The TolC-Like Protein of Neisseria meningitidis Is Required for Extracellular Production of the Repeats-in-Toxin Toxin FrpC but Not for Resistance to Antimicrobials Recognized by the Mtr Efflux Pump System

Transcription of the hlyD-tolC locus in meningococci.

In agreement with the reports by Klee et al. (9) and Wooldridge et al. (32), we identified a 1.403-kb locus in the annotated genome sequence of meningococcal strains MC58 and Z2491 that contains an ORF that would encode a TolC-like protein (data not presented). Analysis of the genome sequence flanking this ORF also revealed a closely linked ORF upstream that encodes a protein similar to the HlyD protein of E. coli. Neither the hlyD-tolC locus nor the individual genes could be identified with the online (www.genome.ou.edu) genome sequence of N. gonorrhoeae strain FA1090 and could not be amplified by PCR from genomic DNA prepared from this strain or from eight other gonococcal strains (FA19, F62, RD5, DGI 4784, UU1, 3115, DGI 1918, and DGI 14804; data not presented). However, the locus was amplified from genomic DNA prepared from meningococcal strain NMB (data not presented), from its acapsular mutant derivative strain M7 (22) (Fig. 1), and from strain 0929 (data not presented). Thus, between the two pathogenic Neisseria species, the hlyD-tolC locus is likely restricted to meningococci, a conclusion supported by the findings of Klee et al. (9). We determined the nucleotide sequence of the hlyD-tolC region that was amplified by PCR with chromosomal DNA from N. meningitidis strain M7. We found that the nucleotide sequence of this 2.95-kb region was virtually identical to the online sequence for strain MC58 (23), with the exception of three single-nucleotide missense mutations in hlyD, which would cause amino acid replacements at positions 28 (K→E), 285 (N→S), and 356 (I→V) in the full-length protein and two silent mutations in hlyD at nucleotide positions 1017 and 1167. In agreement with the predicted HlyD and TolC proteins produced by strains MC58 and Z2491 (14, 23), we determined that the amino acid sequence of the HlyD protein that would be produced by strain M7 (475 amino acids) was 39% identical (62% similarity) to the HlyD protein produced by E. coli over the entire protein, while the TolC proteins of E. coli and N. meningitidis strain M7 (467 amino acids) were 25% identical (45% similarity).

• Open in new tab
• Download powerpoint

FIG. 1.

Organization and transcription of the hlyD-tolC operon in N. meningitidis strain M7. (A) The hlyD and tolC genes are separated by 68 bp in strain M7. A putative promoter element located 56 bp upstream of hlyD is shown. The near consensus −10 and −35 hexamers are separated by 17 bp; nucleotides that differ from the consensus sequence are underlined. (B) The agarose gel profile of RT-PCR products generated using total RNA extracted from exponentially growing N. meningitidis strain M7 is shown. The site of annealing of the HlyD5 (5′-ATGCGGTGGTGAAGATTGAG-3′) and TolC5 (5′-AATCAGCCGAATGTTGCTGC-3′) oligonucleotide primers are shown. Lanes 1, RT-PCR product of the hlyD-tolC sequence obtained when the reaction included RT; 2, same as lane 1 but the RT product was omitted; 3, RT-PCR product of the control rnpB gene transcript that was generated using oligonucleotide primers RnpB1F (5′-CGGGACGGGCAGACAGTCGC-3′) and RnpB1R (5′-GGACAGGCGGTAAGCCGGGTTC-3′). (C) The agarose gel profile of RT-PCR products generated is shown. RNA was prepared from a broth culture of strain M7 at different phases of growth and was subjected to RT-PCR using primers HlyD5 and TolC5. Lanes 1, early log phase; 2, mid-log phase; 3, late log phase; 4, stationary phase. (D) Agarose gel profile of the RT-PCR product of the control rnpB transcript (17). Samples in each lane are the same as those described for panel C.

By sequence analysis, a putative promoter element between the meningococcal hlyD and tolC genes in strain M7 could not be identified, suggesting that the genes were transcriptionally linked and that a promoter element upstream of hlyD is used for the transcription of hlyD and tolC. Indeed, a putative promoter element containing near consensus (5/6 matches) −10 and −35 hexamer sequences, separated by 17 bp, was identified at 56 bp upstream of hlyD (Fig. 1). In order to test whether the ORFs were transcribed together, total RNA was extracted from early-log-, mid-log-, late-log-, and stationary-phase gonococcal base (GCB) broth cultures (containing defined supplements I and II [6] and 4.3% [wt/vol] sodium bicarbonate) of N. meningitidis strain M7, as described previously (16, 17). These RNA preparations were employed in reverse transcription (RT)-PCRs that used a primer (TolC5) annealed within the putative tolC gene (Fig. 1A), while the subsequent PCR used TolC5 and the primer HlyD5 annealed in the upstream hlyD gene. As shown in Fig. 1B, a specific band of 574 bp was observed most predominantly from the RNA prepared from the late-log culture. This result was contrary to an earlier report by Wooldridge et al. (32) that concluded that hlyD and tolC are not transcriptionally linked. In order to ascertain the reason for the discrepant results, we performed RT-PCR experiments using the growth and RT-PCR conditions and primers described by Wooldridge et al. (32), but we were unable to detect RT-PCR products (data not presented). In particular, we found that the operon was maximally expressed by late-log-phase cultures and was poorly expressed by stationary-phase cultures (Fig. 1C), such as that used by Wooldridge et al. (32) to prepare total RNA; as an RNA-loading control, we used RT-PCR to monitor the transcript levels of rnpB (17). Taking these results together, we conclude that the procedures employed by Woolridge et al. (32) were not optimal for detecting transcriptional linkage of hlyD and tolC.

