Skip to main content
PMC home page
Infection and Immunity logoLink to Infection and Immunity
. 2007 Aug 13;75(10):4972–4979. doi: 10.1128/IAI.00763-07

The O Antigen Enables Bordetella parapertussis To Avoid Bordetella pertussis-Induced Immunity▿

Daniel N Wolfe 1, Elizabeth M Goebel 2, Ottar N Bjornstad 3, Olivier Restif 4, Eric T Harvill 1,*
PMCID: PMC2044517  PMID: 17698566

Abstract

Bordetella pertussis and Bordetella parapertussis are closely related endemic human pathogens which cause whooping cough, a disease that is reemerging in human populations. Despite how closely related these pathogens are, their coexistence and the limited efficacy of B. pertussis vaccines against B. parapertussis suggest a lack of cross-protective immunity between the two. We sought to address the ability of infection-induced immunity against one of these pathogens to protect against subsequent infection by the other using a mouse model of infection. Immunity induced by B. parapertussis infection protected against subsequent infections by either species. However, immunity induced by B. pertussis infection prevented subsequent B. pertussis infections but did not protect against B. parapertussis infections. The O antigen of B. parapertussis inhibited binding of antibodies to the bacterial surface and was required for B. parapertussis to colonize mice convalescent from B. pertussis infection. Thus, the O antigen of B. parapertussis confers asymmetrical cross-immunity between the causative agents of whooping cough. We propose that these findings warrant investigation of the relative role of B. parapertussis in the resurgence of whooping cough.

Bordetella pertussis and Bordetella parapertussis are the causative agents of whooping cough (21, 37), a disease that is endemic worldwide despite extensive vaccination efforts (48). Upon the introduction of B. pertussis vaccines more than 50 years ago, the incidence of whooping cough greatly declined in developed countries (9). However, the incidence of this disease has been steadily rising over the past 10 to 20 years in populations despite excellent vaccine coverage (7, 8, 12, 49, 56).

Although disease caused by B. parapertussis may be less severe than that caused by B. pertussis, infections by these bacteria are clinically indistinguishable except for an acute lymphocytosis that is observed only upon infection of naive hosts by B. pertussis (3, 21, 36, 41). Differentiating between these pathogens requires costly laboratory work that does not affect treatment decisions (37), making differential diagnoses rare. When specifically investigated, B. parapertussis has been found to cause anywhere from less than 1 percent to greater than 95 percent of whooping cough cases (reviewed in reference 59). Thus, it is unclear how many of the estimated 50 million annual cases of whooping cough (10) may be due to B. parapertussis and whether or not this pathogen is contributing to the resurgence of this disease.

The coexistence of these two species in human populations creates a paradox to the ecological theory that two closely related immunizing pathogens cannot occupy the same host population if immunity is cross-protective (1, 14, 16, 30). Because B. pertussis and B. parapertussis evolved from a common progenitor, Bordetella bronchiseptica, and share the majority of their virulence factors (42), cross-immunity could be expected. Furthermore, genetic comparisons of B. pertussis and B. parapertussis to B. bronchiseptica suggest that B. parapertussis emerged as a pathogen more recently than B. pertussis (28, 35, 42). Since the mean age at the time of primary B. pertussis infection for unvaccinated children is less than 5 years (9) and infection-induced immunity is protective 5 to 10 years (60), it is likely that B. parapertussis successfully emerged in populations that had some level of immunity to B. pertussis (4).

Although adapted to humans, both organisms efficiently colonize and rapidly grow throughout the murine respiratory tract but are ultimately eliminated from the lower respiratory tract by B- and T-cell-dependent immunity (24, 25, 27, 62). Watanabe and Nagai suggested that an immune response induced by one species conferred efficient protection against infections by either species (58). However, bacterial numbers were quantified only 2 weeks postinoculation, when bacterial numbers were significantly reduced even in the respiratory tracts of naive hosts (25, 58). Additionally, the strain used in that study (18-323) is quite distinct from other B. pertussis strains, as shown by multilocus enzyme electrophoresis and multilocus sequence typing (13, 55), and may actually be more closely related to B. bronchiseptica and B. parapertussis strains than other B. pertussis strains (2, 40). These problems, along with clinical studies showing that B. pertussis vaccines poorly protect against B. parapertussis and that these two organisms coexist in the same populations, led us to investigate cross-protective immunity between these two pathogens.

We sought to determine whether or not B. pertussis and B. parapertussis induced effective cross-immunity using the sequenced prototype strains Tohama I (B. pertussis) (42) and 12822 (B. parapertussis) (42), which are indistinguishable from current isolates by multilocus enzyme electrophoresis and multilocus sequence typing (13, 55). B. parapertussis-induced immunity protected against both species; however, B. pertussis-induced immunity protected only against B. pertussis infection. The asymmetrical cross-immunity appeared to be the result of inefficient binding of B. pertussis-induced antibodies to B. parapertussis. O antigen prevented B. pertussis-induced antibodies from binding to B. parapertussis in vitro and from clearing the bacteria from the respiratory tract in vivo. Together these data provide a molecular basis for the ability of B. parapertussis to avoid B. pertussis-induced immunity, an ability that may be critical to the coexistence of these pathogens in human populations.

MATERIALS AND METHODS

Bacterial strains and growth.

B. pertussis strains 536 and 536NaI (from Duncan Maskell, University of Cambridge, United Kingdom) are streptomycin- and nalidixic acid-resistant derivatives of Tohama I, respectively (51). B. parapertussis strain 12822 has been described previously (20), and 12822G is a gentamicin-resistant derivative of the parent strain (62). B. parapertussis strains CN2591 (wild type) and CN2591Δwbm (O antigen deficient) have previously been described (44). All were maintained on Bordet-Gengou agar (Difco) containing 10% defibrinated sheep blood (Hema Resources) and appropriate antibiotics. Liquid-culture bacteria were grown at 37°C overnight on a roller drum to mid-log phase in Stainer-Scholte broth.

Animal experiments.

