BapC autotransporter protein is a virulence determinant of Bordetella pertussis

Microbial Pathogenesis 51 (2011) 169e177

Contents lists available at ScienceDirect

Microbial Pathogenesis journal homepage: www.elsevier.com/locate/micpath

BapC autotransporter protein is a virulence determinant of Bordetella pertussis Mojtaba Noofeli a, Habib Bokhari b, Paul Blackburn c, Mark Roberts d, John G. Coote a, Roger Parton a, * a

Institute of Infection, Immunity and Inflammation, University of Glasgow, Glasgow, UK Department of Biosciences, COMSATS Institute of Information Technology, Chakshahzad Campus, Islamabad, Pakistan c GE Healthcare, Amersham Place, Little Chalfont, Bucks, UK d Institute of Comparative Medicine, University of Glasgow, Glasgow, UK b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 24 September 2010 Received in revised form 5 April 2011 Accepted 6 April 2011 Available online 4 May 2011

A protein designated Bap-5 (GenBank accession no. AF081494) or BapC (GenBank accession no. AJ277634) has been identified as a member of the Bordetella pertussis autotransporter family and the present work suggests that this protein, like the previously characterised BrkA, is a Bvg-regulated serum resistance factor and virulence determinant. B. pertussis bapC and brkA, bapC mutants were created and, like a brkA mutant, showed greater sensitivity to killing by normal human serum than their parent strains but they were not as sensitive as a bvg mutant. Competition assays also showed an important role for BapC, like BrkA, in virulence of B. pertussis in mice after intranasal infection. Moreover, the bapC and brkA, bapC mutants, like the brkA mutant, were found to be more sensitive to the antimicrobial peptide cecropin P1 than the parent strains. In the genome sequence of B. pertussis strain Tohama, bapC is designated as a pseudogene due, in part, to a frameshift in a poly(C) tract near the 50 end of the gene which creates a truncated BapC protein. Sequence analyses of the bapC region spanning the poly(C) tract of a number of B. pertussis strains showed minor nucleotide and amino acid polymorphisms but it appeared that all had an ORF that would be able to produce BapC. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Bordetella pertussis Autotransporter Serum resistance

1. Introduction The autotransporters are a family of extracellular proteins, found in various Gram-negative bacteria, that have many different functions but appear to have a similar mechanism of export [1e3]. Autotransporters are composed of three main domains: a signal sequence; the passenger domain or a-domain, and a translocator unit or b-domain. The autotransporter polyprotein is exported across the inner membrane using the Sec machinery. The signal sequence is then cleaved, the b-domain inserted into the outer membrane and the passenger domain is translocated to the bacterial cell surface, where it may or may not undergo further processing [1,3]. Before publication of the Bordetella genome sequences [4], four autotransporters had been characterised in Bordetella pertussis, namely the virulence-regulated proteins pertactin, an adhesin [5]; BrkA, a serum resistance factor [6]; tracheal colonisation factor (Tcf), another adhesin [7]; and the product of virulence-activated gene-8

* Corresponding author. Institute of Infection, Immunity and Inflammation, Glasgow Biomedical Research Centre (GBRC), University of Glasgow, 120 University Place, Glasgow G12 8TA, UK. Tel.: þ44 141 3305844; fax: þ44 141 3304600. E-mail address: [email protected] (R. Parton). 0882-4010/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.micpath.2011.04.004

(Vag8) [8]. These proteins have structural homology in their bdomains (c. 30 kDa) but the passenger domains are structurally different, although they all have RGD and (except for pertactin) SGXG motifs. Another member of the B. pertussis autotransporter family was identified in our laboratory when a PCR amplicon with an unexpected sequence was produced using primers directed to the region encoding the b-domain of pertactin in B. pertussis genomic DNA. This sequence was used to identify a gene in B. pertussis strain Tab for what was then the fifth member of the B. pertussis autotransporter family, originally named Bap-5 (GenBank accession no. AF081494). An identical sequence was identified in B. pertussis strain Tohama and was named BapC (GenBank accession no. AJ277634). This latter designation was used subsequently in the Bordetella genome sequences [4]. With the publication of the genome sequences of B. pertussis, Bordetella parapertussis and Bordetella bronchiseptica, it became apparent that the bap5/bapC sequence identified earlier was not the whole bapC gene. In B. bronchiseptica strain RB50, the bapC gene encoded an ORF of 993 amino acids. However, the predicted ORFs in the sequenced strains B. pertussis strain Tohama and B. parapertussis strain 12822 were shown to be truncated at 102 and 100 amino acids, respectively, due to frameshifts [1]. The genome sequences of B. pertussis, B. parapertussis and B. bronchiseptica have revealed that these species have genes for 22 autotransporter proteins, although some of them, especially in

170

M. Noofeli et al. / Microbial Pathogenesis 51 (2011) 169e177

B. pertussis and B. parapertussis, are pseudogenes. Expression of several of these genes, along with other, virulence-related genes, is known to be controlled by the BvgAeBvgS two component regulatory system [1]. In vitro, the Bvg system promotes virulence gene expression at 37  C but the virulent (Bvgþ) phase can be switched off (Bvg phase) by certain “modulating” conditions such as lower temperature or high concentrations of sulphate ions [9,10]. Despite the above finding that bapC in the B. pertussis genome strain Tohama is a pseudogene, our earlier work had suggested that BapC was in fact expressed in the B. pertussis strains that we used and that BapC, like BrkA [6], could function as a serum resistance factor, by interfering with the classical pathway of complement activity ([11] and unpublished observations). In order to determine the relative contributions of these two components to serum resistance and mouse virulence of B. pertussis, single and double mutants were constructed in the same genetic background as the brkA and bvg mutants already available [12,13].

