Opacity proteins increase Neisseria gonorrhoeae fitness in the female genital tract due to a factor under ovarian control

Abstract

The neisserial opacity (Opa) proteins are a family of antigenically distinct outer membrane proteins that undergo phase-variable expression. Opa(+) variants of Neisseria gonorrhoeae strain FA1090 are selected in a cyclical pattern from the lower genital tract of estradiol-treated mice. Here we show that cyclical recovery of Opa(+) gonococci does not occur in ovariectomized mice; therefore, the reproductive cycle plays a role in the selection kinetics in vivo. As predicted by the selection pattern shown by wild-type gonococci, we demonstrated that a constitutive Opa-expressing strain was more fit than an Opa-deficient mutant in the early and late phases of infection. We found no evidence that Opa-mediated colonization selects for Opa(+) variants during murine infection based on adherence assays with cultured murine epithelial cells. We also tested the hypothesis that complement selects for Opa protein expression during infection. Although some Opa(+) variants of a serum-sensitive derivative of strain FA1090 were more resistant to the bactericidal activity of normal human serum, selection for Opa expression was not abrogated in C3-depleted mice. Finally, as previously reported, Opa(+) gonococci were more sensitive to serine proteases. Thus, proteases or protease inhibitors may contribute to the observed in vivo selection pattern. We concluded that Opa proteins promote persistence of N. gonorrhoeae in the female genital tract and that opa gene phase variation allows gonococci to evade or capitalize upon unidentified host factors of the mammalian reproductive cycle. This work revealed an intimate interaction between pathogen and host and provides evidence that hormonally related factors shape bacterial adaptation.

FIG. 1.

For both estradiol-treated and untreated mice there was cyclical recovery of Opa⁺ variants, and estradiol does not select for Opa protein expression in vitro. For mice treated with 17β-estradiol_ws (A) or not treated with estradiol (B) there was cyclical recovery of Opa⁺ variants after the mice were inoculated with a defined suspension containing predominantly Opa⁻ variants. The average fold increases for Opa⁺ variants among vaginal isolates compared to the inoculum are indicated in the histograms, and selection was defined as a ≥2-fold increase (dotted line). The bars indicate standard errors of the means; the asterisks indicate that the P value is <0.03 for a comparison with the value for the middle phase, which was defined as days 3 to 5 for panel A and as day 3 for panel B. The results in panel A are the results of one experiment performed with 6 mice inoculated with 89% Opa⁻ and 11% OpaB variants and are representative of two independent experiments (5 or 6 mice per experiment). The data in panel B are combined data from three experiments in which untreated mice were colonized for 1 to 4 days. The average fold increase was based on the results for 11 mice on day 1, for 5 mice on day 2, for 3 mice on day 3, and for 2 mice on day 1. In each panel the pie charts show the Opa phenotype of a representative mouse, and the colors indicate the different Opa proteins of strain FA1090. The striped patterns show the results for variants that express multiple Opa proteins. (C) Recovery of Opa⁺ isolates from a liquid culture at different stages of growth, expressed as fold increases compared to the inoculum. GCB with no estradiol or with 10, 100, or 1,000 pg/ml 17β-estradiol_ws was inoculated with a bacterial suspension that contained 75% Opa⁻ variants. Concentrations were based on serum estradiol levels in untreated, pellet-treated, and estradiol_ws-treated mice (56). The Opa phenotypes of 96 colonies (zero time) and 36 colonies (6 and 8 h) isolated from each culture at different time points were determined, and the fold changes compared to the inoculum are indicated. No change in the Opa phenotype was observed under any conditions. OD₆₀₀, optical density at 600 nm.

FIG. 2.

Normal (intact) mice, but not Ov⁻ mice, show a cyclical recovery pattern. Five-milligram 17β-estradiol pellets were implanted in members of groups of intact or Ov⁻ mice, and the mice were inoculated with 10⁶ CFU of a suspension containing 70% Opa⁻, 23% OpaB, 3% OpaC, 2% OpaF, and 1% OpaF and OpaI variants. The Opa phenotype of vaginal isolates and the number of gonococci recovered were determined over time. (A) Opa phenotype of vaginal isolates. Selection for Opa expression, which was defined as a ≥2-fold increase in the percentage of Opa⁺ variants compared to the inoculum, occurred in both groups on day 1. However, the subsequent loss and reemergence of Opa⁺ variants that occurred in intact mice did not occur in Ov⁻ mice. (B) Total number of CFU recovered over time. Fluctuations in the total number of CFU recovered from vaginal swab suspensions were observed in intact mice, as reported previously (47), and the dramatic decline that characterized the middle phase occurred on days 5 to 8. Ov⁻ mice showed more uniform recovery of viable gonococci over time. An asterisk indicates a significant difference (P < 0.05) between the numbers of gonococci recovered from groups, and the error bars indicate the standard errors of the means. In the experiment whose results are shown, there were 3 mice per group, except for intact mice on days 11 to 15, for which results for 2 mice are shown. The third mouse was not colonized at these time points. The results were consistent in a second experiment performed with 5 mice per group.

FIG. 3.

OpaB-expressing strain colonizes 17β-estradiol-treated mice better than Opa-deficient mutant. The Opa-deficient strain FA1090opaA-K and the OpaB-expressing strain FA1090opaA-K(B+) were compared to examine differences in growth in liquid culture and the capacity to colonize estradiol-treated mice. (A) Optical density at 600 nm (OD₆₀₀) for GCB cultures inoculated with FA1090opaA-K or FA1090opaA-K(B+) bacteria over time. (B) Average number of CFU recovered by vaginal swabbing following inoculation of estradiol-treated mice with 10⁶ CFU of Opa-deficient or OpaB-expressing bacteria. The error bars indicate the standard errors of the means, and the asterisks indicate significant differences between groups (P < 0.05, Student's t test). (C) Percentage of mice colonized on each day following inoculation with strain FA1090opaA-K and with strain FA1090opaA-K(B+). Significantly more mice were culture positive following inoculation with the constitutive OpaB-expressing strain than following inoculation with the Opa-deficient strain (P = 0.037, log rank test). The results in panels A and B are results from two combined experiments, and each symbol represents 16 mice.

FIG. 4.

OpaB-mediated invasion occurs in immortalized human tissue culture cells but not in murine tissue culture cells. The capacities of strains FA1090opaA-K(B+) and FA1090opaA-K to invade human (ME180 and HEC1B) and mouse (IEC4.1 and BM1.11) cells were compared using the Gm protection assay. A higher number of FA1090opaA-K(B+) CFU than of FA1090opaA-K CFU was recovered from both human cell lines after Gm treatment, and there was 10-fold-greater internalization of the Opa-expressing strain by the CEACAM-expressing ME180 cells than by the CEACAM-negative HEC1B cells. No difference in invasion was detected in either mouse cell line for FA1090opaA-K(B+) or FA1090opaA-K. No difference in the number of cell-associated gonococci was observed between the two strains for any cell line (data not shown). The level of invasion was calculated by comparison with the number of adherent bacteria recovered before Gm treatment (data not shown). The error bars indicate the standard errors of the means for three independent wells. Similar results were obtained in three independent experiments.

FIG. 5.

Expression of some FA1090 Opa proteins is associated with increased resistance to NHS. The bactericidal activity of NHS was tested using different Opa variants of the serum-sensitive derivative FA1090_F62por5-8. (A) Plot of the level of survival of each variant tested versus the dilution of NHS tested. The representative data from several independent assays show that there was an increase in the 50% bactericidal titer compared to the 50% bactericidal titer for Opa⁻ variants for gonococci that express OpaB, OpaE/K, OpaF, OpaA,F, and OpaA,F,E/K but not for variants that express OpaA or OpaA,B. The 50% bactericidal titers were 1.8% (Opa⁻, OpaA, and OpaA,B), 2.2% (OpaB), 3.4% (OpaE/K), 3.8% (OpaF), and ca. 5.1% (OpaA,F and OpaA,F,E/K). (B) Levels of survival of different Opa variants following incubation in 3% NHS. The results are expressed as the average levels of survival when data from at least three independent experiments were combined. The error bars indicate the standard errors of the means, and the asterisks indicate that there was a significant difference (P < 0.0001). The level of survival was calculated by dividing the number of CFU recovered from wells with NHS by the number of CFU recovered from wells with the same concentration of HI-NHS and multiplying by 100.

FIG. 6.

Fluctuations in vaginal C3 levels occur in uninfected mice, but C3 depletion does not abrogate the early selection for Opa⁺ variants. (A) Concentration of C3 in vaginal washes from uninfected mice that were treated with 5-mg 17β-estradiol slow-release pellets. A significantly higher concentration of C3 was present in vaginal washes from days 7 and 8 (late phase) than in vaginal washes from days 3 to 5 (middle phase). The results of two independent experiments in which 10 to 13 mice per day were used were combined. (B) Recovery of Opa⁺ variants in mice treated with CVF to deplete C3 the day prior to inoculation with wild-type bacteria. Control mice were treated with PBS. Mice were challenged with a predominantly Opa⁻ (80%) suspension of wild-type FA1090 gonococci, and the proportion of Opa⁺ variant colonies among colonies isolated from vaginal mucus was determined on days 1 to 3 postinoculation. Selection for Opa expression, as defined as a ≥2-fold increase compared to the inoculum (dotted line), occurred in both groups on day 1 or day 2 after bacterial challenge. The results of a single experiment in which there were 3 or 4 mice per group are shown. (C) Effect of C3 depletion on the recovery of strain FA1090opaA-K. There was no difference in the average numbers of Opa-deficient gonococci recovered from mice treated with CFU (solid line) and mice treated with PBS (dashed line) on days 1 to 3 of infection. The results are the results of a single experiment performed with 7 mice per group, and similar results were obtained in a second independent experiment. In all panels, the error bars indicate the standard errors of the means.

FIG. 7.

Opa⁺ gonococci are more susceptible to growth inhibition by serine proteases. A disk diffusion assay was used to measure the sensitivities of recombinant and wild-type gonococci to trypsin or α-chymotrypsin. Opa⁺ gonococci were more susceptible to growth inhibition when they were exposed to serine proteases. (A) Zones of inhibition for the Opa-deficient mutant FA1090opaA-K and the complemented OpaB-expressing mutant FA1090opaA-K(B+) with different amounts of trypsin and α-chymotrypsin. (B) Zones of inhibition for wild-type Opa-negative gonococci and OpaA, OpaB, and OpaD variants with different amounts of α-chymotrypsin. The average zones of inhibition (mm) are indicated by the bars, and the error bars indicate standard errors. The disk diameter was 6 mm. The asterisks indicate statistical significance (P ≤ 0.05).