Email updates

Keep up to date with the latest news and content from BMC Genomics and BioMed Central.

Open Access Research article

Domain exchange at the 3’ end of the gene encoding the fratricide meningococcal two-partner secretion protein A

Jesús Arenas1*, Kim Schipper2, Peter van Ulsen3, Arie van der Ende24 and Jan Tommassen1

Author Affiliations

1 Department of Molecular Microbiology, Utrecht University, Padualaan 8, Utrecht 3584 CH, The Netherlands

2 Department of Medical Microbiology, Center for Infection and Immunity Amsterdam, Academic Medical Center, P.O. Box 22660, Amsterdam 1100 DD, The Netherlands

3 Department of Molecular Microbiology, Vrije Universiteit, de Boelelaan 1085, Amsterdam 1081 HV, The Netherlands

4 Netherlands Reference Laboratory for Bacterial Meningitis, Academic Medical Center, Amsterdam, The Netherlands

For all author emails, please log on.

BMC Genomics 2013, 14:622  doi:10.1186/1471-2164-14-622


The electronic version of this article is the complete one and can be found online at: http://www.biomedcentral.com/1471-2164/14/622


Received:21 January 2013
Accepted:13 September 2013
Published:14 September 2013

© 2013 Arenas et al.; licensee BioMed Central Ltd.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background

Two-partner secretion systems in Gram-negative bacteria consist of an outer membrane protein TpsB that mediates the secretion of a cognate TpsA protein into the extracellular milieu. TpsA proteins have diverse, often virulence-related functions, and some of them inhibit the growth of related bacteria. In Neisseria meningitidis, several functions have been attributed to the TpsA proteins. Downstream of the tpsB and tpsA genes, several shorter tpsA-related gene cassettes, called tpsC, are located interspersed with intervening open-reading frames (IORFs). It has been suggested that the tpsC cassettes may recombine with the tpsA gene as a mechanism of antigenic variation. Here, we investigated (i) whether TpsA of N. meningitidis also has growth-inhibitory properties, (ii) whether tpsC cassettes recombine with the tpsA gene, and (iii) what the consequences of such recombination events might be.

Results

We demonstrate that meningococcal TpsA has growth-inhibitory properties and that the IORF located immediately downstream of tpsA confers immunity to the producing strain. Although bioinformatics analysis suggests that recombination between tpsC cassettes and tpsA occurs, detailed analysis of the tpsA gene in a large collection of disease isolates of three clonal complexes revealed that the frequency is very low and cannot be a mechanism of antigenic variation. However, recombination affected growth inhibition. In vitro experiments revealed that recombination can be mediated through acquirement of tpsC cassettes from the environment and it identified the regions involved in the recombination.

Conclusions

Meningococcal TpsA has growth-inhibitory properties. Recombination between tpsA and tpsC cassettes occurs in vivo but is rare and has consequences for growth inhibition. A recombination model is proposed and we propose that the main goal of recombination is the collection of new IORFs for protection against a variety of TpsA proteins.

Keywords:
Two-partner secretion; Neisseria meningitidis; Contact-dependent growth inhibition; Gene conversion; Interstrain competition

Background

Neisseria meningitidis is a Gram-negative bacterium that colonizes the human respiratory tract and occasionally causes meningitis and/or sepsis. Disease-related strains can express either one of six different capsular polysaccharides (A, B, C, Y, W135 and X). Polysaccharide-based vaccines have been developed against four of these capsule types. Unfortunately, the serogroup B capsular polysaccharide cannot be used as a vaccine due to its poor immunogenicity, while non-capsular antigens seem to be effective only temporarily and in specific geographic areas due to high antigenic variability in meningococci.

Various outer-membrane components and secreted proteins exhibit variable expression and/or antigenic diversity, which plays an important role in immune escape and has serious implications for the development of effective vaccines [1]. The high variability of these antigens is based on different mechanisms, including gene conversion and slipped-strand mispairing [2].

Gene conversion is the non-reciprocal exchange of DNA fragments situated at different chromosomal locations. It is based upon homologous recombination between the donor and recipient loci [3,4]. Virulence-associated surface structures in bacterial pathogens are often subject to antigenic variation by gene conversion. This has been studied extensively for the pilin locus of Neisseria spp. [5-7]. The retractile type IV pili of Neisseria are involved in attachment to host tissues [8,9], DNA acquisition [10,11] and twitching motility [12]. In addition to the pilin expression locus pilE, the genome of these bacteria contains a repertoire of promoter-less pilS cassettes, with homologous and variable regions relative to pilE. These pilS cassettes can unidirectionally be transferred to the pilE expression locus, resulting in partial alteration of the pilin sequence. Alternatively, this recombination event can switch off the formation of pili by introduction of a premature stop codon in pilE. The frequency of gene conversion seems to be strain and species dependent and can be as high as 0.13 recombination events per cell in the case of gonococcal strain FA1090 [13].

In two-partner secretion (TPS) systems, a large protein, generically called TpsA, is secreted through a specific outer membrane transporter, designated TpsB [14,15]. TpsA is synthesized as a precursor with a signal sequence and transported to the periplasm via the Sec machinery. In the periplasm, it interacts via its N-terminally located TPS domain with TpsB and is transported to the cell surface. TPS systems of N. meningitidis seem of relevance to pathogenicity as suggested by reported roles in adhesion [16], intracellular survival [17] and biofilm formation [18]. Cluster analysis of the sequences of the TpsB proteins and of the TPS domains of the TpsA proteins revealed the presence of three different TPS systems in N. meningitidis[19]. Different isolates of N. meningitidis can contain one to three different TPS systems, and some of these systems can contain two tpsA genes [19]. In the genome sequence of strain MC58, for example, five different tpsA genes were identified, two of system 1 (designated tpsA1a and tpsA1b), two of system 2 (tpsA2a and tpsA2b), and one of system 3 (tpsA3). In contrast, the genome sequences of strains FAM18, 053442 and Z2491 contain only one single tpsA, i.e. a tpsA of system 1 (tpsA1 a.k.a. hrpA) [19-21], which appears most widespread among meningococcal isolates [19]. The tpsB and tpsA genes are situated on specific genetic islands on the chromosome. Downstream of the tpsA genes, several tpsA-related open reading frames (ORFs) are located, generically referred to as tpsC cassettes [21], interspersed with small intervening ORFs (IORFs) (see Figure  1 for examples). As compared to tpsA, all tpsC cassettes are missing extensive portions at the 5’ end including the segment encoding the signal sequence and the TPS domain. They do share sequence similarity with a central part of tpsA but show an entirely different 3’ terminal sequence [21]. From this observation, it was postulated that these tpsC cassettes could substitute the 3’ end of tpsA by gene conversion, thereby causing antigenic variation of TpsA, similarly as in the pilE/S system [2,21].

thumbnailFigure 1. Comparison of the genetic organization of three TPS islands in different meningococcal strains. Each island consists from the 5’ to the 3’ end of a tpsB, a tpsA, and a variable number of tpsC cassettes interspersed with intervening ORFs (IORFs) (open arrows). The tpsA genes and tpsC cassettes can display very different domains of ~450 bp at their 3’ end as indicated by the different colors. Domains with very high sequence similarity are indicated with the same color. Also the IORFs show high heterogeneity and IORFs with high sequence similarity are indicated with the same color. Note that tpsA or tpsC genes with similar sequences at their 3’ end are always followed by IORFs with high sequence similarity. Regions with high sequence similarity between the different islands are indicated by grey shading, indicating gross rearrangements in the organization between the islands. Note that strain MC58 contains five tpsA genes, including two tpsA genes of system 1 (Figure S1 in Additional file 1). The island depicted here contains the tpsA1b gene with locus tag NMB0497.

Recently, Aoki et al. described that certain TpsA proteins of E. coli inhibit the growth of related bacteria [22,23]. This activity is mediated by direct contact between bacteria of the same species through the interaction of TpsA with the conserved outer membrane protein BamA in the target cells [24]. The activity, referred to as contact-dependent growth inhibition (CDI), was also observed in other bacteria [22,25].The CDI activity resides in the C-terminal region of TpsA, which is cleaved off from TpsA after interaction with the receptor and delivered in the cytoplasm of target cells, where it exerts DNase or tRNase activity. Immunity in the producing cells is provided by a small protein encoded by the ORF immediately downstream of tpsA. This immunity protein binds the toxic domain of TpsA and inhibits its activity [22]. Like in the tps gene clusters of N. meningitidis, these tpsA genes are often followed by several 5’-truncated tpsA homologs displaying entirely different 3’ ends [26]. It was proposed that these tpsCs comprise a reservoir of alternative C-terminal regions and that exchange of the 3’ end of tpsA by a tpsC cassette alters the CDI mechanism [26]. However, so far, no study addressed whether such substitutions at the 3’ end of tpsA actually take place.