C57BL/6 mice were obtained from Jackson Laboratories (Bar Harbor, ME) and bred in our Bordetella-free, specific-pathogen-free breeding rooms at The Pennsylvania State University. Four- to 6-week-old mice were sedated with 5% isoflurane (Abbott Laboratories) in oxygen and inoculated by pipetting 50 μl of phosphate-buffered saline (PBS) containing 5 × 105 CFU onto the external nares (25). For challenge experiments, mice were inoculated with 5 × 105 CFU of antibiotic-resistant strains at 28 or 70 days postinoculation. For passive transfer of immune serum, 200 μl of sera from naive or convalescent mice (collected 28 days postinoculation) was intraperitoneally injected just prior to inoculation. All protocols were reviewed by the university IACUC, and all animals were handled in accordance with institutional guidelines.

Bacterial quantification.

Mice were sacrificed immediately after inoculation or 3 days postinoculation to quantify bacteria in the nasal cavities, tracheae, and lungs. Tissues were homogenized in PBS and plated at specific dilutions onto Bordet-Gengou agar containing appropriate antibiotics, and colonies were counted 4 days later. The lower limit of detection was 10 CFU (the lower limit of the y axes).

Splenocyte restimulations.

Splenocytes were isolated by homogenizing spleens through a wire sieve, spinning at 1,500 rpm for 5 min at 4°C, lysing the red blood cells, and then washing the cells with Dulbecco's modified Eagle cell culture medium (HyClone, Logan, UT). Cells (2 × 106) were resuspended in Dulbecco's modified Eagle medium supplemented with 10% fetal calf serum (HyClone), 1 mM sodium pyruvate (HyClone), 100 μg/ml penicillin and streptomycin (HyClone), and 0.005% β-mercaptoethanol and placed into wells of a 96-well plate. Splenocytes were stimulated with medium or 107 heat-killed B. pertussis or B. parapertussis cells. After 3 days, the supernatant was collected and analyzed for gamma interferon (IFN-γ) and interleukin-10 production as specified by the manufacturers of enzyme-linked immunosorbent assay (ELISA) kits (R & D Systems, Minneapolis, MN).

ELISAs.

Bacteria were grown to an optical density of 0.7, heat inactivated, diluted in carbonate buffer, and used to coat 96-well plates. Plates were stored for 1 to 4 weeks at 4°C (wells filled with PBS containing Tween 20 plus 1% bovine serum albumin) before use. For ELISAs using live bacteria, bacteria were grown to an optical density of 0.7, diluted in carbonate buffer, added to each well of a 96-well plate, and incubated for 1 h at 37°C prior to use. A 1:25 dilution (for heat-inactivated bacteria) or a 1:10 dilution (for live bacteria) of serum samples was added to the first wells and serially diluted across the plates. Plates were incubated for 2 h at 37°C in a humidified chamber and washed, and goat anti-mouse (immunoglobulin heavy plus light chains) horseradish peroxidase-conjugated antibodies (Southern Biotech, Birmingham, AL) were added. Plates were incubated at 37°C in a humidified chamber for 1 h and washed, and 2,2′-azino-bis(3-ethylbenz-thiazoline-6-sulfonic acid) in a phospho-citrate buffer and hydrogen peroxide were added to wells, which were incubated at room temperature in the dark for 30 min and read on a plate reader at 405 nm.

Western blots.

Westerns blotting was performed on lysates containing 1 × 108 CFU of heat-killed B. pertussis or B. parapertussis. Lysates were run on 7% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels in denaturing conditions. Membranes were probed with serum from B. pertussis- or B. parapertussis-infected mice at a 1:100 dilution and goat anti-mouse (immunoglobulin heavy plus light chains) horseradish peroxidase-conjugated antibody (Southern Biotech, Birmingham, AL) (at a dilution of 1:15,000) as the detector antibody. The membrane was visualized with ECL Western blotting detection reagents (Amersham Biosciences, Piscataway, NJ).

Statistical analysis.

The means ± standard deviations (error bars) were determined for CFU, IFN-γ production, and antibody titers. Two-tailed, unpaired Student t tests were used to determine statistical significance between groups. Results were also analyzed by nonparametric Mann-Whitney tests with similar significance. All experiments were performed at least twice with similar results.

RESULTS

Cross-immunity between B. pertussis and B. parapertussis is asymmetrical.

The coexistence of B. pertussis and B. parapertussis in humans led us to assess the ability of these pathogens to induce cross-protective immunity. Although considered human-adapted pathogens, both species efficiently colonize the murine respiratory tract (17, 18) and induce sterilizing protective immunity (25), allowing us to test cross-protection in a mouse model. Mice were left uninfected or inoculated with B. pertussis or B. parapertussis. Twenty-eight days later, when strong immune responses had been generated and bacteria had been reduced to less than 1,000 CFU/lung (25, 62), these mice were challenged with antibiotic-resistant strains of B. pertussis or B. parapertussis and sacrificed 3 days later. In comparison to naive mice, mice convalescing from B. pertussis infection (B. pertussis-convalescent mice) harbored antibiotic-resistant B. pertussis numbers that were approximately 20-fold lower in the nasal cavity, 1,000-fold lower in the trachea, and more than 10,000-fold lower in the lungs at 3 days postchallenge (Fig. 1A to C). An immune response to B. parapertussis also conferred effective protection against B. pertussis, as bacterial numbers were approximately 100-fold lower in the nasal cavities, 10,000-fold lower in the tracheae, and 100,000-fold lower in the lungs of B. parapertussis-convalescent mice than in those of naive mice (Fig. 1A to C). Groups of mice were also sacrificed 10 min after inoculation to determine the amount of bacteria delivered by the inoculation, allowing us to quantify the growth or reduction in bacterial numbers from the initial amount deposited in the respiratory tract to the numbers 3 days postinoculation. B. pertussis increased in numbers in the nasal cavities (+100.8 CFU), tracheae (+100.3 CFU), and lungs (+101.6 CFU) of naive mice (Fig. 1D to F). Numbers were reduced in the nasal cavities, tracheae, and lungs of B. pertussis-immune (−100.2, −103.0, and −102.2 CFU, respectively) and B. parapertussis-immune (−100.3, −102.3, and −103.4 CFU, respectively) mice (Fig. 1D to F).