The meningococcal TolC-like protein is required for extracellular production of FrpC but is not needed for resistance to hydrophobic antimicrobials.

Because TolC in E. coli functions in both the export of HlyA through a type I secretion system and as the OMP channel for efflux pumps to remove antimicrobial agents, we asked if the meningococcal TolC-like protein had similar dual properties. For this purpose, we examined a panel of antimicrobials recognized by the AcrA-AcrB-TolC efflux system of E. coli and by the MtrC-MtrD-MtrE efflux system of meningococci (16) and gonococci (2, 5, 8, 18, 19). The MtrC-MtrD-MtrE efflux system was chosen for analysis because of the similarities between TolC and MtrE (see above) and its similarity to the AcrA-AcrB-TolC efflux system.

We previously reported (2) the construction of a nonpolar insertional mutation in the mtrE gene (mtrE::kan) of N. gonorrhoeae strain FA19. DNA from this transformant mutant (strain RD1) was used to transform N. meningitidis strain M7 for resistance to kanamycin (100 μg/ml) as described previously (2, 16). A similar nonpolar mutation was constructed in the meningococcal tolC gene. Briefly, PCR was used to amplify a 1.6-kb sequence from genomic DNA of strain M7, and the purified product was cloned into pBAD. The resulting plasmid construct, pBAD::tolC, was then purified from an E. coli transformant. The promoterless, nonpolar kanamycin resistance cassette (aphA-3) gene from pUC18K (12) was cloned into a unique PstI site located at a position 443 bp within the tolC gene sequence, resulting in pBAD::tolC::kan. The insertion of the cassette was verified by PCR and DNA sequence analysis (data not presented) of the plasmid construct prepared from an E. coli DH5α transformant. A subsequent transformant of meningococcal strain M7 was prepared with this plasmid, and the replacement of the wild-type chromosomal copy of tolC with the inactivated gene in a representative transformant of strain M7 was confirmed by PCR and DNA sequence analysis (data not presented). M7tolC::kan, M7mtrE::kan, and parental strain M7 were then analyzed for their levels of susceptibility to a panel of diverse antimicrobial agents (antibiotics [azithromycin, penicillin, and streptomycin], dyes [crystal violet and ethidium bromide], a nonionic detergent [Triton X-100], and an antimicrobial peptide [LL-37]) recognized by the MtrC-MtrD-MtrE efflux pump, produced by gonococci (6, 8, 18) and meningococci (16). As shown in Table 1, insertional inactivation of mtrE significantly enhanced meningococcal susceptibility to the test hydrophobic agents but did not impact bacterial susceptibility to the hydrophilic antibiotic streptomycin. In contrast, the inactivation of tolC did not impact the levels of meningococcal susceptibility to the hydrophobic antimicrobials or streptomycin (Table 1). Thus, the meningococcal TolC-like protein is not essential for the export of the antimicrobial agents examined, a function that is instead served by MtrE.

Since the 198-kDa RTX toxin FrpC (3, 13, 25) produced by meningococci is similar to the HlyA toxin of E. coli (28, 31), we next asked whether tolC is needed for its extracellular appearance, as is the case for HlyA production by E. coli (31). In preliminary experiments that used sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western immunoblotting to detect FrpC with a rabbit anti-FrpC antiserum (13), we were able to detect FrpC in whole-cell lysates of strains M7 and M7tolC::kan grown under iron-replete or iron-restricted conditions (data not presented). However, while extracellular FrpC could be detected in culture supernatants of strain M7 under iron-replete (Fig. 2, lane 1, and data not presented) and iron-deficient conditions (data not presented), it could not be detected in supernatants from strain M7tolC::kan grown under iron-replete conditions (data not presented and Fig. 2, lane 2) or iron-deficient conditions (data not presented). In other experiments, we were able to detect extracellular FrpC in culture supernatants of strain M7mtrE::kan (data not presented).

• Open in new tab
• Download powerpoint

FIG. 2.

Complementation of the mutation in tolC restores extracellular production of FrpC. Supernatants from meningococcal strains M7 (lane 1), M7tolC::kan (lane 2), and M7tolC::kan tolC⁺ (lane 3) grown in the presence (A) or absence (B) of 1 mM IPTG were solubilized and subjected to SDS-PAGE with a 6% (wt/vol) polyacrylamide gel. FrpC (198 kDa) (3, 13, 24, 25, 32) was detected by immunoblotting using rabbit anti-FrpC and goat anti-rabbit IgG conjugated to horseradish peroxidase. As the levels of total protein in these culture supernatants were frequently <1 μg/ml, an equal volume (10 μl) of culture supernatant from each of the test strains was examined by immunoblotting. In other experiments (data not presented), extracellular FrpC could not be detected even in 10-fold-concentrated supernatants from strain M7tolC::kan.