In the present study, we first investigated whether also the meningococcal TpsA proteins mediate growth inhibition. Then, we investigated whether recombination at the 3’ end of tpsA indeed occurs in vivo and what the frequency, the mechanism, and the function of such recombination might be.

Results

Meningococcal TpsA1 mediates growth inhibition

To investigate whether meningococcal TpsA proteins may possess growth-inhibitory activity, we initially focused on strain FAM18, which contains a single TPS island with one tpsA1 gene and five tpsC cassettes (Figure  1) [2,20]. However, since this strain appeared non-transformable, we switched to strain B16B6, another reference strain of the same clonal complex (cc), i.e. cc11. Sequencing of the complete TPS island of strain B16B6 (Genbank: HQ420265) revealed exactly the same genetic organization and > 99% sequence identity at the nucleotide level with that of FAM18. The TpsA protein of B16B6 shows only three amino-acid substitutions relative to that of FAM18.

Next, we constructed a ΔtpsA-tpsC mutant of strain B16B6 lacking tpsA and all tpsC cassettes and IORFs. In addition, since the presence of a capsule inhibits CDI in E. coli[22], we used capsule-deficient derivatives in all assays described below. When we co-incubated the wild type and the ΔtpsA-tpsC mutant in liquid medium as described for E. coli[23], no growth inhibition of the mutant was observed independent of the media used (TSB, LB, or RPMI), co-incubation time (3–6 hours) or initial ratio of target and killer cells (ranging from 1:1 to 1:1000). It has been reported that meningococcal TpsA affects later stages of biofilm development (48 hours old biofilms) and that its production is increased under anaerobic conditions [18]. Thus, we co-incubated the wild type and the mutant on agar plates in a candle jar with low oxygen concentrations. Under these conditions, growth inhibition of the mutant was observed as a decline in time of the ratio of the colony-forming units (CFU) of the mutant over that of the wild type (Figure  2A, grey bars). In contrast, growth of a ΔtpsB mutant, which is defective in the secretion of TpsA but still contains all IORFs, was not affected when co-incubated with the wild type (Figure  2A, white bars). Also, growth of a ΔtpsC2-5 mutant, which lacks tpsC2 to tpsC6 and the IORFs in between, but still contains IORF1 and IORF2, was not affected when co-incubated with the wild type (Figure  2A, hatched bars), suggesting that protection against the growth-inhibiting properties of TpsA is mediated by IORF1 and/or IORF2.

thumbnailFigure 2. Growth-inhibitory function of meningococcal TpsA. (A) Cells of an unencapsulated derivative of strain B16B6, carrying a plasmid with a chloramphenicol-resistance marker, were mixed 1:1 with cells of ΔtpsA-tpsC, ΔtpsB or ΔtpsC2-5 mutants, all carrying a kan cassette. The suspensions were spotted on GC plates without antibiotics and incubated for various time periods. In the experiment with the ΔtpsC2-5 mutant as the target cells, the suspensions were only spotted after 0 and 48 hours. The ratios of the ΔtpsA-tpsC (grey bars), ΔtpsB (white bars), or ΔtpsC2-5 (hatched bars) mutants over wild-type bacteria in the spots was determined by plating on GC media containing kanamycin or chloramphenicol and counting colony-forming units after overnight incubation. (B) A rifampicin-resistant derivative of the unencapsulated B16B6 strain was mixed 1:1 with the ΔtpsA-tpsC mutant harboring plasmid pFPIORF1, which contains the first IORF of B16B6 under the control of an IPTG-inducible promoter. The bacteria were incubated in the presence or absence of IPTG for 24 hours as above. The ratio of the ΔtpsA-tpsC mutant over the wild-type bacteria was determined as described. Results are means and s.d. of three independent experiments. All the strains tested here did not show differences in viability when grown separately.

To confirm the role of the IORF1 immediately downstream of tpsA in B16B6 in conferring immunity, the corresponding gene was cloned behind an isopropyl-β-D-1-thiogalactopyranoside (IPTG)-inducible promoter on plasmid and introduced into the ΔtpsA-tpsC mutant. Growth of the complemented strain was inhibited when co-incubated with the wild-type strain in the absence but not in the presence of IPTG (Figure  2B). Thus, the IORF indeed confers immunity to growth inhibition mediated by TpsA. In conclusion, the meningococcal TpsA protein has growth-inhibitory activity and the IORF immediately downstream of tpsA provides immunity to this activity.

Structural organization of meningococcal TpsA1 and TpsC proteins

To study possible genetic exchange between tpsA genes and tpsC cassettes, we first extended previous bioinformatic analyses using available genome sequences [19,21], thereby focusing on TPS system 1 because of its widespread presence in meningococcal isolates. The genetic organization of TPS system 1 islands in the available genome sequences of strains α14 [27], α153 [27], α275 [27], 053442 [28], 8013 [29], FAM18 [2], MC58 [30] and Z2491 [31] is depicted in Figure S1 in Additional file 1 with selected examples shown in Figure  1.

Additional file 1. Supplemental information containing one supplemental table and seven supplemental figures.

Format: PDF Size: 1.3MB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

Multiple sequence alignments of the corresponding TpsA proteins (data not shown) revealed that they are modular and consist of three regions: a highly conserved N-terminal region, a semi-variable central region, and a hyper-variable C-terminal region. The N-terminal region of ~500 amino acids shows > 98% identity between all aligned sequences and contains the TPS domain required for interaction with the TpsB protein in the outer membrane. For the central region of ~900 amino-acid residues, two main allelic variants were identified designated core 1 and core 2 (Figure S2A in Additional file 1). These sequences showed > 19% divergence between the groups and < 2% divergence within the groups. The C-terminal regions of ~600 amino-acid residues, where the toxic activity of TpsA resides, are highly variable with sequence divergence ranging from 5-50%. Sequence divergence is particularly high in the last ~150 residues, where five different groups of sequences can be discriminated among the 10 TpsA proteins compared (Figure S2B in Additional file 1). These C-terminal domains show very high sequence conservation within the groups (> 98%) and extremely low, if any, sequence similarity between the groups.

The polypeptides putatively encoded by the tpsC cassettes are variable in length (260–893 residues). In their N-terminal part upstream of a conserved VENN motif (22), they share sequence similarity with TpsA proteins (see Figure S3 in Additional file 1, which shows an alignment of the C-terminal region of TpsA and the putative TpsC proteins of strain FAM18). The corresponding DNA sequences are presumably used for homologous recombination into the tpsA locus. Downstream of the VENN motif and particularly within the C-terminal moieties of ~150 amino-acid residues, the sequences are again hyper-variable (Figure S3 in Additional file 1), some sharing high sequence similarity with C-terminal segments of TpsA proteins and others constituting new groups of C-terminal domains. Sequence comparisons identified 11 different groups of TpsA/TpsC C-terminal sequences (indicated by different colors in Figure  1 and Figure S1 in Additional file 1) with > 90% of similarity within the groups (Additional file 1: Table S1). The results of such sequence analyses are consistent with the occurrence in vivo of the suggested gene conversion at the 3’ end of the tpsA gene [21]. For example, the tpsA of strain Z2491 could have been generated by recombination of a tpsC cassette similar to tpsC3 of strain FAM18 into a tpsA like NMB0497 of MC58 (Figure  1). However, it should be noted that also recombination mechanisms other than gene conversion could be responsible for the observed variation in the 3’ end of tpsA.

The polypeptides encoded by the IORFs located in between tpsA and tpsCs (open arrows in Additional file 1: Figure  1 and Figure S1) putatively encode immunity proteins to confer protection against the toxic activity of the products of the directly upstream located tpsA or tpsC. They can be grouped with > 95% sequence similarity within the groups and very low, if any, similarity between the groups (data not shown). Interestingly, tpsAs and tpsCs with a similar C-terminal sequence are always immediately followed by an IORF from the same sequence group. IORFs of the same group are colored identically in Figure  1 and Figure S1 in Additional file 1, and, to reflect the interconnectivity with the upstream TpsA/TpsC, the same color is used for the IORFs as for the C-terminal ends of the upstream TpsA or TpsC.

