|Home | About | Journals | Submit | Contact Us | Français|
The rates of pilin antigenic variation (Av) of two strains of Neisseria meningitidis were determined using an unbiased DNA sequencing assay. Strain MC58 underwent pilin Av at a rate similar to that of N. gonorrhoeae strain MS11 but lower than that of N. gonorrhoeae strain FA1090. Pilin Av was undetectable in strain FAM18.
Neisseria meningitidis is a Gram-negative diplococcus that colonizes the nasopharynx of approximately 5 to 10% of the population and is usually nonpathogenic but can occasionally enter the bloodstream to cause septicemia and can eventually spread to the meninges, causing meningitis (15). Approximately 500,000 cases of meningococcal meningitis occur every year, with nearly 10% resulting in fatality (2).
Type IV pili (TFP) are long filamentous structures protruding from the bacterial surface and are required for adherence of N. meningitidis to host cells (7). As with the TFP of the closely related pathogen Neisseria gonorrhoeae, the pili are able to undergo antigenic variation (Av). In N. gonorrhoeae, pilin Av occurs as a result of recombination between one of the multiple silent pilS copies and the expressed pilin gene (pilE). The pilS copies share significant regions of homology with pilE yet lack a promoter or ribosome-binding site and the initial 5′ coding sequence. Pilin Av relies on RecA and the RecF-like recombination pathway to catalyze gene conversion, resulting in an altered pilE sequence, carrying part of the pilS donor, and the original unaltered pilS sequence (8, 9).
While the frequency of pilin Av has been measured in N. gonorrhoeae (5, 10, 12), this process has never been quantified in N. meningitidis. Two sequenced strains were picked to measure pilin Av: serogroup B strain MC58 (sequence type [ST-32] complex), isolated from an invasive infection (14), and serogroup C strain FAM18 (ST-11 complex), which was isolated from a patient with septicemia (1). In both strains, the native recA gene was replaced with the very highly conserved N. gonorrhoeae recA6 construct, which allows regulation of expression with IPTG (isopropyl-β-d-thiogalactopyranoside) (17). recA6 strains are RecA+ when grown with IPTG but are RecA− when grown without IPTG (17). These phenotypes were confirmed by measuring the UV sensitivities and DNA transformation competence levels of both strains with or without IPTG, and both strains were shown to be piliated by transmission electron microscopy (data not shown). Bacteria were grown at 37°C with 5% CO2 on gonococcal medium base (GCB; Difco) plus Kellogg supplements I and II (11).
The pilin Av sequencing assay was performed as described previously (5, 10, 12) with slight modifications. Briefly, FAM18 and MC58 were grown on solid GCB with 1 mM IPTG, allowing for the expression of RecA, for 22 h and 12.5 h, respectively, which was estimated to produce 20 generations. For FAM18, little or no pilin Av was expected since the G-quartet-forming sequence required for pilin Av is degenerate in this strain (3). Therefore, two random progenitor colonies were picked from IPTG-enriched medium and passaged on GCB without IPTG. Between 91 and 94 colonies were isolated from each FAM18 progenitor, and the sequence of the pilE gene was determined. For MC58, seven random progenitor colonies were picked and passaged on GCB without IPTG. Between 28 and 47 progeny colonies arising from each of the seven progenitors were isolated, and the pilE gene sequence was determined. In both MC58 and FAM18, the progeny colonies were passaged on GCB two times to ensure colony clonality. A single colony from each sample was isolated, and the pilE gene was PCR amplified as described previously (13).
The primers used for amplification of MC58 pilE were McPilRBS (5′-GCATTTCCTTTCCAATTAGGAG) and MC58SP3A (5′-TTCCGTACGGATAGCTTCGTC). The primers used for amplification of FAM18 pilE were FAMFOR-2 (5′-ATTACGGGTTTACGTTTGCGG) and FAMREV-2 (5′-ACGCACCTACGCCTCACCCTAC). The DNA sequence was determined for each sample (SeqWright, Houston, TX, and the Genomics Core at Northwestern University) and analyzed (MacVector; Symantec Corp.). Colonies that showed pilE sequence changes were reanalyzed to confirm the Av event.