FIG. 1.

FIG. 1.

Open in a new tab

B. pertussis colonization of naive and immunized mice. C57BL/6 mice were inoculated with 5 × 105 CFU of B. pertussis (Bp) or B. parapertussis (Bpp). Immunized and naive mice were challenged 28 days later with 5 × 105 CFU of a nalidixic acid-resistant strain of B. pertussis and sacrificed 3 days after secondary inoculation for the quantification of bacterial numbers in the nasal cavities (A), tracheae (B), and lungs (C). Values are expressed as the log10 means ± standard deviations (SD). For day 3 colonization levels, the difference between each individual value and the mean log10 CFU approximately 10 min after inoculation in the nasal cavities (D), tracheae (E), and lungs (F) was calculated. These data are represented as the mean changes in log10 CFU ± SD. *, P < 0.05; **, P < 0.01 (compared to naive mice).

Upon investigating the ability of B. parapertussis to colonize immunized hosts, it was found that bacterial numbers were approximately 100-fold lower in the nasal cavities and 100,000-fold lower in the tracheae and lungs of B. parapertussis-convalescent mice than in those of naive mice (Fig. 2A to C). However, B. parapertussis numbers were only 5-fold lower in the nasal cavities and approximately 20-fold lower in the tracheae and lungs (Fig. 2A to C) of B. pertussis-convalescent mice than in those of naive mice. B. parapertussis numbers grew over the first 3 days in the nasal cavities (+101.8 CFU), tracheae (+102.2 CFU), and lungs (+101.9 CFU) of naive mice (Fig. 2D to F). Bacterial numbers were reduced in the nasal cavities, tracheae, and lungs of B. parapertussis-immune mice (−100.4, −102.5, and −103.3 CFU, respectively) but increased in the nasal cavities, tracheae, and lungs of B. pertussis-immune mice (+101.7, +101.5, +100.8 CFU, respectively) (Fig. 2D to F). Similar results were obtained when animals were rechallenged on day 70 postinoculation (data not shown), after the primary infection had been completely cleared from the lower respiratory tract. Furthermore, and consistent with previous reports, more efficient, but still incomplete, cross-protection was observed when bacterial numbers were analyzed 14 days postchallenge (data not shown). That being said, B. parapertussis was able to colonize hosts that had previously been infected with, and generated an effective immune response against, B. pertussis.

FIG. 2.

FIG. 2.

Open in a new tab

B. parapertussis colonization of naive and immunized mice. C57BL/6 mice were inoculated with 5 × 105 CFU of B. pertussis (Bp) or B. parapertussis (Bpp). Immunized and naive mice were challenged 28 days later with 5 × 105 CFU of a gentamicin-resistant strain of B. parapertussis and sacrificed 3 days after secondary inoculation for the quantification of bacterial numbers in the nasal cavities (A), tracheae (B), and lungs (C). Values are expressed as log10 means ± standard deviations (SD). For day 3 colonization levels, the difference between each individual value and the mean log10 CFU approximately 10 min after inoculation in the nasal cavities (D), tracheae (E), and lungs (F) was calculated. These data are represented as the mean changes in log10 CFU ± SD. *, P < 0.05; **, P < 0.01.

T-cell, but not antibody, responses are cross-reactive.

The ability of B. parapertussis to infect B. pertussis-immune hosts suggested that the immune response was not effective against B. parapertussis, leading us to examine the cross-reactivity of T-cell and antibody responses. Mice were left uninoculated or inoculated with B. pertussis or B. parapertussis, and their spleens were excised 28 days postinoculation. The response of splenocytes to heat-killed B. pertussis or B. parapertussis was assessed by measuring the amount of IFN-γ produced, since this cytokine has been implicated in protective immunity to both bacteria (31, 32; D. N. Wolfe and E. T. Harvill, unpublished data). Splenocytes from naive and B. parapertussis-infected animals produced small amounts of IFN-γ in response to either species (Fig. 3), although the small amounts of IFN-γ produced by splenocytes from B. parapertussis-infected animals appeared to be protective (Fig. 1 and 2). Splenocytes from B. pertussis-infected mice responded similarly to both species by producing large amounts of IFN-γ (Fig. 3). Similar levels of IL-10 were produced in response to B. pertussis and B. parapertussis, regardless of the primary infection (data not shown). These data indicate that T cells induced by a B. pertussis infection are able to respond to B. parapertussis antigens. Furthermore, IFN-γ can contribute to the recruitment of neutrophils (52), which are important to the clearance of both pathogens (24, 62). Therefore, we quantified the numbers of neutrophils in the lungs of mice after a secondary infection with either species. Approximately 105 neutrophils were observed in the lungs of B. pertussis-convalescent mice upon B. parapertussis challenge compared to only 1 × 104 to 2 × 104 neutrophils in the lungs of naive or B. parapertussis-convalescent mice (data not shown), suggesting that these cells were being recruited to the lungs but were not effectively eliminating B. parapertussis.

FIG. 3.

FIG. 3.

Open in a new tab

IFN-γ production by splenocytes from naive or B. pertussis- or B. parapertussis-immune hosts. C57BL/6 mice were inoculated with 5 × 105 CFU of B. pertussis (Bp) or B. parapertussis (Bpp). Immunized and naive mice were sacrificed 28 days postinoculation, spleens were excised, and splenocytes were exposed to media alone (hatched bars), heat-killed B. pertussis (black bars), or heat-killed B. parapertussis (white bars) for 3 days. The supernatant was collected, and cytokine ELISAs were performed to quantify the levels of IFN-γ produced by splenocytes. Values are expressed as the means ± standard deviations. *, P < 0.05; **, P < 0.01.

