Gonococcal Cervicitis: A Role for Biofilm in Pathogenesis

Abstract

Neisseria gonorrhoeae forms a biofilm in flow cells on glass coverslips as well as on primary cervical epithelial cells. Electron microscopic studies of cervical biopsy specimens from 10 patients with culture-proven N. gonorrhoeae infection revealed evidence of biofilm formation in 3 of the biopsy specimens. These biofilms showed gonococci in networks of bacterial membrane within the biofilm structure. This finding was also observed in biofilms formed over glass cover slips and after infection of primary cervical tissue in vitro. The importance of membranous networks in Neisseria biofilm formation was demonstrated with N. gonorrhoeae strain 1291-msbB, which shows a markedly decreased ability to bleb. This mutant formed significantly less biofilm over glass surfaces and cervical epithelial cells, and complementation showed reversion to wild-type biofilms. Gonoccal biofilms, as part of the cervical infection, may be involved in the mechanisms by which asymptomatic infections, persistence, and increased antibiotic resistance occur.

Neisseria gonorrhoeae, the etiologic agent of the sexually transmitted disease gonorrhea, is a gram-negative diplococcus bacterium. Humans are the only known host of N. gonorrhoeae, with the majority of infections occurring in the urogenital tract [1]. N. gonorrhoeae has been shown previously to be capable of forming biofilms on glass surfaces in continuous-flow chambers as well as on cultured primary urethral epithelial cells and cervical epithelial cells in vitro [2]. On microscopic examination, N. gonorrhoeae biofilms appeared similar in basic structure to the biofilms of other bacteria. Bacteria were embedded in a continuous matrix, and water channels could be seen throughout the biofilm. A unique feature observed in N. gonorrhoeae biofilms was the presence of what appeared to be membranelike material throughout the structure. On the basis of findings of lectin and antibody-binding studies, it appeared that these membranes could be the result of blebbing of the gonococcal outer membrane as well as an accumulation of membranes from dead organisms [2]. Although we have established the ability of N. gonorrhoeae to form a biofilm on human urogenital tract cells in vitro, the ability of N. gonorrhoeae to form a biofilm in its human host has not been demonstrated. To evaluate this possibility, we studied archival cervical biopsy specimens from patients with culture-proven N. gonorrhoeae infection [3]. Using electron microscopy to analyze the biopsy specimens, we observed evidence of biofilm formation by N. gonorrhoeae on the surface of cervical epithelia obtained from these patient samples. The structure of this biofilm closely resembles that observed in flow chambers over human cells, and electron microscopic analyses of gonococcal biofilms grown on cultured cervical epithelial cells indicate that surface blebbing of the gonococcus is extensive.

To substantiate further the role played by membrane formation in gonococcal biofilms, we studied the ability of a N. gonorrhoeae mutant, strain 1291-msbB, to form biofilms in vitro. Other studies with meningococci have shown that this mutation results in a phenotype with reduced membrane blebbing [4]. MsbB is an acyltransferase involved in the biosynthesis of the lipid A portion of lipooligosaccharide (LOS). The lipid A portion of the LOS from the N. gonorrhoeae strain 1291-msbB mutant is pentaacylated, as opposed to the hexaacyl lipid A found in wild-type 1291 [5]. Transmission electron microscopic studies of N. gonorrhoeae strains 1291 and 1291-msbB grown in broth culture show a marked reduction in outer membrane blebbing by 1291-msbB. Biofilm experiments show that strain 1291 forms a significantly thicker and denser biofilm than does the msbB mutant after 48 and 96 h of growth on cervical epithelial cells. These findings indicate that biofilms play a role in cervical gonococcal infection and further implicate outer membrane blebbing as a major constituent of Neisseria biofilm formation.

METHODS

Bacteria and culture conditions

N. gonorrhoeae strain 1291 is a clinical isolate that has been described elsewhere [6], and strain 1291-msbB is a late acyltransferase mutant whose characteristics have been reported elsewhere [5]. The strains were reconstituted from frozen stock cultures and grown at 37°C in 5% CO₂ on GC agar plates supplemented with 1% IsoVitaleX (BBL). Strains 1291 and 1291-msbB expressing green fluorescent protein (GFP) were made by transforming the bacteria with pGFP (pLES98 containing GFP). This plasmid was a gift from V. Clark (University of Rochester).

Cervical biopsy specimens and transmission electron microscopic analysis

Ten cervical biopsy specimens embedded in separate Epon blocks were a gift from B. Evans (Charing Cross Hospital, London). The biopsy specimens were obtained in the mid-1970s from patients at a sexually transmitted disease clinic in London after receipt of informed consent; these specimens were the subject of another study [3]. The Epon blocks were resectioned into 70-nm sections. Samples were viewed on a Hi-tachi H7000 transmission electron microscope.