Replacement of the 3’ end of tpsA is rare in cc11 and cc8 strains

The bioinformatic analysis above is consistent with the occurrence of the suggested gene conversion at the 3’ end of meningococcal tpsA. However, the TPS islands in the strains analyzed could also have been acquired independently by horizontal gene transfer from different sources. To determine whether recombination of tpsC cassettes with the 3’ end of tpsA indeed occurs in vivo, we decided to analyze the TPS system 1 of various isolates from the same cc. If recombination indeed takes place at frequencies comparable, for example, to that reported for gene conversion at the pilE locus, different tpsC cassettes would be expected to be present at the C terminus of TpsA in such isolates. For this purpose, we first focused on cc11, a lineage predominantly associated with invasive disease [32] and therefore useful to evaluate the possible implications of such recombination events during human infection.

The observation that the organization of the TPS island of B16B6 is very similar to that of FAM18 (see above) in spite of the very different isolation years of these strains being 23 years apart (Additional file 2) already suggested that the tps loci might actually be rather stable. We then decided to analyze a large panel of 277 cc11 disease isolates collected in The Netherlands between 1960 and 2008 (Additional file 2). We used PCR to assess for possible replacements of the 3’ end of tpsA by tpsC cassettes as outlined in detail in Figure S4 in Additional file 1; the results for individual isolates are listed in Additional file 2: Table S2. With only three exceptions, which are described in detail in the next section, all isolates showed the presence of the same cassette at the 3’ end of tpsA as in FAM18. These results demonstrate that the tpsA locus in cc11 strains is remarkably stable. A possible explanation for the observed stability of tpsA is that the cassette present at the C terminus of TpsA in cc11 isolates is maintained because it is required for the highly invasive phenotype of these strains or because of reduced interbacterial competition during infection. To investigate this possibility, we also analyzed six cc11 carrier isolates using similar procedures (Figure S4 in Additional file 1), which revealed again the same organization as that in FAM18, suggesting that the 3’ end of tpsA is also stably maintained in these isolates and that the lack of exchange is not favored by host invasion.

Additional file 2: Table S2. listing the meningococcal isolates used in this study and summarizing the results of PCR and Western blot analyses.

Format: PDF Size: 520KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

For comparison, we determined the frequency of another well-known mechanism of genetic variation, i.e. phase-variation by slipped-strand mispairing at direct DNA repeats, in cc11 strains. Thus, the expression of the autotransporter NalP and the outer membrane Opa proteins, which both exhibit phase-variable expression [33,34], was evaluated by Western blotting in a panel of 56 cc11 isolates, including both carrier and disease isolates (Additional file 2: Table S2). Of these isolates, 32 were positive for NalP expression (57%) and 23 for Opa expression (41%). Thus, the frequency of recombination at the 3’ end of the tpsA gene is clearly much lower than that of another genetic mechanism of variation used by this pathogen during infection.

Finally, we analyzed strains from another hyperinvasive cc, i.e. cc8, using the same strategy. The TPS island of cc8 strain NMB was recently reported to have a similar organization as that in FAM18 [18]. We sequenced tpsA of another cc8 strain, strain 2996 (Genbank: HQ420264), and found 99% identity at the nucleotide level with tpsA of FAM18. PCR analysis targeted for various fragments in the downstream region (Figure S4 in Additional file 1) indicated the same organization of tpsCs as in FAM18. Next, a panel of 91 cc8 disease isolates was analyzed (Figure S4 in Additional file 1); all of them appeared to contain the same cassette at the 3’ end of tpsA as in FAM18. Thus, recombination at the 3’ end of the tpsA locus did not occur in the isolates examined and is therefore rare, also in cc8 strains.

Characterization of the TPS island in the deviating cc11 isolates

The three cc11 isolates that, according to the PCR analysis, deviated from FAM18 at the 3’ end of tpsA, i.e. isolates 2001044, 2020041 and 348 (Additional file 2), all produced a TpsA as evidenced by Western blotting (data not shown). Further PCR analysis of the TPS island of isolates 2001044 and 2020041 (Figure S4 in additional file 1) indicated that the tpsC1 cassette had recombined into the tpsA locus with loss of the intervening DNA as depicted in Figure  3A. Sequence analysis of the entire TPS island of isolate 2001044 (Genbank: HQ420262) confirmed this suggestion. Of note, tpsC1 of FAM18 shows 99% sequence identity between bp 195–1628 with a segment of tpsA (bp 4218–5651) (Figure S5A in Additional file 1). Comparison of the upstream and downstream sequences indicated that the recombination resulting in the tpsA of isolate 2001044 had occurred in this shared region. The sequence downstream of tpsA in isolate 2001044 showed > 99% identity with the region downstream of tpsC1 in FAM18 (Figure  3A). This region was also similar in isolate 2020041 (data not shown). Thus, the results for these isolates demonstrate that replacement of the 3’ end of tpsA by tpsC cassettes indeed occurs in vivo, albeit at a very low frequency.

thumbnailFigure 3. Genetic organization of the deviant TPS organization in cc11 isolates 2001044 and 348 and consequences in growth-inhibition activity. (A) Comparison of the TPS organization in cc11 reference strain FAM18 and the clinical isolate 2001044. (B) Comparison of the genetic organization of the TPS island in clinical isolate 348 and those in cc11 reference strain FAM18 and cc4 reference strain Z2491. In both panels, regions of high sequence similarity between the islands are indicated by grey shading. Note that the TPS organization in isolate 2001044 could have been generated by intragenomic recombination, whilst that in isolate 348 must have implicated horizontal gene transfer. (C) Unencapsulated derivatives of strains B16B6 and its ΔtpsA-tpsC mutant, both carrying an erythromycin-resistance marker, were mixed 1:1 with an unencapsulated derivative of strain 2001044 carrying a chloramphenicol-resistance marker, spotted on GC plates and incubated as described in the legend to Figure  2. The ratios of 2001044 bacteria over B16B6 (left) or B16B6ΔtpsA-tpsC mutant (right) bacteria were determined after 0 hours (gray bars) or 24 hours (white bars) of incubation by plating on GC media containing erythromycin or chloramphenicol.

Sequence analysis of the complete TPS island of the third deviant cc11 strain, isolate 348 (Genbank: HQ420263), revealed more complex rearrangements (Figure  3B). The encoded TpsA has a central core 2 region like in strain Z2491 instead of a core 1 as in FAM18 (Figure S2A in Additional file 1), and it contains a C terminal end identical to that of TpsC1 of Z2491 (Figure  3B). Thus, compared to Z2491, the 3’ end of tpsA is replaced by tpsC1 in isolate 348. Sequence comparisons (Figure S5B) suggest that the recombination event took place between bp 4174–4605 of tpsA and bp 910–1341 of tpsC1, where these sequences show high sequence identity.

The IORF downstream of tpsA in isolate 348 is identical to that downstream of tpsC1 of Z2491 (Figure  3B). The subsequent tpsC1 in 348 showed 94% identity within the first 858 bp with the corresponding 5’ fragment of tpsC2 of Z2491 (Figure  3B and Figure S5C in Additional file 1). The rest of this tpsC1 was identical to the 3’ end of tpsC1 of FAM18 (Figure  3B and Figure S5C in Additional file 1), differing only in one single bp. At the boundary, an 18-bp sequence is shared between the three genes (Figure S5D in Additional file 1), which was likely involved in the recombination event; apparently, only short stretches of 100% identity are required for this. The remaining part of the TPS island of isolate 348 showed > 99% identity with the corresponding region of FAM18 (Figure  3B). Thus, it appears that the tpsA of isolate 348 was generated by recombination between a tpsA and a tpsC1similar to those of Z2491. This tpsA gene was then introduced in the cc11 strain by recombination at the 5’ end in the DNA encoding the TPS domain of tpsA or in tpsB and at the 3’ end within tpsC1.