To determine the ability of antibodies induced by one species to recognize the other, sera from B. pertussis- and B. parapertussis-convalescent mice were analyzed by ELISA and Western blotting. B. pertussis was recognized by sera raised against either species, although titers were approximately five times higher in B. pertussis-induced sera (titer ∼ 1,000) than in B. parapertussis-induced sera (titer ∼ 200) (Fig. 4A). B. parapertussis was recognized by B. parapertussis-induced antibodies (titer ∼ 7,500), but titers in B. pertussis-induced sera were approximately 150-fold lower and near the lower limit of detection (titer ∼ 50) (Fig. 4B), suggesting that the antibodies did not efficiently bind to B. parapertussis-coated plates. Western blots showed that antibodies induced by B. pertussis infection recognized antigens of both B. pertussis and B. parapertussis (Fig. 4C). The recognition of denatured, isolated B. parapertussis proteins in this assay is in contrast to the lack of recognition of whole bacteria in the ELISA (Fig. 4B), suggesting that shared protein antigens may not be exposed on the surface. B. parapertussis-induced antibodies also bound protein antigens of both species, as well as a large region of between 17 and 36 kDa present in B. parapertussis but not B. pertussis. This region was absent from an isogenic mutant of B. parapertussis lacking O antigen, and B. pertussis does not express an O antigen, indicating that this smear represents lipopolysaccharide containing O antigen (Fig. 4D). These data indicate that antibody responses to B. pertussis and B. parapertussis differ in the recognition of O antigen.

FIG. 4.

FIG. 4.

Open in a new tab

Recognition of B. pertussis and B. parapertussis by antibodies from convalescent-phase serum. C57BL/6 mice were inoculated with 5 × 105 CFU of B. pertussis (Bp) or B. parapertussis (Bpp), and serum was collected from these mice 28 days later. ELISA was performed on serum from naive (NS), B. pertussis-infected (Bp IS), and B. parapertussis-infected (Bpp IS) mice to quantify titers of antibodies specific for B. pertussis (A) or B. parapertussis (B). The dashed line represents the lower limit of detection. Antibody titers from the ELISA are quantified as the log10 means of the end point titers ± standard deviations. *, P < 0.05; **, P < 0.01. Western blots of pooled B. pertussis-induced (C) and pooled B. parapertussis-induced (D) serum were used to determine if serum raised against one species recognized specific antigens of B. pertussis, B. parapertussis, or the B. parapertussis Δwbm strain (BppΔO-ag).

Transfer of B. parapertussis-specific antibodies to B. pertussis-immune mice inhibits colonization by B. parapertussis.

If B. parapertussis escapes cross-immunity by avoiding antibody recognition, adding B. parapertussis immune serum should confer protection to B. pertussis-immune mice. Mice were inoculated with B. pertussis and 28 days later were given a passive transfer of naive serum, B. pertussis-immune serum, or B. parapertussis-immune serum and challenged with B. parapertussis. Transferring naive or B. pertussis-induced serum had no measurable effect on B. parapertussis numbers, but transferring B. parapertussis immune serum to B. pertussis-immune hosts resulted in an approximately 100-fold reduction in bacterial numbers (Fig. 5). These data support the hypothesis that the lack of recognition by B. pertussis-induced antibodies allows B. parapertussis to avoid the immune response induced by its sister species.

FIG. 5.

FIG. 5.

Open in a new tab

Effect of passive transfer of immune serum to B. pertussis-immune mice on B. parapertussis numbers. C57BL/6 mice were inoculated with 5 × 105 CFU of B. pertussis. Twenty-eight days later, these mice were passively transferred sera from naive (naive serum), B. pertussis-infected (Bp serum), or B. parapertussis-infected (Bpp serum) mice and then challenged with B. parapertussis. B. parapertussis-convalescent mice (Bpp immune) were also challenged with B. parapertussis for comparison. CFU were quantified 3 days after secondary inoculation. Numbers of bacteria are expressed as log10 means ± standard deviations. *, P < 0.05; **, P < 0.01.

O antigen inhibits the binding of B. pertussis-induced antibodies to B. parapertussis.

We next sought to determine how B. parapertussis avoids recognition by antibodies induced by such a closely related pathogen. B. parapertussis, but not B. pertussis, expresses O antigen (5), which has been shown to prevent the recognition of surface protein antigens by antibodies in other bacterial models (39, 53, 54, 57). We hypothesized that O antigen blocked the binding of B. pertussis-induced antibodies to common antigens. Live bacteria were used in ELISAs to determine if B. pertussis-induced antibodies bound more efficiently to intact B. parapertussis in the absence of O antigen. Sera from both B. pertussis- and B. parapertussis-infected mice recognized B. pertussis (titers ∼ 1,000 and 600, respectively) (Fig. 6A). Titers of B. pertussis-induced antibodies that bound to live B. parapertussis were not significantly higher than the lower limit of detection, but titers of antibodies binding to live B. parapertussis lacking O antigen were approximately 15-fold higher (Fig. 6B). Thus, O antigen inhibits the binding of B. pertussis-induced antibodies to the surface of B. parapertussis.

FIG. 6.

FIG. 6.

Open in a new tab

Antibody recognition of live B. pertussis and B. parapertussis and O-antigen-deficient B. parapertussis by immune serum. C57BL/6 mice were inoculated with 5 × 105 CFU of B. pertussis (Bp) or B. parapertussis (Bpp), and serum was collected 28 days later. Sera from B. pertussis (Bp serum)- or B. parapertussis-infected mice (Bpp serum) or naive sera were examined by ELISA for their ability to bind to live B. pertussis (A) or B. parapertussis (B; white bars) or the O-antigen-deficient B. parapertussis (BppΔO-ag) (B; hatched bars). The dashed line represents the lower limit of detection. Antibody titers are expressed as the end point titers ± standard deviations. **, P < 0.01.

O antigen enables B. parapertussis to colonize B. pertussis-immune hosts.

