Secondary Logo

Journal Logo


Experimental Gonococcal Urethritis and Reinfection with Homologous Gonococci in Male Volunteers


Author Information
  • Free


Urethral gonorrhea appears to confer little protection against reinfection, which is so common that it accounts for 30% to 50% of the gonorrhea treated in clinics for sexually transmitted disease. 1–3 It is not known whether reinfection takes place because the initial infection fails to produce a protective immune response, or because antigenic variations among gonococcal strains permit them to evade protective immunity elicited by prior infections.

The question bears directly on efforts to develop gonococcal vaccines. If protective immunity is not induced by infection, how can it be induced by vaccination? What would be the nature of the protective immunity? What antigens, and how many, would be required for a successful vaccine?

The failure of three protein vaccines to provide protection 4–6 has led to the suggestion that an effective vaccine may not be possible. 7 However, a pilin vaccine did induce protection in a small controlled trial with men, 8 and multiple gonococcal infections appear to afford some protection from recurrent infections in women. 9 Finally, untreated gonococcal urethritis resolves in men after several weeks, 2 although the relation between resolution and immunity is unknown.

We designed a two-phase study to learn whether urethritis caused in male volunteers by intraurethral inoculation of gonococci would protect volunteers from reinfection with the same organism. The study called for infection of volunteers in the first phase, followed by challenge of these same volunteers as well as naïve volunteers in the second phase. Phase 1 used an intraurethral inoculum of 57,000 piliated, transparent (opacity protein–negative [Opa]) MS11mkC Neisseria gonorrhoeae. The MS11mkC organisms are highly infectious for the male urethra, 10 and the inoculum was chosen on the basis of previous studies for its likelihood of producing infection in all the volunteers. 10–12 Volunteers were allowed to remain infected for at least 5 days to encourage development of immune response.

In the phase 2 challenge, 2 weeks after treatment of the initial infection, an inoculum of 7,100 organisms was used. This inoculum was expected to produce infections in approximately 50% of the volunteers. 10–12 If the initial infection could induce meaningful immunity, we reasoned, its induction should be established by at least 5 days after infection, 13,14 and its effect maintained for at least 2 weeks.

Materials and Methods

Challenge Organism

N gonorrhoeae MS11mkC is a lipooligosaccharide (LOS) variant of MS1mk (1B, nonargenine-hypoxanthine-uracil auxotype) used as the challenge organism in a series of human experiments that provided the background for the current study. As few as 250 of this LOS variant can cause urethritis and discharge. 10

Several previous iterations of the same basic experimental protocol 10–12 were used to derive the relation of the intraurethral dose of MS11mkC to its infectivity, and the results were used to calculate inocula for the two phases of the current study. A large inoculum (104) was chosen for phase 1 in order to infect all of the volunteers. Analysis of the inoculum used showed that it contained 57,000 organisms.

For phase 2, an inoculum close to the ID50 for this variant (2.3 × 103; 95% CI, 2.7 × 102–8.2 × 103) was chosen because it promised to show induced immunity more readily than a larger inoculum. 8 Although a smaller inoculum (e.g., one productive of approximately 20% infections) might have shown more modest levels of immunity, it would have required infecting many more volunteers to show a significant difference in the infection rates between naïve and reinfected volunteers. Analysis of the phase 2 inoculum showed that it contained 7,100 organisms.

Study Design

Male volunteers 18 to 50 years old were recruited through local advertisement. They provided a medical history, in which each denied prior gonococcal or meningococcal disease, or any sexually transmitted disease. Each volunteer received a physical examination with laboratory evaluations that included urinalysis, complete blood count, serum chemistry, renal and liver function tests, complement activity by CH50 assay, and serologic tests for syphilis, HIV, and hepatitis B virus. Urethral swabs were cultured for N gonorrhoeae and Trichomonas vaginalis. Infections caused by Chlamydia trachomatis were sought by GeneProbe (GenProbe, San Diego CA). Criteria for exclusion were physiologic, urogenital, or cardiac abnormalities; abnormal complement tests 15,16; and positive results for serologic tests or cultures. Volunteers also were excluded if they had allergies to penicillin and ciprofloxacin, a recent (within 1 year) history of gonococcal or meningococcal infections, or a history of antibiotic use within 7 days before the first study day.

The studies were conducted in an access-controlled area at the Kimbrough Army Community Hospital, Fort Meade, Maryland. A physician was quartered with the volunteers, who were continuously supervised from the time of the inoculation until the time of treatment.

A literature review was used to plan for a duration of the phase 1 infection that would be sufficient for the development of an immune response while minimizing the risk of local complications. Cohen et al 13 found a fourfold increase in serum immunoglobulin G (IgG) and IgA during experimental infections of 5 days duration. Reports from the preantibiotic and early antibiotic eras described the onset of local complications as generally slow in typical urethral infections, with a fibroblastic response occurring only after 5 to 10 days of acute symptomatic infection. 17 The time after infection between generation of an immune response and generation of a fibroblastic response provided a window of safety, a time during which the infection could safely remain untreated. Because the average time from inoculation of MS11mkC into the male urethra to the onset of symptoms is approximately 48 hours, 10–12 we limited the time between inoculation and treatment to a maximum of 7 days.

In phase 1, volunteers were asked to delay treatment until day 7 after inoculation. However, they were treated earlier if medically indicated or on their request. In phase 2, the volunteers were treated as soon as clear signs of infection (i.e., dysuria with leukorrhea) were present. 12

Inoculation of Volunteers

Volunteers were inoculated with piliated (P+) transparent (Opa) N gonorrhoeae strain MS11mkC as in previous studies. 10–12 The organisms were hydrated from the lyophilized state, passaged, and prepared as described previously. 12 Colonies of P+, Opa were suspended in tryptic soy broth, and the density of the organisms was adjusted spectrophotometrically to an OD650 of 0.5 (≈5 × 108 gonococci/ml). The organisms were diluted in tryptic soy broth so that 0.2 ml contained the desired inoculum. Viability and colonial variation were assessed by culturing the challenge suspension immediately before and after the volunteers were inoculated.

A flexible 8-F pediatric catheter (Mentor Health Care Products Corporation, Santa Barbara, CA) was cut to a length of 8 cm and attached to a 1-ml syringe (Becton Dickinson, Franklin Lakes, NJ). The inoculum was drawn through the catheter into the syringe. The catheter then was inserted 5 cm into the urethra of supine volunteers, and 0.2 ml of inoculum was instilled slowly. After removal of the catheter, each volunteer was asked to compress his penile meatus digitally for 30 minutes and to refrain from urination for 2 hours.

Evaluation of Volunteers for Infections

Urine specimens, collected 2 hours after inoculation and daily thereafter, were processed as described previously. 10–12 Cultures, urinary sediments, and urethral exudates were stained with the enhanced Gram stain (CMS Microbiologicals, Decatur GA), and cultures were confirmed to be N gonorrhoeae by the oxidase test (Becton Dickinson Microbiology Systems, Sparks, MD). Experimental infection of the male urethra with MS11mkC is reproducible. 10–12 Subjects remain asymptomatic for a variable incubation, and generally have an “eclipse phase,” during which organisms can no longer be isolated from urine sediment. Infected individuals then begin shedding Opa+ organisms, with development of dysuria, followed 24 to 48 hours later with visible discharge. 10–12 Subjects who resist infection do not resume shedding of organisms, nor do other symptoms develop after the eclipse phase.

Volunteers were given daily physical examinations. Blood was collected for serologic testing before the first challenge, at 3-day intervals to day 36, and on day 80. Incubation was defined as the time from inoculation to presentation of dysuria or urethral discharge. Infection was defined as the development of dysuria or discharge. 10–12 Resistance to infection was defined as the absence of these signs and symptoms.

Diagnosis of infection was confirmed by the observation of more than three polymorphonuclear neutrophils per high-power field, with Gram-negative intracellular diplococci, in a volunteer’s urethral discharge, or positive results from a urethral swab culture for N gonorrhoeae after presentation with dysuria or discharge. The attending physician’s diagnosis of gonococcal urethritis was made independently of the urine sediment culture results according to standard Centers for Disease Control guidelines. 18

In phase 1, all infected volunteers were treated orally with 500 mg ciprofloxacin (Bayer, Westhaven, CT) either on request, when deemed medically indicated by the attending physician, or on day 7 after inoculation. Volunteers were compensated to the same degree regardless how long their infection persisted. In phase 2, the volunteers were treated immediately on diagnosis or, if asymptomatic, on day 7 after inoculation. Urethral swabs taken 24 hours after treatment were cultured to confirm clearing of the infection.

Western Blot Analysis

Outer membrane proteins (OMP) were extracted from colonies grown on GCDC agar with use of a previously described 19 modification of the method by Blake and Gotschlich. 20 Protein concentration was determined by the method of Lowry et al. 21 Outer membrane protein extracts were separated electrophoretically through preparative 13.1% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels (250 μg/gel) 22 and electroblotted onto nitrocellulose membranes. 12,23 A strip that contained molecular weight standards was cut from the blots and stained with India ink for visualization of the proteins. 24 The blots were cut into vertical strips and incubated overnight with serum diluted 1 to 40 or neat urine. 25,26 Bound antibodies were detected by incubating the blots for 2 hours with alkaline-phosphatase–labeled goat antihuman IgG, IgM, or IgA (Kirkegaard and Perry, Gaithersburg, MD), followed by fast red, 2-napthol MX phosphate substrate (Sigma, St Louis, MO). Opa and pilin proteins were identified on control strips cut from the same blot with sera from rabbits that had been immunized, respectively, with a conserved Opa or Pil peptide. Captured rabbit antibodies were detected with alkaline-phosphatase–labeled goat antirabbit IgG (Kirkegaard and Perry) followed by the same phosphate substrate. The molecular weights of the OMP components were estimated by comparison with marker proteins of known molecular weight (BioRad, Richmond CA). The MS11mkC organisms extracted for the assay expressed only OpaK (Mr:30), the predominant (≈60%) Opa made by this strain when grown on agar. 19

Enzyme-Linked Immunoabsorbent Assay

Antibody titers to LOS were determined with an enzyme-linked immunoabsorbent assay using flat-bottomed microtiter plates (Nunc Maxisorp, Nalge Nunc International, Naperville, IL) that had been incubated at 4 °C overnight with 2 μg per well of MS11mkC LOS 27 diluted in PBS, then blocked for 2 hours with casein/bovine serum albumin filler. Sera were added in dilutions of 1 to 10 in the top row of wells, and twofold serial dilutions were made in the columns of wells. For each set of a volunteer’s sera, the day 1 serum was diluted into a column of wells coated only with filler, as a negative control. Serum from a rabbit immunized with MS11mkC was used as a positive control.

The plates were incubated at 37 °C for 2 hours. Captured human antibodies were detected with alkaline phosphatase–conjugated goat antihuman IgG, IgM, or IgA, followed by pNP substrate (Kirkegaard and Perry); captured rabbit antibodies with alkaline phosphatase-labeled goat antirabbit IgG (Kirkegaard and Perry) followed using the same substrate. The antibody titer was taken as the reciprocal of the highest serum dilution that gave an absorbance twice that of the negative control. Titers had to increase at least fourfold to be considered significant. 13


Although 20 volunteers were recruited for phase 1, only 15 reported for its initiation. Of the 15 volunteers, 14 (93%) became infected after challenge with 57,000 OpaP+ MS11mkC (Figure 1), with a median incubation period of 54 hours (range, 19–69 hours).

Fig. 1
Fig. 1:
Phase 1 quantitative cultures of urinary sediments from the 14 volunteers (Vol) who were infected with 57,000 MS11mkC gonococci during this phase. Panels A, B, and C are the infection profiles of the eight volunteers who subsequently resisted reinfection during phase 2 (Resistant). Panels D and E are the infection profiles of the six volunteers who were reinfected by inoculation of 7,100 organisms during phase 2 (Reinfected). The (+) symbol notes the transition time of the recovered organisms from the Opa to the Opa+ phenotype. The (←) symbol notes the time of treatment with 500 mg of ciprofloxacin.

Volunteers 66 and 77 requested treatment at 69 hours (day 3) because of low tolerance for their symptoms. Volunteer 66 developed a discharge at 58 hours, dysuria at 66 hours, and positive results from a culture at 68 hours. Volunteer 77 developed dysuria at 40 hours, and a discharge at 44 hours that grew N gonorrhoeae. Two other volunteers (68 and 74) were treated at 56 and 95 hours, respectively, at the attending physician’s discretion. Volunteer 68 developed dysuria and a culture-positive discharge at 20 hours, and difficulty urinating by 56 hours, whereas volunteer 74 developed dysuria and discharge at 19 hours and a localized maculopapular rash on his penis at 95 hours.

As shown in Figure 1, 11 volunteers remained untreated until the termination of the first challenge. Volunteer 76 did not shed gonococci at any time during the phase 1 study. Regarded as having remained uninfected, he was treated on day 7. This volunteer was not included in the phase 2 rechallenge experiment. Another participant, volunteer 64, experienced dysuria on day 3, and a discharge that contained polymorphonuclear neutrophils and gram-negative diplococci by day 4, but despite the continuing presence of polymorphonuclear neutrophils and organisms in his urinary sediments, gonococci could not be cultured after day 4.

Fourteen days after all the volunteers in phase 1 had been treated (21 days after the initial challenge), the 14 phase 1 volunteers who had become infected and a control group of 10 naïve volunteers were challenged with 7,100 MS11mkC gonococci. Six (43%) of the 14 phase 1 volunteers and 5 (50%) of the 10 naïve volunteers became infected during phase 2. All four of the phase 1 volunteers treated early were reinfected during phase 2. Eight (80%) of the 10 volunteers who remained infected for the entire duration of phase 1 resisted reinfection (P = 0.32, Fisher exact test for the comparison between naïve and fully experienced phase 2 volunteers).

The infections that followed the phase 2 inoculations were indistinguishable from those that followed the phase 1 inoculations. All the volunteers considered to be infected in either phase experienced a purulent penile discharge and shed Opa+ organisms. Most experienced dysuria. No asymptomatic infections were found. The development of symptoms was not related to the number of organisms shed in urine over time, but the symptoms usually began during or after the Opa phenotype transition from Opa to Opa+ (Figure 1). Clinical diagnosis usually was made by the time shed gonococci reached 105 to 106 organisms/ml of urinary sediment.

The volunteers reinfected with the lower inoculum in phase 2 had incubations similar to those after inoculation with the higher inoculum in phase 1 (median of 54 hours in phase 1 and 52 hours in phase 2), whereas the naïve phase 2 volunteers had longer incubations (range, 69–139 hours; median, 93 hours).

Serum Antibodies

The preinoculation serum from each volunteer contained IgG antibodies that reacted with various OMPs on Western immunoblot analysis (Figure 2). All the phase 1 and phase 2 volunteers had serum IgG that reacted with a 35-kDa protein before infection, and all but volunteers 65 and 68 had preinfection serum IgG that reacted with a 29-kDa protein. On the basis of the visually assessed density of bands in Figure 2, the eight volunteers who resisted reinfection appeared to have more antibodies to more proteins than six volunteers who became reinfected.

Fig. 2
Fig. 2:
Western immunoblot showing the binding of volunteers’ serum IgG to gonococcal outer membrane proteins on day 1, before infection. Each volunteer presented with a unique panel of prestudy antibodies reactive with a variety of outer membrane proteins. Resistance to reinfection could not be correlated with antibodies to any particular protein band. Panel A: phase 1 volunteers who resisted reinfection during phase 2. Panel B: phase 1 volunteers who were reinfected during phase 2. Panel C: naïve phase 2 volunteers who resisted infection. Panel D: naïve phase 2 volunteers who became infected. std: standard molecular weight markers (kDa). cs: control strip that was exposed to rabbit antisera specific for Opa (O) and Pil (P) proteins. ▸ indicates the area of the blot where lipooligosaccharides are found. The numbers are those of the volunteers.

Most of the volunteers (18/24) had some pilus IgG at the time of enrollment. However, neither the presence nor the amount of pilus IgG consistently distinguished the phase 1 volunteers who resisted reinfection during phase 2 from those who did not. However, the phase 1 volunteers (72, 75, and 78) whose sera contained the most abundant pilus and 29-kDa-reactive IgG before the first inoculation resisted reinfection during phase 2.

Only 4 of the 14 phase 1 volunteers, (two [70 and 79] who resisted reinfection and two [65 and 74] who did not) had sera that reacted appreciably in a western blot analysis with a band that corresponded in molecular weight to Opa K (30 kDa) (Figure 2). Four of the 10 naïve phase 2 volunteers (102, 104, 105, and 106), three of whom were infected by the challenge, also had detectable IgG antibodies reactive with the 30-kDa band in their preinoculation sera (Figure 2).

Figure 3 shows a Western immunoblot comparing serum (IgG) and urine (IgM) antibodies from volunteers who showed an increase in band intensity from day 1 to day 21. It can be observed that IgM reacted with fewer protein bands than IgG, and that the antigens with which the two antibody classes reacted were not always the same. However, both immunoglobin classes reacted with multiple LOS bands (Figure 3A, arrow).

Fig. 3
Fig. 3:
Western immunoblots from all the volunteers who showed an increase in antibodies or made new antibodies. Of these, only volunteer 74 was reinfected in the phase 2 challenge. Western blots of gonococcal outer membrane proteins were reacted with day 1 (first day of the phase 1 infection) and day 21 (first day of the phase 2 infection) serum (panel A) and urine (panel B) from the volunteers. (Panel A) Lane 1: IgM day 1. Lane 2: IgM day 21. Lane 3: IgG day 1. Lane 4: IgG day 21. (Panel B) Lane 1: IgG day 1. Lane 2: IgG day 21. Lane 3: IgA day 1. Lane 4: IgA day 21. std: standard molecular weight markers (kDa). cs: control strip that was exposed to rabbit antisera specific for Opa (O) and Pil (P) proteins. ▸ indicates the area of the blot where lipooligosaccharides are found.

Urine Antibodies

All of the phase 1 and phase 2 volunteers had preexisting urine IgA that bound the 29 -kDa and 35-kDa OMPs (Figure 3 B). Urinary IgG and IgA changed little over the course of phase 1, although four phase 1 volunteers, all of whom resisted reinfection in phase 2, acquired urinary IgA antibodies that had been absent on day 1 (Figure 3B).

Lipooligosaccharide Antibodies

Five of the phase 1 volunteers who resisted reinfection, and one volunteer (74) who became reinfected had visible increases in serum antibodies that bound Western blot bands corresponding in electrophoretic mobility to LOS (Figure 3 A). The LOS antibody titers therefore were determined by ELISA in all the sera collected in the study. The visible increase in the intensity of bound LOS antibodies seen on Western immunoblot was found to correlate with a fourfold or greater increase in serum LOS antibody titers by ELISA (Table 1). The phase 1 volunteers who resisted infection during phase 2 were significantly more likely to have had a fourfold or greater increase in LOS-reactive IgG during phase 1 than those who did not resist reinfection (P = 0.026, Fisher exact test, Table 1). The volunteers who resisted reinfection had a geometric mean LOS IgG titer at the time of reinoculation ten times greater than those who became reinfected (Figure 4). The difference in LOS IgG titers between the two groups persisted through day 80 (Figure 4). The increase in LOS IgG preceded the increase in IgM for four of the eight volunteers who resisted reinfection.

Table 1
Table 1:
Increase of Serum Lipooligosaccharide Antibodies Over Day 1 Levels*
Fig. 4
Fig. 4:
Geometric mean lipooligosaccharide IgG titers, as determined by enzyme-linked immunoabsorbent assay (see Methods section) in the sera of phase 1 volunteers who did (⋄) or did not (□) become reinfected during phase 2.


This study was designed to discover whether clinical gonorrhea would protect against infection on rechallenge with the same strain of N gonorrhoeae, as might occur naturally when a man exposes himself to the same partner after he has been treated. Under the experimental conditions used, the initial infection did not provide protection against reinfection.

The experimental infections were similar to typical, community-acquired gonococcal urethritis. Incubation periods in both phases of the study were comparable with those seen in natural infections. 27,28 Rechallenged volunteers from phase 1 did become symptomatic a day sooner than the naïve volunteers during phase 2, but such a difference is consistent with observed variations in incubation times of experimental gonorrhea, 10–12,29 and longer incubations have been noted for initial cases of community-acquired gonorrhea than for repeat cases. 30

Although infection rates in the naïve and rechallenged groups were quite similar (5 of 10 and 6 of 14 volunteers, respectively), the study did not have sufficient power to detect a modest level of protection. Of the 20 volunteers recruited for phase 1, only 15 reported for its initiation. For logistic reasons, we were not able to recruit additional volunteers or to ask the missing five volunteers to report later. Power calculations for the study were based on the assumption that 20 volunteers would participate in phase 1.

We also assumed that all of the phase 1 volunteers would remain infected for 7 days. Only 2 of the 10 volunteers who completed phase 1 without early treatment became reinfected during phase 2. This 20% rate of reinfection was not different from the 50% reinfection rate among the 10 naïve volunteers. However, if we had been able to study 20 volunteers infected for the full 7 days, the study would have had the power to determine whether such a reduction in the risk of reinfection was significant.

The average risk of developing gonorrhea from intercourse with an infected female is only 22%. 31 Infected women may shed organisms only intermittently, 32 but it is not known whether the relative inefficiency of female-to-male transmission is caused by intermittent shedding from the infected cervix, lack of expression by the organism of virulence factors needed to infect a particular man’s urethral epithelium, or differences in the resistance of consorting males. It also is not known whether resistance is a function of induced mucosal and/or serum antibodies, or of some other natural immunity.

Whatever the nature of resistance, it can be overcome in experimental infections by increasing the inoculum. 8 Indeed, Brinton et al 8 found that inoculating healthy male volunteers with a vaccine composed of gonococcal pili and contaminating LOS shifted the infectivity dose–response curve to the right, increasing the ID50 by approximately one log. The inoculum encountered during heterosexual intercourse is not known, although it certainly is variable and probably fairly low. 31,32 Roughly 80% of men can resist it. 31

For the reasons outlined in the Methods section, we chose an inoculum for phase 1 that would overwhelm the background rate of resistance and yield 100% infections, and an inoculum for phase 2 that would produce a 50% infection rate. The use of a lower inoculum in phase 2 might have produced a more modest increase in resistance, but this would have required doubling the number of participants, which was not feasible.

It is not clear why all four of the volunteers treated early during phase 1 became reinfected in phase 2. One explanation might be that these four had very low levels of resistance at the outset of phase 1, so that an intensely inflammatory infection rapidly developed, causing them to be treated early. If the initial infection did not induce much increase in resistance, these volunteers also would have had insufficient resistance to prevent reinfection with the lower inoculum used in phase 2. In contrast, 8 of the 10 volunteers who were able to tolerate infection for the full 7 days of phase 1 might well have been able to resist an inoculum of 7,100 de novo, and we merely documented this in phase 2. Under this scenario, resistance was not boosted by the phase 1 infection, and we studied how differences in natural resistance affected the inoculum needed to cause infection.

Alternatively, it may require 5 or more full days of active infection for an increase in resistance to develop. Five days of experimental infection were adequate to induce fourfold increases in serum IgG and IgA gonococcal antibodies, respectively, for 9 and 7 of the 10 volunteers studied by Cohen et al. 13 In eight of our volunteers, including one volunteer (74) treated at 95 hours, 5 days also was adequate for the induction of LOS antibodies. Early treatment, however, could have aborted the development of resistance, leaving the volunteers susceptible to reinfection.

Care was taken to minimize the potential for complications. MS11mkC is a Por serovar 1B, a Por type less likely to cause disseminated infection. 33 Because disseminated infections have been associated with complement deficiencies, volunteers were screened for complement sufficiency. 15,16,34 The fully informed volunteers received continuous monitoring and treatment as needed by a physician who was present during the entire study. To minimize trauma in phase 1, urethral swabs were taken during treatment only to confirm infection or its absence, and after treatment only as a test of cure. Finally, limiting the duration of the infection to 5 days minimized the risk of local complications from a fibroblastic response. 17

The 14-day interval between treatment of the phase 1 volunteers and initiation of phase 2 would have avoided induced gonococcal serum or urethral IgA and IgM antibodies because they usually are no longer detectable 14 days after successful treatment. Serum and urethral IgG antibodies, however, are detectable at least 28 days after treatment. 13,14

Resistance to a second challenge was the outcome used to judge whether the initial infection had induced an immune response. The uncertainty caused by the smaller than planned number of volunteers in the study prompted a detailed analysis of humoral immune responses. The prechallenge sera from our volunteers, all of whom denied a history of gonococcal infection, contained antibodies reactive with many gonococcal OMPs. The presence of such antibodies in the serum of individuals without known prior experience with gonococci has been noted by others. 25,35,36 It has been appreciated for many years that the sera of normal, uninfected persons have “natural antibodies” that react with “heat-stable” (i.e., protein) antigens of assorted gram-negative bacteria. 37 These antibodies develop early in life; are present in the IgG, IgA, and IgM antibody classes, and react with gonococcal antigens. 37,38 Furthermore, N gonorrhoeae are known to share surface molecules with other Neisseria species that are common commensals or occasional pathogens. 8,25,39–42

Conversely, Lammel et al 35 and Zak et al 36 found that gonococcal OMP antibodies were more abundant in patients with a history of prior gonococcal infections than in those without such a history. They also found when gonococcal OMP antibodies were present in the sera of blood donors and laboratory personnel, OMP IgG antibodies mostly bound higher (>40 kDa) molecular weight proteins. This would suggest that the lower-mass OMP antibodies found in the sera from many of our volunteers might have originated during prior gonococcal exposures. 35,36 None of our participants had infections at the time of inoculation, and none acknowledged prior gonococcal infections. However, such a history must be accepted with skepticism. As Zenilman has pointed out, people tell physicians what they think physicians want to hear regarding their sexual histories. 43 This is particularly true when payment for participation in a study is being offered. Therefore, although no evidence exists to show that the antibodies existing before inoculation were induced by an earlier gonococcal infection, there is no proof to the contrary.

Most of the volunteers had preinoculation antibodies that bound 29-kDa and 35-kDa OMPs. We did not try to identify these proteins, but the apparent mass of the 35-kDa protein is within the range of Por proteins. 36 Epidemiologic data are contradictory as the whether gonococcal infections can induce Por-specific protection against reinfection. Plummer et al 9 found evidence of decreasing susceptibility to gonococcal infections with increasing age and exposure in a population of Kenyan prostitutes, 65% of whom were HIV-positive. They ascribed this to a sharply reduced risk of reinfection with four of five studied serovars. 9 The serovar that was not protective was 1B-1. Fox et al 44 could not confirm this protection in a population of North Carolinians, 75% of whom were men. They pointed out that Plummer et al 9 did not adequately account for changes in the prevalence of serovars among the community or partner choices. 44 Buchanan et al 45 also found no evidence of a serovar-specific reduction in the risk of gonococcal infections among women. They did, however, find evidence of a Por-specific reduction in the risk of salpingitis after infection, a finding that Plummer et al 46 could not confirm.

The apparent mass of the 29-kDa protein was slightly less than that of OpaC, the lowest mass Opa made by MS11mkC. 19 Abundant antibodies against this unidentified protein appear to be associated with resistance to infection by the lower inoculum used in phase 2.

Despite the abundance of antibodies in many preinoculation sera, we were not able to find clear evidence that antibodies against any specific gonococcal protein increased during these experimental infections. Lammel et al 35 also found no apparent increase in abundance or specificity of antibodies to gonococcal OMPs during convalescence from naturally acquired gonococcal infections in men and women. Gonococcal urethritis, however acquired, does not appear to induce an appreciable increase in preexisting gonococcal OMP antibodies.

Urinary OMP antibodies paralleled those in the preinoculation sera, but had a more narrow range of specificity. Both McMillan et al 14 and Kearns et al 47 found that gonococcal-reactive IgA was present in urethral exudates. The latter investigators wondered whether these were cross-reacting antibodies that had been elicited originally by other organisms. The proportion of infected men in the former study who had IgA in their exudates did not increase over time before treatment. 14

The protocol used in this study has previously been used to document that N gonorrhoeae must express both pili and paraglobosyl LOS to infect the male urethra. 10,12,48 Opa expression appears to be needed for the development of symptoms. 12,49 Although many of our volunteers had some pilus IgG at the time of challenge, the presence of such antibodies did not appear to protect them from infection during either phase. This is consistent with the failure of an immunogenic pilus vaccine to prevent gonococcal urethritis in a large field trial. 4

Induction of serum LOS antibodies during the phase 1 infection was associated with resistance to reinfection with the lower phase 2 inoculum. The strength of this association was surprising. Induction of serum LOS antibodies during urethral gonorrhea has been noted before, 35,50 but it is not clear how these serum antibodies would mediate protection at the urethral mucosa. They may be only surrogate markers for mucosal antibodies, or perhaps they are translocated to the mucosal surface during the early stages of inflammation. McMillan et al 14 reported that they could absorb much of the gonococcal IgA in urethral exudates from infected men using Neisseria lactamica, a Neisseria that frequently colonizes the pharynges of infants, and that shares LOS, but not protein structures with pathogenic Neisseria.40 They also reported what appeared to be transudative IgG in the urethral exudates. Although exudative IgA (and IgM) antibodies should have declined by the start of phase 2, they may have been rapidly recalled. 14,47

We did not try to identify the LOS structure or structures that induced these antibodies. However, the paraglobosyl LOS that gonococci must make to infect the male urethra 10,12 shares its glycose structure with human paraglobosyl glycosphingolipids 41,42 and would seem to be an unlikely immunogen. Gonococci do make other LOS molecules, and it has been proposed that a conserved basal region 42 structure recognized by monoclonal antibody (mAb) 2C7 and dependent on β-chain substitution of Hep2 can induce potentially protective antibodies. 51–53 However, Ms11mkC does not make this LOS structure, 54,55 nor bind mAb 2C7. 56

The increase in LOS IgG before that of IgM in the serum of phase 1 volunteers suggested that they had had previous community-based encounters with Neisseria, many of which share at least some LOS structure, including those that are not normally pathogenic. 40-42,57

The antibody response to Opa during gonococcal infection is quite complex. 36 Serum Opa antibodies do not appear to protect against cervical infection in women, although they may help to prevent ascending infection of the fallopian tubes. 46 Only four of our phase 1 volunteers, and four of the phase 2 volunteers appeared to have serum antibodies that recognized Opa K, the predominant Opa made by MS11mkC, 19 but we would not have detected antibodies to the nine other Opas expressed less frequently by MS11mkC during growth on agar. 19,36 Nevertheless, because we challenged the volunteers with an Opa strain, it is unlikely that Opa antibodies were involved in resistance.

Although we were not able to demonstrate protection from reinfection in this study, the results suggest that immunity to reinfection is more complex than anticipated by our experimental design. Perhaps, as suggested by Plummer et al, 46 protection is cumulative and increases as each exposure to N gonorrhoeae or other Neisseria adds new elements of immunity to those already in place.


1. Beller M, Middaugh J, Gellin B, Ingle D. The contribution of reinfection to gonorrhea incidence in Alaska, 1983–1987. Sex Transm Dis 1992; 19: 41–46.
2. Hill J. Experimental infection with Neisseria gonorrhoeae. Am J Syphilis Gonorrhea Vener Dis 1942; 27: 733–771.
3. Noble RC, Kirk NM, Slagel WA, Vance BJ, Somes GW. Recidivism among patients with gonococcal infection presenting to a venereal disease clinic. Sex Transm Dis 1977; 4: 39–43.
4. Boslego JW, Tramont EC, Chung RC, et al. Efficacy trial of a parenteral gonococcal pilus vaccine in men. Vaccine 1991; 9: 154–162.
5. Greenberg L, Diena BB, Ashton FA, et al. Gonococcal vaccine studies in Inuvik. Can J Public Health 1974; 65: 29–33.
6. Tramont EC. Gonococcal vaccines. Clin Microbiol Rev 1989; 2: S74–S77.
7. Rice PA, Gulati S, McQuillen DP, Ram S. Is there protective immunity to gonococcal disease? In: Abstracts of the Tenth International Pathogenic Neisseria Conference; 1996; Baltimore, MD:3–8.
8. Brinton CC Jr, Wood SW, Brown A, et al. The development of a neisserial pilus vaccine for gonorrhea and meningococcal meningitis. In: Robbins JB, Hill JC, Sadoff JC, eds. Bacterial Vaccines. Seminars in Infectious Diseases. Vol 4. New York: Thieme-Stratton, 1982: 140–159.
9. Plummer FA, Simonsen JN, Chubb H, et al. Epidemiologic evidence for the development of serovar-specific immunity after gonococcal infection. J Clin Invest 1989; 83: 1472–1476.
10. Schneider H, Cross AS, Kuschner RA, et al. Experimental human gonococcal urethritis: 250 Neisseria gonorrhoeae MS11mkC are infectious. J Infect Dis 1995; 172: 180–185.
11. Schneider H, Schmidt KA, Skillman DR, et al. Sialylation lessens the infectivity of Neisseria gonorrhoeae MS11mkC. J Infect Dis 1996; 173: 1422–1427.
12. Schneider H, Griffiss JM, Boslego JW, Hitchcock PJ, Zahos KM, Apicella MA. Expression of paragloboside-like lipooligosaccharides may be a necessary component of gonococcal pathogenesis in men. J Exp Med 1991; 174: 1601–1605.
13. Cohen IR, Kellogg DS Jr, Norins LC. Serum antibody response in experimental human gonorrhoeae. Br J Vener Dis 1969; 45: 325–327.
14. McMillan A, McNeillage G, Young H. Antibodies to Neisseria gonorrhoeae: a study of the urethral exudates of 232 men. J Infect Dis 1979; 140: 89–95.
15. Ellison RT III, Curd JG, Kohler PF, Reller LB, Judson FN. Underlying complement deficiency in patients with disseminated gonococcal infection. Sex Transm Dis 1987; 14: 201–204.
16. Lee TJ, Utsinger PD, Snyderman R, Yount WJ, Sparling PF. Familial deficiency of the seventh component of complement associated with recurrent bacteremic infections due to Neisseria. J Infect Dis 1978; 138: 359–367.
17. Harkness AH. The pathology of gonorrhoea. Br J Vener Dis 1948; 24: 137–147.
18. Guidelines for treatment of sexually transmitted diseases. Centers for Disease Control and Prevention. MMWR Morb Mortal Wkly Rep 1998; 47(No. RR-1):59–60.
19. Schmidt KA, Deal CD, Kwan M, Thattassery E, Schneider H. Neisseria gonorrhoeae MS11mkC opacity protein expression in vitro and during human volunteer infectivity studies. Sex Transm Dis 2000; 27: 278–283.
20. Blake MS, Gotschlich EC. Purification and partial characterization of the opacity-associated proteins of Neisseria gonorrhoeae. J Exp Med 1984; 159: 452–462.
21. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 1951; 193: 265–276.
22. Laemmli UK. Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature (London) 1970; 227: 680–685.
23. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 1979; 76: 4350–4354.
24. Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning Cold Spring Harbor. New York: Cold Spring Harbor Laboratory Press, 1989.
25. Hicks CB, Boslego JW, Brandt B. Evidence of serum antibodies to Neisseria gonorrhoeae before gonococcal infection. J Infect Dis 1987; 155: 1276–1281.
26. Griffiss JM, Jarvis GA, O’Brien JP, Eads MM, Schneider H. Lysis of Neisseria gonorrhoeae initiated by binding of normal human IgM to a hexosamine-containing lipooligosaccharide epitope(s) is augmented by strain-specific, properdin-binding–dependent alternative complement pathway activation. J Immunol 1991; 147: 298–305.
27. Barnes RC, Holmes KK. Epidemiology of gonorrhea: current perspectives. Epidemiol Rev 1984; 6: 1–30.
28. Sherrard J, Barlow D. PPNG at St. Thomas’ Hospital: a changing provenance. Int J STD AIDS 1993; 4: 330–332.
29. Cohen MS, Cannon JG, Jerse AE, et al. Human experimentation with Neisseria gonorrhoeae: rationale, methods, and implications for the biology of infection and vaccine development. J Infect Dis 1994; 169: 532–537.
30. Schofield CBS. Some factors affecting the incubation period and duration of symptoms of urethritis in men. Br J Vener Dis 1982; 58: 184–187.
31. Holmes KK, Johnson DW, Trostle HJ. An estimate of the risk of men acquiring gonorrhea by sexual contact with infected females. Am J Epidemiol 1970; 91: 170–174.
32. Johnson DW, Holmes KK, Kvale PA, Halverson CW, Hirsch WP. An evaluation of gonorrhea case findings in the chronically infected female. Am J Epidemiol 1969; 90: 438–448.
33. Sandström EG, Knapp JS, Reller LB, et al. Serogrouping of Neisseria gonorrhoeae: correlation of serogroup with disseminated gonococcal infection. Sex Transm Dis 1984; 11: 77–80.
34. Petersen BH, Lee TJ, Snyderman R, Brooks GF. Neisseria meningitidis and Neisseria gonorrhoeae bacteremia associated with C6, C7, or C8 deficiency. Ann Intern Med 1979; 90: 917–920.
35. Lammel CJ, Sweet RL, Rice PA, et al. Antibody–antigen specificity in the immune response to infection with Neisseria gonorrhoeae. J Infect Dis 1985; 152: 990–1001.
36. Zak K, Diaz JL, Jackson D, Heckels JE. Antigenic variation during infection with Neisseria gonorrhoeae: detection of antibodies to surface proteins in sera of patients with gonorrhea. J Infect Dis 1984; 149: 166–174.
37. Cohen IR, Norins LC. Natural human antibodies to Gram-negative bacteria: immunoglobulins G, A, and M. Science 1966; 152: 1257–1259.
38. Cohen IR. Natural and immune human antibodies reactive with antigens of virulent Neisseria gonorrhoeae: Immunoglobulins G, M, and A. J Bacteriol 1967; 94: 141–148.
39. Cannon JG, Black WJ, Nochamkin I, Stewart PW. Monoclonal antibody that recognizes an outer membrane antigen common to the pathogenic Neisseria species but not to most nonpathogenic Neisseria species. Infect Immun 1984; 43: 994–999.
40. Kim JJ, Mandrell RE, Griffiss JM. Neisseria lactamica and Neisseria meningitidis share lipooligosaccharide epitopes, but lack common capsular and class 1, 2, and 3 protein epitopes. Infect Immun 1989; 57: 602–608.
41. Mandrell RE, Griffiss JM, Macher BA. Lipooligosaccharides (LOS) of Neisseria gonorrhoeae,Neisseria meningitidis have components that are immunochemically similar to precursors of human blood group antigens: carbohydrate sequence specificity of the mouse monoclonal antibodies that recognize cross-reacting antigens on LOS and human erythrocytes. J Exp Med 1988; 168: 107–126.
42. Griffiss JM, Schneider H. The chemistry and biology of lipooligosaccharides: the endotoxins of bacteria of the respiratory and genital mucosae. In: Brade H, Morrison DC, Opal S, Vogel S, eds. Endotoxin in Health and Disease. New York: Marcel Dekker, 1999: 179–194.
43. Zenilman JM. New paradigms for sexually transmitted diseases surveillance and field studies. Sex Transm Dis 2000; 24: 310–311.
44. Fox KK, Weiner DH, Davis RH, Sparling PF, Cohen MS. Longitudinal evaluation of serovar-specific immunity to Neisseria gonorrhoeae. Am J Epidemiol 1999; 149: 353–358.
45. Buchanan TM, Eschenbach DA, Knapp JS, Holmes KK. Gonococcal salpingitis is less likely to recur with Neisseria gonorrhoeae of the same principal outer membrane protein antigenic type. Am J Obstet Gynecol 1980; 138: 978–980.
46. Plummer FA, Chubb H, Simonsen JN, et al. Antibodies to opacity proteins (Opa) correlate with a reduced risk of gonococcal salpingitis. J Clin Invest 1994; 93: 1748–1755.
47. Kearns DH, O’Reilly RJ, Lee L, Welsh BG. Secretory IgA antibodies in the urethral exudate of men with uncomplicated urethritis due to Neisseria gonorrhoeae. J Infect Dis 1973; 127: 99–101.
48. Swanson J, Robbins K, Barrera O, et al. Gonococcal pilin variants in experimental gonorrhea. J Exp Med 1987; 165: 1344–1357.
49. Swanson J, Barrera O, Sola J, Boslego JW. Expression of outer membrane protein II by gonococci in experimental gonorrhea. J Exp Med 1988; 168: 2121–2129.
50. Ward ME, Glynn AA. Human antibody response to lipopolysaccharides from Neisseria gonorrhoeae. J Clin Path 1972; 25: 56–59.
51. Gulati S, McQuillen DP, Mandrell RE, Jani DB, Rice PA. Immunogenicity of Neisseria gonorrhoeae lipooligosaccharide epitope 2C7, widely expressed in vivo with no immunochemical similarity to human glycosphingolipids. J Infect Dis 1996; 174: 1223–1237.
52. Gulati S, McQuillen DP, Sharon J, Rice PA. Experimental immunization with a monoclonal anti-idiotope antibody that mimics the Neisseria gonorrhoeae lipooligosaccharide epitope 2C7. J Infect Dis 1996; 174: 1238–1248.
53. Erwin AL, Haynes PA, Rice PA, Gotschlich EC. Conservation of the lipooligosaccharide synthesis locus Igt among strains of Neisseria gonorrhoeae: requirement for IgtE in synthesis of the 2C7 epitope and of the β-chain of strain 15253. J Exp Med 1996; 184: 1233–1241.
54. Yamasaki R, Nasholds W, Schneider H, Apicella MA. Epitope expression and partial structural characterization of F62 lipooligosaccharide (LOS) of Neisseria gonorrhoeae: IgM monoclonal antibodies (3F11 and 1–1-M) recognize nonreducing terminus of F62 LOS components. Mol Immunol 1991; 28: 1233–1242.
55. Kerwood DE, Schneider H, Yamasaki R. Structural analysis of lipooligosaccharide produced by Neisseria gonorrhoeae strain MS11MK (variant A). Biochemistry 1992; 32: 12760–12768.
56. Mandrell RE, Schneider H, Apicella MA, Zollinger WD, Rice PA, Griffiss JM. Antigenic and physical diversity of Neisseria gonorrhoeae lipooligosaccharides. Infect Immun 1986; 54: 63–69.
57. Schneider H, Hale TL, Zollinger W, Seid RC Jr, Hammack CA, Griffiss JM. Heterogeneity of molecular size and antigenic expression within the lipooligosaccharides of individual strains of Neisseria gonorrhoeae and Neisseria meningitidis. Infect Immun 1984; 45: 544–549.
© Copyright 2001 American Sexually Transmitted Diseases Association