Immunoelectron microscopy of cervical cell biopsy specimens embedded in Epon

Thin sections were cut and treated with 1% hydrogen peroxide to etch the Epon. The antigen retrieval process was performed by heating the sections in 0.01 mol/L citrate buffer. The sections were blocked in 5% normal goat serum. The samples were incubated overnight at 4°C with the anti-H.8 monoclonal antibody 2C3 (murine IgG). Monoclonal antibody 2C3 (gift from P. Rice, University of Massachusetts) recognizes a highly conserved gonococcal outer membrane protein, H.8, that has been shown to be specific for pathogenic Neisseria [7]. The sections were washed and then incubated in goat anti–mouse IgG secondary antibody conjugated with ultrasmall gold particles. The sections were washed and then fixed with 2.5% glutaraldehyde. After washing, a Aurion R-Gent SE-EM silver enhancement kit (Electron Microscopy Sciences) was used to enhance the size of the gold beads. Contrast stain was applied with 5% uranyl acetate and lead citrate. The sections were viewed on a Hitachi H7000 transmission electron microscope.

Processing of broth-grown bacteria for transmission electron microscopy

Overnight plate cultures of strain 1291 and 1291-msbB were inoculated into GC broth supplemented with 1% IsoVitaleX to an optical density (OD)₆₀₀ of 0.05. The samples were shaken at 37°C until an OD₆₀₀ of 0.6 was achieved. Five milliliters of each culture was then added to an equal volume of 4% paraformaldehyde in PBS and fixed overnight at 4°C. The bacterial cells were treated with 1% osmium tetroxide according to standard protocols and dehydrated through a graded ethanol series. The samples were embedded in Epon acrylic resin, and thin sections of the embedded bacteria were mounted on nickel grids and stained with 5% uranyl acetate and lead citrate for contrast. The samples were viewed with a Hitachi H7000 transmission electron microscope.

Biofilm growth in a continuous-flow chamber

Collagen-coated glass coverslips (22 by 50 mm) were seeded with human papillomavirus (HPV) E6/E7–transformed cervical epithelial cells [8] before use in the biofilm experiments. Cells were grown in defined keratinocyte serum-free medium (d-KSFM; Gibco) until they reached confluency. The cervical cells were stained with a 1:5000 dilution of CellTracker Orange (Invitrogen) so that they could be distinguished later from the GFP-expressing bacteria by confocal microscopy. The coverslips were placed in a watertight flow chamber with the cell monolayer facing the interior of the chamber. The bacteria were inoculated into the flow chamber at an MOI of 100:1 in 3 mL of the cell culture medium that would be used during biofilm growth. This medium was composed of d-KSFM, McCoy’s 5A medium, and Hybridoma-SFM (Gibco) at a 1:1:2 ratio. The bacteria were allowed to adhere to the cervical cells for 1 h at 37°C in a 5% CO₂ incubator. At this time, the cell culture medium, diluted 1:5 in PBS (pH 7.4), 50 μg/mL sodium bicarbonate, 100 μmol/L sodium nitrite, and 1% IsoVitaleX, was perfused through the chambers at a rate of 150 μL/min. The flow chambers were kept at 37°C in an environmental chamber. At the conclusion of the biofilm growth experiment, effluent was cultured to assure culture purity. Microscopic images of the biofilms were then obtained.

Laser scanning confocal microscopy of biofilms

Confocal microscopic images of biofilms grown on human cervical epithelial cells in continuous-flow chambers were obtained using a Nikon C1 scanning confocal microscope. The bacteria within the biofilm chamber were visualized at a magnification of ×20, and the images were compiled as cross-sections of a z-series.

Complementation of the msbB mutant

The proB gene (NGO849) in the sequenced FA1090 strain of N. gonorrhoeae is unannotated. We sequenced this region in strain 1291 and found that it does encode for an intact proB gene. We used this region to insert a copy of msbB to complement the 1291-msbB mutant. A 1084-bp portion of the proB gene from strain 1291 was amplified by means of polymerase chain reaction (PCR). This fragment was cloned into pGEM-T (Promega), and the resulting plasmid was designated pCTS30. PCR was used to amplify a blunt-ended fragment containing the spectinomycin cassette and its upstream promoter but not the downstream transcription terminator from pSPECR. After PCR, the fragment was treated with T4 polynucleotide kinase (NEB) and ligated into pCTS30 that had been digested with BsaBI (blunt ended). The resulting construct was named pCTS32. The msbB gene was PCR amplified from strain 1291, using primers that generated an AflII site at the 5′ end and a SmaI site at the 3′ end. The resulting fragment was cloned into pGEM-T, followed by subcloning into AflII/SmaI-digested pCTS32. This construct was designated pCTS35 and sequenced. The plasmid was transformed into the 1291-msbBgfp mutant. Transformants were selected on GC agar containing spectinomycin and supplemented with proline. Integration of the construct was confirmed by PCR.