If the blocking of antibody binding by O antigen allowed the evasion of B. pertussis-induced immunity, it would be expected that B. pertussis immunity would confer protection against O-antigen-deficient B. parapertussis. Mice were left uninfected or inoculated with B. pertussis or B. parapertussis and 28 days later challenged with wild-type or O-antigen-deficient B. parapertussis. Again, immunity to B. pertussis did not protect against wild-type B. parapertussis (Fig. 2 and 7). The O-antigen-deficient strain of B. parapertussis was present at approximately 106, 106, and 105 CFU in the nasal cavities, tracheae, and lungs of naive mice, respectively, but bacterial numbers were approximately 10-fold lower in the nasal cavities and 100,000-fold lower in the tracheae and lungs of B. pertussis-convalescent mice (Fig. 7). Thus, O antigen is required for B. parapertussis to evade B. pertussis-induced immunity and colonize B. pertussis-immune hosts.

FIG. 7.

FIG. 7.

Open in a new tab

Ability of the O-antigen-deficient strain of B. parapertussis to colonize B. pertussis-immune hosts. C57BL/6 mice were inoculated with 5 × 105 CFU of B. pertussis (Bp) or B. parapertussis (Bpp). Immunized and naive mice were challenged 28 days later with 5 × 105 CFU of B. parapertussis (white bars) or the O-antigen-deficient strain of B. parapertussis (BppΔO-ag; hatched bars). Mice were sacrificed 3 days after secondary inoculation for the quantification of bacterial numbers in the nasal cavities (A), tracheae (B), and lungs (C). All values are expressed as log10 means ± standard deviations. *, P < 0.05; **, P < 0.01.

DISCUSSION

B. parapertussis likely emerged into populations in which B. pertussis was endemic (4), making the ability to evade B. pertussis-induced immunity important to invading these populations. Here we have shown that B. parapertussis evades immunity to B. pertussis but that the evasion of cross-immunity is asymmetric. O antigen, which has been shown to mask surface antigens in other systems (39, 53, 54, 57), inhibits the binding of B. pertussis-induced antibodies and allows B. parapertussis to colonize B. pertussis-immune hosts. The observed asymmetric cross-immunity between B. pertussis and B. parapertussis is likely to contribute to their coexistence as endemic human pathogens.

Previous work has suggested efficient cross-immunity between these two species (58), but that study used a B. pertussis strain that, it has been suggested, is more similar to B. bronchiseptica and B. parapertussis than other B. pertussis strains (13, 55). More recent work has suggested that infection by an attenuated strain of B. pertussis may protect against subsequent B. parapertussis infections (38), but that study also observed that B. parapertussis numbers did not grow even in naive mice, whereas multiple studies have shown that B. parapertussis grows rapidly in naive mice (18, 25, 62). The attenuated strain used was also defective in producing pertussis toxin (38), a toxin that inhibits antibody production (6). Since low titers of B. parapertussis-specific antibodies in the serum of B. pertussis convalescent mice enabled the evasion of cross-immunity, pertussis toxin-mediated inhibition of antibody production may facilitate the evasion of immunity. It is important to point out that these two studies are in contrast to the voluminous research on B. pertussis vaccines, which show greatly diminished protection against B. parapertussis relative to protection against B. pertussis (11, 19, 29, 50, 61).

B. parapertussis O antigen limits the ability of antibodies from B. pertussis-immunized hosts to recognize surface antigens that are common between these two pathogens. An additional contributor to the asymmetrical cross-immunity could be that B. pertussis infection does not induce an antibody response to the appropriate protective antigen(s). B. pertussis does not express O antigen and thus does not induce O-antigen-specific antibodies. O antigen is a key protective antigen for a number of bacteria (23, 46), and antibodies against it may be important to protection against B. parapertussis. While O-antigen variation contributes to the coexistence of related bacterial species (43), we propose that O antigen is not essential to the infection of humans by Bordetella, as evidenced by its absence from B. pertussis strains, but the maintenance of O antigen by B. parapertussis may have allowed it to emerge in the niche of a B. pertussis-endemic population.

Host immunity has profound effects on the persistence of pathogens within host populations (15, 45, 47). Cross-reactive immunity between two species causes immune-mediated competition that may ultimately displace one from the host population. Despite the apparent asymmetrical cross-immunity between B. pertussis and B. parapertussis, these two species appear to circulate in out-of-phase cycles (26), and positive antibody titers to each have been observed in more than 50 percent of a population (33, 34). The question is, If there is asymmetrical cross-immunity between B. pertussis and B. parapertussis, why has the latter not displaced the former from human populations?

Whooping cough caused by B. parapertussis may be milder (3, 36); thus, assuming less coughing translates to less transmission, it is possible that B. parapertussis transmits less effectively than B. pertussis. This theory is supported by seroprevalence data indicating that approximately 60 percent of a population was seropositive for B. parapertussis but that over 90 percent of that population was seropositive for B. pertussis (33, 34). Furthermore, immunity induced by B. pertussis wanes after 5 to 10 years (60). The decay of immunity to B. parapertussis is not known, but cross-reacting immunity is likely to decay more rapidly than autologous immunity, potentially allowing subsequent infections by different strains or species (22).

The data presented here are evidence of a novel pattern of asymmetric cross-immunity conferred by B. pertussis and B. parapertussis that is consistent with the poor cross-protection conferred by vaccines (11, 19, 29, 50, 61) and that would affect the circulation of these pathogens throughout their host population. The evasion of B. pertussis-induced immunity by B. parapertussis may have allowed the latter to invade a population in which the former was endemic, leading to the observed coexistence. Importantly, it is possible that vaccination has changed, or is changing, the immune-mediated interactions between these two organisms. The effects of vaccines on B. parapertussis are unclear, but some have suggested that the prevalence of this species may be increasing despite, or even because of, vaccination efforts (29). In the face of these epidemiological and immunological complexities, it is important to establish the relative prevalence of B. pertussis and B. parapertussis as etiological agents in the ongoing resurgence of whooping cough.

Acknowledgments

The authors have no conflicting financial interests, and this work has not been submitted elsewhere.

This study was supported by NIH grant AI 053075 (E.T.H.).

