These authors contributed equally to this work.
Characterization of Helicobacter pylori factors that control transformation frequency and integration length during inter-strain DNA recombination
Article first published online: 23 NOV 2010
© 2010 Blackwell Publishing Ltd
Volume 79, Issue 2, pages 387–401, January 2011
How to Cite
Humbert, O., Dorer, M. S. and Salama, N. R. (2011), Characterization of Helicobacter pylori factors that control transformation frequency and integration length during inter-strain DNA recombination. Molecular Microbiology, 79: 387–401. doi: 10.1111/j.1365-2958.2010.07456.x
- Issue published online: 11 JAN 2011
- Article first published online: 23 NOV 2010
- Accepted manuscript online: 4 NOV 2010 07:50AM EST
- Accepted 27 October, 2010.
- Top of page
- Experimental procedures
- Supporting Information
Helicobacter pylori is a genetically diverse bacterial species, owing in part to its natural competence for DNA uptake that facilitates recombination between strains. Inter-strain DNA recombination occurs during human infection and the H. pylori genome is in linkage equilibrium worldwide. Despite this high propensity for DNA exchange, little is known about the factors that limit the extent of recombination during natural transformation. Here, we identify restriction-modification (R-M) systems as a barrier to transformation with homeologous DNA and find that R-M systems and several components of the recombination machinery control integration length. Type II R-M systems, the nuclease nucT and resolvase ruvC reduced integration length whereas the helicase recG increased it. In addition, we characterized a new factor that promotes natural transformation in H. pylori, dprB. Although free recombination has been widely observed in H. pylori, our study suggests that this bacterium uses multiple systems to limit inter-strain recombination.
- Top of page
- Experimental procedures
- Supporting Information
Helicobacter pylori infects over half of the world human population and induces chronic gastritis, peptic ulcers and gastric malignancies (Kusters et al., 2006). Genome sequencing and comparative genomic hybridization analyses of H. pylori isolates have demonstrated its genetic diversity, with as much as 25% of the chromosome composed of strain variable genes (Gressmann et al., 2005; Oh et al., 2006). This high level of genomic plasticity is likely a consequence of DNA recombination between H. pylori strains facilitated by its natural competence for DNA uptake. Humans are on occasion infected with multiple genetically distinct H. pylori isolates and extensive inter-strain recombination has been documented (Kersulyte et al., 1999; Falush et al., 2001; Salama et al., 2007). Furthermore, H. pylori is one of approximately 100 bacterial species that inhabit the human stomach (Bik et al., 2006). Horizontal gene transfer occurs through a type IV secretion system conserved in all H. pylori strains (Hofreuter et al., 1998; 2000; Karnholz et al., 2006). Although H. pylori may benefit from increasing genetic exchange during human infection, limiting the extent of recombination may also be beneficial to avoid subversion of the genome by DNA from other strains or species.
Several barriers to gene transfer between unrelated strains or species have been identified in bacteria. The mismatch repair system recognizes differences in nucleotide sequence and decreases genetic exchange between Escherichia coli and Salmonella typhimurium (Rayssiguier et al., 1989; Roberts and Cohan, 1993; Zahrt and Maloy, 1997). In Haemophilus influenzae and Neisseria gonorrhoeae, specific DNA sequences are required for uptake by the natural transformation apparatus (Chen and Dubnau, 2004). Limiting the length of DNA recombined into the chromosome is also a partial barrier to gene transfer. The average length was measured in Neisseria meningitides (1.5–9.9 kpb) (Linz et al., 2000) and in Bacillus subtilis (3.4 kbp and some fragments smaller than 2 kbp) (Zawadzki and Cohan, 1995). Previous work in H. pylori estimated an integration length of 417 bp from a small data set of human paired isolate multi-locus sequence typing (Falush et al., 2001), an observation that was further refined with direct measurement of integration length (1.3–3.9 kbp) (Kulick et al., 2008; Lin et al., 2009).
Restriction modification (R-M) systems are present in over 90% of the bacterial genomes sequenced (Roberts et al., 2009) and may provide another barrier to gene transfer among strains and species. Foreign DNA entering the cell is destroyed because restriction endonucleases (REase) cleave DNA at specific sites unless protected by the activity of the corresponding DNA methyltransferase (MTase). The efficiency of genomic DNA restriction varies between species: transformation frequency is not significantly effected in the naturally competent bacteria Haemophilus influenzae, Bacillus subtilis and Streptococcus pneumoniae (Stuy, 1976; Bron et al., 1980; Lacks and Springhorn, 1984). In contrast, indirect evidence suggests that restriction plays a large role in limiting recombination in Pseudomonas stutzeri (Berndt et al., 2003) and Staphylococcus aureus (Corvaglia et al., 2010). Restriction is generally assumed to occur on double-stranded DNA, raising the interesting possibility that the structure of DNA transported by competence may determine the efficiency of restriction. Recent work in H. pylori suggests that DNA is converted to a single strand upon transit through the competence machinery (Stingl et al., 2010).
Type II R-M systems consist of a site-specific REase to cleave DNA and a site-specific MTase to protect DNA. Type II REases effectively limit transformation by plasmid or chromosomal DNA in H. pylori (Aras et al., 2002; Humbert and Salama, 2008). H. pylori encodes numerous type II R-M systems, but these enzymes can be strain-specific (Alm et al., 1999), phase-variable due to di-nucleotide repeats in the coding sequence (Salaun et al., 2004) and are often inactivated due to mutation. Consequently, while each H. pylori strain harbours approximately two dozen R-M systems (Lin et al., 2001), only a small number are functional. Only four of 16 type II REases are biochemically active in H. pylori strain J99 (Kong et al., 2000) and four of 14 are active in strain 26695 (Lin et al., 2001). Given the heterogeneity of R-M systems, a major challenge is to determine the extent to which they limit genetic exchange among unrelated strains.
Natural transformation in H. pylori also requires RecA, the master regulator of homologous recombination (Thompson and Blaser, 1995; Amundsen et al., 2008), implying that at least a sub-set of homologous recombination factors are required. H. pylori encodes several factors that may be capable of generating the 3′ ssDNA tail needed for strand invasion of the recipient DNA. The AddAB helicase-nuclease is functionally related to the RecBCD complex in E. coli that promotes recombinational repair of dsDNA breaks. In H. pylori, the role of AddAB in recombination during natural transformation is controversial with the ΔaddA mutant showing enhanced (Marsin et al., 2008), decreased (Wang and Maier, 2009) or no change (Amundsen et al., 2008) in transformation frequency. In E. coli, the 5′-3′ exonucleases RecJ and exonuclease VII (xseA and xseB) (Chase and Richardson, 1974) form 3′ ssDNA tails and RecJ is required for recombination in the absence of RecBCD (Lovett and Kolodner, 1989). Finally, the function of the NucT nuclease is unclear, but it has also been implicated in H. pylori natural transformation (O'Rourke et al., 2004).