Processing of biofilm samples for light microscopy

The coverslips with biofilm attached were removed from the flow chambers and fixed with 1% osmium tetroxide in 100% perfluorocarbon for 1 h. The slides were washed with 100% perfluorocarbon followed by 100% ethanol. The samples were embedded in Epon resin. One-micrometer sections were cut and stained with Richardson’s blue for light microscopy. The sections were viewed on an Olympus BX51 scope, and images were acquired with the attached Olympus DP70 digital camera.

Bacterial attachment assay

HPV E6/E7–transformed cervical epithelial cells were grown until confluent in a 24-well cell culture plate in d-KSFM. After confluent growth was achieved, N. gonorrhoeae strain 1291 or strain 1291-msbB, taken from an overnight plate culture, was inoculated onto the cells in 100 μL of medium at an MOI of 100:1. The transformed cervical cells and bacteria were incubated together for 1 h at 37°C. After incubation, the medium was removed from the cell monolayer, and the cells were washed 3 times with 1 mL of PBS. Then, 1 mL of 0.5% saponin in PBS was added to the cell monolayer, and the plate was incubated at 37°C for an additional 30 min. The cervical cells and attached bacteria were collected, serially diluted in PBS, and plated onto GC agar. After 24 h at 37°C, plate counts of colony-forming units from bacteria collected with transformed cervical cells were compared with the bacterial inoculum plate counts to determine the adherence ratio.

RESULTS

Other studies have shown that N. gonorrhoeae has the ability to form a biofilm on glass as well as on primary human genital tract epithelial cells in flow chambers. To determine whether N. gonorrhoeae is also capable of forming a biofilm in the female human genital tract, we acquired archival cervical biopsy specimens embedded in Epon that had been obtained from women with culture-proven N. gonorrhoeae infection. These blocks were resectioned and carefully examined for signs of gonococcal biofilm formation. Three of the 10 cervical biopsy specimens showed evidence of biofilm formation on cervical epithelial cell surfaces.

Figure 1A and 1B show transmission electron micrographs from these cervical biopsy specimens. The squamous epithelial cell layer can be seen in the lower portions of figure 1A–1C. In figure 1A, a cluster of gonococci appear above the epithelial cell surface. Large amounts of membranelike material, presumably produced by the gonococci, can be visualized intercalated among the bacterial cells. The organisms in this image extend almost 11 μm above the epithelial cell surface. This image is similar in appearance to those from other electron microscopic studies of biofilms seen in flow chambers grown over genital epithelial cells [2].

Figure 1.

Electron micrographs of cervical biopsy specimens from patients with gonococcal cervicitis. The squamous epithelial cell layer can be seen in the lower portion of panels A–C; above is a cluster of gonococci, which appear to be part of a biofilm over this epithelial cell surface. The solid arrows in panels A and B indicate blebs on the surface of the organisms, and dashed arrows indicate membranous structures intercalated between the organisms. Panel C shows a control section of a biopsy specimen where biofilm formation could not be identified. Scale bars in panels A–C indicate 1 μm. Panel D shows immunostaining of an Epon-embedded cervical biopsy specimen from a woman with gonococcal cervicitis demonstrating the edge of a cluster of gonococci, which appear to be part of a biofilm on the cell surface. The section was stained with the anti– gonococcal H.8 monoclonal antibody 2C3 after an antigen retrieval process consisting of heat and oxidation of the Epon with hydrogen peroxide. The top portion of panel D shows the mass of organisms on the epithelial cell surface, and the bottom portion shows a higher-magnification view demonstrating the colloidal gold labeling.