The pilin Av frequency was determined for the progeny of each progenitor by dividing the total number of detected pilin Av events by the number of progeny of each set, resulting in two values for FAM18 and seven values for MC58 (Table (Table1).1). The pilin Av rate was determined by dividing the pilin Av frequency by the number of generations for each sample grown in the presence of IPTG, as determined by a colony assay at the time of harvest. After growth for the same time period, the total numbers of generations for MC58 and FAM18 grown under RecA induction were approximately 19 and 23, respectively.
MC58 possesses eight pilS copies in a single locus directly upstream of pilE, designated pilS1 to pilS8 (GenBank accession numbers NMB0019 to NMB0026, respectively) (16). FAM18 possesses two pilS sequences in a single locus, designated pilS1 and pilS2 (GenBank accession numbers NMC0002 and NMC0003, respectively) (1). All MC58 pilin Av events were confirmed by comparing the sequence of the altered pilE to the eight pilS sequences, and each was matched to a pilS sequence donor.
As predicted, no pilin Av was detected in FAM18 within the 185 progeny colonies analyzed (Table (Table1).1). Therefore, the pilin Av frequency is less than the detection limit of 5.4 × 10−3 events/CFU and the rate is less than the detection limit of 2.3 × 10−4 events/CFU/generation. This result is consistent with observations from clinical ST-11 isolates which have a conserved pilE sequence (4).
In contrast, pilin Av in MC58 was detected in five of the seven sets, with the highest frequency and rate belonging to set A, which had a frequency of 0.2 events/CFU and a rate of 1 × 10−2 events/CFU/generation (Table (Table1).1). The median frequency of pilin Av in MC58 was 0.03 events/CFU, and the median rate was 1.6 × 10−3 events/CFU/generation. Using the Wilcoxon rank sum test, the rates of pilin Av of MC58 were statistically reduced relative to the rates previously reported for N. gonorrhoeae strain FA1090 grown for 20 generations (10), with a P value of <0.05. In contrast, the rates of pilin Av of MC58 and gonococcal strain MS11, which was shown to have a reduced level of pilin Av relative to that of strain FA1090 from the same study (10), were not statistically different from each other.
The pilS donor was determined for all 13 MC58 samples with a variant pilE gene. Five pilin Av events were the result of recombination with pilS3. One pilin Av event was the result of recombination with pilS1. One event was the result of recombination with pilS8. Five events could have been the result of recombination with pilS1, pilS2, or pilS3 in areas of these pilS copies that are identical. One event was the result of recombination with either pilS5 or pilS7. No recombination events with either pilS4 or pilS6 were detected. While these data suggest that there is a nonrandom distribution of donor pilS copies, which has been shown in N. gonorrhoeae (4, 9), further studies with more samples will have to be performed to verify this.
These results definitively demonstrate that N. meningitidis strain MC58 undergoes pilin Av at a rate similar to that of N. gonorrhoeae strain MS11 yet reduced relative to that of N. gonorrhoeae strain FA1090. N. meningitidis strain FAM18, which lacks a well-defined G-4 quartet and encodes class II pilin (6, 18), undergoes pilin Av at a rate not detectable by the assay described here and is unlikely to undergo antigenic variation at all. These data clearly show that major differences in strain-to-strain pilin Av rates exist, an observation previously made for N. gonorrhoeae (10). The greatly reduced level of pilin Av in FAM18 suggests that this strain is not dependent upon pilin Av for its survival and spread from host to host and that other class II pilin gene-expressing strains (4) must have a relationship with the host immune system different from that of class I pilin-expressing strains. Whether there are differences in transmission, pilus function, or interactions with the host remains to be determined.
We are grateful to Paul Duffin and Paul Schook for technical assistance and Laty Cahoon for helpful discussions.
This work was supported by grants R01 AI044239, R01 AI055977, and R37 AI033493 to H.S.S. R.A.H. was partially supported by grant F32 AI065091.
Published ahead of print on 14 May 2010.