We thank Anne Buboltz for critical reading and discussion of the manuscript. We also thank Pamela Correll and Sandeep Prabhu for the use of materials and equipment.

Editor: D. L. Burns

Footnotes

▿

Published ahead of print on 13 August 2007.

REFERENCES

  • 1.Anderson, R. M. 1995. Evolutionary pressures in the spread and persistence of infectious agents in vertebrate populations. Parasitology 111(Suppl.):S15-S31. [DOI] [PubMed] [Google Scholar]
  • 2.Arico, B., R. Gross, J. Smida, and R. Rappuoli. 1987. Evolutionary relationships in the genus Bordetella. Mol. Microbiol. 1:301-308. [DOI] [PubMed] [Google Scholar]
  • 3.Bergfors, E., B. Trollfors, J. Taranger, T. Lagergard, V. Sundh, and G. Zackrisson. 1999. Parapertussis and pertussis: differences and similarities in incidence, clinical course, and antibody responses. Int. J. Infect. Dis. 3:140-146. [DOI] [PubMed] [Google Scholar]
  • 4.Bjornstad, O. N., and E. T. Harvill. 2005. Evolution and emergence of Bordetella in humans. Trends Microbiol. 13:355-359. [DOI] [PubMed] [Google Scholar]
  • 5.Burns, V. C., E. J. Pishko, A. Preston, D. J. Maskell, and E. T. Harvill. 2003. Role of Bordetella O antigen in respiratory tract infection. Infect. Immun. 71:86-94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Carbonetti, N. H., G. V. Artamonova, C. Andreason, E. Dudley, R. M. Mays, and Z. E. Worthington. 2004. Suppression of serum antibody responses by pertussis toxin after respiratory tract colonization by Bordetella pertussis and identification of an immunodominant lipoprotein. Infect. Immun. 72:3350-3358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Celentano, L. P., M. Massari, D. Paramatti, S. Salmaso, A. E. Tozzi, and the EUVAC-NET Group. 2005. Resurgence of pertussis in Europe. Pediatr. Infect. Dis. J. 24:761-765. [DOI] [PubMed] [Google Scholar]
  • 8.Centers for Disease Control and Prevention. 2002. Pertussis—United States, 1997-2000. JAMA 287:977-979. [PubMed] [Google Scholar]
  • 9.Cherry, J. D. 1984. The epidemiology of pertussis and pertussis immunization in the United Kingdom and the United States: a comparative study. Curr. Probl. Pediatr. 14:1-78. [DOI] [PubMed] [Google Scholar]
  • 10.Crowcroft, N. S., C. Stein, P. Duclos, and M. Birmingham. 2003. How best to estimate the global burden of pertussis? Lancet Infect. Dis. 3:413-418. [DOI] [PubMed] [Google Scholar]
  • 11.David, S., R. van Furth, and F. R. Mooi. 2004. Efficacies of whole cell and acellular pertussis vaccines against Bordetella parapertussis in a mouse model. Vaccine 22:1892-1898. [DOI] [PubMed] [Google Scholar]
  • 12.de Melker, H. E., J. F. Schellekens, S. E. Neppelenbroek, F. R. Mooi, H. C. Rumke, and M. A. Conyn-van Spaendonck. 2000. Reemergence of pertussis in the highly vaccinated population of The Netherlands: observations on surveillance data. Emerg. Infect. Dis. 6:348-357. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Diavatopoulos, D. A., C. A. Cummings, L. M. Schouls, M. M. Brinig, D. A. Relman, and F. R. Mooi. 2005. Bordetella pertussis, the causative agent of whooping cough, evolved from a distinct, human-associated lineage of B. bronchiseptica. PLoS Pathog. 1:e45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Ferguson, N. M., A. P. Galvani, and R. M. Bush. 2003. Ecological and immunological determinants of influenza evolution. Nature 422:428-433. [DOI] [PubMed] [Google Scholar]
  • 15.Gupta, S., M. C. Maiden, I. M. Feavers, S. Snee, R. M. May, and R. M. Anderson. 1996. The maintenance of strain structure in populations of recombining infectious agents. Nat. Med. 2:437-442. [DOI] [PubMed] [Google Scholar]
  • 16.Gupta, S., and M. C. J. Maiden. 2001. Exploring the evolution of diversity in pathogen populations. Trends Microbiol. 9:181-185. [DOI] [PubMed] [Google Scholar]
  • 17.Harvill, E. T., P. A. Cotter, and J. F. Miller. 1999. Pregenomic comparative analysis between Bordetella bronchiseptica RB50 and Bordetella pertussis Tohama I in murine models of respiratory tract infection. Infect. Immun. 67:6109-6118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Harvill, E. T., A. Preston, P. A. Cotter, A. G. Allen, D. J. Maskell, and J. F. Miller. 2000. Multiple roles for Bordetella lipopolysaccharide molecules during respiratory tract infection. Infect. Immun. 68:6720-6728. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Heininger, U., J. D. Cherry, K. Stehr, S. Schmitt-Grohe, M. Uberall, S. Laussucq, T. Eckhardt, M. Meyer, J. Gornbein, et al. 1998. Comparative efficacy of the Lederle/Takeda acellular pertussis component DTP (DTaP) vaccine and Lederle whole-cell component DTP vaccine in German children after household exposure. Pediatrics 102:546-553. [DOI] [PubMed] [Google Scholar]
  • 20.Heininger, U., P. A. Cotter, H. W. Fescemyer, G. Martinez de Tejada, M. H. Yuk, J. F. Miller, and E. T. Harvill. 2002. Comparative phenotypic analysis of the Bordetella parapertussis isolate chosen for genomic sequencing. Infect. Immun. 70:3777-3784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Heininger, U., K. Stehr, S. Schmitt-Grohe, C. Lorenz, R. Rost, P. D. Christenson, M. Uberall, and J. D. Cherry. 1994. Clinical characteristics of illness caused by Bordetella parapertussis compared with illness caused by Bordetella pertussis. Pediatr. Infect. Dis. J. 13:306-309. [DOI] [PubMed] [Google Scholar]