The higher magnification in figure 1B allows the diplococcal morphology of the organisms to be more clearly seen. In addition, numerous blebs can be identified on the outer membrane of the bacteria, as well as some that have detached from the bacterial surface. Extensive blebbing in these cervical biopsy samples had been noted and reported previously [3]. Membranous structures can also be seen surrounding the bacteria in this image. Figure 1C shows a control section in a biopsy specimen in which gonococcal biofilms could not be visualized. To confirm further that the organisms in the biofilms were N. gonorrhoeae, we performed immunoelectron microscopy using pathogenic Neisseria-specific monoclonal antibody 2C3 on the cervical biopsy sections. The cervical biopsy section used for immunoelectron microscopy in figure 1D shows the edge of a cluster of gonococci, which appear to be part of a biofilm on the cell surface. The bottom panel in figure 1D is a higher magnification of the top panel, in which the colloidal gold surface labeling on the bacteria can be clearly seen. These results support the presence of N. gonorrhoeae in the biofilm structures found in the cervical biopsy specimens.

A common feature we have seen in both in vitro and in vivo N. gonorrhoeae biofilm structures is the presence of what appear to be large amounts of membrane fragments among the bacteria. These fragments can reach 10–15 μm in length, and in flow chamber biofilm studies, they are bound by an antibody that recognizes terminal sugars found on gonococcal LOS [2]. They are most likely formed by blebbing of the gonococcal outer membrane, followed by fusion of the blebbed material. We have constructed a N. gonorrhoeae mutant with a deletion of the acyltransferase gene, msbB, which is responsible for the addition of a secondary acyl substitution on the lipid A portion of LOS. An msbB deletion modifies the lipid A portion of LOS, resulting in a pentaacyl instead of a hexaacyl lipid A structure [5].

Figure 2 shows transmission electron micrographs of both N. gonorrhoeae strain 1291 and strain 1291-msbB cells that were grown in broth culture. The surface of strain 1291 is ruffled and reveals the propensity of the outer membrane to form blebs. Similar bleb formation can be seen on the surface of the gonococci in figure 1. In contrast, the surface of the msbB mutant is smooth, with no obvious blebs being formed on the membrane surface. This suggests that an altered lipid A may impede proper bleb formation and would decrease the amount of gonococcal membrane material that can be incorporated into a biofilm structure.

Figure 2.

Transmission electron micrograph of strains 1291 and 1291-msbB Neisseria gonorrhoeae grown in GC broth.

To determine whether the deletion of msbB had any effect on biofilm formation, we inoculated both N. gonorrhoeae strains 1291-gfp and 1291-msbBgfp onto separate cervical cell monolayers in flow chambers at an MOI of 100:1. After a 1-h incubation to allow the bacteria to attach to the cells, medium was perfused through the chambers at a rate of 150 μL/min for 48 h. After 48 h, the biofilm was analyzed by laser scanning confocal microscopy. A total of 10 z-series were acquired from 2 separate chambers each of strain 1291-gfp and 1291-msbBgfp biofilms. The z-series were examined using COMSTAT software to look for differences in the physical features of the biofilms [9]. COMSTAT is a program for quantifying 3-dimensional biofilm structures that analyzes z-series acquired by laser scanning confocal microscopy. This experiment was repeated a total of 3 times.

Strain 1291-gfp biofilm had significantly more biomass than the 1291-msbBgfp biofilm (P = .004). Likewise, strain 1291-gfp biofilm was significantly thicker, on average, than strain 1291-msbBgfp biofilm (P = .002). Complementing the msbB mutation by introducing a single copy of msbB on the chromosome resulted in formation of biofilm that did not differ significantly from strain 1291 in total biomass (P = .96) or average thickness (P = .88). Student’s t test was used to determine the statistical significance of these differences.

The z-series for strain 1291-gfp, 1291-msbBgfp, and the complemented 1291-msbBgfp mutant biofilms were analyzed by Volocity high-performance 3-dimensional imaging software (version 3.6.1; Improvision), which gives a 3-dimensional image of the biofilm. This is done by compiling all of the sections from a z-series into a single image. The results are seen in figure 3. The GFP-expressing biofilm bacteria appear green, and the underlying cervical epithelial cells, stained with CellTracker Orange, appear red. Strain 1291-gfp formed a denser and thicker biofilm on the cervical cell surface than strain 1291-msbBgfp. The complemented msbB mutant formed a biofilm that is essentially identical in appearance to that of strain 1291. To assure that the membrane lipid modification did not alter the ability of 1291-msbB to adhere to cervical cells, adherence studies were performed on strain 1291 and 1291-msbB. The findings indicated that these strains did not differ in their abilities to adhere to cervical cells (data not shown).

Figure 3.

Three-dimensional images of strain 1291-gfp (A), strain 1291-msbBgfp (B), and complemented strain 1291-msbBgfp biofilms (C), rendered by Volocity software using stacked vertical z-series acquired from 48-h biofilms grown over cervical epithelial cells. The cervical epithelial cells were stained with CellTracker Orange and appear red; green fluorescent protein– expressing biofilm bacteria appear green.