2. Results 2.1. Construction of bapC mutants In initial studies, bapC mutants of B. pertussis strain Tab and our laboratory strain of Tohama were constructed by replacement of the bapC gene by allelic exchange with the bapC gene disrupted with a kanamycin-resistance (Kmr) cassette ([11] and Section 4.2). PCR was carried out on DNA extracted from selected transconjugant colonies to confirm that the wild-type bapC gene in the B. pertussis strains had been successfully replaced by the mutated bapC gene from the plasmid. PCR with primers BAPCF and BAPCR1 (Section 4.3) produced an amplicon of expected size (3.5 kbp) for bapC::Kmr and there was no evidence of a 2.2 kbp amplicon indicative of the native bapC gene present in the parent strains (data not shown). Southern blot analysis, using genomic DNA from the parent and bapC mutant strains digested with SacI, showed that a bapC-specific probe (Section 4.3) hybridised to a fragment of c. 5.2 kbp in DNA preparations from the parent strains whereas it hybridised to a fragment of c. 6.5 kbp in both mutant strains (data not shown). The size difference (1.3 kbp) corresponded to the size of the inserted Kmr cassette. Expression of bapC in the parent and mutant strains was investigated by RT-PCR with primers BAPCF and BAPCR2 (Section 4.4), expected to amplify a 505 bp fragment from the 30 region of bapC. The RT-PCR result (Fig. 1) indicated that bapC was expressed (505 bp product) in the Tab and Tohama parent strains (lanes 3 and 5) but not in their corresponding bapC mutants (lanes 4 and 6), or in the bvg mutant strain BP338 bvg used as a control (lane 7). The lack of transcript with BP338 bvg clearly indicates that BapC expression is regulated by bvg. Lanes 1 and 2 show PCR products, with the

same primers, obtained with genomic DNA from the Tab bapC mutant (505 bp þ 1300bp Kmr cassette) and its parent strain (505 bp), respectively. Next, bapC and brkA, bapC mutants were created in the same genetic background as a brkA mutant already available [12] to compare directly the role of BapC and BrkA. To do this, the bapC genes in B. pertussis strain BP338 (a Tohama derivative) and in the BP338 brkA mutant (BP2041) were replaced, by allelic exchange, with a bapC gene disrupted with a tetracycline-resistance (Tcr) cassette (section 4.2). This was used instead of a Kmr cassette which had been used previously to create the bapC mutants of strains Tab and Tohama because strain BP338 brkA had been created by Tn5 insertion, which encodes Kmr, into the brkA gene [12]. After conjugation between the Escherichia coli mobilising strain SM10(lpir) carrying suicide plasmid pSS1129 bapC::Tcr and streptomycinresistant derivatives of BP338 and BP338 brkA, PCR was carried out on DNA extracted from selected transconjugant colonies to confirm that the wild-type bapC gene in the B. pertussis strains had been successfully replaced by the mutated bapC gene from the plasmid. PCR with primers BAPCF and BAPCR1 produced an amplicon of expected size (3.5 kbp) for bapC::Tcr and there was no evidence of a 2.2 kbp amplicon indicative of the native bapC gene present in the parent strains (data not shown). The presence of the Tcr gene in the amplified 3.5 kbp bapC::Tcr fragment, after gel extraction, was shown by PCR amplification of the 1.3 kbp Tcr cassette with primers TCF1 and TCR1 (Section 4.3) (data not shown). Southern blot analysis, using genomic DNA from the parent and bapC mutant strains digested with XhoI, showed that a bapC-specific probe hybridised to an approximately 7.7 kbp fragment in DNA preparations from both parent strains, whereas it hybridised to an approximately 9 kbp fragment in both mutant strains, BP338 bapC and BP338 brk, bapC (data not shown). The size difference (1.3 kbp) corresponds to the size of the inserted Tcr cassette.

2.2. Serum resistance of the parent and mutant strains A preliminary test was performed to compare the number of B. pertussis survivors when exposed to phosphate-buffered saline (PBS) or to a 1/40 dilution of the heat-inactivated (56  C for 30 min) pooled human serum. There was no significant killing, and no significant difference (P > 0.05) between these two controls in a time-course study over 120 min (data not shown). This revealed that any agglutination by antibodies present in the human sera did not significantly affect the bacterial counts. After exposure to a 1/40 dilution of the unheated pooled human serum for up to 45 min (Fig. 2A), the B. pertussis BP338 parent strain survived much better than the bapC, brkA or brkA, bapC mutants. This better survival was also evident after longer exposure to normal human serum, for up to 120 min (data not shown). As shown in Fig. 2A, the brkA, bapC

Fig. 1. RT-PCR with primers BAPCF and BAPCR2 using RNA prepared from B. pertussis wild-type strains, bapC mutants and B. pertussis strain BP338 bvg. M ¼ markers, lane 1, PCR control with genomic DNA from B. pertussis Tab bapC; lane 2, PCR control with genomic DNA from B. pertussis Tab; lanes 3e7, RT-PCR with RNA from the B. pertussis strains: lane 3, BP Tab; lane 4, Tab bapC; lane 5, Tohama; lane 6, Tohama bapC; lane 7, BP338 bvg.

M. Noofeli et al. / Microbial Pathogenesis 51 (2011) 169e177

A

100

Survivors%

10 BP338 Tab Tab bapC BP338 brkA

1

BP338 bapC BP338 brkA, bapC BP338 bvg 0.1

0.01 0

15

30

45

Time (min)

B

100 BP338

mutant of strain BP338 was more serum sensitive than either the brkA or bapC single mutants of this strain but it was not as sensitive as the BP338 bvg mutant (BP347). The Tab parent and bapC mutant showed serum resistances similar to that of their corresponding BP338 strains (Fig. 2A). The greater susceptibility of BP338 bvg could suggest that other bvgregulated factors, in addition to BapC and BrkA, might confer or contribute to serum resistance in these B. pertussis strains. The involvement of the Bvg regulatory system in BapC and BrkA expression was confirmed by comparing the serum sensitivity of the parent and double mutant strains when grown under modulating conditions, in 40 mM MgSO4, to induce the Bvg-phase, and under non-modulating conditions (Bvgþ-phase). With the 1/40 dilution of normal human serum, the parent strains BP338 and Tab grown under modulating conditions were markedly more serum sensitive than when grown under non-modulating conditions (Fig. 2B). Similarly, the brkA, bapC double mutant of strain BP338 was more sensitive when grown under modulating conditions. However, the mutant strain BP338 bvg was still the most sensitive to serum killing whether grown in modulating or non-modulating conditions, suggesting that the modulating conditions were not completely eliminating expression of all the bvg-regulated factors.