A possible explanation for the low frequency of recombination observed at the 3’ end of tpsA in vivo is that such recombination results in the concomitant loss of the downstream IORF, which provides immunity to the TpsA of the parental strain. Thus, within a microbial community such as a biofilm or a microcolony, the growth of a recombinant would be inhibited by its congeners. To investigate whether recombination indeed confers a growth disadvantage in the presence of a wild-type strain, we performed growth-inhibition experiments with the recombinant isolate 2001044 and strain B16B6. The growth of the deviating isolate was drastically reduced when co-cultured with B16B6 (Figure  3C, left bars), presumably because it is not protected against TpsA of B16B6. The growth of strain 2001044 was not inhibited when the strain was co-cultured with the ΔtpsA-tpsC mutant of B16B6, which does not produce TpsA (Figure  3C, right bars), demonstrating that the TpsA of strain B16B6 is responsible for the growth inhibition of isolate 2001044. In the latter experiment (Figure  3C, right bars), rather the growth of the ΔtpsA-tpsC mutant appeared to be inhibited, indicating that the recombinant TpsA produced by isolate 2001044 has growth-inhibitory activity against which the ΔtpsA-tpsC mutant of B16B6 is not protected presumably because it lacks all immunity proteins including that encoded by the IORF immediately downstream of the tpsC1 cassette. Consistent with this notion is the observation that the growth of the ΔtpsB mutant of B16B6, which does not secrete TpsA but has retained all IORFs, was not inhibited by isolate 2001044 (data not shown). To further confirm that the growth of isolate 2001044 is inhibited in the presence of B16B6 because it lacks the IORF immediately downstream of tpsA of B16B6 (Figure  3A), the strain was transformed with plasmid pFPIORF1, which contains this IORF under lac promoter control, and competition experiments with B16B6 were performed in the presence or absence of IPTG. The growth inhibition of the plasmid-containing 2001044 target cells was reduced by 3.1 log units (mean of three independent experiments) in the presence as compared to the absence of IPTG. Together, these data support the hypothesis that recombination at the tpsA locus confers a selective disadvantage within a genetically similar community because of the loss of the relevant immunity protein.

Comparison of the frequency of exchange at the pilE and tpsA loci

The results described above confirmed that the 3’ end of tpsA can be replaced by tpsC cassettes in vivo but at a much lower frequency than anticipated considering the reported high frequency of recombination at the pilE locus. However, it was reported recently that also gene conversion at the pilE locus is not detectable in cc8 and cc11 isolates [35,36]. Although this can be explained by the nature of the pilE and pilS genes present in these strains (vide infra), we cannot exclude the possibility that cc8 and cc11 strains contain some additional genetic defect that reduces the recombination frequency also at the tpsA locus. To rule out this possibility, we analyzed another clonal lineage, which is known to be proficient in recombination at the pilE locus, i.e. cc32 [35], even though this analysis is more complicated by the presence of two tpsA1 copies in strains of this cc, such as MC58 (Additional file 1: Figure S1).

We examined 50 cc32 disease isolates. First, the region corresponding to the pilE locus was amplified by PCR as described [35]. Three isolates did not yield an amplicon, suggestive of loss of the pilE locus, whilst two yielded a smaller amplicon than expected suggesting recombination with a pilS (Additional file 2). In a subsequent PCR, one of the primers annealed to a variable region in pilE. Of the 47 isolates positive for pilE, 13 did not yield an amplicon in this PCR (Additional file 2), demonstrating recombination relative to the pilE sequence of MC58 in at least 28% of the strains.

Next, recombination at the tpsA1 loci was examined by PCR as outlined in detail in Additional file 1: Figure S6. This analysis revealed that the tpsA1b locus corresponding to NMB0497 of strain MC58 was stably maintained in all isolates. However, the tpsA1a locus corresponding to NMB1779 was detected in the initial PCR in only 37/50 isolates suggesting that either recombination had occurred at this tpsA1 locus in the remaining 13 isolates or the tpsA1 locus was not duplicated in these isolates. Subsequent PCRs targeting the boundaries of the TPS islands (Additional file 1: Figure S6) revealed the absence of the duplication in 11 of these isolates. In the remaining two isolates, the duplication was detected and thus, apparently, recombination had occurred at the tpsA1a locus. Subsequent PCRs (outlined in Figure S6 in Additional file 1) revealed the presence in one of these isolates of a tpsA1 similar to that of FAM18, i.e. with a central core different from that of the tpsA1 genes of MC58. Therefore, this tpsA1 must have been acquired in this isolate by horizontal gene transfer. In the other isolate, we detected that the tpsC5 located downstream of tpsA1b had recombined into the tpsA1a locus (Figure S6 in Additional file 1). Thus, in 50 isolates containing in total 89 tpsA1 genes, only a single case of tpsA/C exchange was detected (i.e. 1.1%), in agreement with the low frequency detected in cc11 and cc8 isolates and much lower than the frequency of gene conversion at the pilE locus in the same set of strains. Consistent with our results, also in the newly available genome sequence of cc32 isolate H44/76 [37], the organization of the tpsA1-containing islands is similar to that in MC58 whilst the pilE gene is different.

Recombination within TPS sequences under laboratory conditions

One of the models for gene conversion at the pilE locus implicates the acquisition of pilS sequences from the environment [38]. Hence, we investigated whether we could induce recombination at the 3’ end of tpsA by exogenously supplying tpsC cassettes. In preliminary experiments, we incubated purified chromosomal DNA from strain B16B6 with the homologous (B16B6) or heterologous (α153 and α14) strains. PCR analysis of randomly picked colonies failed to provide evidence for recombination at the 3’ end of tpsA (data not shown), indicating that such events are rare consistent with the low recombination frequencies found in vivo. Then, a kanamycin-resistance (kan) cassette was inserted between the 5’ end of tpsC2 and the 3’ end of the TPS island (Figure  4A). When chromosomal DNA of the resulting mutant, designated B16B6ΔtpsC2-5, was used to transform the parent strain B16B6, the kan cassette was found at the same position as in the original mutant in all 150 transformants examined as evidenced by PCRs. Apparently, the high sequence identity at either side of the kan cassette forces reciprocal exchange by homologous recombination. When strain α14 was transformed, recombination had occurred within the 5’ region of the disrupted tpsC2 cassette of the donor DNA and the tpsA gene in the recipient DNA in all 19 kanamycin-resistant transformants examined as revealed in PCRs and subsequent sequence analysis (Figure  4B). Recombination in the various transformants had taken place at a variety of small stretches of sequence identity ranging in size from 5 to 23 bp (Figure S7 in Additional file 1). These results demonstrate that tpsC sequences can recombine into the 3’ end of a tpsA gene when exogenously supplied and that recombination does not occur at a specific site but can take place at any short stretch of sequence identity within a homologous region.

thumbnailFigure 4. In vitro recombination in the TPS islands. (A) Organization of the TPS islands in strain B16B6 and its mutant derivative B16B6ΔtpsC2-5 where a kan cassette replaces several tpsC cassettes and IORFs. (B) Organization of the TPS island in recombinant strain α14-tpsA* (middle) obtained after transformation of strain α14 (top) with chromosomal DNA from strain B16B6ΔtpsC2-5 (bottom). (C) Organization of the TPS island in recombinant strain α153-tpsA* (middle) obtained after transformation of strain α153 (top) with chromosomal DNA from strain α14-tpsA*#1 (bottom). The organization of the TPS island of strain α153 is not completely depicted because it is not located on a single contig in the available genome sequence.

Next, chromosomal DNA from one of the α14-tpsA* recombinants (#1 in Additional file 1: Figure S7) was used to transform strain α153, which has a different central core region in the tpsA gene (in Additional file 1: Figure S2). Two kanamycin-resistant transformants were obtained and PCR analysis and subsequent sequencing (data not shown) revealed that the disrupted tpsA of the donor DNA had recombined into the TPS island of α153 (Figure  4C). Thus, not only can tpsC cassettes recombine into the 3’ region of tpsA genes but also the central region of the tpsA gene can be replaced as was observed in vivo in the case of cc11 isolate 348.

Discussion

TpsA1 proteins of N. meningitidis have reported roles in adhesion [16], intracellular survival [17] and biofilm formation [18]. In several other bacteria, TpsA proteins have been described that inhibit the growth of related bacteria probably in competition for the same niche [22,23,25,26]. This CDI activity resides in the C-terminal domain, and an immunity protein encoded by the gene downstream of tpsA protects the producing strain. Together, the C-terminal domain of TpsA and the immunity protein function as a toxin/antitoxin pair. We demonstrated here that also the TpsA1 proteins of N. meningitidis have growth-inhibitory activity and that the IORF immediately downstream of the tpsA gene confers immunity.