In addition to using laser scanning confocal microscopy to analyze biofilm formation, we studied the strain 1291 and 1291-msbB biofilms by light microscopy (data not shown). The bacteria were inoculated into chambers over a cervical cell monolayer as described above at an MOI of 100:1. After the initial attachment period of 1 h, medium was perfused through the chambers. After 96 h, the coverslips with attached biofilm were removed and processed for light microscopy. The strain 1291 biofilm contained dense clusters of bacteria that formed a thick structure over the surface of the cervical epithelial cells. In contrast, the 1291-msbB biofilm contained no large bacterial clusters. The bacteria appeared scattered over the cervical cell surface and were unable to form any kind of structure resembling that seen in strain 1291.

DISCUSSION

We have shown previously that N. gonorrhoeae is able to form biofilms on glass surfaces as well as on cultured urogenital tract epithelial cells in flow chambers [2]. In this report, we provide evidence that N. gonorrhoeae can form biofilm structures within the human female genital tract as well. We used electron microscopy to identify N. gonorrhoeae biofilm formation on cervical cell biopsy specimens obtained from women with culture-proven gonococcal infections. The biofilms formed in vivo were structurally similar to those grown on cultured cervical cells in flow chambers. In particular, large amounts of membranous extensions were seen intercalated among the bacteria in both the in vivo and in vitro biofilms.

Staining with antibodies specific for epitopes found within the outer membrane or LOS of N. gonorrhoeae suggests that this membranous material is composed of gonococcal outer membrane. A strain 1291-msbB mutant of N. gonorrhoeae with a severely compromised ability to form outer membrane blebs produced biofilms on cultured transformed cervical cells that were significantly reduced in thickness and overall biomass, compared with the 1291 parent strain biofilms. However, the 1291-msbB bacteria were able to attach to cultured transformed cervical cells just as well as 1291 wild-type bacteria, and the growth rates in GC broth were the same for both strains (data not shown). In addition, N. gonorrhoeae strain FA1090 modA13 had an increased blebbing phenotype compared with the FA1090 parent strain. This mutant formed a biofilm of significantly greater thickness and biomass than that formed by the wild-type strain (data not shown). These results further suggest that outer membranes released by N. gonorrhoeae within a biofilm are an integral component of biofilm formation and that hypo- or hyperblebbing of the outer membrane directly affects biofilm thickness and density. Although it is well accepted that biofilms in general have an exopolysaccharide framework for their matrix, in addition to a host of other products produced by the cell or acquired from the surrounding environment, we have not yet explored what this framework is composed of in N. gonorrhoeae biofilms [10, 11]. However, our findings suggest that incorporation of outer membranes resulting from blebbing of the gonococcus is crucial to proper biofilm formation.

In recent years, the resistance of N. gonorrhoeae to antibiotic treatment has increased to a point where many previously used antibiotics are no longer efficacious in treating the disease [12]. The ability of N. gonorrhoeae to form biofilms in the female genital tract further exacerbates this dilemma. Biofilm formation in bacteria has been shown to prevent complete access of antibiotics to all regions of the biofilm. The diverse microenvironments found in biofilms as a result of pH differences, oxygen variance, water channels, and varied growth rates among the embedded bacteria also contribute to the difficulty of treating biofilms with antimicrobial agents [13]. Furthermore, horizontal gene transfer among bacteria in a biofilm can increase the spread of antibiotic resistance genes [14]. In addition to antibiotic resistance, biofilm formation by N. gonorrhoeae in the female urogenital tract may also contribute to the high incidence of asymptomatic infections in women [1]. Although in men gonococcal infections tend to be more cell invasive, leading to a robust inflammatory response, in most instances women are unaware they are infected. This lack of symptoms can lead to a greater likelihood of infecting additional sexual partners as well as to more severe complications of N. gonorrhoeae infection, including pelvic inflammatory disease.

Because of the ability of N. gonorrhoeae to form biofilms and the refractive nature of biofilms to antibiotic treatment, alternative strategies to resolve N. gonorrhoeae infection should be pursued. Interesting recent research has involved attempts to interfere with quorum sensing in bacteria to prevent biofilm formation. Much of this research has focused on inhibition of N-acyl homoserine lactones and LuxR-type proteins, which would prevent a quorum-sensing cascade necessary for biofilm formation [15–17]. An alternative strategy to the prevention of biofilms could be rapid dispersal of existing biofilms. A recent discovery that nitric oxide has the ability to disperse Pseudomonas aeruginosa biofilms may lead to future methods of treating biofilm-based infections [18].