BP338 + MgSO4

10

Survivors%

171

2.3. Mouse-virulence of the parent and mutant strains

Tab 1

Tab + MgSO4 BP338 brkA, bapC

0.1

BP338 brkA, bapC + MgSO4 BP338 bvg

0.01 0

15

30

45

BP338 bvg + MgSO4

Time (min) Fig. 2. Survival of B. pertussis strains in a 1/40 dilution of normal human serum. A, Comparison of parent and mutant strains. B, Comparison of selected strains grown in modulating (with MgSO4) and non-modulating conditions. Each point represents the percentage survivors derived from plate counts in triplicate.

To compare the virulence of individual strains directly, competition assays were performed. At 7 days after infection of the mice, lung counts were done and the relative numbers of each strain were determined. There were significant differences between the numbers of each pair of bacteria recovered, depending on the mixtures used. The ratio of the two strains recovered from each mouse, compared to the input ratio, was used to determine a competitive index (see legend to Fig. 3). Results in Fig. 3 show that, for example, when BP338 bapC was compared with bvg mutant strain of BP338, competitive indices of approximately 0.2 were obtained from all mice, indicating that fewer colonies of BP338 bvg were recovered, approximately 1/5 the number of those of BP338 bapC. Thus BP338 bvg is much less able to persist in the mouse lung than the bapC mutant (P < 0.05). The

1 0 .9

C o m p etitive in d ex (C I)

0 .8 0 .7 0 .6 0 .5 0 .4 0 .3 0 .2 0 .1 0

BP338 bapC BP338 bvg

BP338 bapC BP338 brkA, bapC

BP338 brkA, bapC BP338 bvg

BP338 BP338 bvg

BP338 brkA BP338 bapC

BP338 brkA BP338 brkA, bapC

BP338 BP338 brkA

BP338 BP338 brkA, bapC

BP338 BP338 bapC

Fig. 3. Competition assays of virulence for mice. Different strains (e.g. strain 1 and strain 2) of B. pertussis were mixed in a 1:1 ratio (input ratio, confirmed by viable counts) and groups of 5 or 10, 4-week old mice were infected intranasally. At 7 days, lung counts (colony-forming units, CFU) on BG agar plates with and without appropriate antibiotics were used to determine the output ratio of the two strains. Each point represents the competitive index (CI) for each mouse, calculated as: CI ¼ 1/(strain1: strain 2 output CFU/strain1: strain 2 input CFU).

172

M. Noofeli et al. / Microbial Pathogenesis 51 (2011) 169e177

in vivo analysis revealed that the bvg mutant strain BP338 bvg was the least virulent of the strains. The data also indicated that BP338, the parent strain in this study, was far better in colonising the mouse lung than either of its bapC or brkA single mutants (P < 0.05). Moreover, the double mutant was less virulent than either of the single mutants (P < 0.05), although it was not so low in virulence as the bvg mutant.

2.4. Resistance to antimicrobial peptides The susceptibility of the different B. pertussis parent and mutant strains to inhibition by the antimicrobial peptides cecropin P1 and protamine was determined by a radial diffusion assay. Different responses were obtained. Compared with the BP338 parent strains, the bapC and brkA mutant strains were more susceptible to inhibition by cecropin P1 and the double mutant was even more sensitive (Fig. 4A). The bvg mutant was marginally more sensitive than the brkA,bapC double mutant, but this was not statistically significant (P > 0.05). The Tab parent and bapC mutant showed resistances similar to those of their corresponding BP338 strains. In contrast, the strain BP338 bvg was the most resistant to inhibition by protamine and the parent strain was the most sensitive (Fig. 4B). Again, the Tab parent and bapC mutant showed resistances similar to those of their corresponding BP338 strains.

A

2.5. Sequence variation in the bapC gene The bapC gene in the genome sequence of B. pertussis strain Tohama was reported to be a pseudogene (GenBank accession No. NP_881344). A frameshift in a poly(C) tract altered the amino acid sequence and resulted in premature termination of the protein at amino acid 102 due to a UAA translational stop codon (Fig. 5). In the present study and in an earlier work [11] using strain BP338, a Tohama derivative, and B. pertussis strains Tab and our laboratory strain of Tohama, evidence was presented that the BapC protein was in fact expressed. To address this anomaly, the region encompassing the poly(C) tract of a number of B. pertussis strains was sequenced using primers MNF and MNR (Section 4.3) which amplified a 408 bp region at the 50 -end of the bapC gene. This was done to determine any variations in different B. pertussis strains and to compare them with the published Tohama genome sequence. The strains chosen were: B. pertussis BP338 (a Tohama derivative), BP338 brkA (BP2041), and wild-type strains Tohama, Tab, 18-323 and PICU. Analysis of the region spanning the poly(C) tract in B. pertussis strains showed a polymorphism varying from 13 to 15 C nucleotides (Fig. 5A). All of the B. pertussis strains except 18-323 showed 13 C nucleotides, as in the unmutated gene, compared with 14 C nucleotides in the Tohama genome strain, whereas strain 18-323 exhibited 15 C nucleotides in this region. The translated region spanning the poly(C) tract (Fig. 5B) showed premature termination of the BapC protein in the Tohama genome strain with the

30

Zone size (mm)

25 20 15 10 5 0 BP338 brkA, bapC

BP338 bapC

BP338 brkA

BP338

Tab

Tab bapC BP338 bvg

B. pertussis strains

B

25

Zone size (mm)

20

15

10

5

0 BP338 brkA, bapC

BP338 bapC

BP338 brkA

BP338

Tab

Tab bapC

BP338 bvg

B. pertussis strains

Fig. 4. Susceptibility of B. pertussis strains to the antimicrobial peptides cecropin P1 (A) and protamine (B) in a radial diffusion assay. Data represent the mean and standard error of the mean size of the zone of inhibition from three determinations.