In order to verify that it was the loss of TolC production and not a polar effect of the mutation that was responsible for the loss of extracellular FrpC production in the tolC mutant strain, we performed a complementation test that utilized the neisserial insertional complementation system (20). This system allows for the expression of a gene at a second site (between lctP and aspC) on the meningococcal chromosome. Briefly, primers with PacI and PmeI restriction sites were designed to amplify tolC such that the PCR product contained 45 bp of DNA sequence upstream of tolC (including the ribosome binding site), the full coding sequence, and 250 bp of downstream DNA. The primer used in the forward direction was tolCF-Pac1 (5′-CGTTAATTAAATCCGCCGGTTCGGACATG-3′), and the reverse primer was tolCR3PmeI (5′-CGGTTTAAACAGCTCAACCCGATTGAGAAG-3′). The PCR product was then cloned into pGCC4 (20) that had been digested with PacI and PmeI (20), which contains an isopropyl-β-d-thiogalactoside (IPTG)-regulated lac promoter, and E. coli DH5α transformants were obtained by selection for resistance to the kanamycin that was included in the LB agar at 50 μg/ml. Plasmid DNA from a representative transformant was prepared and digested with NotI to liberate an 8.1-kb fragment containing the wild-type tolC gene from strain M7 and the flanking plasmid-borne ermC, lac promoter, lacI^q, aspC, and lctP gene sequences (20). An agarose gel-purified NotI fragment was then used to transform strain M7tolC::kan for resistance to erythromycin (1 μg/ml). The insertion of a wild-type copy of tolC between lctP and aspC in a representative transformant was verified by PCR and DNA sequencing (data not presented). In order to determine whether the presence of the copy of tolC could restore extracellular production of FrpC in M7tolC::kan, we analyzed culture supernatants from strains M7 and M7tolC::kan and a representative transformant (complemented strain) M7tolC::kan tolC⁺. For this purpose, meningococci were grown overnight on GCB agar (with or without 1 mM IPTG) that lacked glucose (normally added in defined supplement I [6]), and standard iron-replete conditions were employed. After they were grown overnight, bacteria were scraped from the agar and resuspended in 1 ml of GCB broth, incubated at 37°C for 30 min, and vortexed, and culture supernatants were collected by centrifugation. The culture supernatants from the test strains were examined for FrpC levels by SDS-PAGE and immunoblotting. As shown in Fig. 2, the tolC gene-complemented strain, but not strain M7tolC::kan (Fig. 2A and B, lane 2), produced extracellular FrpC (Fig. 2A, lane3) in the presence of IPTG, and this extracellular production was IPTG dependent (Fig. 2B, lane 3). In the presence of IPTG, the level of FrpC produced by the complemented strain was similar to that of wild-type strain M7 (Fig. 2A, compare lanes 1 and 3). Thus, we conclude from these data that extracellular FrpC production by meningococci requires the TolC-like protein.

Our results indicate that the meningococcal TolC-like protein is not required for the efflux of hydrophobic agents recognized by the mtr gene system in meningococci (16) but is needed for extracellular production of the RTX-like toxin FrpC (Fig. 2). In contrast to an earlier report (32), we determined that the tandemly linked hlyD and tolC genes are coexpressed (Fig. 1), a finding that supports the notion that they are organized as an operon. We also found that the hlyD-tolC operon in meningococci is maximally expressed during the late log phase of growth (Fig. 1), when levels of free iron would be depleted. However, the significance of this observation is unclear as it is not yet known if conditions of iron restriction, which would be expected in vivo, are important in controlling the hlyD-tolC operon expression or how iron limitation impacts FrpC production, as has been reported by others (24-26).

While FrpC production in an infant rat model for septicemia could not be correlated with virulence (3) and it lacks cytotoxic activity, there is reason to believe, as emphasized earlier by Wooldridge et al. (32), that FrpC production may be of importance during meningococcal carriage, infection, and/or disease. First, FrpC is expressed during human meningococcal disease and is highly antigenic (13), as determined by the presence of high levels of immunoglobulin G (IgG) and IgA in convalescent-phase sera from patients recovering from meningococcal infection. Second, frpC alleles are present in numerous capsular serogroup B and C strains. The absence of FrpC in other Neisseria, particularly gonococci, has raised the question as to whether its function is unique for meningococcal disseminated disease (32). While the hlyD and tolC genes are present in numerous clinical isolates of the major clonal lineages of meningococci (9, 13) and hlyB and hlyD (32) are required for FrpC export, it is not yet clear if this type I secretion system secretes other proteins important for meningococcal pathogenesis. Accordingly, it is important to define the physiologic and genetic control mechanisms that regulate this export system in meningococci.

Nucleotide sequence accession number.

The nucleotide sequence of the hlyD gene in meningococcal strain M7 is available at GenBank under accession number 1015055.