The donor and recipient DNA initiate pairing during synapsis and duplex DNA structures are formed and extended by branched-DNA-specific helicases (Sharples et al., 1999). In E. coli, RuvA and the RuvB helicase bind DNA junctions. The RuvAB complex and the helicase RecG act in overlapping pathways driving branch migration of Holliday junctions and promoting recombination (West, 1994; Whitby and Lloyd, 1995). To complete the recombination reaction, the RuvC endonuclease resolves junctions by dual-strand incision across the branch point to release nicked duplexes (Dunderdale et al., 1991).
In this study, we analyse transformation frequency and integration length following inter-strain DNA recombination. We show that only a small number of H. pylori type II REases actively participate in DNA restriction, resulting in reduced transformation frequency and integration length. We find that the nuclease NucT, the resolvase RuvC and the helicase RecG alter integration length. Finally, we characterize a novel factor that promotes natural transformation, a putative RuvC homologue, DprB. Our results suggest that both R-M systems and the homologous recombination machinery control the extent of genetic exchange among unrelated strains in H. pylori.
- Top of page
- Experimental procedures
- Supporting Information
Identification of active type II REases and methyltransferases
We took a bioinformatic approach to define the REases that actively restrict genetic exchange between H. pylori strains. Restriction of transformation depends on two conditions: (i) the REase must be active in the strain taking in the DNA (recipient) (ii) donor DNA must lack the corresponding protective methylation, due to absence of the appropriate MTase activity in donor cells. We focused on type II R-M systems because they are the most common and well-understood R-M systems whereas H. pylori type I and type III R-M system activities are for the most part uncharacterized and their target sites unknown.
We analysed two H. pylori strains for which whole genome sequences are available: NSH57 (Baldwin et al., 2007), a G27-derivative (Baltrus et al., 2009), was assessed for active REases and strain J99 (Alm et al., 1999) was assessed for MTase activity. To identify active REases in the recipient NSH57, we searched for large inactivating deletions and frameshift mutations in known and predicted type II REase genes and found that 16/20 were likely inactivated by mutation (Fig. 1). A REase cannot be active in an H. pylori strain unless the chromosomal DNA is protected by methylation at its recognition site. To test for methylation of target sites corresponding to the four apparently active REases, chromosomal DNA of strain NSH57 was digested with commercially available methylation sensitive restriction enzymes. All four sites were protected from digestion (Fig. 1, 17–20), suggesting that these REases are active in strain NSH57.
To determine which type II DNA MTases are active in donor strain J99, we used sequence analysis, published work (Kong et al., 2000; Xu et al., 2000; Lin et al., 2001), and the in vitro restriction enzyme protection assay described above. Out of the four REases predicted to be active in NSH57, sites TCNNGA and TGCA are not protected by DNA methylation in J99 and are susceptible to cleavage by enzymes R.Hpy188III and R.HpyCH4V respectively (Fig. 1, 19–20). These two enzymes are predicted to be active in NSH57, suggesting that restriction should occur during genetic exchange with J99.
A type II MTase cannot be inactivated if the corresponding REase is active (Humbert and Salama, 2008). To determine if R.Hpy188III and R.HpyCH4V are active, we showed that their corresponding MTases are essential unless the corresponding REase is first inactivated. No transformants were obtained with deletion cassettes for MTase genes hpy188IIIM and hpyCH4VM unless the corresponding REase was first inactivated by mutation. Thus we conclude that REases R. Hpy188III and R.HpyCH4V are active in NSH57. Taken together, these results indicate that only two of the 20 type II REases present in NSH57 should restrict J99 genomic DNA (gDNA) during genetic exchange.
Type II REases limit genetic exchange in H. pylori
To model acquisition of a whole gene, we assessed exchange of a chloramphenicol resistance gene (cat) within gene cagH from the cag pathogenicity island. This locus is 96% conserved in sequence between strains J99 and NSH57 and there are 5 restriction sites for R.Hpy188III and R.HpyCH4V in cat. The transformation frequency of J99 (homeologous) DNA was approximately fivefold to sevenfold lower than for NSH57 (homologous) DNA over a 25-fold range of DNA concentration (Fig. 2A), showing that restriction had occurred. Both the Δhpy188IIIR mutant and the ΔhpyCH4VR mutant showed the same restriction barrier as wild-type (WT) indicating that their contributions to restriction are redundant for the cagH locus (Fig. 2B). To alleviate restriction of J99 gDNA in recipient NSH57 bacteria, the double mutant Δhpy188IIIR::aphA3ΔhpyCH4VR::erm was constructed and will be referred to as ΔΔhpyR(188III/CH4V). In the ΔΔhpyR(188III/CH4V) double mutant, there was a twofold difference in transformation frequency between homeologous and homologous DNA (Fig. 2B and C), indicating that the REases R.Hpy188III and R.HpyCH4V are the main barriers to genetic exchange between strains NSH57 and J99. This remaining twofold difference between homologous and homeologous transformation in the ΔΔhpyR(188III/CH4V) double mutant may be explained either by the action of other R-M systems (type I and III REases were not included in this analysis) or by the sequence divergence that exists between donor and recipient DNA (approximately 4% at this locus).