M. Noofeli et al. / Microbial Pathogenesis 51 (2011) 169e177

173

Fig. 5. CLUSTAL W (1.83) multiple sequence alignment of the 50 -end of the bapC gene (A) from different B. pertussis strains and the corresponding N-terminal region of the BapC protein (B) encompassing the poly(C) tract, starting at nucleotide no. 196 in the bapC gene of B. pertussis Tohama genome sequence [GenBank accession no. NP_881344].

frameshift. This was not evident in the other strains. Taking account of the poly(C) and flanking regions, the sequences of the strains investigated here were in-frame and should all express BapC. Some minor differences were noted. Strain 18-323 had an extra proline at amino acid 95, and leucine replacement by phenylalanine at amino acid 96, and strain PICU had a phenylalanine in place of serine at amino acid 96. 3. Discussion Bordetella species have the capacity to encode 22 autotransporter proteins of which a few (BrkA, pertactin, SphB1, TcfA and Vag8) have been assigned functions in host interaction and virulence [4]. From the Bordetella genome data, an autotransporter designated BapC was predicted as a protein of 998 amino acids in B. bronchiseptica (GenBank accession No. NP_888576) but in B. pertussis Tohama (GenBank accession No. NP_881344) and B. parapertussis (GenBank accession No. NP_884815) the predicted proteins were truncated at 102 and 100 amino acids, respectively, due to frameshifts [1]. In previous studies from our laboratory [11] and from our RT-PCR studies (Fig. 1), it appeared that the bapC gene was expressed in our B. pertussis strains Tohama and Tab and that bapC expression was regulated by the BvgAeBvgS two component regulatory system. In view of this apparent discrepancy, further genetic and functional characterisation of BapC was undertaken. The rationale for creating BP338 bapC and BP338 brkA, bapC mutants was to determine the exact role of BapC in the same genetic background as the existing BrkA mutation [12,13]. This study has confirmed that BapC, like BrkA [14], plays a role in protecting B. pertussis from serum killing since the bapC mutants of

B. pertussis strains Tab and BP338 (a nalidixic acid resistant derivative of Tohama [12,13]) were more sensitive than their parents to killing by normal human serum. The BP338 brkA, bapC double mutant was significantly more sensitive to killing by serum than the single brkA or bapC mutants in the same genetic background. Moreover, killing of B. pertussis BP338 and its brkA and bapC single and brkA, bapC double mutants, grown under modulating conditions, was significantly greater when exposed to normal human serum compared to those grown under non-modulating conditions. In addition, the double mutant was not as sensitive as the bvg mutant of BP338 strain (BP347). These data suggest that although BapC, in addition to BrkA, seems to play a role in conferring resistance on B. pertussis to killing by complement, other as yet unidentified bvg-regulated factors may also be involved in serum resistance. A range of surface-associated or released virulence factors is employed by B. pertussis to survive in its host, and to delay or evade the immune effector mechanisms deployed against it. B. pertussis does show some sensitivity to complement killing and this is attributed in part to the absence of O-side chain on its surface lipooligosaccharide [15,16]. B. bronchiseptica and B. parapertussis strains, that possess lipopolysaccharide and express O antigens, are completely resistant to naïve serum whereas O-chain defective mutants are very sensitive to complement [16,17]. In the absence of O antigens, B. pertussis does therefore require other mechanisms to resist serum killing. However, resistance is complex and appears to involve several different mechanisms [18e20]. The mechanism of BapC action in resistance to complement was not examined in this study and further investigation would be required to determine, for example, whether it can bind to or inhibit key components of the various complement pathways.

174

M. Noofeli et al. / Microbial Pathogenesis 51 (2011) 169e177

Lysis of Gram-negative bacteria by complement is due to the insertion of the C9 component of the membrane attack complex into the membranes, which is analogous to the mechanism of killing by some antimicrobial peptides in that lysis is also dependent on insertion of the peptide into the membranes [21]. In a previous study, BrkA in B. pertussis was shown to confer resistance to at least one antimicrobial peptide, cecropin P1, and appeared to contribute to sensitivity to protamine [22]. In the present study, the BP338 brkA, bapC mutant was found to be more susceptible to killing by cecropin P1 than the single brkA or bapC mutants, but less susceptible than BP338 bvg. The order of susceptibility to cecropin P1 in B. pertussis strains was approximately the same as serum sensitivity, suggesting a possible role for BrkA and BapC proteins in resistance to killing by some antimicrobial peptides and the possibility that the same mechanisms may be involved. Moreover, since BP338 brkA, bapC was not as sensitive as the bvg mutant, this again suggests that B. pertussis factors other than BrkA and BapC can confer resistance to cecropin P1, just like resistance to complement. Fernandez and Weiss [22] found the same order of resistance to cecropin P1 with some of the same strains: BP338 > BP338 brkA (BP2041) > BP338 bvg (BP347). In the present study, the B. pertussis wild-type strains were more sensitive than their brkA or bapC mutant or the brkA, bapC double mutant to killing by protamine. The bvg mutant was least susceptible, as reported previously [22]. Overall, therefore, the evidence suggests that BapC, like BrkA, seems to confer resistance to the antimicrobial peptide, cecropin P1, but may possibly contribute to sensitivity to protamine. As suggested by Fernandez and Weiss [22], BrkA, due to its surface location on the bacterium and by the negative charge (pI 5.89) of its passenger domain, may serve as a barrier to prevent damage by the cationic peptide cecropin P1. The passenger domain of BapC is also negatively charged (pI 5.13) and may have a similar action. The reason for the greater sensitivity of the wild-type, brkA, bapC and brkA; bapC mutants to another cationic peptide, protamine, is unknown. Fernandez and Weiss [22] have suggested that the presence of a crystalline porin structure which is unique to the avirulent form of B. pertussis [23] or the product(s) of a bvgrepressed gene might explain the resistance of the bvg mutant strain to killing by protamine. Competition assays also showed an important role for BapC, like BrkA [20,24], in virulence of B. pertussis strains in a mouse model of infection. It is well-known that BP338 bvg has greatly reduced virulence in the mouse model compared to the parent strain [24]. The order of virulence determined by the competition assays was approximately the same as the order of resistance to serum killing and to the antimicrobial peptide (cecropin P1). Taken together, these findings suggest that BapC, like BrkA, is an important virulence determinant of B. pertussis and also that the resistance to complement and antimicrobial peptides may have important roles in virulence. The full-length BapC and BrkA proteins are 63% similar over their last 300 amino acids but 36.4% over their passenger domains (adomains) as determined by BLASTp analysis (data not shown). The passenger domains of BapC and BrkA proteins contain one and two arg-gly-asp (RGD) motifs, respectively, that may promote binding to integrins and are important in adhesion to mammalian cells [25]. The RGD motif in BapC is located halfway between the predicted signal sequence and processing site, similar to the position of the second RGD in BrkA, (positions 418 and 490 in BapC and BrkA, respectively). BapC and BrkA also have one and two potential glycosaminoglycan (SGXG) attachment sites, respectively, which have been suggested may be involved in the mechanism for serum resistance of BrkA in B. pertussis, possibly by inhibition of C9 polymerization [6]. The b-domains of BapC and BrkA share a C-terminal outer membrane localisation motif (FHA/LGYRYS/TW/F), which