  • 22.Kawai, H., T. Aoyama, Y. Murase, C. Tamura, and A. Imaizumi. 1996. A causal relationship between Bordetella pertussis and Bordetella parapertussis infections. Scand. J. Infect. Dis. 28:377-381. [DOI] [PubMed] [Google Scholar]
  • 23.Kingsley, R. A., and A. J. Baumler. 2000. Host adaptation and the emergence of infectious disease: the Salmonella paradigm. Mol. Microbiol. 36:1006-1014. [DOI] [PubMed] [Google Scholar]
  • 24.Kirimanjeswara, G. S., L. Agosto, M. J. Kennett, O. N. Bjornstad, and E. T. Harvill. 2005. Pertussis toxin inhibits neutrophil recruitment to delay antibody-mediated clearance of Bordetella pertussis. J. Clin. Investig. 115:3594-3601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Kirimanjeswara, G. S., P. B. Mann, and E. T. Harvill. 2003. Role of antibodies in immunity to Bordetella infections. Infect. Immun. 71:1719-1724. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Lautrop, H. 1971. Epidemics of parapertussis. 20 years’ observations in Denmark. Lancet i:1195-1198. [DOI] [PubMed] [Google Scholar]
  • 27.Leef, M., K. L. Elkins, J. Barbic, and R. D. Shahin. 2000. Protective immunity to Bordetella pertussis requires both B cells and CD4+ T cells for key functions other than specific antibody production. J. Exp. Med. 191:1841-1852. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Letowska, I., and W. Hryniewicz. 2004. Epidemiology and characterization of Bordetella parapertussis strains isolated between 1995 and 2002 in and around Warsaw, Poland. Eur. J. Clin. Microbiol. Infect. Dis. 23:499-501. [DOI] [PubMed] [Google Scholar]
  • 29.Liese, J. G., C. Renner, S. Stojanov, B. H. Belohradsky, and the Munich Vaccine Study Group. 2003. Clinical and epidemiological picture of B. pertussis and B. parapertussis infections after introduction of acellular pertussis vaccines. Arch. Dis. Child. 88:684-687. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Lipsitch, M. 1999. Bacterial vaccines and serotype replacement: lessons from Haemophilus influenzae and prospects for Streptococcus pneumoniae. Emerg. Infect. Dis. 5:336-345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Mahon, B. P., B. J. Sheahan, F. Griffin, G. Murphy, and K. H. Mills. 1997. Atypical disease after Bordetella pertussis respiratory infection of mice with targeted disruptions of interferon-gamma receptor or immunoglobulin mu chain genes. J. Exp. Med. 186:1843-1851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Mahon, B. P., and K. H. Mills. 1999. Interferon-gamma mediated immune effector mechanisms against Bordetella pertussis. Immunol. Lett. 68:213-217. [DOI] [PubMed] [Google Scholar]
  • 33.Maixnerova, M. 2003. The 2001 serological survey in the Czech Republic—parapertussis. Cent. Eur. J. Public Health 11(Suppl.):S23-S24. [PubMed] [Google Scholar]
  • 34.Maixnerova, M. 2003. The 2001 serological survey in the Czech Republic—pertussis. Cent. Eur. J. Public Health 11(Suppl.):S17-S22. [PubMed] [Google Scholar]
  • 35.Makinen, J., J. Mertsola, H. Soini, H. Arvilomni, M. K. Viljanen, N. Guiso, and Q. He. 2003. PFGE and pertactin gene sequencing suggest limited genetic variability within the Finnish Bordetella parapertussis population. J. Med. Microbiol. 52:1059-1063. [DOI] [PubMed] [Google Scholar]
  • 36.Mastrantonio, P., P. Stefanelli, M. Giuliano, Y. Herrera Rojas, M. Ciofi degli Atti, A. Anemona, and A. E. Tozzi. 1998. Bordetella parapertussis infection in children: epidemiology, clinical symptoms, and molecular characteristics of isolates. J. Clin. Microbiol. 36:999-1002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Mattoo, S., and J. Cherry. 2005. Molecular pathogenesis, epidemiology, and clinical manifestations of respiratory infections due to Bordetella pertussis and other Bordetella subspecies. Clin. Microbiol. Rev. 18:326-382. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Mielcarek, N., A. S. Debrie, D. Raze, J. Bertout, C. Rouanet, A. B. Younes, C. Creusy, J. Engle, W. E. Goldman, and C. Locht. 2006. Live attenuated B. pertussis as a single-dose nasal vaccine against whooping cough. PLoS Pathog. 2:e65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Morona, R., and L. Van den Bosch. 2003. Lipopolysaccharide O antigen chains mask IcsA (VirG) in Shigella flexneri. FEMS Microbiol. Lett. 221:173-180. [DOI] [PubMed] [Google Scholar]
  • 40.Musser, J. M., E. L. Hewlett, M. S. Peppler, and R. K. Selander. 1986. Genetic diversity and relationships in populations of Bordetella spp. J. Bacteriol. 166:230-237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Olson, L. 1975. Pertussis. Medicine (Baltimore) 54:427-469. [DOI] [PubMed] [Google Scholar]
  • 42.Parkhill, J., M. Sebaihia, A. Preston, L. D. Murphy, N. Thomson, D. E. Harris, M. T. Holden, C. M. Churcher, S. D. Bentley, K. L. Mungall, A. M. Cerdeno-Tarraga, L. Temple, K. James, B. Harris, M. A. Quail, M. Achtman, R. Atkins, S. Baker, D. Basham, N. Bason, I. Cherevach, T. Chillingworth, M. Collins, A. Cronin, P. Davis, J. Doggett, T. Feltwell, A. Goble, N. Hamlin, H. Hauser, S. Holroyd, K. Jagels, S. Leather, S. Moule, H. Noberczak, S. O'Neil, D. Osmond, C. Price, E. Rabbinowitsch, S. Rutter, M. Sanders, D. Saunders, K. Seeger, S. Sharp, M. Simmonds, J. Skelton, R. Squares. S. Squares, K. Stevens, L. Unwin, S. Whitehead, B. G. Barrell, and D. J. Maskell. 2003. Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat. Genet. 35:32-40. [DOI] [PubMed] [Google Scholar]