Type II REases decrease integration length during homeologous gDNA transformation
Although our results reveal a barrier to transformation mediated by type II R-M systems, transformants were obtained in the presence of active restriction. This observation suggested that restriction enzymes are only partially effective, which would reduce the length of DNA available for integration without completely blocking integration. We measured the length of donor-derived DNA incorporated in the recipient genome following inter-strain recombination by mapping the integration end-points on each side of the cagH::cat site. Over 600 single nucleotide polymorphisms (SNPs) exist between NSH57 and J99 DNA in the 15 kbp surrounding the cagH::cat site, which corresponds to one SNP every 25 bp on average (Fig. 3A). A 5–7 kbp DNA region was amplified on each side of the selective marker and the resulting DNA was digested with five restriction enzymes that have unique sites in either the donor or recipient DNA (Fig. 3A). Based on the restriction fragment pattern obtained, the location of integration end-points could be mapped to a rough interval and then pinpointed with sequencing from a primer flanking the interval. Since gDNA purified in vitro may not be reflective of the DNA taken up by H. pylori in its natural environment, we co-cultured donor strains in which the competence apparatus was inactivated to prevent their transformation with recipient strains and measured transformation frequency. We found similar transformation frequencies between purified gDNA and co-culture (Table 1). In addition, there was no significant difference in integration length when the source of genomic DNA was either purified gDNA or co-cultured cells (Table 1). Therefore, we pooled co-culture and gDNA data sets to evaluate integration length.
|Source of DNAa||Transformation frequency (×10−6)b Mean (SD)||Integration length (bp)|
|gDNA 1 ng||NSH57||12.4 (6.19)||22.5 (7.81)||9||4295e||5766f||17|
|J99||1.79 (0.773)||9.10 (3.85)|
|Co-culture||NSH57||10.8 (3.17)||4.17 (2.13)||3||3592e||5572f||8|
|J99||2.04 (0.255)||2.51 (0.381)|
We assessed 25 independent transformants each of NSH57 WT bacteria and the restriction mutant ΔΔhpyR(188III/CH4V) (Fig. 3B and C). As has been reported in previous studies (Kulick et al., 2008; Lin et al., 2009), some transformants (7 out of 57, 12%) had short regions of interspersed DNA from the recipient strain. This phenomenon may be caused by multiple strand invasion events during homologous recombination (Hulter and Wackernagel, 2008; Lin et al., 2009) so we chose to omit these clones from our analysis. The mean integration length of the ΔΔhpyR(188III/CH4V) mutant was 5704 bp, range 1716–11 920 bp (Fig. 3C) which represents a statistically significant increase in length of over 1.6 kbp over WT which had a mean integration length of 4070 bp, range 1022–8397 bp (t-test P = 0.03). This indicates that R-M systems decrease the length of DNA that is incorporated in the chromosome during genetic exchange in addition to lowering transformation frequency. We attempted to correlate integration end-points with target sites for REases R.Hpy188III and R.HpyCH4V, which are found at comparable frequencies at this locus (42 TCNNGA and 52 TGCA within 16 kbp). However, the large number of restriction sites masked any significant correlation (Fig. 4).
We took an alternate approach to quantify integration length in the complete absence of restriction. A mosaic gDNA was isolated from NSH57 harbouring approximately 10 kbp of J99 derived DNA surrounding the cagH::cat site. This gDNA is thus methylated by NSH57, thereby protecting restriction sites that are unprotected on donor gDNA isolated from J99; however, the sequence differences surrounding cagH::cat remain. NSH57 WT bacteria were transformed with either of two distinct mosaic gDNA templates or NSH57 gDNA and no difference in transformation frequency was observed [mean transformation frequency for mosaic gDNA 3.88 × 10−6 ± 2.32 × 10−6 transformants per ng of gDNA (n = 8) versus 1.74 × 10−6 ± 5.64 × 10−7 for WT (n = 4)]. This observation indicates that the 4% divergence between H. pylori strain NSH57 and J99 does not limit genetic exchange, suggesting that active type I or type III R-M systems form the additional barrier to genetic exchange observed in the ΔΔhpyR(188III/CH4V) double mutant.
Following transformation with the two mosaic gDNAs, 13 of the 18 clones analysed (72%) displayed a reduced integration length as compared with the starting 10 kbp region (Fig. 3D). Two other clones (10%) showed regions of interspersed DNA from the recipient strain and were not included in the analysis. The mean integration length was 5717 bp and is not statistically different from the value obtained previously for the ΔΔhpyR(188III/CH4V) mutant (5704 bp, P = 0.99). This number is an underestimation because integration length in the absence of restriction can reach 12 kbp (Fig. 3C), but our analysis was limited to the 10 kbp region derived from strain J99 that provides the SNPs differentiating donor and recipient DNA. These data suggest that additional H. pylori factors limit integration length to 5–6 kbp in the absence of restriction.
A diverse group of recombination factors have little influence on transformation frequency
To identify additional H. pylori factors that affect integration length, we carried out a limited screen of candidate genes in recipient bacteria. We assessed transformation frequency in cells lacking a variety of enzymes that process DNA including nucT (O'Rourke et al., 2004), addA (Amundsen et al., 2008; Wang and Maier, 2009), ruvC (Loughlin et al., 2003), recG (Kang et al., 2004), recJ and xseA. Transformation efficiency in H. pylori varies over at least 100-fold between strains but also between experiments (Levine et al., 2007; Amundsen et al., 2008; Marsin et al., 2008; Wang and Maier, 2009). In cells lacking the nucleases nucT, addA, recJ and xseA there was no significant effect on transformation frequency, as indicated by t-test P > 0.05 (Fig. 5A). To determine whether RecJ is required in the absence of AddAB, as it is in the absence of RecBCD in E. coli (Lovett and Kolodner, 1989), we constructed the ΔaddA ΔrecJ double mutant and observed no significant change in transformation frequency as compared with WT (Fig. 5A). We also measured transformation frequency in cells lacking the helicases recG and ruvB and the resolvase ruvC, but found no significant change in their absence (Fig. 5B). We note that our observations differ from published studies of the ΔnucT mutant (O'Rourke et al., 2004), the ΔrecG mutant (Kang et al., 2004), ΔruvB mutant (Kang and Blaser, 2008) and the ΔruvC mutant (Loughlin et al., 2003). We observed a high degree of variability in transformation frequency in these mutants and suggest that this variability results from the accumulation of unrepaired DNA damage and related poor growth (Dorer et al., 2010), as well as unknown strain differences.
The nuclease NucT reduces integration length
To address how homologous recombination factors control integration length, we used the restriction polymorphism assay to map integration end-points in several mutants. We first assessed cells lacking nucleases thought to be involved in the pre-synaptic stages of recombination. No change in integration length was observed for the ΔxseA mutant, the ΔrecJ mutant, the ΔaddA mutant or the ΔrecJ ΔaddA double mutant (Table 2). In contrast, in the ΔnucT mutant, the length of the DNA integrated in the chromosome following homeologous DNA recombination was increased downstream of the selective marker as compared with WT (mean WT 1856 bp versus 3023 bp in mutant, P = 0.03) (Table 2). Complementation reversed this increase in the length of DNA integrated in the chromosome (Table 2). These observations are consistent with the proposed role of NucT in natural competence (O'Rourke et al., 2004).
The helicase RecG and the resolvase RuvC have opposing effects on integration length
To address the role of post-synaptic recombination factors in controlling integration length, we mapped integration end-points in the ΔruvB mutant, the ΔrecG mutant and the ΔruvC mutant. The ΔruvB mutant showed no change in integration length, whereas the ΔrecG mutant showed reduced integration length upstream of the selective marker as compared with WT (mean WT 2099 bp versus 1298 bp in ΔrecG, P = 0.01) (Table 2). The decrease in integration length in the ΔrecG mutant was reversed by complementation (Table 2). We conclude that the RecG helicase but not RuvB increased integration length by extending the heteroduplex region formed between donor and recipient DNA.
In the ΔruvC mutant, the overall integration length was increased by more than 1.5 kbp as compared with WT (mean WT 4070 bp versus 5598 bp for ΔruvC, P = 0.03) (Table 2) and complementation restored integration length to near WT levels (Table 2). This result is consistent with a role for RuvC in limiting the migration of heteroduplex junctions by cleaving and resolving the donor and recipient DNA molecules.
DprB, a YqgF homologue, is required for natural transformation
As the RuvC resolvase limits integration length, we decided to characterize a ruvC homologue, HPG27_316 for its role in natural transformation. HPG27_316 lies immediately downstream of another cytosolic factor that is required for high levels of natural transformation, dprA (Ando et al., 1999) and the two genes were recently shown to be co-transcribed (Sharma et al., 2010), thus we refer to this gene as dprB. To confirm the role of dprA in natural transformation, we measured transformation frequency in the ΔdprA mutant and were unable to recover transformants of an antibiotic resistance gene (transformation frequency < 7 × 10−7). Our inability to obtain transformants in the ΔdprA mutant differs from the reported two-log decrease in transformation frequency of a point mutation (Ando et al., 1999). We suggest that the requirement for DprA may vary in different strain backgrounds or that loss of DprA drastically shortens integration length and transformation of the point mutation requires shorter integration length than transformation with a whole gene. DprB shares four conserved motifs and considerable sequence homology with yqgF from E. coli (30% identity, E-value 7e-05). YqgF was predicted to possess analogous catalytic activity to that of RuvC based on secondary structure (Aravind et al., 2000). Although yqgF is essential in E. coli (Freiberg et al., 2001), HPG27_316 could be deleted and resulted in decreased transformation frequency (Fig. 5B) and reduced growth (data not shown). The defects in transformation frequency and replication were both reversed by complementation (Fig. 5B and data not shown). DprB differs from RuvC in two important ways: first, DprB does not affect integration length (Table 2), and second, it does not affect recovery from treatment with either UV light (Fig. 6) or with DNA damaging anti-microbials (minimum inhibitory concentration of ciprofloxacin WT = 0.146 ± 0.03 µg ml−1, ΔdprB = 0.125 ± 0.01 µg ml−1) (Loughlin et al., 2003). These data indicate that DprB is not required for DNA repair, but is important for both natural transformation and growth.
- Top of page
- Experimental procedures
- Supporting Information
In this study, we uncover multiple factors that control the length of DNA integrated during transformation. We demonstrate that type II R-M systems are the main defence against foreign DNA in H. pylori, since they limit genetic exchange between unrelated strains by both destroying foreign DNA and decreasing the length of integrated DNA. Additionally, a diverse group of DNA processing enzymes including RecG, RuvC and NucT control integration length, but have little effect on transformation frequency. Finally, we identify DprB, a putative resolvase that is required for natural transformation but not DNA repair. Taken together, our studies indicate that R-M systems and homologous recombination control the length of DNA integrated into the chromosome, thus maintaining H. pylori genetic content.
Helicobacter pylori type II REases decrease the frequency of transformation 4- to 200-fold (Aras et al., 2002; Humbert and Salama, 2008). As transformation frequency is reduced but not blocked by restriction, we suggest that cleavage of incoming DNA is partial and limited to only a fraction of restriction sites. Partial cleavage of DNA may be achieved by limiting the concentration of REase. This endogenous DNA damage is likely countered by DNA repair, which is required for efficient growth of H. pylori in broth culture (Dorer et al., 2010). DNA binding proteins may also limit the accessibility of REases to the DNA molecule and prevent cleavage of the DNA (Polach and Widom, 1999). Finally, many questions remain about the temporal and spatial events surrounding REase cleavage. Double-stranded DNA taken up by natural competence is likely converted to single-stranded DNA, possibly during transit (Stingl et al., 2010), but single-stranded DNA is generally a poor substrate for REases (Wilson and Murray, 1991). One possibility is that the conversion to single-stranded DNA may limit cutting. Alternatively, REases may act before double-stranded DNA is converted to single-stranded. Although restriction is globally maintained in H. pylori, REase activity has a dual influence on fitness that results in positive and negative selection (Fig. 1).
Some naturally competent organisms rely on a species-specific DNA uptake sequence to distinguish self DNA from foreign DNA (Chen and Dubnau, 2004). We propose that H. pylori uses a species-independent approach, employing type II R-M systems to destroy unmethylated DNA. Our analysis shows that restriction is not fully effective, reducing transformation frequency approximately fourfold to eightfold or simply decreasing integration length, suggesting that several different REase cuts may be required to fully block integration of foreign DNA. As H. pylori strains are more likely to carry protective methylation against some of the REases, this requirement for multiple cuts may allow H. pylori to sample DNA from other H. pylori strains while preventing incorporation of DNA from more distant species that likely do not carry multiple protective methylations. Furthermore, recognition sites for R-M systems are not homogenously distributed in the H. pylori chromosome (O. Humbert, unpublished results), raising the interesting possibility that some chromosomal loci are more likely to be exchanged than others.
We find that both R-M systems and DNA processing enzymes are important for limiting the length of DNA integrated into the genome, but they have differing effects.The resolvase RuvC and endonuclease NucT limit the length of DNA integrated into the genome, whereas the helicase RecG increases integration length. In contrast to REases, these DNA processing enzymes have little effect on the overall transformation frequency. WT cells show multiple peaks of transformation frequency that vary over two logs during the growth cycle (Baltrus and Guillemin, 2006). Transformation frequency can be extremely variable in the tested mutants (Fig. 5), possibly due to accumulation of unrepaired DNA damage and related poor growth (Dorer et al., 2010). The variability in tranformation frequency has been previously documented for the ΔaddA mutant, with groups reporting an increase (Marsin et al., 2008), decrease (Wang and Maier, 2009) and no change (Amundsen et al., 2008) in transformation frequency. In addition, the ΔrecG mutant has been reported to have a hyper-recombination phenotype (Kang et al., 2004) but in our strain background, we do not observe this phenotype. We also do not consistently observe a decrease in transformation frequency of the ΔruvC mutant, in contrast to a previous report (Loughlin et al., 2003). Control of integration length, but no consistent effect on transformation frequency suggests that DNA processing enzymes act after synapsis, effecting only resolution of the recombinants. REases appear to act prior to synapsis and either cleave DNA inside the antibiotic resistance marker, thus preventing its recombination, or cleave near the marker, resulting in decreased integration length (Fig. 7).
Restriction enzymes mediate the tightest control of integration length prior to synapsis (Fig. 7) and this activity can prevent integration of whole genes by cleaving DNA to prevent integration or reduce integration length. The processing of 5′ single-stranded DNA appears dispensable for recombination during natural transformation in H. pylori, since recombination frequency and integration length were unchanged in the 5′-3′ exonuclease mutants. RecA loading by AddA/B is also not required, despite the requirement for RecA for natural transformation. One possibility is that dsDNA may be converted to ssDNA during transport through the competence machinery (Berge et al., 2002; Stingl et al., 2010). ssDNA is cooperatively bound by the conserved master regulator of homologous recombination RecA and by the natural competence specific protein DprA, both of which contribute to natural transformation in H. pylori (Thompson and Blaser, 1995; Ando et al., 1999; Mortier-Barriere et al., 2007). We suggest that DprB also acts prior to synapsis. It is unclear whether homologues of DprB function as resolvases, since the E. coli homologue YqgF is monomeric in solution, but all known Holliday junction resolvases function as dimers (Liu et al., 2003). The monomeric RuvC homologues may instead have a structural role for the initiation and stabilization of branched-DNA junctions during strand exchange. Based on coexpression with DprA (Sharma et al., 2010) and homology to the RuvC DNA binding protein, we speculate that DprB may modulate DprA binding to DNA or stabilize a pre-synaptic DNA structure, thereby increasing the efficiency of transformation.
Our data suggest that homologous recombination proteins act after synapsis to fine-tune integration length. During synapsis, a D-loop is formed between donor and recipient DNAs and the helicase RecG binds the three-stranded junction to drive migration and extend the heteroduplex region. The resolvase RuvC cleaves the junctions to separate donor from recipient DNA and limits integration length. RecG and RuvC therefore represent opposing forces that control how much of the donor DNA will be incorporated into the recipient DNA. Our data suggest that the branched-DNA specific helicase RecG increases integration length but the RuvB helicase does not. The absence of an integration length phenotype in the ΔruvB mutant may be due to the need for the entire RuvABC complex for resolution. In E. coli RuvAB is required for the delivery of RuvC to Holliday junctions (Shah et al., 1994). Mutation of RuvB could then result in the loss of both helicase and resolvase activities, which have opposing effects on integration length and may cancel each other out. The NucT nuclease also acts in this pathway to decrease integration length, but further study of its role in recombination is needed to place it in this pathway. As none of these recombination factors has any consistent effect on transformation frequency, we suggest that they act post-synaptically to modify the D-loop structure. Our data suggest that, except for RecA, the classical homologous recombination pathway makes only a minor contribution to recombination of DNA substrates acquired by natural competence.
Nucleotide divergence can decrease natural transformation in the absence of restriction. The divergence in DNA sequence between H. pylori strains did not affect transformation frequency, indicating that this level of non-homology (4% on average) is not a barrier to recombination. Mismatch repair genes have not been identified in H. pylori (Tomb et al., 1997), although mutation rate in the H. pylori strain used for these studies is equivalent to that found in E. coli (Baltrus et al., 2008; Dorer et al., 2010), suggesting that H. pylori has some capacity for repairing mismatches. Natural transformation with mosaic gDNA also showed evidence of interspersed sequence of the recipient (ISR) in some clones. These ISR were previously hypothesized to originate from multiple recombination events at a single locus, involving discrete DNA molecules processed by REases (Lin et al., 2009); however, we found ISR in the complete absence of restriction and conclude that R-M systems are not essential for ISR formation.
In conclusion, we demonstrated that both R-M systems and DNA processing enzymes control genetic exchange during natural transformation in H. pylori. REases and DNA processing enzymes likely act at different stages to limit recombination of foreign DNA. We have found that MTases tend to show higher functional conservation and their presence may allow transfer of DNA into a strain that has retained the cognate REase. DNA that passes the restriction barrier is then subject to homologous recombination proteins that further modulate its incorporation into the host chromosome. Together, these factors balance genetic diversification against the subversion of the genome by DNA from unrelated H. pylori strains.
- Top of page
- Experimental procedures
- Supporting Information
Bacterial strains and growth conditions
Helicobacter pylori strains (Table S1) were grown on solid horse blood agar (HB) plates containing 4% Columbia agar base (Oxoid), 5% defibrinated horse blood (HemoStat Laboratories), 0.2% β-cyclodextrin (Sigma), 10 µg ml−1 vancomycin (Sigma), 5 µg ml−1 cefsulodin (Sigma), 2.5 U ml−1 polymyxin B (Sigma), 5 µg ml−1 trimethoprim (Sigma), and 8 µg ml−1 amphotericin B (Sigma) at 37°C either under a microaerobic atmosphere generated using a CampyGen sachet (Oxoid) in a gas pack jar or in an incubator equilibrated with 14% CO2 and 86% air. For liquid culture, H. pylori was grown in Brucella broth (Difco) containing 10% fetal bovine serum (BB10, Invitrogen) with shaking in a gas pack jar containing a CampyGen sachet. For resistance marker selections, bacterial media were additionally supplemented with 15 µg ml−1 chloramphenicol (Cm, Sigma), 25 µg ml−1 kanamycin (Kan, Fisher Scientific) 2.5 µg ml−1 erythromycin (Ery, Fisher Scientific) or 36 µg ml−1 metronidazole (Mtz, Sigma).
DNA manipulations, such as restriction digestion, PCR and agarose gel electrophoresis, were performed according to standard procedures (Ausubel et al., 1997). H. pylori genomic DNA (gDNA) was prepared by Wizard genomic DNA preparation kits (Promega). Primers used for PCR and sequencing are described in Table S2. Plasmid DNA (Table S3) was isolated and prepared from E. coli using Qiagen Maxiprep kit (Qiagen). The FHCRC Genomics Shared Resource performed the sequencing of plasmid DNA and PCR products and the resulting sequences were analysed using Sequencher (Gene Codes Corporation).
Generation of H. pylori knockout isogenic mutants
Knockout alleles were constructed in H. pylori NSH57 using a vector-free allelic replacement strategy to generate alleles in which a non-polar kanamycin resistance (aphA3) cassette (Menard et al., 1993), an erm cassette conferring resistance to erythromycin (Lampson and Parisi, 1986; Dailidiene et al., 2006), or a chloramphenicol acetyl transferase (cat) resistance cassette fused to a sucrose sensitivity marker (sacB) (Copass et al., 1997; Humbert and Salama, 2008) replaced 80–90% of the coding sequence of the gene while preserving the start and stop codons. The primers used for this procedure are designated as 1 through 4 and are given in Table S2. After natural transformation with the appropriate PCR product and selection on Kan-, Ery- or Cm-containing media, four clones were evaluated by PCR to confirm replacement of the WT allele with the null allele. The ΔrecJ::kanΔaddA852-2540 double mutant was generated by transforming strain ΔrecJ::kanΔaddA::catsacB with a PCR product digested with SspI (New England Biolabs) and ligated with T4 DNA ligase (Invitrogen) to delete a 1.7 kbp intergenic region in addA. Transformants were selected on sucrose-containing HB plates, screened on Cm-containing media and checked by PCR to confirm the addA deletion. Urease activity and flagella-based motility were confirmed for all the clones generated. Single clones were used for transformation experiments.
Generation of H. pylori complemented mutants
Constructs for chromosomal complementation at the rdxA locus were made by cloning each gene individually into pLC292 (Terry et al., 2005), which were then introduced into H. pylori NSH57 by natural transformation and selection on Mtz-containing media (Dailidiene et al., 2006). Each gene was amplified using primers -XbaI and -SalI (Table S2) from H. pylori NSH57 gDNA using high-fidelity Taq polymerase (Platinum Taq, Invitrogen). The resulting PCR product was digested with XbaI and SalI (New England Biolabs), ligated into pLC292, and electroporated in E. coli strain DH10B or XA90 (Ezaz-Nikpay et al., 1994) for pOH10 (Table S3). All inserted genes were fully sequenced and contained the expected nucleotide sequences.
To generate knockout and complemented mutant strains of H. pylori, bacteria were freshly grown for 24–32 h on HB plates, transferred as patches onto fresh plates and grown for an additional 6–8 h. DNA (plasmid or PCR product) was diluted as appropriate in distilled water and 10 µl was added to each patch and incubated overnight. The mixture was harvested from the plate surface, resuspended in 350 µl phosphate-buffered saline (PBS) and plated onto selective HB plates.
To assess the frequency of natural transformation, recipient H. pylori bacteria freshly grown on HB plates were resuspended in 350 µl BB10 media and used to inoculate a 5 ml liquid culture grown for 6–8 h. The optical density at 600 nm (OD600) of this culture was measured and the culture was diluted back to OD600 0.015 to reach logarithmic phase of growth (OD600∼1) after overnight incubation. One hundred microlitres of recipient bacteria was dispensed in a flat-bottom 96-well plate and transformed in duplicates or triplicates with 10 µl of 1 ng µl−1 donor gDNA. Donor gDNA was constructed by inserting the cat resistance cassette at bp 483 in gene cagH of H. pylori strain NSH57 and J99 (hpG27-499 and jhp0489 respectively). To measure transformation of the ΔdprA mutant, donor gDNA was isolated from the G27 cag2::aphA3-sacB clone (Pinto-Santini and Salama, 2009). After 3 h incubation, 50 µl and 5 µl of the mixture were plated on Cm or Kan HB plates and 20 µl of a 10−5 dilution was plated on plain HB plates to determine the total number of viable bacteria. Transformation frequency was calculated as the number of Cm or Kan resistant colonies per colony-forming unit.
In the co-culture experiment, NSH57 and J99 ΔcomB10::ermΔcagH::cat were used as donor strains and to maximize DNA released in the culture media, we grew donor bacteria to stationary phase before mixing them with the recipient strain. ΔcomB10 strains show no detectable transformation (Dorer et al., 2010) ensuring unidirectional transformation in the co-culture assay. Recipient strains NSH57 hp0203-hp0204::aphA3 (Langford et al., 2006) and Δhpy188IIIR::aphA3ΔhpyCH4VR::erm were grown to logarithmic phase as described above and mixed at equal volume with the donor strains in a flat-bottom 96-well plate. After 3 h co-incubation, 100 µl of the mixture was plated on Cm + Kan HB plates to select for recombinant clones and 20 µl of a 10−5 dilution was plated on Kan HB plates to determine the total number of recipient bacteria.
Mapping of integration end-points
Chromosomal DNA of the transformants was prepared and 5–7 kbp of the regions upstream and downstream of the cat marker were amplified by PCR using primer pairs -6FcagH/cagH::cat-3 and cagH::cat-4/5RcagH (Table S2) respectively. The resulting PCR products were purified with the DNA clean and concentrator-5 kit (Zymo Research) and digested with the appropriate restriction enzymes for a minimum of 4 h (New England Biolabs) or sequenced by the FHCRC Genomics Shared Resource.
Sensitivity to UV and antimicrobial agents
UV sensitivity assays were carried out as described previously (Amundsen et al., 2008). For antimicrobial sensitivity testing, H. pylori were grown overnight in liquid culture to OD600 = 0.3, and 200 µl was plated on solid medium lacking all other antimicrobials, and incubated for 30 min in a CO2 incubator. E-test strips (AB Biodisk) were then placed on the plates, which were further incubated for two days and read according to the manufacturer's instructions.
A t-test was used to compare the mean of integration lengths or transformation frequency between WT bacteria and mutant clones and those comparisons resulting in a P-value of < 0.05 were considered significant. All statistical analyses were performed using the SAS version 9.1 software (SAS Institute, Cary, NC, USA).
In silico genomic analysis
Helicobacter pylori sequences were retrieved from the H. pylori genome browser http://hpylori.ucsc.edu/. For H. pylori strain NSH57, the sequence of the parent strain G27 was used (Baltrus et al., 2009). The distribution of restriction sites and single nucleotide polymorphism was analysed with Sequencher (Gene Codes Corporation).
- Top of page
- Experimental procedures
- Supporting Information
We thank Jutta Fero for generating the mutants ΔrecJ::kan and ΔaddA::kan, Sarah Talarico for help with the statistical analyses, Douglas Berg (Washington University) for providing the erm cassette and Brendan Cormack (Johns Hopkins University) for E. coli XA90. This work is supported by grants AI054423 (N.R.S.) and DK080894 (M.S.D.) from the NIH. The contents of this study are solely the responsibility of the authors and do not necessarily represent the official views of the NIAID or NIDDK.
- Top of page
- Experimental procedures
- Supporting Information
- 1999) Genomic-sequence comparison of two unrelated isolates of the human gastric pathogen Helicobacter pylori. Nature 397: 176–180. , , , , , , et al. (
- 2008) Helicobacter pylori AddAB helicase-nuclease and RecA promote recombination-related DNA repair and survival during stomach colonization. Mol Microbiol 69: 994–1007. , , , , , , and (
- 1999) HP0333, a member of the dprA family, is involved in natural transformation in Helicobacter pylori. J Bacteriol 181: 5572–5580. , , , and (
- 2002) Helicobacter pylori interstrain restriction-modification diversity prevents genome subversion by chromosomal DNA from competing strains. Nucleic Acids Res 30: 5391–5397. , , , and (
- 2000) Survey and summary: Holliday junction resolvases and related nucleases: identification of new families, phyletic distribution and evolutionary trajectories. Nucleic Acids Res 28: 3417–3432. , , and (
- 1997) Short Protocols in Molecular Biology. New York: John Wiley & Sons. , , , , , , and (
- 2007) Identification of Helicobacter pylori genes that contribute to stomach colonization. Infect Immun 75: 1005–1016. , , , , , , and (
- 2006) Multiple phases of competence occur during the Helicobacter pylori growth cycle. FEMS Microbiol Lett 255: 148–155. , and (
- 2008) Natural transformation increases the rate of adaptation in the human pathogen Helicobacter pylori. Evolution 62: 39–49. , , and (
- 2009) The complete genome sequence of Helicobacter pylori strain G27. J Bacteriol 191: 447–448. , , , , , , et al. (
- 2002) Uptake of transforming DNA in Gram-positive bacteria: a view from Streptococcus pneumoniae. Mol Microbiol 45: 411–421. , , , , and (
- 2003) DNA restriction is a barrier to natural transformation in Pseudomonas stutzeri JM300. Microbiology 149: 895–901. , , and (
- 2006) Molecular analysis of the bacterial microbiota in the human stomach. Proc Natl Acad Sci USA 103: 732–737. , , , , , , et al. (
- 1980) Restriction and modification in B. subtilis: effects on transformation and transfection with native and single-stranded DNA. Mol Gen Genet 179: 103–110. , , , and (
- 1974) Exonuclease VII of Escherichia coli. Mechanism of action. J Biol Chem 249: 4553–4561. , and (
- 2004) DNA uptake during bacterial transformation. Nat Rev Microbiol 2: 241–249. , and (
- 1997) Introduction of unmarked mutations in the Helicobacter pylori vacA gene with a sucrose sensitivity marker. Infect Immun 65: 1949–1952. , , and (
- 2010) A type III-like restriction endonuclease functions as a major barrier to horizontal gene transfer in clinical Staphylococcus aureus strains. Proc Natl Acad Sci USA 107: 11954–11958. , , , , , and (
- 2006) Contraselectable streptomycin susceptibility determinant for genetic manipulation and analysis of Helicobacter pylori. Appl Environ Microbiol 72: 5908–5914. , , , and (
- 2010) DNA damage triggers genetic exchange in Helicobacter pylori. PLoS Pathog 6: e1001026. , , and (
- 1991) Formation and resolution of recombination intermediates by E. coli RecA and RuvC proteins. Nature 354: 506–510. , , , , , and (
- 1994) Construction of an overproduction vector containing the novel srp (sterically repressed) promoter. Protein Sci 3: 132–138. , , , and (
- 2001) Recombination and mutation during long-term gastric colonization by Helicobacter pylori: estimates of clock rates, recombination size, and minimal age. Proc Natl Acad Sci USA 98: 15056–15061. , , , , , , and (
- 2001) Identification of novel essential Escherichia coli genes conserved among pathogenic bacteria. J Mol Microbiol Biotechnol 3: 483–489. , , , , , and (
- 2005) Gain and loss of multiple genes during the evolution of Helicobacter pylori. PLoS Genet 1: e43. , , , , , , et al. (
- 1998) Natural competence for DNA transformation in Helicobacter pylori: identification and genetic characterization of the comB locus. Mol Microbiol 28: 1027–1038. , , , and (
- 2000) Genetic competence in Helicobacter pylori: mechanisms and biological implications. Res Microbiol 151: 487–491. , , , , and (
- 2008) Frequent integration of short homologous DNA tracks during Acinetobacter baylyi transformation and influence of transcription and RecJ and SbcCD DNases. Microbiology 154: 3676–3685. , and (
- 2008) The Helicobacter pylori HpyAXII restriction-modification system limits exogenous DNA uptake by targeting GTAC sites but shows asymmetric conservation of the DNA methyltransferase and restriction endonuclease components. Nucleic Acids Res 36: 6893–6906. , and (
- 2008) Repair and antirepair DNA helicases in Helicobacter pylori. J Bacteriol 190: 4218–4224. , and (
- 2004) Effect of host species on recG phenotypes in Helicobacter pylori and Escherichia coli. J Bacteriol 186: 7704–7713. , , , , and (
- 2006) Functional and topological characterization of novel components of the comB DNA transformation competence system in Helicobacter pylori. J Bacteriol 188: 882–893. , , , , , and (
- 1999) Emergence of recombinant strains of Helicobacter pylori during human infection. Mol Microbiol 31: 31–43. , , and (
- 2000) Functional analysis of putative restriction-modification system genes in the Helicobacter pylori J99 genome. Nucleic Acids Res 28: 3216–3223. , , , , , , and (
- 2008) Mosaic DNA imports with interspersions of recipient sequence after natural transformation of Helicobacter pylori. PLoS ONE 3: e3797. , , , , , and (
- 2006) Pathogenesis of Helicobacter pylori infection. Clin Microbiol Rev 19: 449–490. , , and (
- 1984) Cloning in Streptococcus pneumoniae of the gene for DpnII DNA methylase. J Bacteriol 157: 934–936. , and (
- 1986) Naturally occurring Staphylococcus epidermidis plasmid expressing constitutive macrolide-lincosamide-streptogramin B resistance contains a deleted attenuator. J Bacteriol 166: 479–483. , and (
- 2006) In vitro and in vivo complementation of the Helicobacter pylori arginase mutant using an intergenic chromosomal site. Helicobacter 11: 477–493. , , , , and (
- 2007) Plastic cells and populations: DNA substrate characteristics in Helicobacter pylori transformation define a flexible but conservative system for genomic variation. FASEB J 21: 3458–3467. , , , , , , et al. (
- 2009) Natural transformation of Helicobacter pylori involves the integration of short DNA fragments interrupted by gaps of variable size. PLoS Pathog 5: e1000337. , , , , , and (
- 2001) Comparative genomics of the restriction-modification systems in Helicobacter pylori. Proc Natl Acad Sci USA 98: 2740–2745. , , , and (
- 2000) Frequent interspecific genetic exchange between commensal Neisseriae and Neisseria meningitidis. Mol Microbiol 36: 1049–1058. , , , and (
- 2003) Solution structure of the hypothetical protein YqgF from Escherichia coli reveals an RNAse H fold. J Biomol NMR 27: 389–392. , , and (
- 2003) Helicobacter pylori mutants defective in RuvC Holliday junction resolvase display reduced macrophage survival and spontaneous clearance from the murine gastric mucosa. Infect Immun 71: 2022–2031. , , , , and (
- 1989) Identification and purification of a single-stranded-DNA-specific exonuclease encoded by the recJ gene of Escherichia coli. Proc Natl Acad Sci USA 86: 2627–2631. , and (
- 2008) Unveiling novel RecO distant orthologues involved in homologous recombination. PLoS Genet 4: e1000146. , , , , and (
- 1993) Overproduction of 3'-aminoglycoside phosphotransferase type I confers resistance to tobramycin in Escherichia coli. Antimicrob Agents Chemother 37: 78–83. , , , , , and (
- 2007) A key presynaptic role in transformation for a widespread bacterial protein: DprA conveys incoming ssDNA to RecA. Cell 130: 824–836. , , , , , , et al. (
- 2006) The complete genome sequence of a chronic atrophic gastritis Helicobacter pylori strain: evolution during disease progression. Proc Natl Acad Sci USA 103: 9999–10004. , , , , , , et al. (
- 2004) Evidence for the active role of a novel nuclease from Helicobacter pylori in the horizontal transfer of genetic information. J Bacteriol 186: 2586–2593. , , , , and (
- 2009) Cag3 is a novel essential component of the Helicobacter pylori Cag type IV secretion system outer membrane subcomplex. J Bacteriol 191: 7343–7352. , and (
- 1999) Restriction enzymes as probes of nucleosome stability and dynamics. Methods Enzymol 304: 278–298. , and (
- 1989) The barrier to recombination between Escherichia coli and Salmonella typhimurium is disrupted in mismatch-repair mutants. Nature 342: 396–401. , , and (
- 1993) The effect of DNA sequence divergence on sexual isolation in Bacillus. Genetics 134: 401–408. , and (
- 2009) REBASE – a database for DNA restriction and modification: enzymes, genes and genomes. Nucleic Acids Res 38: D234–D236. , , , and (
- 2007) Genetic analysis of Helicobacter pylori strain populations colonizing the stomach at different times postinfection. J Bacteriol 189: 3834–3845. , , , , , , and (
- 2004) The diversity within an expanded and redefined repertoire of phase-variable genes in Helicobacter pylori. Microbiology 150: 817–830. , , , and (
- 1994) Genetic recombination in E. coli: RuvC protein cleaves Holliday junctions at resolution hotspots in vitro. Cell 79: 853–864. , , and (
- 2010) The primary transcriptome of the major human pathogen Helicobacter pylori. Nature 464: 250–255. , , , , , , et al. (
- 1999) Holliday junction processing in bacteria: insights from the evolutionary conservation of RuvABC, RecG, and RusA. J Bacteriol 181: 5543–5550. , , and (
- 2010) Composite system mediates two-step DNA uptake into Helicobacter pylori. Proc Natl Acad Sci USA 107: 1184–1189. , , , , and (
- 1976) Restriction enzymes do not play a significant role in Haemophilus homospecific or heterospecific transformation. J Bacteriol 128: 212–220. (
- 2005) Chemotaxis plays multiple roles during Helicobacter pylori animal infection. Infect Immun 73: 803–811. , , , and (
- 1995) Isolation of the Helicobacter pylori recA gene and involvement of the recA region in resistance to low pH. Infect Immun 63: 2185–2193. , and (
- 1997) The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388: 539–547. , , , , , , et al. (
- 2009) A RecB-like helicase in Helicobacter pylori is important for DNA repair and host colonization. Infect Immun 77: 286–291. , and (
- 1994) The processing of recombination intermediates: mechanistic insights from studies of bacterial proteins. Cell 76: 9–15. (
- 1995) Branch migration of three-strand recombination intermediates by RecG, a possible pathway for securing exchanges initiated by 3'-tailed duplex DNA. EMBO J 14: 3302–3310. , and (
- 1991) Restriction and modification systems. Annu Rev Genet 25: 585–627. , and (
- 2000) Identification of type II restriction and modification systems in Helicobacter pylori reveals their substantial diversity among strains. Proc Natl Acad Sci USA 97: 9671–9676. , , , and (
- 1997) Barriers to recombination between closely related bacteria: MutS and RecBCD inhibit recombination between Salmonella typhimurium and Salmonella typhi. Proc Natl Acad Sci USA 94: 9786–9791. , and (
- 1995) The size and continuity of DNA segments integrated in Bacillus transformation. Genetics 141: 1231–1243. , and (
- Top of page
- Experimental procedures
- Supporting Information
|MMI_7456_sm_TableS1-3.pdf||62K||Supporting info item|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.