Downstream of the tpsA gene, the TPS islands of N. meningitidis contain a variable number of putative genes, called tpsC cassettes, interspersed with IORFs. The tpsC cassettes putatively encode N-terminally truncated TpsA homologs, which present entirely different C-terminal domains. It was suggested that these tpsC cassettes could recombine with tpsA by gene conversion possibly as a mechanism of immune evasion [19-21,31]. Also in several toxin/antitoxin systems, silent genes encoding truncated toxins are present and it was suggested that recombination of these silent genes with the toxin gene alters the CDI mechanism [26]. However, up to date no direct experimental evidence for such recombination events has been reported.

Initial bioinformatic analysis indeed suggested extensive recombination between tpsA genes and tpsC cassettes, as many cases were found of sequences that are present at the 3’ end of a tpsC cassette in one genome and at the 3’ end of a tpsA gene in another genome (Figure  1, and Additional file 1: Figure S1 and Table S1). However, because of the different phylogenetic origins of these strains, these observations do not directly prove that recombination does occur in vivo. Therefore, to demonstrate that the expected recombination events indeed occur and to determine their frequency, we turned to the analysis of large collections of strains of the same clonal complexes. Initially, we focused on strains of cc11 and cc8, and we obtained evidence that tpsC cassettes can indeed recombine in vivo into the tpsA locus, but the observed frequency was dramatically lower than the frequency of antigenic variation by slipped-strand mispairing of other surface structures in the same collection of strains. In cc8 and cc11 isolates, gene conversion at the pilE locus is not detectable either [35,36]. This is, however, not due to a general recombination defect also affecting the tpsA locus but to the nature of the pilE and pilS genes present in these strains. N. meningitidis strains can elaborate one of two classes of pili. Whereas class I pili are closely related to gonococcal pili, class II pili present remarkably different features, i.e. lack of conserved regions in between the semi- and hyper-variable regions, deletion in a hyper-variable region, and no homology to class I pili in the flanking sequences [39]. The cc8 and cc11 strains contain a class II pilE[35] and the genome of FAM18 shows only two pilS cassettes, which exhibit strong homology to type-I pilin genes [39]. Therefore, pilE/S recombination is not to be expected and the low recombination frequency at the tpsA locus must have a different cause. Consistent with this hypothesis, we also found a very low frequency of recombination in the tpsA locus in cc32 isolates, which are proficient in recombination at pilE. Considering that the strains we analyzed are from hyperinvasive lineages, the low frequency of recombination at the 3’ end of tpsA compared to known mechanisms of antigenic variation suggests that the selection of new TpsA variants in vivo is not driven by immune pressure. Also, although N. meningitidis apparently has a large repertoire of alternate cassettes that could replace the killing module at the C terminus of TpsA, it apparently has only a low tendency to change it. This suggests that replacement of the C-terminal cassette of TpsA may even be detrimental to the bacteria. In the nasopharynx, N. meningitidis appears in microcolonies [40]. Obviously, within such a microcolony, the substitution of the C terminus of TpsA with the concomitant loss of the immunity protein as found in some cc11 isolates (Figure  3) would be detrimental as the protection against other cells within the community would be lost.

Although the frequency is low, the postulated recombination at the 3’ end of tpsA was confirmed both in vivo and in vitro. Figure  5A illustrates a model for the recombination mechanism as deduced from our work. We postulate that the minimal recombination unit consists of a tpsC, an IORF and the 5’ region of the next tpsC. Homologous recombination occurs in the borders of the recombination unit in the donor DNA with equivalent sequences in the recipient DNA. During the recombination event, a variable number of additional tpsCs, together with their downstream IORFs, can be incorporated or lost in the recipient DNA, which may result in extensive modification of the tpsC repertoire and may or may not alter the 3’ end of the tpsA gene. The sequences of the IORFs located immediately downstream of tpsAs or tpsCs with similar 3’ ends are highly conserved and show no sequence similarity with other IORFs. In the recombination models, the IORFs always remain associated with the upstream tpsC and will therefore confer immunity to the producing strain when this tpsC is recombined into the tpsA expression locus (Figure  3). Besides this process, the central region of the tpsA can also be replaced in which case the recombination at the 5’ end occurs within the DNA encoding the TPS domain or in tpsB or even further upstream, and at the 3’ end in the 5’ region of a downstream tpsC. No functional activity has been assigned to this central core region. Perhaps, it is only required for the efficient secretion and surface exposure of the killing module. However, considering its size and variability, it may serve additional functions. Besides mediating growth inhibition, meningococcal TpsA proteins are also involved in adhesion [16], intracellular survival [17] and biofilm formation [18]. Further experiments should clarify if the core region has a direct role in these additional TpsA activities.

thumbnailFigure 5. Recombination model for the TPS system. (A) Recombination model requiring double crossover. The recipient and donor DNA are indicated in the first and the second line, respectively, and carry in the depicted example a different central core region in tpsA. IORFs are not depicted. Within the 3’ region, the tpsA genes share sequences with segments at the 5’ end of tpsCs (black boxes). The minimal recombination units (indicated by double-headed arrows) are formed by the 5’ region of a tpsC, an IORF, and the 5’ region of the immediately downstream tpsC. The recombination products show the recipient DNA organization after double crossovers with donor DNA sequences. The double-headed lines show the recombination sites where the crossovers were established. (B) Recombination model requiring an intra-chromosomal single crossover and the product of such an event evolved from the organization of the recipient DNA shown in the first line of panel A.

Neisseria shows autolytic behavior and has the capacity to take up Neisserial DNA fragments by recognition of specific DNA uptake sequences [38]. The acquisition of pilS sequences from the environment is used for recombination of pilS to pilE[38,41]. The presence of DNA uptake sequences within the TPS islands, mostly located at the 5’ end of tpsC cassettes (Additional file 1: Figure S1), is consistent with the hypothesis that donor DNA for recombination in tpsA can be acquired from the medium. Our in vitro recombination studies offer experimental support for this mechanism. Obviously, for exchange of the central core region of tpsA, uptake of donor DNA from the environment is a necessity. Another model explains pilE/S gene conversion based on intracellular recombination substrates and assumes that DNA replication producing two chromosomes precedes recombination by double crossover [41,42]. Yet another model implicates the generation of an extrachromosomal circular intermediate generated by a single crossover (Figure  5B) [42,43]. Both models could explain the TPS organization in cc11 isolates in which the tpsC1 cassette has apparently recombined into the tpsA locus (Figure  3A). It is interesting to note that tpsC5 of FAM18 is the most distal tpsC in many TPS regions (colored blue in Figure S1in Additional file 1). The absence of a downstream tpsC cassette for establishing recombination may prevent its movement to another position in the tpsC repertoire. In the genome of strain α14, this tpsC has recombined into the tpsA gene with concomitant loss of all intervening tpsCs (Figure S1 in Additional file 1), which can be explained through a single crossover recombination event. Thus, like in the pilE/pilS system [13], recombination of tpsCs into the tpsA locus can be mediated by three different mechanisms involving uptake of DNA from the environment, recombination between sister chromosomes, or an intra-chromosomal single crossover event.

Where exactly the recombination between tpsA and tpsC cassettes takes place, has remained unclear thus far. Based on in silico analysis, it was previously predicted to be mediated by multiple direct repeats located in the 3’ half of tpsA and the 5’ end of tpsCs[21]. In contrast, Poole and colleagues suggested that these events take place at a more specific location, i.e. the VENN domain that is present in many tpsAs and tpsCs [26], including tpsA and tpsCs of N. meningitidis (indicated in Additional file 1: Figure S3 ). The results presented here show that the recombination can take place in small shared stretches of different length (5 to 23 bp, in Additional file 1: Figure S7) that are broadly distributed along a larger homology domain (in Additional file 1: Figure S3), in line with what was reported in the pilE/S system [13].

Although the postulated recombination at the 3’ end of TpsA was demonstrated in this study, the frequency was very low. If the recombination frequency is that low and recombination may even be detrimental, why have most N. meningitidis isolates generated such a large collection of tpsC cassettes? Possibly, the importance of the collection is not in the tpsC cassettes, but in the associated IORFs that, together, confer immunity to a large variety of TpsAs produced by competitors in the nasopharynx. Consistent with the proposed minor role of the tpsC cassettes is that we detected many tpsCs with internal stop codons, frameshift mutations or extensive truncations at the 3’ end removing the homology domain for recombination into the tpsA locus (Additional file 1: Table S1 and Figure S1).