  • 43.Pennington, J. E., G. J. Small, M. E. Lostrom, and G. B. Pier. 1986. Polyclonal and monoclonal antibody therapy for experimental Pseudomonas aeruginosa pneumonia. Infect. Immun. 54:239-244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Preston, A., A. G. Allen, J. Cadisch, R. Thomas, K. Stevens, C. M. Churcher, K. L. Badcock, J. Parkhill, B. Barrell, and D. J. Maskell. 1999. Genetic basis for lipopolysaccharide O-antigen biosynthesis in bordetellae. Infect. Immun. 67:3763-3767. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Restif, O., and B. T. Grenfell. 2006. Integrating life history and cross-immunity into the evolutionary dynamics of pathogens. Proc. R. Soc. B 273:409-416. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Robbins, J. B., C. Chu, and R. Schneerson. 1992. Hypothesis for vaccine development: protective immunity to enteric diseases caused by nontyphoidal salmonellae and shigellae may be conferred by serum IgG antibodies to the O-specific polysaccharide of their lipopolysaccharides. Clin. Infect. Dis. 15:346-361. [DOI] [PubMed] [Google Scholar]
  • 47.Rohani, P., D. J. Earn, and B. T. Grenfell. 1999. Opposite patterns of synchrony in sympatric disease metapopulations. Science 286:968-971. [DOI] [PubMed] [Google Scholar]
  • 48.Rohani, P., C. J. Green, N. B. Mantilla-Beniers, and B. T. Grenfell. 2003. Ecological interference between fatal diseases. Nature 422:885-888. [DOI] [PubMed] [Google Scholar]
  • 49.Skowronski, D. M., G. De Serres, D. MacDonald, W. Wu, C. Shaw, J. Macnabb, S. Champagne, D. M. Patrick, and S. A. Halperin. 2002. The changing age and seasonal profile of pertussis in Canada. J. Infect. Dis. 185:1448-1453. [DOI] [PubMed] [Google Scholar]
  • 50.Stehr, K., J. Cherry, U. Heininger, S. Schmitt-Grohe, M. Uberall, S. Laussucq, T. Eckhardt, M. Meyer, R. Englehardt, and P. Christenson. 1998. A comparative efficacy trial in Germany in infants who received either the Lederle/Takeda acellular pertussis component DTP (DTaP) vaccine, the Lederle whole-cell component DTP vaccine, or DT vaccine. Pediatrics 101:1-11. [DOI] [PubMed] [Google Scholar]
  • 51.Stibitz, S., and M. S. Yang. 1991. Subcellular localization and immunological detection of proteins encoded by the vir locus of Bordetella pertussis. J. Bacteriol. 173:4288-4296. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Sun, K., S. L. Salmon, S. A. Lotz, and D. W. Metzger. 2007. Interleukin-12 promotes gamma interferon-dependent neutrophil recruitment in the lung and improves protection against respiratory Streptococcus pneumoniae infection. Infect. Immun. 75:1196-1202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.van der Ley, P., P. de Graaff, and J. Tommassen. 1986. Shielding of Escherichia coli outer membrane proteins as receptors for bacteriophages and colicins by O-antigenic chains of lipopolysaccharide. J. Bacteriol. 168:449-451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.van der Ley, P., O. Kuipers, J. Tommassen, and B. Lugtenberg. 1986. O-antigenic chains of lipopolysaccharide prevent binding of antibody molecules to an outer membrane pore protein in Enterobacteriaceae. Microb. Pathog. 1:43-49. [DOI] [PubMed] [Google Scholar]
  • 55.van der Zee, A., F. Mooi, J. Van Embden, and J. Musser. 1997. Molecular evolution and host adaptation of Bordetella spp.: phylogenetic analysis using multilocus enzyme electrophoresis and typing with three insertion sequences. J. Bacteriol. 179:6609-6617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.von Konig, C. H., S. Halperin, M. Riffelmann, and N. Guiso. 2002. Pertussis of adults and infants. Lancet Infect. Dis. 2:744-750. [DOI] [PubMed] [Google Scholar]
  • 57.Voorhis, D. L., S. Dillon, S. B. Formal, and R. R. Isberg. 1991. An O antigen can interfere with the function of the Yersinia pseudotuberculosis invasin protein. Mol. Microbiol. 5:317-325. [DOI] [PubMed] [Google Scholar]
  • 58.Watanabe, M., and M. Nagai. 2001. Reciprocal protective immunity against Bordetella pertussis and Bordetella parapertussis in a murine model of respiratory infection. Infect. Immun. 69:6981-6986. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Watanabe, M., and M. Nagai. 2004. Whooping cough due to Bordetella parapertussis: an unresolved problem. Expert Rev. Anti-Infect. Ther. 2:447-454. [DOI] [PubMed] [Google Scholar]
  • 60.Wendelboe, A. M., A. Van Rie, S. Salmaso, and J. A. Englund. 2005. Duration of immunity against pertussis after natural infection or vaccination. Pediatr. Infect. Dis. J. 24(5S):S58-S61. [DOI] [PubMed] [Google Scholar]
  • 61.Willems, R. J., J. Kamerbeek, C. A. Geuijen, J. Top, H. Gielen, W. Gaastra, and F. R. Mooi. 1998. The efficacy of a whole cell pertussis vaccine and fimbriae against Bordetella pertussis and Bordetella parapertussis infections in a respiratory mouse model. Vaccine 16:410-416. [DOI] [PubMed] [Google Scholar]
  • 62.Wolfe, D. N., G. Kirimanjeswara, and E. T. Harvill. 2005. Clearance of Bordetella parapertussis from the lower respiratory tract requires humoral and cellular immunity. Infect. Immun. 73:6508-6513. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES