. R. (2003). Experimentally revised repertoire of putative contingency loci in Neisseria meningitidis strain MC58: evidence for a novel mechanism of phase variation. Mol Microbiol 50, 245–257.[CrossRef][Medline]

Monod, M., Denoya, C. & Dubnau, D. (1986). Sequence and properties of pIM13, a macrolide-lincosamide-streptogramin B resistance plasmid from Bacillus subtilis. J Bacteriol 167, 138–147.[Medline]

Moxon, E. R., Rainey, P. B., Nowak, M. A. & Lenski, R. E. (1994). Adaptive evolution of highly mutable loci in pathogenic bacteria. Curr Biol 4, 24–33.[Medline]

Pizza, M., Scarlato, V., Masignani, V. & 33 other authors (2000). Identification of vaccine candidates against serogroup B meningococcus by whole-genome sequencing. Science 287, 1816–1820.[Abstract/Free Full Text]

Plested, J. S., Makepeace, K., Jennings, M. P. & 10 other authors (1999). Conservation and accessibility of an inner core lipopolysaccharide epitope of Neisseria meningitidis. Infect Immun 67, 5417–5426.[Abstract/Free Full Text]

Richardson, A. R. & Stojiljkovic, I. (1999). HmbR, a hemoglobin-binding outer membrane protein of Neisseria meningitidis, undergoes phase variation. J Bacteriol 181, 2067–2074.[Abstract/Free Full Text]

Richardson, A. R. & Stojiljkovic, I. (2001). Mismatch repair and the regulation of phase variation in Neisseria meningitidis. Mol Microbiol 40, 645–655.[CrossRef][Medline]

Richardson, A. R., Yu, Z., Popovic, T. & Stojiljkovic, I. (2002). Mutator clones of Neisseria meningitidis in epidemic serogroup A disease. Proc Natl Acad Sci U S A 99, 6103–6107.[Abstract/Free Full Text]

Rolfsmeier, M. L., Dixon, M. J., Pessoa-Brandao, L., Pelletier, R., Miret, J. J. & Lahue, R. S. (2001). Cis-elements governing trinucleotide repeat instability in Saccharomyces cerevisiae. Genetics 157, 1569–1579.[Abstract/Free Full Text]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sarkari, J., Pandit, N., Moxon, E. R. & Achtman, M. (1994). Variable expression of Opc outer membrane protein in Neisseria meningitidis is caused by size variation of a promoter containing poly-cytidine. Mol Microbiol 13, 207–217.[Medline]

Saunders, N. J., Jeffries, A. C., Peden, J. F., Hood, D. W., Tettelin, H., Rappuoli, R. & Moxon, E. R. (2000). Repeat-associated phase variable genes in the complete genome sequence of Neisseria meningitidis strain MC58. Mol Microbiol 37, 207–215.[CrossRef][Medline]

Saunders, N. J., Moxon, E. R. & Gravenor, M. B. (2003). Mutation rates: estimating phase variation rates when fitness differences are present and their impact on population structure. Microbiology 149, 485–495.[CrossRef][Medline]

Strauss, B. S., Sagher, D. & Acharya, S. (1997). Role of proofreading and mismatch repair in maintaining the stability of nucleotide repeats in DNA. Nucleic Acids Res 25, 806–813.[Abstract/Free Full Text]

Tettelin, H., Saunders, N. J., Heidelberg, J. & 39 other authors (2000). Complete genome sequence of Neisseria meningitidis serogroup B strain MC58. Science 287, 1809–1815.[Abstract/Free Full Text]

Tompkins, J. D., Nelson, J. L., Hazel, J. C., Leugers, S. L., Stumpf, J. D. & Foster, P. L. (2003). Error-prone polymerase, DNA polymerase IV, is responsible for transient hypermutation during adaptive mutation in Escherichia coli. J Bacteriol 185, 3469–3472.[Abstract/Free Full Text]

Tran, H. T., Keen, J. D., Kricker, M., Resnick, M. A. & Gordenin, D. A. (1997). Hypermutability of homonucleotide runs in mismatch repair and DNA polymerase proofreading yeast mutants. Mol Cell Biol 17, 2859–2865.[Abstract]

van der Ende, A., Hopman, C. T. P. & Dankert, J. (2000). Multiple mechanisms of phase variation of PorA in Neisseria meningitidis. Infect Immun 68, 6685–6690.[Abstract/Free Full Text]

Wagner, J., Etienne, H., Janel-Bintz, R. & Fuchs, R. P. (2002). Genetics of mutagenesis in E. coli: various combinations of translesion polymerases (Pol II, IV and V) deal with lesion/sequence context diversity. DNA Repair 1, 159–167.[CrossRef][Medline]

Wierdl, M., Dominska, M. & Petes, T. D. (1997). Microsatellite instability in yeast: dependence on the length of the microsatellite. Genetics 146, 769–779.[Abstract/Free Full Text]

Received 22 March 2004; revised 8 June 2004; accepted 8 June 2004.