consists preferentially of amino acid residues with hydrophobic side chains such as phenylalanine or tryptophan in the last nine residues. The overall similarity of these domains is perhaps not surprising given that the function of these moieties is to form a b-barrel. Generally, the C-terminus domains of autotransporters are composed of 250e300 amino acid residues, which all show some homology but vary in their sequences [1]. Analysis of full-length BapC in B. pertussis, B. parapertussis and B. bronchiseptica using SignalP v.3.0 (released 2004) revealed a signal peptide with a maximum cleavage site probability between amino acid residues Ala38 and Gln39 (data not shown). A database search of the upstream flanking region of bapC using the B. pertussis Tohama genome sequence (BLAST search at http://www.Sanger.ac.uk/projects/B_ pertussis) indicated two potential BvgA binding sites (TTTCATA and TTTCGTA) upstream of the bapC gene at positions 230 bp and 144 bp, respectively (data not shown), from the start of the predicted signal peptide (contig BX640419) between a probable ammonium transporter gene (amtB, locus_tag BP2737) and the bapC gene. These two potential BvgA binding sites upstream of the predicted translational site of bapC are also similar to the characteristic heptameric sequence TTTC(C/T)TA identified by Kinnear et al. [26] upstream of prnA. A rho-independent terminator sequence downstream of the translational stop codon of the bapC gene was found. This terminator adopts a hairpin or loop-shape secondary structure, consisting of a GC-rich stem-loop region followed by a run of U residues, which is considered responsible for RNA polymerase terminating mRNA synthesis. The bapC gene of the B. pertussis genome strain Tohama has been reported to be a pseudogene, with two frame shifts, one in a homopolymeric tract (HPT) of 13 cytosine (polyC) and another in a homopolymeric tract of 11 guanine residues, both at the 50 -end of the a-domain (B. pertussis genome locus_tag BP2738). Gogol et al. [27] examined a collection of 90-geographically separate isolates of B. pertussis for phase variation at HPTs including those in bapC, and found (G) alleles varying from G8 to G12 using colony PCR/LDR (polymerase chain reaction/ligase detection reaction). In most of these cases, mixed allelic content was verified by sequencing of individual strains. The HPTs in BapC varied widely across the strain collection and allelic polymorphisms were detected even within a single round of culture. It is suggested that the limitation of genetic diversity in many pathogens such as Bordetella species might be overcome by high frequency phase variation to adapt to the hostile and changing host environment. Reversible expansion or contraction of HPTs is one of the most common mechanisms of phase variation to evade the immune system [27]. To address this further, the 50 -end of the bapC gene was sequenced from chromosomal amplicons of a selection of B. pertussis strains and the nucleotide and deduced amino acid sequences were aligned. Sequencing was performed in the regions of the poly(C) and poly(G) tracts and revealed polymorphisms varying from 13e15 (C) (Fig. 5A) and 9e12 (G) nucleotides (data not shown). However no difference was found between the numbers of (C) nucleotides in strains Tab, Tohama, PICU, BP338 and BP338 brkA compared with the published sequence without the frameshift (locus_tag BP2738). It can therefore be concluded from the above observations that, although the entire bapC gene was not sequenced in these different B. pertussis strains, there are some minor nucleotide and amino acid changes in some strains in the poly(C) and poly (G) tracts but it appears likely that all would be able to produce some form of BapC, unlike the B. pertussis Tohama genome strain. We have also cloned the bapC gene from B. pertussis strain BP338, a Tohama derivative and expressed it in E. coli expression strain BL21 (DE3) pLysS. Peptide fingerprinting of a w90 kDa band from this strain showed that it corresponded to the unprocessed form of BapC protein published in the B. pertussis genome (locus_tag BP2738) with 46%

M. Noofeli et al. / Microbial Pathogenesis 51 (2011) 169e177

175

Table 1 .B. pertussis strains used in this study. Species/strains

Genotype/phenotype

Source

B. pertussis Tab B. pertussis Tab bapC B. pertussis Tohama

Wild-type bapC::Kmr; Nalr, Smr Wild-type

B. pertussis Tohama bapC B. pertussis BP338

bapC::Kmr; Nalr, Smr Nalr derivative of B. pertussis Tohama; Parent of BP2041 and BP347 BP338 brkA1::Tn5 ( Kmr), Nalr BP338 bvgS1::Tn5 (Kmr), Nalr BP338 bapC::Tcr; Nalr, Smr BPM2041 bapC::Tcr; Nalr, Kmr, Smr Wild-type Wild-type

Clinical isolate, Glasgow (1979) Bokhari [11], PhD Thesis, University of Glasgow Weiss, A. Dept. of Molecular Genetics, Biochemistry & Microbiology, University of Cincinnati, USA Bokhari [11], PhD Thesis, University of Glasgow Weiss, A. Dept. of Molecular Genetics, Biochemistry & Microbiology, University of Cincinnati, USA Weiss, A. (as above)

B. pertussis BP2041 (BP338 brkA) B. pertussis BP347 (BP338 bvg) B. pertussis BP338 bapC B. pertussis BP338 brkA, bapC B. pertussis 18-323 (NCTC 10739) B. pertussis PICU

known-peptide coverage (data not shown). This clearly showed that BapC would indeed be expressed in B. pertussis strain BP338. Thus, the data supported our findings that B. pertussis strains Tohama (our laboratory strain and Tohama derivative BP338) and Tab do, in fact, produce BapC protein. Taken together, these data suggest that many B. pertussis strains are capable of producing BapC protein and that it is an important virulence factor in this species.

4. Materials and methods 4.1. Bacterial strains and growth media B. pertussis strains used in this study are shown in Table 1. They include B. pertussis Tohama and Tab wild-type strains, B. pertussis Tohama derivative BP338 and its brkA (BP2041) and bvg (BP347) mutant derivatives created by transposon Tn5 mutagenesis [12]. B. pertussis was grown routinely at 37  C on Bordet Gengou (BG) agar (Difco) plates containing 15% v/v horse blood for 48-72h. For liquid culture, Cyclodextrin Liquid [28] or Stainer-Scholte (SS) [29] medium in shake flasks were inoculated from such BG plates and incubated for 48e72 h at 37  C with shaking at 150e200 rpm. For growth in modulating (Bvg) conditions, the NaCl in SS medium was replaced with 40 mM MgSO4. E. coli strain SM10(lpir) [thi thr leu tonA lacY supE recA::RP4-2-Tc::Mu; Kmr ] [30,31] was grown overnight at 37  C on Luria-Bertani (LB) agar (LB broth: tryptone, 10 g; yeast extract, 5 g; sodium chloride, 10 g; per litre; solidified with 12 g/L micro agar (Duchefa Biochemie)). For liquid culture, LB broth in shake flasks was inoculated from such plates and incubated overnight at 37  C. Except where stated, the following antibiotics were used at the concentrations indicated: streptomycin (Sm), 100 mg/ml; nalidixic acid (Nal), 40 mg/ml, kanamycin (Km), 40 mg/ml; cephalexin (Cfx), 40 mg/ml; ampicillin (Amp),100 mg/ml and tetracycline (Tc), 20 mg/ml. 4.2. Construction of bapC mutants The bapC gene was amplified from BP338 genome using the primers BAPCF and BAPCR1 and cloned into a suicide vector pSS1129 [30]. It was then disrupted either with a kanamycinresistance (Kmr) cassette amplified from plasmid pUC4K using primer KANACOI (Section 4.3), which contained an NcoI site and annealed to two sites which flank the Km cassette, or with a tetracycline-resistance (Tcr) cassette amplified from pBR322 using primers TCF and TCR containing NcoI sites (Section 4.3). After cloning the amplified Kmr or Tcr cassette into plasmid pGEMT it was

Weiss, A. (as above) This study This study NCTC Dr N.K. Fry, Health Protection Agency, London NW9 5HT

removed by NcoI treatment and inserted into the NcoI restriction site at position 1244 of bapC in vector pSS1129 to give bapC::Kmr or bapC::Tcr. Replacement of the bapC chromosomal gene with its in vitro-altered counterpart was performed by homologous recombination in spontaneous Sm-resistant derivatives of B. pertussis strains Tohama, Tab or BP338 and BP2041 (BP338 brkA). Bacterial conjugation was carried out by plate mating on BG agar plates between these strains and E. coli SM10(lpir) (a plasmid mobilising strain) [31] carrying pSS1129 containing bapC::Kmr or bapC::Tcr. To select for the first (single) crossover, where the suicide plasmid is incorporated into the B. pertussis chromosome, bacterial suspensions were plated onto BG agar containing Cfx, Nal and Km or Tc (5 mg/ml). The plates were incubated at 37  C for 5e6 days to select the exconjugants. The Cfx was included to select against growth of the E. coli donor, as B. pertussis is naturally resistant to Cfx. The second crossover was obtained by selecting for the loss of the integrated plasmid. The exconjugants were plated on BG agar supplemented with Sm, Nal, and Km or Tc (5 mg/ml), and were incubated at 37  C for 5e6 days. The rspL gene of the suicide vector pSS1129 encodes Sm sensitivity and only those bacteria that have lost the plasmid will be able to grow. To confirm the loss of integrated plasmid, the survivors of the above selection were grown on BG agar containing Amp, resistance to which is encoded by the suicide plasmid (pSS1129). Survivors of Sm selection that were Amps, indicating that they had lost the vector, but which were Kmr or Tcr and Smr, were presumed to be B. pertussis in which the bapC allele had been replaced by the bapC::Kmr or bapC::Tcr allele. This was confirmed by PCR and Southern blotting (Sections 4.3 and 4.5). 4.3. PCR Primer sequences used for PCR in this study were: BAPCF (50 to 30) ATGGCACCTCGCCTTCGATTCGCGTCCAAG and BAPCR1 (30 to 50 ) AGGTGGAACGTCCAAGGCAAGGTCAGCTTG for amplification of a 2.2 kbp region of the bapC gene [accession no. NP 881344]; MNF (50 to 30 ) ATGAATGACAGAAAATCCAATAGC and MNR (30 to 50 ) GCTGTCACGCACGGTGAGCGAACG for amplification of a 408 bp region at the 50 -end of the bapC gene encompassing the poly(C) tract; TCF (50 to 30 ) AATC/CATGGTTCTCATGTTTGACAGCTTATCATCG and TCR (30 to 50 ) ACGC/CATGGTTTGCGCATTCACAGTTCTCCGC for amplification of the tetracycline-resistance cassette from pBR322, with NcoI recognition sites (underlined); KANACOI C/ CATGGCCGTCGACCTGCAGG for amplification of the kanamycinresistance cassette from pUC4K, also with a NcoI recognition site (underlined). The following thermocycling parameters were used: initial activation step at 95  C for 10 min followed by 30 cycles of

176

M. Noofeli et al. / Microbial Pathogenesis 51 (2011) 169e177

a denaturation at 94  C for 1 min, an annealing at 60  C for 1 min and an extension at 72  C for 1e3.5 min depending on length of sequence, and finally a final extension step at 72  C for 10 min (modified from [32]). A Hybaid thermal cycler was used for all reactions. For DNA sequencing, a MegaBACE1000 (96 capillary) sequencer, which used Big Dye (Applied Biosystems) and ET-Dye Terminator (Amersham Bioscience) chemistries, was employed by the Molecular Biology Support Unit (MBSU) at the University of Glasgow. The resultant sequences derived from electropherograms were analyzed using Chromas (version 1.45) and BioEdit version 5.0.6 and then aligned using ClustalW, a multiple sequence alignment tool. 4.4. RT-PCR The RNeasy mini kit (Qiagen) was used for the extraction of total RNA from B. pertussis cells and any trace of genomic DNA that had copurified with the total RNA was removed by DNase treatment according to the manufacturer’s instructions. The reverse transcription-PCR (RT-PCR) was performed using the Omniscript Kit (Qiagen) according to the manufacturer’s instructions using a two step reaction. cDNA synthesis was done with 1 cycle at 39  C for 60 min followed by 1 cycle at 93  C for 2 min. Primers BAPCF (Section 4.3) and BAPCR2 (30 to 50 , GCTACGTCAGCTCATAATTGATGCTG) were used for both RT-PCR and subsequent PCR to amplify a 505 bp fragment of the bapC gene using the thermocycling parameters given in Section 4.3. The products were analysed on a 0.7% agarose gel by electrophoresis in TBE buffer using a horizontal submarine electrophoresis tank (E-C Apparatus Corporation) and detected by using a UV transilluminator (model TM-40, UVP Inc., California, USA). 4.5. Southern blot analysis Chromosomal DNA was isolated from B. pertussis using a WizardÒ genomic DNA purification kit (Promega). The DNA was digested with XhoI, separated on a 0.7% (w/v) agarose gel, and transferred to a positively-charged nylon membrane (Hybond-N, Amersham). The blot was probed with a 2.2 kbp region of the bapC gene which had been amplified with primers BAPCF and BAPCR1 (Section 4.3) and labelled with digoxigenin-dUTP using the PCR Dig Probe Synthesis kit (Roche). After standard hybridization and posthybridization washes, the membrane was incubated in blocking buffer for 60 min and the hybridised bands detected with antidigoxigenin-alkaline phosphatase conjugate. 4.6. Serum killing assay Blood was collected from adult volunteers who were not actively engaged in working with B. pertussis and had no recent history of bordetella infection. The blood was allowed to clot at 37  C for 1 h and then placed on ice for 1e2 h and the clear serum was collected after centrifugation at 10000  g for 5 min and pooled. Heat inactivation, when appropriate, was carried out at 56  C for 30 min. Aliquots were stored at 80  C until used. A modification of the method of Barnes and Weiss [14] was used for the serum killing assay. BG agar cultures of B. pertussis strains were harvested after 20e24 h and suspended to a concentration of c. 109 CFU/ml in warm (37  C) SS medium. A volume of 500 ml of bacterial suspension, 475 ml of SS medium and 25 ml of normal or heat-inactivated pooled human serum were mixed and incubated in a 37  C water bath for up to 120 min. The mixture was then placed on ice for 5 min before diluting 1 in 10 in phosphatebuffered saline (PBS) (128 mM NaCl, 2.7 mM KCl, 1.5 mM KH2PO4, 5 mM K2HPO4 [pH 7.4]) with 10 mM EDTA to stop the complement

reaction. Further dilutions were made in SS medium before plating in triplicate on BG agar to determine the viable counts. Percentage survival was calculated from the mean number of CFU/ml after serum treatment compared to the mean number from a heatinactivated serum control (non-killing control). Statistical analysis was performed using Student’s t-test. P values <0.05 were assumed to be significant. 4.7. Mouse infection Four-week-old female CD1 mice (Harlan Olac, Bicester, Oxfordshire, UK), were used for infection experiments. B. pertussis suspensions were prepared in a 1% w/v Casamino acids solution (Casamino acids (Difco), 10g; MgCl2.6H2O 0.1g; NaCl, 5g; CaCl2, 0.016g; per litre, pH 7.1) from bacteria grown at 37  C for 24h on BG agar. For competition assays, groups of 5 or 10 mice were inoculated intranasally, under light halothane anaesthesia, with a 2-strain mixture containing c. 1  105 CFU/ml of each strain (confirmed by colony count). After 7 days, mice were sacrificed and their lungs were removed and homogenized in 10 ml of PBS. Dilutions were plated on BG agar with appropriate antibiotics to determine the numbers of survivors of each strain in the two strain mixture. Statistical analysis was performed using Student’s t-test. 4.8. Sensitivity to antimicrobial peptides A radial diffusion method [33] was used to determine the effect of antimicrobial peptides on B. pertussis strains. Bacteria grown for 24 h on BG agar were harvested in modified SS medium [13] to an optical density at 600 nm of c. 0.2, and 0.2 ml of this suspension was added to 10 ml of molten (52  C) 1% (w/v) agarose (type I; low electroendosmosis (Sigma) in modified SS medium and 0.15% (w/v) bovine serum albumin (Sigma) (SS-agarose). The agarose was dispensed into petri dishes and was allowed to harden. Holes (3 mm in diameter) were made with an aspirator punch and 5 ml of peptide (Sigma) serially diluted in sterile, filtered distilled water was placed therein. After incubation for 4h at room temperature to allow for diffusion, a 10-ml overlay of SS-agarose without bacteria was added. The resultant zones of inhibition were measured after 48h with a metric scale under a stereomicroscope. Student’s t-test was employed to analyse the data. Acknowledgements We are grateful to Alison Weiss and Norman Fry for provision of B. pertussis strains. Mojtaba Noofeli wishes to thank his sponsors, the Ministry of Health, Iran, and the Razi Vaccine Institute for their support for this work. Habib Bokhari wishes to thank the Commonwealth Association for their kind support. References [1] Henderson IR, Navarro-Garcia F, Desvaux M, Fernandez RC, Ala’Aldeen D. Type V protein secretion pathway: the autotransporter story. Microbiol Mol Biol Rev 2004;68:692e744. [2] Dautin N, Bernstein HD. Protein secretion in gram-negative bacteria via the autotransporter pathway. Annu Rev Microbiol 2007;61:89e112. [3] Nishimura K, Tajima N, Yoon YH, Park SY, Tame JR. Autotransporter passenger proteins: virulence factors with common structural themes. J Mol Med 2010; 8:451e8. [4] Parkhill J, Sebaihia M, Preston A, Murphy LD, Thomson N, Harris DE, et al. Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 2003;35:32e40. [5] Charles I, Fairweather N, Pickard D, Beesley J, Anderson R, Dougan G, et al. Expression of the Bordetella pertussis P.69 pertactin adhesin in Escherichia coli: fate of the carboxy-terminal domain. Microbiol 1994;140(Pt 12):3301e8. [6] Fernandez RC, Weiss AA. Cloning and sequencing of a Bordetella pertussis serum resistance locus. Infect Immun 1994;Vol. 62:4727e38.

M. Noofeli et al. / Microbial Pathogenesis 51 (2011) 169e177 [7] Finn TM, Stevens LA. Tracheal colonization factor: a Bordetella pertussis secreted virulence determinant. Mol Microbiol 1995;16:625e34. [8] Finn TM, Amsbaugh DF. Vag8, a Bordetella pertussis bvg-regulated protein. Infect Immun 1998;66:3985e9. [9] Beier D, Gross R. Regulation of bacterial virulence by two-component systems. Curr Opin Microbiol 2006;9:143e52. [10] Cummings CA, Bootsma HJ, Relman DA, Miller JF. Species- and strain-specific control of a complex, flexible regulon by Bordetella BvgAS. J Bacteriol 2006; 188:1775e85. [11] Bokhari H. Characterisation and secretion mechanism of Bordetella pertussis autotransporter proteins. University of Glasgow. Ph.D. thesis, 2002. [12] Weiss AA, Hewlett EL, Myers GA, Falkow S. Tn5-induced mutations affecting virulence factors of Bordetella pertussis. Infect Immun 1983;42:33e41. [13] Weiss AA, Melton AR, Walker KE, Andraos-Selim C, Meidl JJ. Use of the promoter fusion transposon Tn5 lac to identify mutations in Bordetella pertussis vir-regulated genes. Infect Immun 1989;57:2674e82. [14] Barnes MG, Weiss AA. BrkA protein of Bordetella pertussis inhibits the classical pathway of complement after C1 deposition. Infect Immun 2001;69:3067e72. [15] Byrd DW, Roop RM, Veit HP, Schurig GG. Serum sensitivity and lipopolysaccharide characteristics in Bordetella bronchiseptica, B. pertussis and B. parapertussis. J Med Microbiol 1991;34:159e65. [16] Burns VC, Pishko EJ, Preston A, Maskell DJ, Harvill ET. Role of Bordetella O antigen in respiratory tract infection. Infect Immun 2003;71:86e94. [17] Barnes MG, Weiss AA. Activation of the complement cascade by Bordetella pertussis. FEMS Microbiol Lett 2003;220:271e5. [18] Pishko EJ, Betting DJ, Hutter CS, Harvill ET. Bordetella pertussis acquires resistance to complement-mediated killing in vivo. Infect Immun 2003;71: 4936e42. [19] Berggard K, Lindahl G, Dahlback B, Blom AM. Bordetella pertussis binds to human C4b-binding protein (C4BP) at a site similar to that used by the natural ligand C4b. Eur J Immunol 2001;31:2771e80. [20] Elder KD, Harvill ET. Strain-dependent role of BrkA during Bordetella pertussis infection of the murine respiratory tract. Infect Immun 2004;72:5919e24.

177

[21] Reddy KV, Yedery RD, Aranha C. Antimicrobial peptides: premises and promises. Int J Antimicrob Agents 2004;24:536e47. [22] Fernandez RC, Weiss AA. Susceptibilities of Bordetella pertussis strains to antimicrobial peptides. Antimicrob Agents Chemother 1996;40:1041e3. [23] Kessel M, Brennan MJ, Trus BL, Bisher ME, Steven AC. Naturally crystalline porin in the outer membrane of Bordetella pertussis. J Mol Biol 1988;5;203(1): 275e8. [24] Weiss AA, Goodwin MS. Lethal infection by Bordetella pertussis mutants in the infant mouse model. Infect Immun 1989;57:3757e64. [25] Wells TJ, Tree JJ, Ulett GC, Schembri MA. Autotransporter proteins: novel targets at the bacterial cell surface. FEMS Microbiol Lett 2007;274:163e72. [26] Kinnear SM, Boucher PE, Stibitz S, Carbonetti NH. Analysis of BvgA activation of the pertactin gene promoter in Bordetella pertussis. J Bacteriol 1999;181: 5234e41. [27] Gogol EB, Cummings CA, Burns RC, Relman DA. Phase variation and microevolution at homopolymeric tracts in Bordetella pertussis. BMC Genomics 2007;8:122. [28] Imaizumi A, Suzuki Y, Ono S, Sato H, Sato Y. Effect of Heptakis(2,6-ODimethyl)beta-Cyclodextrin on the production of pertussis toxin by Bordetella pertussis. Infect Immun 1983;41:1138e43. [29] Stainer DW, Scholte MJ. A simple chemically defined medium for the production of phase 1 Bordetella pertussis. J Gen Microbiol 1971;63:211e20. [30] Stibitz S. Use of conditionally counterselectable suicide vectors for allelic exchange. Methods Enzymol 1994;235:458e65. [31] Simon R, Priefer U, Puhler A. A broad host range mobilization system for in vivo genetic engineering, transposon mutagenesis in Gram-negative bacteria. Bio/Technology 1983;1:784e91. [32] Packard ER, Parton R, Coote JG, Fry NK. Sequence variation and conservation in virulence-related genes of Bordetella pertussis isolates from the UK. J Med Microbiol 2004;53:355e65. [33] Lehrer RI, Rosenman M, Harwig SS, Jackson R, Eisenhauer P. Ultrasensitive assays for endogenous antimicrobial polypeptides. J Immunol Methods 1991; 137:167e73.