Conclusions

Several TpsA proteins produced by Gram-negative bacteria inhibit the growth of other related bacteria probably in competition for the same niche. The toxic activity resides in its C-terminal region while an immediately downstream located IORF encodes an antitoxin protein. Here, we showed that also meningococcal TpsA mediates growth inhibition and that the downstream IORF confers immunity to the producing strain. We also demonstrated that tpsC cassettes can recombine into the tpsA locus leading to the presentation of a different toxic module at the C terminus of TpsA. We obtained evidence for such recombination both in vivo during infection and under laboratory conditions. However, new TpsA variants are generated only at a very low frequency possibly due to negative selection against recombination as consequence of the concomitant loss of the immunity protein resulting in sensitivity to the TpsA protein produced by the congeners in the same microbial community. The tpsC cassettes for recombination can be acquired from the environment and recombine into the tpsA gene or in the downstream region, potentially resulting in an increased or decreased number of tpsC cassettes and associated IORFs. We propose that the main goal of recombination is the collection of new IORFs for protection against a variety of TpsA proteins. Further studies on the mechanisms of TpsA-mediated fratricide might eventually enable us to fight these pathogens with their own weapons.

Methods

Bacterial strains and growth conditions

A panel of 424 N. meningitidis strains was examined here (Additional file 2). Disease isolates were from The Netherlands Reference Laboratory for Bacterial Meningitis (NRLBM) in Amsterdam and carrier strains from the Institut für Hygiene und Mikrobiologie, Würzburg. Isolates were classified in clonal groups based upon sequence analysis of conserved house-keeping genes and comparison of the results with the data on the Neisseria Multi Locus Sequence Typing website (http://pubmlst.org/neisseria/ webcite) [44]. Included in this study were also reference strains B16B6 [45], 2996 [46], FAM18 [2], α14 and α153 [27] (Additional file 2). Strains were grown at 37°C on GC medium (Oxoid) supplemented with Vitox (Oxoid) and antibiotics as appropriate (erythromycin 7 μg/ml; kanamycin, 100 μg/ml; chloramphenicol, 5 μg/ml; rifampicin, 50 μg/ml) in candle jars or in tryptic soy broth (TSB; Beckton Dickinson) at 37°C with constant shaking at 110 rpm for 7 to 8 hours. Bacteria were inactivated for 30 minutes at 56°C when appropriate. E. coli strain DH5α was grown in LB medium containing 100 μg/ml of kanamycin or ampicillin when required.

PCR amplification and sequencing

Bacteria, grown for 12 hours on plate, were resuspended in de-ionized water and boiled for 10 minutes. The supernatant obtained after removal of the cell debris by centrifugation was used as template DNA for PCRs [19]. PCRs were performed using 2 μl of extracted DNA, 0.25 μM of different primer combinations (Additional file 3), 0.5 U Taq DNA polymerase, 200 μM dNTPs and PCR buffer (all from Promega). For PCR products to be sequenced, the Expand High Fidelity PCR System (Roche) was used. PCR conditions consisted of 10 minutes incubation at 95°C, 30 cycles of 1 minute at 95°C, 0.5 minutes at 58°C, and 2.5 minutes at 72°C, and finally 10 minutes incubation at 72°C. The PCR products were analyzed on 1% agarose gels and visualized with ethidium bromide.

Additional file 3: Table S3. listing the PCR primers used in this study.

Format: PDF Size: 104KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

For sequencing of the TPS islands, several overlapping DNA fragments of variable length (568–2500 bp) were amplified, purified using the Wizard SV Gel and PCR Clean-Up System (Promega), and sequenced using internal primers at Macrogen (Seoul, Korea). When required, PCR products were cloned into the TOPO TA vector (Invitrogen). The plasmids were purified with the plasmid extraction kit I (Omega) and the inserts were sequenced using M13 universal primers. PCR reactions were carried out independently and repeated at least twice for each fragment. Sequences were assembled using the SeqMan II software (DNAstart Inc.).

For assessment of gene conversion at the pilE locus, we first determined the presence of the pilE locus by PCR with primers 5’pilE and 3’pilE, which anneal upstream and downstream of pilE, respectively [35]. In a second PCR, we combined primer 5’pilE with a primer (TTAGCTGGCATCACTTGCG) that anneals with a hypervariable region of pilE in MC58.

DNA sequence analyses

The tpsA and tpsC sequences available in the genome sequence databases of N. meningitidis strains α14, α153, α275 [27], 053442 [28], 8013 [29], FAM18 [2], MC58 [30], and Z2491 [31] were used. Comparative analysis within TPS regions was performed using Clone Manager software. All sequences were aligned using MAFFT server version 6 (http://align.bmr.kyushu-u.ac.jp/mafft/online/server/ webcite). Amino-acid distances were determined by phylogenetic analysis using the Neighbor-joining method with the available MEGA software version 4.0 (http://www.megasoftware.net/ webcite).

Generation of DNA constructs and mutants

To obtain a ΔtpsA-tpsC knockout mutant, DNA fragments h1, corresponding to the tpsB gene, and i, located downstream of the TPS island, were amplified from chromosomal DNA of strain B16B6 using the primers listed in Additional file 3. Both PCR products were purified, digested with proper restriction enzymes (Additional file 3), and sequentially cloned into pUC21. Then, a kan cassette was inserted between the cloned fragments yielding plasmid pKOtpsA-tpsC. To obtain the ΔtpsC2-5 mutant, DNA fragment h2, encompassing most of tpsC2 and its upstream IORF, was amplified, purified, and exchanged for fragment h1 on pKOtpsA-tpsC using restriction enzymes NotI and DraIII yielding plasmid pKOtpsC2-5. Plasmid pKO-tpsB1 [19] was used to obtain a tpsB mutant. The knockout constructs were purified with the plasmid extraction kit I and used to transform [47] strain B16B6 where the target genes were disrupted by homologous recombination. Appropriate PCRs were used to verify the presence of the correct mutations in kanamycin-resistant transformants and, where appropriate, the loss of TpsA or TpsB expression was confirmed on Western blots.

The unencapsulated mutant of serogroup B strain B16B6, designated BB-1, was previously described [48]. Unencapsulated mutants of derivatives of B16B6 and of serogroup C isolate 2001044 were obtained by gene replacement as described [48]. Recombination of the knockout constructs into the chromosome resulted in replacement of the complete capsule locus or only the siaD gene by an erythromycin- or chloramphenicol-resistance cassette, respectively. Correct mutations were verified by appropriate PCRs. The absence of serogroup B capsule in the B16B6 derivatives was also confirmed with monoclonal antibody anti-MBPS [49].

For complementation experiments, DNA fragment v corresponding to the IORF downstream of tpsA was amplified from chromosomal DNA of strain B16B6 using the primers listed in Additional file 3: Table S3. The PCR product was purified and cloned via NdeI and AatII digestion into pEN11-Imp [50] to yield plasmid pFPIORF1. The correct sequence of the inserted DNA fragment was confirmed.

Electrophoresis and immunoblotting

For SDS-PAGE, 8% (w/v) polyacrylamide gels were used. Proteins were visualized with Coomassie Brilliant Blue G250 or transferred to nitrocellulose membranes. The blots were incubated with blocking buffer [phosphate-buffered saline with 0.5% (w/v) non-fat dried milk (Protifar, Nutricia) and 0.1% (v/v) Tween 20 (Merck)]. Washes were done with blocking buffer without Protifar. All incubations were carried out for 1 hour at room temperature with constant shaking. After incubation with polyclonal antiserum directed against the TPS domain of TpsA1 [19] or NalP (anti-1669) [51] or with monoclonal antibodies for the specific detection of OpaJ and OpaA (15–1.P5.5) [52] or OpaD and OpaB (MN20E12.70) [53] at working dilutions in blocking buffer, the blots were developed with horseradish-peroxidase-conjugated goat anti-rabbit-IgG or anti-mouse-IgG (Biosource International) and the Pierce ECL Western Blotting Substrate.

Growth inhibition assay

The unencapsulated derivative of strain B16B6, BB-1, was transformed with plasmid pEN300 [51] containing a chloramphenicol-resistance marker to be able to discriminate between bacteria from co-cultures. A spontaneous rifampicin-resistant derivative of BB-1 was used for the same purpose in complementation assays, where the target cells carried a chloramphenicol-resistance marker on plasmid pFPIORF1. Bacteria from overnight cultures were inoculated in antibiotic-free TSB either containing or not 0.25 mM IPTG and grown with shaking until they reached an optical density at 600 nm of ~3. Subsequently, bacteria were mixed 1:1 and drops were spotted on antibiotic-free GC medium plates either with or without 0.25 mM IPTG and incubated at 37°C in candle jars in a humid atmosphere containing ~5% CO2. After different time periods, the bacteria were scraped from the plates, resuspended in TSB, and the ratio of the different bacteria in the co-culture was determined by plating on selective GC plates containing the appropriate antibiotics and overnight incubation. The numbers of viable cells in the initial cultures were similarly determined.

Abbreviations

Bp: Base pairs; CDI: Contact-dependent growth inhibition; IORF: Intervening open reading frame; IPTG: Isopropyl-β-D-1-thiogalactopyranoside; Kan: Kanamycin-resistance; ORF: Open reading frame; PCR: Polymerase chain reaction; TPS: Two-Partner Secretion; TSB: Tryptic soy broth.

Competing interests

The authors declare they have no competing interests.

Authors’ contributions

JA, PvU, AvdE, and JT conceived and designed the experiments; JA and KS performed the experiments; JA, AvdE, and JT analyzed the data; JA and JT wrote the paper. All authors have read and approved the manuscript.

Acknowledgements

We would like to thank Drs U. Vogel and H. Claus for various materials used in this study and Dr. M. Arenas for his help in genomic analyses. This work was supported by the Netherlands Organization for Health Research and Development (ZonMw).

References

  1. Lo H, Tang CM, Exley RM: Mechanisms of avoidance of host immunity by Neisseria meningitidis and its effect on vaccine development.

    Lancet Infect Dis 2009, 9:418-427. PubMed Abstract | Publisher Full Text OpenURL

  2. Bentley SD, Vernikos GS, Snyder LAS, Churcher C, Arrowsmith C, Chillingworth T, Cronin A, Davis PH, Holroyd NE, Jagels K, Maddison M, Moule S, Rabbinowitsch E, Sharp S, Unwin L, Whitehead S, Quail MA, Achtman M, Barrell B, Saunders NJ, Parkhill J: Meningococcal genetic variation mechanisms viewed through comparative analysis of serogroup C strain FAM18.

    PLoS Genet 2007, 3:e23. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  3. Santoyo G, Romero D: Gene conversion and concerted evolution in bacterial genomes.

    FEMS Microbiol Rev 2005, 29:169-183. PubMed Abstract OpenURL

  4. van der Woude MW, Baumler AJ: Phase and antigenic variation in bacteria.

    Clin Microbiol Rev 2004, 17:581-611. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  5. Andrews TD, Gojobori T: Strong positive selection and recombination drive the antigenic variation of the PilE protein of the human pathogen Neisseria meningitidis.

    Genetics 2004, 166:25-32. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  6. Hill SA, Davies JK: Pilin gene variation in Neisseria gonorrhoeae: reassessing the old paradigms.

    FEMS Microbiol Rev 2009, 33:521-530. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  7. Kline KA, Sechman EV, Skaar EP, Seifert HS: Recombination, repair and replication in the pathogenic Neisseriae: the 3 R’s of molecular genetics of two human-specific bacterial pathogens.

    Mol Microbiol 2003, 50:3-13. PubMed Abstract | Publisher Full Text OpenURL

  8. Nassif X, Lowy J, Stenberg P, O’Gaora P, Ganji A, So M: Antigenic variation of pilin regulates adhesion of Neisseria meningitidis to human epithelial cells.

    Mol Microbiol 1993, 8:719-725. PubMed Abstract | Publisher Full Text OpenURL

  9. Rytkönen A, Johansson L, Asp V, Albiger B, Jonsson AB: Soluble pilin of Neisseria gonorrhoeae interacts with human target cells and tissue.

    Infect Immun 2001, 69:6419-6426. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  10. Rudel T, Facius D, Barten R, Scheuerpflug I, Nonnenmacher E, Meyer TF: Role of pili and the phase-variable PilC protein in natural competence for transformation of Neisseria gonorrhoeae.

    Proc Natl Acad Sci USA 1995, 92:7986-7990. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  11. Wolfgang M, van Putten JPM, Hayes SF, Koomey M: The comP locus of Neisseria gonorrhoeae encodes a type IV prepilin that is dispensable for pilus biogenesis but essential for natural transformation.

    Mol Microbiol 1999, 31:1345-1357. PubMed Abstract | Publisher Full Text OpenURL

  12. Merz AJ, So M, Sheetz MP: Pilus retraction powers bacterial twitching motility.

    Nature 2000, 407:98-102. PubMed Abstract | Publisher Full Text OpenURL

  13. Criss AK, Kline KA, Seifert HS: The frequency and rate of pilin antigenic variation in Neisseria gonorrhoeae.

    Mol Microbiol 2005, 58:510-519. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  14. Jacob-Dubuisson F, Locht C, Antoine R: Two-partner secretion in Gram-negative bacteria: a thrifty, specific pathway for large virulence proteins.

    Mol Microbiol 2001, 40:306-313. PubMed Abstract | Publisher Full Text OpenURL

  15. Jacob-Dubuisson F, Fernandez R, Coutte L: Protein secretion through autotransporter and two-partner pathways.

    Biochim Biophys Acta 2004, 1694:235-257. PubMed Abstract | Publisher Full Text OpenURL

  16. Schmitt C, Turner D, Boesl M, Abele M, Frosch M, Kurzai O: A functional two-partner secretion system contributes to adhesion of Neisseria meningitidis to epithelial cells.

    J Bacteriol 2007, 189:7968-7976. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  17. Talà A, Progida C, De Stefano M, Cogli L, Spinosa MR, Bucci C, Alifano P: The HrpB-HrpA two-partner secretion system is essential for intracellular survival of Neisseria meningitidis.

    Cell Microbiol 2008, 10:2461-2482. PubMed Abstract | Publisher Full Text OpenURL

  18. Neil RB, Apicella MA: Role of HrpA in biofilm formation of Neisseria meningitidis and regulation of the hrpBAS transcripts.

    Infect Immun 2009, 77:2285-2293. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  19. van Ulsen P, Rutten L, Feller M, Tommassen J, van der Ende A: Two-partner secretion systems of Neisseria meningitidis associated with invasive clonal complexes.

    Infect Immun 2008, 76:4649-4658. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  20. Klee SR, Nassif X, Kusecek B, Merker P, Beretti JL, Achtman M, Tinsley CR: Molecular and biological analysis of eight genetic islands that distinguish Neisseria meningitidis from the closely related pathogen Neisseria gonorrhoeae.

    Infect Immun 2000, 68:2082-2095. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  21. van Ulsen P, Tommassen J: Protein secretion and secreted proteins in pathogenic Neisseriaceae.

    FEMS Microbiol Rev 2006, 30:292-319. PubMed Abstract | Publisher Full Text OpenURL

  22. Aoki SK, Diner EJ, de T’Kint Roodenbeke C, Burgess BR, Poole SJ, Braaten BA, Jones AM, Webb JS, Hayes CS, Cotter PA, Low DA: A widespread family of polymorphic contact-dependent toxin delivery systems in bacteria.

    Nature 2010, 468:439-442. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  23. Aoki SK, Pamma R, Hernday AD, Bickham JE, Braaten BA, Low DA: Contact-dependent inhibition of growth in Escherichia coli.

    Science 2005, 309:1245-1248. PubMed Abstract | Publisher Full Text OpenURL

  24. Aoki SK, Malinverni JC, Jacoby K, Thomas B, Pamma R, Trinh BN, Remers S, Webb J, Braaten BA, Silhavy TJ, Low DA: Contact-dependent growth inhibition requires the essential outer membrane protein BamA (YaeT) as the receptor and the inner membrane transport protein AcrB.

    Mol Microbiol 2008, 70:323-340. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  25. Anderson MS, Garcia EC, Cotter PA: The Burkholderia bcpAIOB genes define unique classes of two-partner secretion and contact dependent growth inhibition systems.

    PLoS Genet 2012, 8:e1002877. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  26. Poole SJ, Diner EJ, Aoki SK, Braaten BA, de T’Kint Roodenbeke C, Low DA, Hayes CS: Identification of functional toxin/immunity genes linked to contact-dependent growth inhibition (CDI) and rearrangement hotspot (Rhs) systems.

    PLoS Genet 2011, 7:e1002217. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  27. Schoen C, Blom J, Claus H, Schramm-Glück A, Brandt P, Müller T, Goesmann A, Joseph B, Konietzny S, Kurzai O, Schmitt C, Friedrich T, Linke B, Vogel U, Frosch M: Whole-genome comparison of disease and carriage strains provides insights into virulence evolution in Neisseria meningitidis.

    Proc Natl Acad Sci USA 2008, 105:3473-3478. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  28. Peng J, Yang L, Yang F, Yang J, Yan Y, Nie H, Zhang X, Xiong Z, Jiang Y, Cheng F, Xu X, Chen S, Sun L, Li W, Shen Y, Shao Z, Liang X, Jin Q: Characterization of ST-4821 complex, a unique Neisseria meningitidis clone.

    Genomics 2008, 91:78-87. PubMed Abstract | Publisher Full Text OpenURL

  29. Rusniok C, Vallenet D, Floquet S, Ewles H, Mouzé-Soulama C, Brown D, Lajus A, Buchrieser C, Médigue C, Glaser P, Pelicic V: NeMeSys: a biological resource for narrowing the gap between sequence and function in the human pathogen Neisseria meningitidis.

    Genome Biol 2009, 10:R110. PubMed Abstract | BioMed Central Full Text | PubMed Central Full Text OpenURL

  30. Tettelin H, Saunders NJ, Heidelberg J, Jeffries AC, Nelson KE, Eisen JA, Ketchum KA, Hood DW, Peden JF, Dodson RJ, Nelson WC, Gwinn ML, DeBoy R, Peterson JD, Hickey EK, Haft DH, Salzberg SL, White O, Fleischmann RD, Dougherty BA, Mason T, Ciecko A, Parksey DS, Blair E, Cittone H, Clark EB, Cotton MD, Utterback TR, Khouri H, Qin H, et al.: Complete genome sequence of Neisseria meningitidis serogroup B strain MC58.

    Science 2000, 287:1809-1815. PubMed Abstract | Publisher Full Text OpenURL

  31. Parkhill J, Achtman M, James KD, Bentley SD, Churcher C, Klee SR, Morelli G, Basham D, Brown D, Chillingworth T, Davies RM, Davis P, Devlin K, Feltwell T, Hamlin N, Holroyd S, Jagels K, Leather S, Moule S, Mungall K, Quail MA, Rajandream MA, Rutherford KM, Simmonds M, Skelton J, Whitehead S, Spratt BG, Barrell BG: Complete DNA sequence of a serogroup A strain of Neisseria meningitidis Z2491.

    Nature 2000, 404:502-506. PubMed Abstract | Publisher Full Text OpenURL

  32. Claus H, Maiden MC, Wilson DJ, McCarthy ND, Jolley KA, Urwin R, Hessler F, Frosch M, Vogel U: Genetic analysis of meningococci carried by children and young adults.

    J Infect Dis 2005, 191:1263-1271. PubMed Abstract | Publisher Full Text OpenURL

  33. Belland RJ, Morrison SG, Carlson JH, Hogan DM: Promoter strength influences phase variation of neisserial opa genes.

    Mol Microbiol 1997, 23:123-135. PubMed Abstract | Publisher Full Text OpenURL

  34. Saunders NJ, Jeffries AC, Peden JF, Hood DW, Tettelin H, Rappuoli R, Moxon ER: Repeat-associated phase variable genes in the complete genome sequence of Neisseria meningitidis strain MC58.

    Mol Microbiol 2000, 37:207-215. PubMed Abstract | Publisher Full Text OpenURL

  35. Cehovin A, Winterbotham M, Lucidarme J, Borrow R, Tang CM, Exley RM, Pelicic V: Sequence conservation of pilus subunits in Neisseria meningitidis.

    Vaccine 2010, 28:4817-4826. PubMed Abstract | Publisher Full Text OpenURL

  36. Helm RA, Seifert HS: Frequency and rate of pilin antigenic variation of Neisseria meningitidis.

    J Bacteriol 2010, 192:3822-3823. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  37. Piet JR, Veld RAG H i ‘t, van Schaik BDC, van Kampen AHC, Baas F, van de Beek D, Pannekoek Y, van der Ende A: Genome sequence of Neisseria meningitidis serogroup B strain H44/76.

    J Bacteriol 2011, 193:2371-2372. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  38. Seifert HS, Ajioka RS, Marchal C, Sparling PF, So M: DNA transformation leads to pilin antigenic variation in Neisseria gonorrhoeae.

    Nature 1988, 336:392-395. PubMed Abstract | Publisher Full Text OpenURL

  39. Aho E, Botten J, Hall R, Larson M, Ness J: Characterization of a class II pilin expression locus from Neisseria meningitidis: evidence for increased diversity among pilin genes in pathogenic Neisseria species.

    Infect Immun 1997, 65:2613-2620. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  40. Sim RJ, Harrison MM, Moxon ER, Tang CM: Underestimation of meningococci in tonsillar tissue by nasopharyngeal swabbing.

    Lancet 2000, 356:1653-1654. PubMed Abstract | Publisher Full Text OpenURL

  41. Gibbs CP, Reimann BY, Schultz E, Kaufmann A, Haas R, Meyer TF: Reassortment of pilin genes in Neisseria gonorrhoeae occurs by two distinct mechanisms.

    Nature 1989, 338:651-652. PubMed Abstract | Publisher Full Text OpenURL

  42. Howell-Adams B, Seifert HS: Molecular models accounting for the gene conversion reactions mediating gonococcal pilin antigenic variation.

    Mol Microbiol 2000, 37:1146-1158. PubMed Abstract | Publisher Full Text OpenURL

  43. Howell-Adams B, Wainwright LA, Seifert HS: The size and position of heterologous insertions in a silent locus differentially affect pilin recombination in Neisseria gonorrhoeae.

    Mol Microbiol 1996, 22:509-522. PubMed Abstract | Publisher Full Text OpenURL

  44. Jolley KA, Chan MS, Maiden MC: mlstdbNet - distributed multi-locus sequence typing (MLST) databases.

    BMC Bioinforma 2004, 5:86. BioMed Central Full Text OpenURL

  45. Frasch CE, Chapman SS: Classification of Neisseria meningitidis group B into distinct serotypes. II. Extraction of type-specific antigens for serotyping by precipitin techniques.

    Infect Immun 1972, 6:127-133. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  46. Poolman JT, Hopman CT, Zanen HC: Colony variants of Neisseria meningitidis strain 2996 (B:2b:P1.2): influence of class-5 outer membrane proteins and lipopolysaccharides.

    J Med Microbiol 1985, 19:203-209. PubMed Abstract | Publisher Full Text OpenURL

  47. van der Ley P, Poolman JT: Construction of a multivalent meningococcal vaccine strain based on the class 1 outer membrane protein.

    Infect Immun 1992, 60:3156-3161. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  48. Arenas J, Nijland R, Rodriguez FJ, Bosma TNP, Tommassen J: Involvement of three meningococcal surface-exposed proteins, the heparin-binding protein NhbA, the α-peptide of IgA protease, and the autotransporter protease NalP, in initiation of biofilm formation.

    Mol Microbiol 2013, 87:254-268. PubMed Abstract | Publisher Full Text OpenURL

  49. Frosch M, Görgen I, Boulnois GJ, Timmis KN, Bitter-Suermann D: 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 USA 1985, 82:1194-1198. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  50. Bos MP, Tefsen B, Geurtsen J, Tommassen J: Identification of an outer membrane protein required for the transport of lipopolysaccharide to the bacterial cell surface.

    Proc Natl Acad Sci USA 2004, 101:9417-9422. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  51. van Ulsen P, van Alphen L, ten Hove J, Fransen F, van der Ley P, Tommassen J: A Neisserial autotransporter NalP modulating the processing of other autotransporters.

    Mol Microbiol 2003, 50:1017-1030. PubMed Abstract | Publisher Full Text OpenURL

  52. Wang JF, Caugant DA, Morelli G, Koumare B, Achtman M: Antigenic and epidemiologic properties of the ET-37 complex of Neisseria meningitidis.

    J Infect Dis 1993, 167:1320-1329. PubMed Abstract | Publisher Full Text OpenURL

  53. de Jonge MI, Vidarsson G, van Dijken HH, Hogerhout P, van Alphen L, Dankert J, van der Ley P: Functional activity of antibodies against the recombinant OpaJ protein from Neisseria meningitidis.

    Infect Immun 2003, 71:2331-2340. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL