INFECTION AND IMMUNITY, July 1998, p. 3223–3231 0019-9567/98/$04.0010 Copyright © 1998, American Society for Microbiology. All Rights Reserved.
Vol. 66, No. 7
Immune Responses against Major Outer Membrane Antigens of Neisseria meningitidis in Vaccinees and Controls Who Contracted Meningococcal Disease during the Norwegian Serogroup B Protection Trial E. WEDEGE,1* E. A. HØIBY,2 E. ROSENQVIST,1
Department of Vaccinology and Department of Bacteriology,2 National Institute of Public Health, N-0403 Oslo, Norway Received 19 September 1997/Returned for modification 8 January 1998/Accepted 22 April 1998
Sera from vaccinees and controls who contracted serogroup B meningococcal disease during the blinded and open parts of a two-dose protection trial in Norway were compared for antigen-specific and bactericidal antibodies against vaccine strain 44/76 (B:15:P1.7,16). From 16 of 20 (80%) vaccinees and 26 of 35 (74%) controls, one or more serum samples (n 5 104) were collected during the acute phase (1 to 4 days), early convalescent phase (5 to 79 days), and late convalescent phase (8 to 31 months) after onset of disease. Binding of immunoglobulin G (IgG) to the major outer membrane antigens (80- and 70-kDa proteins, class 1, 3, and 5 proteins, and lipopolysaccharide [LPS]) on immunoblots was measured by digital image analysis. Specific IgG levels in vaccinees increased from acute to early convalescent phases, followed by a decline, while controls showed a small increase over time. Vaccinees had significantly higher levels than controls against class 1 and 3 porins and LPS in acute sera, against all antigens during early convalescence, and against class 1 and 3 porins in the later sera. Vaccinees who were infected with strains expressing subtype P1.7,16 proteins demonstrated a level of IgG binding to protein P1.7,16 with early-convalescent-phase sera that was fourfold higher than that of those infected with other strains. Bactericidal titers in serum against the vaccine strain were 192-fold higher for vaccinees than those for controls during early convalescence, but similarly low levels were found during late convalescence. A vaccine-induced anamnestic response of specific and functional antibody activities was thus shown, but the decrease in protection over time after vaccination indicated that two vaccine doses did not induce sufficient levels of long-term protective antibodies. the survivors showed significant complement deficiencies (26, 30). During the first year of the open part (1991 to 1992), in which 137,800 vaccinees and 98,600 controls participated, the corresponding numbers with group B disease were 8 and 11, respectively (5). The latter vaccine failures had all been vaccinated in 1988 to 1989; a decrease in protection over time after vaccination was also observed in the blinded part of the trial (6, 50). From most of the patients, one or more serum samples were collected at different times after onset of disease. In the present study, digital image analysis was used to measure the immunoglobulin G (IgG) binding intensities of these sera to the major OMV vaccine antigens on immunoblots. The aim of our study was to compare the quantitated IgG responses of vaccine failures and unvaccinated controls and to analyze the possible associations between these antibody specificities and the bactericidal activities in serum. Our results demonstrated that the group B vaccine had induced immunological memory but that two doses were not sufficient to obtain long-term protective antibody levels. A preliminary immunoblot study of sera from nine of the patients has been published previously (59).
Serogroup B meningococcal disease is a major health problem in many countries throughout the world. Serogroup B polysaccharide vaccine is poorly immunogenic in humans (66), probably because of its structural similarity to sialic acid residues on human cells (20). Therefore, vaccines based on noncapsular surface antigens have been developed and used in several trials (6, 7, 19, 21, 56). In Norway, the high incidence of meningococcal disease, which is caused mainly by B:15:P1.7,16 strains of the ET-5 complex (8, 12, 33), led to a placebo-controlled double-blind protection trial between 1988 and 1991. An outer membrane vesicle (OMV) vaccine from a representative epidemic strain (strain 44/76), which was adsorbed to aluminum hydroxide (24), was given in two doses to 88,800 secondary school students, while 83,000 received the placebo preparation of aluminum hydroxide. After 29 months of observation time, the point estimate for protection against group B meningococcal disease was 57.2% (P 5 0.012) (6). From June 1991, the study continued as an open trial in which 49,000 of the previous placebo controls accepted vaccination (5). In this part of the trial, the 64,600 nonparticipants in the blinded part served as additional controls, since they were proven to have the same risk of contracting meningococcal disease as those given the placebo (5). In the blinded part of the trial, 12 vaccinees and 24 controls contracted systemic group B meningococcal disease. None of
MATERIALS AND METHODS Sera. During the blinded part of the protection trial (1988 to 1991), all meningococcal disease cases were identified by K numbers and procedures were established to collect acute and convalescent-phase sera (32). In the present study, acute-phase sera were defined as sera obtained 1 to 4 days after onset of disease, and early-convalescent-phase sera were defined as those obtained 5 to 79 days after onset. In addition, late-convalescent-phase sera, which were obtained 8 to 31 months after disease and were previously analyzed for immune deficiencies and antibody levels (18, 26, 30, 34), were included in the study. Patients in the subsequent open part of the trial (1991 to 1992) were identified by X numbers; from these late-convalescent-phase sera were not collected. A total of
* Corresponding author. Mailing address: Department of Vaccinology, National Institute of Public Health, P.O. Box 4404 Torshov, N-0403 Oslo, Norway. Phone: 47 22 04 26 99. Fax: 47 22 04 23 01. E-mail: [email protected]
WEDEGE ET AL.
104 serum samples was obtained from 10 of 12 (83%) vaccine failures and 22 of 24 (92%) placebo controls in the blinded part and from 6 of 8 (75%) vaccine failures and 4 of 11 (36%) controls in the open part of the trial. The control group included placebo controls from both parts of the trial and nonparticipants from the open part. From one control, three different acute-phase sera were obtained, while two different early-convalescent-phase samples were collected from 2 vaccinees and 10 controls. Strains. Meningococcal isolates from patients were typed with serotype- and subtype-specific monoclonal antibodies by a dot blot method (64). Nontypeable and nonsubtypeable strains were studied for expression of class 1, 2, or 3 porins in sodium dodecyl sulfate gels (37). Serogroups of culture-negative patient strains were determined by analyzing paired sera for IgG and IgM antibodies against purified polysaccharides by enzyme-linked immunosorbent assay (ELISA) (32). Reference strains M1080 (B:1:P1.1,7) and S3032 (B:NT:P1.12,16) (22) were used to determine the variable region (VR) specificities of the class 1 protein responses on immunoblots as described previously (34, 51, 59). The class 4 proteindeficient mutant of vaccine strain 44/76 was a gift from Milan Blake while he was at Rockefeller University, New York, N.Y. Immunoblotting and quantification of immunoreactive bands. Immunoblotting was carried out as described elsewhere (64) with OMVs from vaccine strain 44/76 as the antigen (24). After electrotransfer of 12% acrylamide gels (7 by 8 cm) to nitrocellulose filters, strips loaded with OMV protein (corresponding to approximately 1.8 mg of protein before electrotransfer) were incubated overnight with 1:200 dilutions of human sera without and with 0.15% Empigen BB (Albright & Wilson, Whitehaven, United Kingdom) to restore partial antigenicity of OMV proteins (40, 61). IgG binding was detected with 1:500 dilution of rabbit anti-human IgG conjugated to horseradish peroxidase (Dakopatts, Glostrup, Denmark). All filters were stained with substrate for 10 min, and care was taken to blot all sera under identical experimental conditions. The intensities of binding of IgG from each serum to the seven OMV antigens (the 80- and 70-kDa high-molecular-mass components, the major class 1, 3, 4, and 5 proteins [including Opc], and lipopolysaccharide [LPS]) was recorded by both visual and digital scanning. By visual determination, the immunoreactive bands were scored on a scale from 0 to 4, where scores between 0 and 1.5, between 2 and 2.5, and between 3 and 4 represented no to weak binding, medium binding, and strong binding, respectively. After storage of the strips in plastic covers in the dark for 3 to 4 years, they were photographed under identical contrast enhancement conditions with a Sony CCD-IRIS black and white video camera equipped with a green filter. Image analysis was performed with the 1-D module of Cream Software from Kem-En-Tec A/S, Copenhagen, Denmark. Band densities were recorded in arbitrary units as integrated peak areas between baseline and curve. When IgG binding to any antigen was higher in the presence of Empigen BB, this value was used in the further analyses. Control strips from each blot, which were incubated with monoclonal antibodies against the class 1, 3, and 4 and Opc proteins, served to identify the positions of the major OMV antigens. ELISA. Binding of IgG to OMVs was analyzed by ELISA (49). Levels of IgG were measured in 1:200 serum dilutions and were recorded as optical densities at 405 nm after 30 min of incubation with substrate. Bactericidal assay. Serum bactericidal activities in microtiter plates were assayed in the presence of 25% human complement and 70 to 80 CFU of strain 44/76-SL (31). Approximately 10% of the colonies of this strain expressed the Opc protein strongly (51). A twofold dilution series of each serum, starting at a 1:2 dilution, was examined. The complement source alone gave no reduction in numbers of CFU after 30 min of incubation and showed no reaction with OMVs on immunoblots. The bactericidal titer (BT) was recorded as the reciprocal of the final serum dilution causing more than 50% killing of the inoculum. A titer of 1 was assigned to sera with less than 50% killing at a 1:2 dilution. Statistical analyses. Differences in binding of IgG to OMV components between sera from vaccine failures and controls were compared by the MannWhitney rank sum test, and correlations were assessed by the Spearman rank order correlation test or regression analyses. Differences in proportions of observations were compared by the chi-square test or the Fisher exact test for five or fewer observations. P values of #0.05 were considered significantly different. The statistical analyses were performed with a SigmaStat program from Jandel GmbH, Erkrath, Germany.
RESULTS Case isolates. The serogroup B isolates, which caused disease among the 20 vaccinees and 35 controls during the blinded and open parts of the Norwegian protection trial, were analyzed on dot blots and immunoblots with available monoclonal typing reagents and by gel electrophoresis (Table 1). The majority of strains expressed the serotype 15 protein; 15 of 20 strains (75%) among vaccinees and 24 of 35 strains (69%) among controls carried this protein. All disease isolates, except the single B:16:2 strain, expressed a class 1 protein. Approximately 30% of both vaccine failures and controls contracted strains expressing P1.7 and/or P1.16 epitopes on their class 1
INFECT. IMMUN. TABLE 1. Typing characteristics of serogroup B isolates causing systemic meningococcal disease among vaccine failures and controls during the Norwegian protection trial (1988 to 1992) No. of patients infected
Serotype and subtype of disease isolate
15:P1.7,16 15:P1.7h,16a 15:P1.7,16cb 15:P1.7 15:P1.7h 15:P1.7,2 15:P1.5,2 15:P1.12 15:P1.12,13 15:P1.12,13ac 15:2d 4:P1.15 4:P1.7h,4e 4:2d 8:P1.15 16:2f NT:P1.16g Lost/unknownh
7 1 1 0 0 0 0 0 1 4 1 0 1 1 0 0 0 3
7 3 1 2 1 1 5 1 1 2 0 1 0 1 1 1 1 6
P1.7 , masked P1.7 epitope (62). Point mutation in P1.16 epitope (52). One threonine less than the P1.13 epitope (65). d Class 1 protein of unknown subtype. e This strain showed subtype P1.4,14 on dot blots. f No class 1 protein detected. g Class 3 protein of unknown serotype. h Strains lost or negative after culturing. PCR identified porA from one vaccinee and three controls (13). b c
proteins. This number increased to approximately 50% when strains with different P1.7 and P1.16 variants, i.e., the masked P1.7 epitope (P1.7h) and P1.16c (52, 62), were taken into account. Neither the observation that more vaccinees than controls (25 versus 11%) were infected with strains expressing subtype P1.12 nor the observation that no vaccine failure contracted 15:P1.2,5 strains, the second-most-frequent strain causing disease among controls, were significant. Lost or culturenegative strains, which were shown by other methods to be serogroup B (32), accounted for approximately 15% of the causative strains in each patient group. These data indicated that vaccinees and unvaccinated controls were infected to the same extent with vaccine-like strains. Identification of the class 1 proteins of some of the unknown or lost strains by PCR of cerebrospinal fluids (13) did not change this result. Patient sera. A total of 104 patient serum samples (37 from vaccine failures and 67 from controls) were analyzed by immunoblotting (Table 2). They represented one or more samples, which were drawn at different times after disease onset, from 16 of the 20 (80%) vaccinees and from 26 of the 35 (74%) controls. The sera were divided into three groups according to the time of sampling after disease onset, namely, acute-phase sera (1 to 4 days), early-convalescent-phase sera (5 to 79 days), and late-convalescent-phase sera (8 to 31 months). Mean values for collection from vaccine failures were 1.7 days for acutephase sera, 23.8 days for early-convalescent-phase sera, and 17.1 months for late-convalescent-phase sera. These numbers were not significantly different from the corresponding numbers (1.7, 24.6, and 18.8, respectively) of the control group. Paired acute-phase and early-convalescent-phase sera were ob-
Serum after onset of disease (mean [days]) No. of sera n Fold in(%)b creasec
Range (median)a n Fold in(%)b creasec
Range (median)a n (%)b
Class 3 protein
IgG binding of the following OMV antigens:
n Fold in(%)b creasec Fold increasec n Fold in(%)b creasec
Class 5 proteins
Range (median)a Range (median)a
LPS n (%)b
Class 1 protein
0–983 (243) 3 (23) 367–3,806 (1,929d,f) 13 (93) 0–8,186 (2,020 f) 7 (54) 3,424–19,687 (11,481d,f) 14 (100)
2.8 223–5,311 (1,016 f) 7 (54) 3,235–6,021 (4,555d, f) 14 (100)
235–1,345 (561 f) 4 (31) 694–6,031 (1,598d, f) 12 (86)
0–2,761 (10) 4 (31) 0–1,592 (747 f) 8 (57)
0–950 (199) 2 (15) 228–3,561 (2,028d, f) 13 (93)
0–1,906 (429) 757–6,617 (4,066)
3.3 737–6,349 (3,511d, f) 8 (80)
453–5,551 (1,859d, f) 8 (80)
3 (20) 17 (57)
0–3,034 (326) 5 (50)
0–8,432 (1,106) 0–11,369 (3,604e)
3 (20) 18 (60)
48–1,619 (334) 0–1,558 (576)
149–542 (388) 0–3,046 (437) 0 (0) 0 (0)
6.4 0–373 (171) 0–772 (165)
3 (20) 3 (10) 1.1
0–905 (10) 0–1,565 (64) 0 (0) 2 (7)
0 (0) 10 (33)
0–275 (177) 0–1,819 (328e)
0–5,188 (602e) 13 (59)
1.0 106–2,976 (711e)
3 (14) 4.7
0–1,831 (10) 8 (36)
2.5 0–8,669 (806e)
6 (27) 1.6
0–1,247 (289e) 6 (27)
TABLE 2. Digital scanning analysis of binding intensity of IgG to immunoblotted OMV antigens with acute-phase and early- and late-convalescent-phase sera from vaccinees and controls contracting group B meningococcal disease during the Norwegian protection trial
Vaccine failures Acute phase (1.7) Early convalescent phase (23.8) Late convalescent phase (17.1 mo)
Controls Acute phase (1.7) Early convalescent phase (24.6) Late convalescent phase (18.8 mo)
a Densitometric values are integrated peak areas in arbitrary units. b n is number of sera that have values greater than cutoff values. Cutoff points differentiate between visually weak and more distinctly stained bands. For 80-kDa, 70-kDa, and class 5 proteins, cutoffs 5 500; for class 1 and 3 proteins, cutoffs 5 1,000; and for LPS, cutoff 5 2,000. c Compared to levels of corresponding acute-phase sera. d Significantly different (P , 0.05) from median values of acute-phase sera from vaccine failures. Significantly different (P , 0.05) from median values of acute-phase sera from controls. Significantly different (P , 0.05) from median values of corresponding control sera. e
tained from 10 (50%) and 12 (34%) of the vaccinees and controls, respectively. Comparison between visual and scanning quantitation of immunoreactive bands. The intensities of binding of IgG of all 104 sera to the different OMV antigens (80- and 70-kDa proteins, class 1, 3, and 5 proteins, and LPS) were first scored visually, on a scale from 0 to 4, and then quantitated by digital scanning after storage in the dark for 3 to 4 years. The strips showed no evidence of fading under these storage conditions. A comparison was made between visual scoring and densitometric scanning of the immunoreactive bands (data not shown). For the two vaccine antigens with molecular masses of 80 and 70 kDa, the correlation coefficients between these two measurements were 0.90 and 0.84, respectively. The corresponding coefficients were 0.79, 0.94, and 0.88 for the class 1, 3, and 5 proteins, respectively, and 0.93 for LPS (all P , 0.001). A good correlation between visual scoring and digital scanning was thus found for the binding of IgG to the major OMV vaccine antigens. This even applied to the class 5 (including Opc) proteins and LPS, which often showed two or more bands on blots. In these cases, the sum of the band intensities was used. In previous reports on antibody responses, which were measured by visual scoring of immunoblots, a distinction was made between weakly stained bands, which were ignored, and those of medium or stronger staining (34, 51). From plots between visual scores and the corresponding densitometric values (data not shown), the approximate scan value in arbitrary units which discriminated visually weak from medium stained bands was obtained for each OMV antigen. These cutoff values depended on the band widths of the immunoreactive antigens. For the more narrow immunoreactive bands, shown by the 80and 70-kDa and class 5 proteins, cutoff values were about 500, whereas they were about 1,000 for the broader class 1 and 3 protein bands. For the even broader LPS bands, cutoff was about 2,000. Scan values higher than the cutoffs corresponded to visually distinct immunoreactive bands. Since the scan values depend somewhat on the contrast conditions used to acquire the image of the strips, all strips were photographed under identical conditions. Comparison of IgG levels determined by ELISA and by digital scanning of all immunoreactive bands. Total levels of IgG against OMVs in the patient sera were analyzed by ELISA (data not shown). In this method, the coating OMV antigen is less denatured than in immunoblotting and more antibodies against conformational epitopes will be detected. Since digital image analysis of blots gave a quantitative measurement of IgG levels, it was of interest to examine if there was any association between the sum of scanned bands with the IgG levels measured in ELISA. For this analysis, all immunoreactive bands on the blots were included. No serum with a high ELISA value and a low scan value was found. The correlation coefficients for the two methods were 0.77 and 0.59 (P , 0.001), respectively, for all sera from vaccinees and controls. These results indicated that the intensities of the band patterns on blots reflected the IgG levels of the corresponding sera. Binding of IgG to major OMV antigens with sera from vaccine failures and controls. Table 2 shows the densitometric analyses of binding of IgG to the individual OMV antigens with the 104 serum samples taken at different times after onset of disease from vaccinees and controls. The results are given as ranges, medians, fold increases relative to the levels of the acute-phase sera and as the significance of differences between and within the two patients groups. Also, the number of sera with band intensities higher than the cutoff values, which represented visually distinct bands (see above), is presented. For most antigens, large individual variations in antibody responses were observed. Because the amounts of the various antigens in 3225
WEDEGE ET AL.
FIG. 1. Binding of IgG to OMV vaccine protein P1.7,16 at different times after onset of disease in vaccine failures infected with serogroup B meningococci of different subtypes. Sera were collected during acute phase (1 to 4 days), early convalescent phase (5 to 79 days), and late convalescent phase (8 to 31 months) after onset of disease from vaccinees infected with strains expressing either P1.7, P1.16, or variant epitopes thereof (P1.7/16) or with strains expressing other or unknown subtypes (P1.n/x). IgG binding was measured by digital scanning of immunoblots; bars represent median scan values in arbitrary units. Only earlyconvalescent-phase sera showed significant differences in antibody binding (P 5 0.003). conv., convalescent.
the OMVs are different and because their transfer to the filters during electrotransfer may be dissimilar, care should be taken in comparing IgG levels between the different antigens. 80- and 70-kDa antigens. The 80- and 70-kDa antigens constitute 3% or less of the total protein content of OMVs (23, 24). The 70-kDa protein reacts with a monoclonal antibody against iron-regulated protein FrpB. The nature of the 80-kDa antigen is unknown (23), but it is strongly immunogenic after vaccination with both our OMV vaccine (51, 59) and a recombinant PorA OMV vaccine (16, 46). During the acute phase, IgG binding to the 80-kDa protein was low for both vaccinees and controls. It increased by more than 10-fold during early convalescence of the vaccine failures, and the levels were significantly higher than those for the controls (P , 0.0001). Nearly all early-convalescent-phase sera from vaccinees displayed a distinct band, compared to 33% of the controls. Antibody levels of the vaccinees then decreased and were not significantly different from those of the late-control-phase sera. The controls showed a slight increase in antibody binding over time after disease. Both early- and late-convalescent-phase sera had significantly higher levels (P , 0.008) than acutephase sera. The 70-kDa antigen also elicited a significantly higher response among vaccinees than controls during early convalescence (P 5 0.001). The apparent high fold increase was less relevant because of the very low level of median binding of the acute-phase sera. This was the only antigen for which a significant increase in antibody levels after disease was not demonstrated within the two patient groups. Class 1 protein. Vaccinees responded significantly higher to the P1.7,16 protein at all times after disease compared to controls (P , 0.009). During early convalescence of vaccinees, IgG binding increased by approximately threefold, with little change thereafter (Table 2). The majority (86%) of the earlyconvalescent-phase sera from vaccinees had IgG bands higher than the cutoff, whereas only two control sera (7%) showed similar distinct bands. They were obtained 9 and 12 days after onset of disease from the one individual (K-177) contracting a NT:P1.16 strain (Table 1). Control sera demonstrated a slight, significant increase in IgG binding over time (P 5 0.04). We also analyzed if infection of vaccinees with meningococci expressing vaccine-like class 1 proteins would boost vaccine-
induced antibodies against the P1.7,16 class 1 protein (Fig. 1). Acute-phase sera from vaccinees, which were infected with strains carrying P1.7, P1.16, or their epitope variants (P1.7/16), demonstrated low levels of binding to class 1 OMV protein that were similar to those for sera from vaccinees infected with strains of other subtypes (P1.n/x). During early convalescence, the median antibody level of the P1.7/16 group increased by ninefold and they showed fourfold higher binding than the other group (P 5 0.003). The levels of the P1.7/16 group then decreased, and with the late-phase sera there was no significant difference between the two groups of vaccinees. We also compared the class 1 protein responses of the P1.1x/n vaccine group and those of all controls (Table 2) during early convalescence. The controls had significantly lower (P , 0.001) levels than the vaccine group, suggesting that the vaccine had primed for a partial subtype-specific response against the P1.7,16 protein. Previously, we reported that the P1.16 epitope, which is located on loop 4 of the P1.7,16 protein, seemed to be more immunogenic than the P1.7 epitope on loop 1 following vaccination (34, 51, 59). This was also found to be the case for the patient sera. Late-convalescent-phase sera with immunoreactive class 1 porin bands higher than the cutoff (Table 2) were incubated with the two meningococcal reference strains M1080 (B:1:P1.7,1) and S3032 (B:NT:P1.12,16). Each of these expresses one of the subtype epitopes of the P1.7,16 protein. By visual examination, five of the eight serum samples from the vaccinees and four of the six serum samples from controls bound distinctly to the class 1 protein of S3032, but not to M1080, implying antibodies directed against the P1.16 epitope. All but one of the P1.16-positive sera were from individuals infected with strains expressing subtype P1.7,16, or its variants, or with culture-negative strains, which were demonstrated by PCR to express vaccine-like class 1 proteins (13). Thus, infection of both vaccinees and controls with vaccine-like strains mainly induced class 1 porin antibodies directed against the P1.16 epitope. Class 3 protein. Also for the class 3 porin, the responses among the vaccine failures were significantly higher at all times after disease compared to controls (P , 0.009). Totals of 54 and 100%, respectively, of the acute- and early-convalescentphase sera from the vaccinees had antibody levels higher than the cutoff, whereas no control sera showed a corresponding band intensity (Table 2). During early convalescence of the vaccinees, class 3 porin antibodies increased by 4.5-fold (P 5 0.001), and the levels decreased only slightly with time. Control sera demonstrated a significant increase in the level of antibody binding from the acute to the late convalescent phases (P , 0.001). Only eight late-convalescent-phase sera from controls showed class 3 porin bands higher than the cutoff (Table 2). The majority of these controls had been infected with serotype 15 strains. Since distinct bands were not seen with their earlier sera, this suggested a late development of class 3 porin antibodies. Class 5 proteins. Antibodies against class 5 proteins include those binding to the Opc protein as well as to the P5.5 Opa protein in the vaccine (24). Vaccinees responded significantly higher to the class 5 proteins during early convalescence (P 5 0.0001) compared to controls (Table 2). Their antibody levels increased by eightfold (P 5 0.0001) compared to a less-thantwofold increase for the controls. Both groups had similar levels in the acute- and late-convalescent-phase sera. Control sera showed a significant increase in their anti-class 5 protein levels from the acute to the late convalescent phases (P 5 0.01).
VOL. 66, 1998
IMMUNE RESPONSES OF PATIENT SERA
TABLE 3. BTs of early- and late-convalescent-phase sera from vaccine failures and controls contracting serogroup B meningococcal disease in the Norwegian protection trial Source of serum
Vaccine failures Early-convalescentphase sera Late-convalescentphase sera
FIG. 2. Binding of IgG to L3 and L8 bands of LPS in the OMV vaccine at different times after onset of disease in vaccine failures (vacc) and controls (cont). For serum samples and IgG measurements, see the legend to Fig. 1. Acute- and early-convalescent-phase sera from vaccinees had significantly higher L8 antibody levels than controls (P , 0.002), whereas only early-convalescentphase sera showed significant differences for the L3 band (P 5 0.02). conv., convalescent.
LPS. Blots generally showed one or two bands of equal width in the LPS region. The one of higher molecular weight (L3 band) was recognized by monoclonal antibodies specific for immunotypes L3, L3,7, and L3,7,9. The lower-molecularmass band (L8 band) bound monoclonal antibodies specific for immunotypes L8 and L1,8,10 as well as two monoclonal antibodies specific for the conserved inner core (2, 3). Some sera with strong LPS bands also gave a band moving slightly behind the L3 band. The LPS antibody levels listed in Table 2 are the sum of these bands. Vaccinees responded to infection with significantly higher levels in both acute (P 5 0.02)- and earlyconvalescent (P , 0.0001)-phase sera compared to controls. They showed an increase of approximately sixfold in IgG binding, but during the late phase the two patient groups had similar low levels. For the controls, the anti-LPS responses of both early (P 5 0.01)- and late (P 5 0.04)-convalescent-phase sera were significantly higher than those for acute-phase sera. IgG binding to each of the two LPS bands was also examined (Fig. 2). L8 bands were significantly stronger (P , 0.002) during both acute- and early-convalescent phases of vaccinees compared to controls, whereas the L3 antibody level was only significantly higher (P 5 0.02) during early convalescence. No significant differences between the two patient groups were detected with the late-convalescent-phase sera. For the vaccine failures, antibody levels against the L3 and L8 bands increased by 7.5-fold (P 5 0.001) and 2.4-fold (P 5 0.004), respectively, from the acute phase to the early convalescent period. Class 4 protein and other OMV antigens. Many sera showed a band in the narrow region between the class 3 and Opc bands, which corresponded to the position of the class 4 protein as shown by a class 4 protein-specific monoclonal antibody. Several experiments (data not shown) indicated that an unknown antigen comigrated with the class 4 protein and thus interfered with the estimation of antibody levels. In addition, bands other than those represented by the major OMV antigens were seen on many OMV blots. They were mainly positioned between the 70-kDa antigen and class 1 porin or between the class 5 proteins and LPS and were generally more distinct with sera from vaccine failures than with those from controls. Bactericidal activity of patient sera. Bactericidal activities of all but two early- and late-convalescent-phase sera were determined with variant 44/76-SL expressing low levels of Opc pro-
Controls Early-convalescentphase sera Late-convalescentphase sera
No. of sera
Range of BTsa (median)
No. (%) with: BT of 1b
BT $ 128
1–1,024 (192 )
BT against strain 44/76-SL. BT of 1 implies no bactericidal activity at a 1:2 serum dilution. c Significantly different from corresponding control sera by the Mann-Whitney rank sum (P , 0.001). d Significantly different from corresponding control sera by the Fisher exact test (P , 0.002). b
tein. Acute-phase sera were not analyzed because of the regular presence of antibiotics. During early convalescence, the median level of bactericidal antibodies was nearly 200-fold higher among vaccine failures than among controls (P , 0.001), whereas there was no significant difference between the late-convalescent-phase samples (Table 3). The change in titers over time after disease within each patient group was not significant. Totals of 75 and 40% of the vaccinees’ early- and late-convalescent-phase sera, respectively, showed high titers (BT $ 128), whereas less than 10% of the control sera had similar titers. Significantly more controls than vaccinees (P 5 0.002) had no serum bactericidal activity (BT 5 1) for the early-convalescent-phase sera. Two serum samples, which were drawn at different times during early convalescence from 2 vaccinees and 10 controls (see Materials and Methods), allowed us to examine the variation of bactericidal antibody levels within that time period. Six of the paired control sera had no titers (BT 5 1). All but one of the remaining pairs showed a decrease in activity (Fig. 3). The high BTs of the sera from the two vaccine failures (K-80 and X-100) decreased within a period of 5 to 6 weeks. Thus, they decreased from BTs of 512 to 64 between days 11 and 45 after onset of disease for K-80 and from BTs of 1,024 to 128 between days 19 and 60 for X-100. These titers were higher throughout the study period than those of the paired control
FIG. 3. BTs against vaccine strain 44/76-SL of paired sera collected after onset of disease. Individuals designated K-80 and X-100 were vaccine failures, and K-168, K-177, K-188, and X-78 were controls who suffered from group B meningococcal disease during the protection trial.
WEDEGE ET AL.
TABLE 4. Correlations between BTs in serum and binding of IgG to the major OMV vaccine components of convalescent-phase sera from vaccine failures and controlsa Antigen
80-kDa protein 70-kDa protein Class 1 protein Class 3 protein Class 5 protein LPS
Spearman rank order correlation coefficientb Vaccine failures
0.46 (P 5 0.03) NSc 0.61 (P 5 0.003) NS 0.64 (P 5 0.001) NS
NS NS 0.45 (P , 0.001) NS NS 0.46 (P , 0.001)
a IgG binding was determined by digital scanning. Values are for early- and late-convalescent-phase sera. b Statistical analyses performed within each patient group. c NS, not significant (P . 0.05).
sera (K-168, K-188, and X-78), which showed similar decreases in the lower BT range. In contrast, BTs for the sera from the remaining control (K-177) increased from 32 to 128 between days 9 and 12 after onset of disease. This individual was the only control with a distinct class 1 protein response on immunoblots. Convalescent-phase sera were also tested for killing activities against strain 44/76-1, a variant expressing high levels of Opc proteins (51). Only a single serum sample from a control, who was infected with a B:15:P1.2,5 strain, showed distinctly higher titers with this variant, that is BT of 256 with 44/76-1 versus BT of 4 with 44/76-SL. On blots, this serum sample demonstrated a medium-stained Opc band. These results suggested that infection in general did not induce high levels of bactericidal antibodies against the Opc protein. Correlation between bactericidal activity and immunoreactive band intensities. The association between BTs of convalescent-phase sera and their IgG binding to the individual OMV antigens, which was measured by densitometric scanning, was examined (Table 4). Significant correlation coefficients between the BTs and antibody levels against the 80-kDa, class 1, and class 5 proteins were observed for all early- and late-convalescent-phase sera from the vaccine failures. A similar analysis of control sera showed significant correlations for class 1 protein and LPS. For the two LPS components, antibodies against the L3 band demonstrated a significant relation with BTs for both vaccinees (rs 5 0.61 [P 5 0.003]) and controls (rs 5 0.47 [P , 0.001]). No significant correlation was found for the L8 antibodies in vaccinees, whereas the controls only showed low correlation (rs 5 0.28 [P 5 0.042]) for these antibodies. Multiple-linear, forward-stepwise, or best-subsets regression analyses of control sera gave the same results as those found by the Spearman rank test (Table 4). For the vaccine group, multiple-linear regression analysis demonstrated that none of the specific antibody levels made any significant contribution to BTs. Forward-stepwise regression showed a significant contribution of class 5 protein antibodies (P 5 0.002), while bestsubsets regression analysis indicated that this specificity as well as the antibody levels against class 1 protein and LPS significantly predicted the BTs (P 5 0.002 to 0.042). DISCUSSION In this report, sera from vaccinees and unvaccinated controls who contracted group B meningococcal disease during the blinded and first year of the open part of the Norwegian protection trial (1988 to 1992) were compared for specific and
functional antibodies. Together with an analysis of the case isolates, the results allowed us to draw some conclusions as to why the vaccine failed to protect some individuals in the trial. Strains which caused disease among the vaccine failures and controls showed the same distribution of sero- and subtypes (Table 1), confirming previous antigenic and genetic studies of isolates from the blinded part of the trial (9, 25). This was also true when the results from a recent PCR analysis of spinal fluids from four of the patients, whose isolates were lost or culture negative, were included (13). Because the proportions of class 5 proteins and major LPS immunotypes were also similar for isolates from vaccinees and controls in the blinded part of the trial (25), our data support the conclusion that the disease of the vaccinees was not likely to be caused by a predominance of escape strains. Sera were collected from the majority of vaccinees and controls who were infected during the protection trial. These samples were divided into acute-, early-, and late-convalescentphase sera, corresponding to approximately 2 days, 24 days, and 18 months, respectively, after onset of disease for both patient groups. The levels of binding of IgG antibodies of these sera to each of the major vaccine antigens, i.e., the 80-kDa, 70-kDa (FrpB), class 1 (subtype P1.7,16), class 3 (serotype 15), and class 5 proteins and LPS, were measured by digital image analysis of immunoblots. The levels of antibodies to class 4 protein were not determined, because an antigen of an unknown nature interfered with the measurements. Such antigens have been described for gonococci and meningococci (1, 4). Although both vaccinated and unvaccinated control patients showed large individual variations in their specific antibody levels, statistical analysis established that the vaccinees had been primed for a higher IgG response. In agreement with other reports (10, 17, 18, 28, 29, 38, 39, 45, 48), the specific antibody levels of both vaccinees and controls were low during the acute phase, i.e., approximately 2 days after onset of disease. Only the responses against the two porins and LPS were somewhat higher for the vaccine failures than for the controls (Table 2). However, about 3 weeks later, the vaccinees showed higher responses to all major vaccine components compared to the controls. This distinct increase in specific antibody levels was not observed for the control group. Higher antiporin activities during convalescence of vaccine failures compared to nonvaccinees have also been reported previously (18, 28, 29, 67). The responses of the vaccinees waned from the early to the late convalescent period, so that their late-convalescentphase sera, which were collected about 1.5 years after onset of disease, had the same low levels of specific antibodies to most OMV components as the acute-phase sera. Only the class 1 and 3 porin antibody levels remained high during the late phase of the vaccinees. These porin specificities were also the only ones that were significantly higher than those for the controls during late convalescence. The reduction in antiporin levels during late convalescence, which has been described by others (28, 29, 34, 43), was less evident from our analyses (Table 2). This may be due to an underestimation by the scanning method of strong signals due to color saturation at the single serum dilution used for the blotting experiments (60). For the controls, the slow and modest increases after onset of disease in antibody binding to all major OMV components, except for the 70-kDa antigen (Table 2), was probably caused by carriage of meningococci, which was prevalent in this teenage group (14), or of other antigen-related organisms. The ratio between the L3 and L8 bands of LPS in the vaccine has been estimated to be approximately 3:1 in silver-stained sodium dodecyl sulfate gels (1a, 2). This, together with the
VOL. 66, 1998
specific IgG levels (Fig. 2), indicated that LPS components in the L8 band were more immunogenic after onset of disease in vaccinees than those in the L3 band, containing the host-like lacto-N-neotetraose structure (35, 44). Similar results were also found from blotting studies of healthy vaccinees’ sera (51). Our demonstration of a priming effect of the vaccine rests on an immunoblot analysis, and the ability to measure antibodies against conformational epitopes by this method has been questioned. However, different studies suggest that the blotting method may not be limited only to detection of antibodies against linear epitopes. We found significant correlations between antibody levels, which were measured as all immunoreactive bands on the blots, and the corresponding IgG levels by ELISA with native OMVs. Only a few sera showed detergentdependent binding to blotted class 1 and 5 proteins when Empigen BB was present as a refolding agent (40, 61). Antibodies against native proteins have been shown to assist in refolding of denatured proteins (11, 15), and bovine serum albumin in the incubation buffer can increase macromolecular interactions (58). It is thus feasible that some renaturation of blotted antigens may occur during the overnight incubation through different molecular mechanisms. BTs of early-convalescent-phase sera from vaccine failures were significantly higher than those for controls (Table 3), supporting a previous study (67). However, the bactericidal antibodies declined rather rapidly during this period for both vaccinees and controls (Fig. 3). From statistical analyses, the contributions of specific OMV antibody levels to the bactericidal activity of the convalescent-phase sera were somewhat different for the two patient groups (Table 4). BTs of vaccinees’ sera correlated with antibody levels against the 80-kDa, class 1, and class 5 proteins, while levels against class 1 protein and LPS contributed to the killing activity of control sera. Thus, significant associations between BTs and IgG responses to all major OMV antigens, except for the iron-regulated 70kDa protein, were demonstrated. Since the meningococcal cells in the bactericidal assays were cultivated in an iron-rich medium, it is likely that this protein was not sufficiently expressed to achieve a critical epitope density for the bactericidal antibodies. That these specific antibody activities may contribute to functional activity has also been demonstrated in other studies. Class 1 porins elicit bactericidal antibodies both after disease and vaccination (31, 34, 41, 46, 51, 54, 55, 63, 68), and class 5 proteins induce such activity after vaccination (68). Opc protein also gives rise to bactericidal antibodies in vaccinees, but few convalescent-phase sera have corresponding high activity (51, 53). The latter finding was confirmed from the analysis of the larger number of convalescent-phase sera in the present study. For the 80-kDa antigen, a correlation between BTs and antibody levels has been observed for healthy vaccinees (51). Bactericidal LPS activity in convalescent-phase sera has been demonstrated (27), whereas conflicting results have been presented for vaccine-induced antibodies (51, 57). In the case of bactericidal antibodies against the two LPS components, Moran et al. (42) demonstrated higher titers against L8 than L3 variant strains with convalescent-phase sera from patients with group B meningococcal disease. That we found no contribution of L8 antibodies to BTs may be consistent with the lack of expression of the L8 epitope on the test strain (2). The priming effect of the vaccine was also demonstrated by stronger binding of IgG to vaccine P1.7,16 protein with sera from vaccinees who were infected with strains expressing vaccine-like class 1 porins compared to other vaccinees (Fig. 1). From immunoblot analysis, the specificity of the P1.7,16-positive sera was mainly directed against the P1.16 region in VR2
IMMUNE RESPONSES OF PATIENT SERA
of this protein. A similar result was found when such sera were studied by ELISA with recombinant antigens with inserts of the P1.7 or P1.16 region (34). Because neither of the methods demonstrated activities against the P1.7 (VR1) region, this suggested that VR2 played a major role in inducing antibodies to nonconformational epitopes on the class 1 protein. In addition, others have demonstrated a VR2 response in inhibition ELISA after onset of disease caused by B:15:P1.7,16 strains (10, 36). We also found induction of VR2-specific antibodies after OMV vaccination of healthy volunteers (51, 59). These results were somewhat different from those obtained after vaccination of human volunteers with a hexavalent PorA OMV vaccine, based on our vaccine strain 44/76, in which the P1.7,16 antibodies were directed against VR1 and/or VR2, as shown by peptide ELISA (54). Similar but not always identical specificities were found for the P1.7,16-specific bactericidal antibodies employing isogenic 44/76 strains with deletions of VR1 or VR2 (54, 55). It is possible that our methods will not allow detection of conformation-dependent VR1 antibodies. However, it might be argued that the hexavalent PorA vaccine, which does not contain other major outer membrane proteins (16), may elicit antibodies with epitope specificities partly different from those induced with our OMV vaccine or after onset of meningococcal disease. It is also possible that deletions of the VR1 or VR2 region in the bactericidal test strains (55) altered the conformation of the class 1 protein so that other epitopes became accessible. Previously, antibody responses of healthy volunteers after one, two, and three doses of the OMV vaccine were expressed as percentages of sera with distinct IgG binding to the individual OMV antigens after visual scoring of band intensities (51). From the number of patient sera with specific antibody levels higher than the cutoff values (Table 2) which roughly differentiated between visually weak and more strongly stained bands, a comparison with the responses of healthy vaccinees was possible. Acute-phase sera from vaccine failures showed more distinct antibody binding to class 1 and class 3 proteins and LPS than prevaccination sera from healthy volunteers. Responses of early-convalescent-phase sera from vaccine failures were comparable to those found 6 weeks after the third dose of healthy vaccinees. Acute-phase sera from controls had roughly the same responses as prevaccination sera, but their early-convalescent-phase sera showed little similarity to any of the postvaccination sera. This was most evident for the class 1 and class 3 porin activities, which were low during convalescence of the controls. On the whole, group B meningococcal disease of unvaccinated individuals was reflected in low levels of specific and functional antibodies against strain 44/76. An important question is why some vaccinees contracted group B meningococcal disease during the protection trial. These patients did not suffer from general immune deficiencies (26, 30), nor did we find any evidence for the predominance of escape strains. The present report demonstrates that the vaccine had primed the vaccine failures to respond with higher specific and bactericidal antibody levels compared to unvaccinated controls. However, these antibody activities decreased with time after onset of disease. A decrease in protection with time after vaccination was observed during the blinded part of the trial (6, 50), and all eight vaccine failures in the open part of the trial (1991 to 1992) had been vaccinated in 1988 to 1989 (5). Taken together, these results suggest that the protective antibody levels decreased below a critical limit over time after the two-dose vaccination. It is therefore encouraging that revaccination with a third vaccine dose induces longer-lasting and more-cross-reactive bactericidal antibodies (47, 51) and may thus extend the duration of protection.
WEDEGE ET AL.
ACKNOWLEDGMENTS We are most grateful to T. E. Michaelsen, National Institute of Public Health, Oslo, Norway, for collecting the late-convalescentphase sera; to A. Halstensen, Haukeland Hospital, University of Bergen, Bergen, Norway, for sending us sera from seven patients; and for the cooperation of clinical staffs in many other hospitals throughout Norway to obtain sera. We also thank E. Holten, Akershus Central Hospital, Nordbyhagen, Norway, and L. O. Frøholm, National Institute of Public Health, Oslo, Norway, for supplying the patient isolates and J. T. Poolman, RIVM, Bilthoven, The Netherlands; W. D. Zollinger, Walter Reed Army Institute of Research, Washington, D.C.; C. E. Frasch, Center for Biologics Evaluation and Research, Rockville, Md.; and C. T. Sacchi, Adolfo Lutz Institute, Sao Paulo, Brazil, for the generous gift of specific monoclonal antibodies. K. Bolstad, A. M. Klem, and E. Rønnild are thanked for skillful technical assistance. REFERENCES 1. Andersen, S. R., G. Bjune, E. A. Høiby, T. E. Michaelsen, A. Aase, U. Rye, and E. Jantzen. 1997. Outer membrane vesicle vaccines made from shortchain lipopolysaccharide mutants of serogroup B Neisseria meningitidis: effect of the carbohydrate chain length on the immune response. Vaccine 15: 1225–1234. 1a.Andersen, S. R. Personal communication. 2. Andersen, S. R., G. Bjune, J. Lyngby, K. Bryn, and E. Jantzen. 1995. Shortchain lipopolysaccharide mutants of serogroup B Neisseria meningitidis of potential value for production of outer membrane vesicle vaccines. Microb. Pathog. 19:159–168. 3. Andersen, S. R., J. Kolberg, E. A. Høiby, E. Namork, D. A. Caugant, L. O. Frøholm, E. Jantzen, and G. Bjune. 1997. Lipopolysaccharide heterogeneity and escape mechanisms of Neisseria meningitidis: possible consequences for vaccine development. Microb. Pathog. 23:139–155. 4. Barrera, O., and J. Swanson. 1987. Antigens that are similar in apparent molecular weight to gonococcal outer membrane protein III. J. Clin. Microbiol. 25:1155–1158. 5. Bjune, G., E. A. Høiby, P. Djupesland, E. Rosenqvist, E. Wedege, P. Aavitsland, and L. O. Frøholm. 1994. Outer membrane vesicle vaccine against serogroup B meningococcal disease in Norway, p. 750–751. In J. Einhorn, C. E. Nord, and S. R. Norrby (ed.), Recent advances in chemotherapy. Proceedings of the 18th International Congress of Chemotherapy, Stockholm, Sweden. American Society for Microbiology, Washington, D.C. 6. Bjune, G., E. A. Høiby, J. K. Grønnesby, Ø. Arnesen, J. H. Fredriksen, A. Halstensen, E. Holten, A. K. Lindbak, H. Nøkleby, E. Rosenqvist, L. K. Solberg, O. Closs, J. Eng, L. O. Frøholm, A. Lystad, L. Bakketeig, and B. Hareide. 1991. Effect of outer membrane vesicle vaccine against group B meningococcal disease in Norway. Lancet 338:1093–1096. 7. Boslego, J., J. Garcia, C. Cruz, W. Zollinger, B. Brandt, S. Ruiz, M. Martinez, J. Arthur, P. Underwood, W. Silva, E. Moran, W. Hankins, J. Gilly, J. Mays, and the Chilean National Committee for Meningococcal Disease. 1995. Efficacy, safety, and immunogenicity of a meningococcal group B (15:P1.3) outer membrane protein vaccine in Iquique, Chile. Vaccine 9:821– 829. 8. Bøvre, K., L. O. Frøholm, P. Gaustad, E. Holten, and E. A. Høiby. 1983. Some agent characteristics and their coexistence related to the occurrence and severity of systemic meningococcal disease in Norway, winter 1981–1982. NIPH Ann. 6:75–84. 9. Brooks, J. L., E. Rosenqvist, G. Bjune, P. R. Lambden, and J. E. Heckels. 1995. Comparison of the class-1 outer membrane protein from B:15:P1.16 Neisseria meningitidis strains isolated from patients previously immunized with a serogroup B outer membrane protein vaccine in Norway. Microb. Pathog. 17:425–430. 10. Bundell, K. R., A. F. Fox, M. Guiver, D. M. Jones, and J. T. Poolman. 1987. The development of an inhibition enzyme-linked immunosorbent assay to detect total antibody to the P1.16 subtype antigen of Neisseria meningitidis in human sera. Serodiagn. Immunother. Infect. Dis. 1:353–361. 11. Carlson, J. D., and M. L. Yarmush. 1992. Antibody assisted protein refolding. Bio/Technology 10:86–91. 12. Caugant, D. A., L. O. Frøholm, K. Bøvre, E. Holten, C. E. Frasch, L. F. Mocca, W. D. Zollinger, and R. K. Selander. 1986. Intercontinental spread of a genetically distinctive complex of clones of Neisseria meningitidis causing epidemic disease. Proc. Natl. Acad. Sci. USA 83:4927–4931. 13. Caugant, D. A., E. A. Høiby, L. O. Frøholm, and P. Brandtzæg. 1996. Polymerase chain reaction for case ascertainment of meningococcal meningitis: application to the cerebrospinal fluids collected in the course of the Norwegian meningococcal serogroup B protection trial. Scand. J. Infect. Dis. 28:149–153. 14. Caugant, D. A., E. A. Høiby, P. Magnus, O. Scheel, T. Hoel, G. Bjune, E. Wedege, J. Eng, and L. O. Frøholm. 1994. Asymptomatic carriage of Neisseria meningitidis in a randomly sampled population. J. Clin. Microbiol. 32: 323–330. 15. Chavez, L. G., and D. G. Benjamin. 1978. Antibody as an immunological
20. 21. 22. 23.
26. 27. 28. 29. 30. 31.
probe for studying the refolding of bovine serum albumin. J. Biol. Chem. 253:8081–8086. Claassen, I., J. Meylis, P. van der Ley, C. Peeters, H. Brons, J. Robert, D. Borsboom, A. van der Ark, I. van Straaten, P. Roholl, B. Kuipers, and J. Poolman. 1996. Production, characterization and control of a Neisseria meningitidis hexavalent class 1 outer membrane protein containing vesicle vaccine. Vaccine 14:1001–1008. Danelli, M. G. M., C. M. A. Alves, R. C. Bastos, N. M. Batoreu, D. E. Barroso, J. M. Peralta, and C. E. Frasch. 1996. Human immune response to epitopes on the meningococcal outer membrane class 5 protein following natural infection. FEMS Immunol. Med. Microbiol. 15:159–168. Delvig, A. A., E. Wedege, T. E. Michaelsen, E. A. Høiby, P. Brandtzæg, and E. Rosenqvist. 1996. Immune responses to linear epitopes on the PorB protein of Neisseria meningitidis in patients with systemic meningococcal disease. Microbiology 142:2491–2498. De Moraes, J. C., B. A. Perkins, M. C. C. Camargo, N. T. R. Hidalgo, H. A. Barosa, C. T. Sacchi, I. M. Land Gral, V. L. Gattas, H. D. G. Vasconcelos, B. D. Plikaytis, J. D. Wenger, and C. V. Broome. 1992. Protective efficacy of a serogroup B meningococcal vaccine in Sao Paulo, Brazil. Lancet 340:1074– 1078. Finne, J., M. Leinonen, and P. H. Ma ¨kela ¨. 1983. Antigenic similarities between brain components and bacteria causing meningitis. Implications for vaccine development and pathogenesis. Lancet ii:355–357. Frasch, C. E., G. Coetzee, J. M. Zahradnik, H. A. Feldman, and H. J. Koornhof. 1983. Development and evaluation of group B serotype 2 protein vaccines: report of a group B field trial. Med. Trop. (Marseille) 43:177–180. Frasch, C. E., W. D. Zollinger, and J. T. Poolman. 1985. Serotype antigens of Neisseria meningitidis and a proposed scheme for designation of serotypes. Rev. Infect. Dis. 7:504–510. Fredriksen, J. H., E. Griffiths, E. M. Grotterød, E. A. Høiby, E. Rosenqvist, P. Stevenson, and E. Wedege. 1994. Characterization of high molecular weight components in MenB-vaccine “Folkehelsa,” an outer membrane vesicle vaccine against group B meningococcal disease, p. 818–824. In C. J. Conde-Glez, S. Morse, P. Rice, F. Sparling, and E. Calderon (ed.), Pathobiology and immunobiology of Neisseriaceae. Instituto Nacional de Salud Publica, Cuernavaca, Morelos, Mexico. Fredriksen, J. H., E. Rosenqvist, E. Wedege, K. Bryn, G. Bjune, L. O. Frøholm, A. K. Lindbak, B. Møgster, E. Namork, U. Rye, G. Stabbetorp, R. Winsnes, A. Aase, and O. Closs. 1991. Production, characterization and control of MenB-vaccine “Folkehelsa”: an outer membrane vesicle vaccine against group B meningococcal disease. NIPH Ann. 14:67–80. Frøholm, L. O., D. A. Caugant, E. Holten, E. A. Høiby, E. Rosenqvist, and E. Wedege. 1991. Meningococcal strains isolated from teenage patients during the serogroup B vaccination trial in Norway: serotyping, serosubtyping, immunotyping and clonal analysis. NIPH Ann. 14:139–146. Garred, P., T. E. Michaelsen, G. Bjune, A. Thiel, and A. Svejgaard. 1993. A low serum concentration of mannan-binding protein is not associated with serogroup B or C meningococcal disease. Scand. J. Immunol. 37:468–470. Griffiss, J. M., B. L. Brandt, D. D. Broud, D. K. Goroff, and C. J. Baker. 1984. Immune responses of infants and children to disseminated infections with Neisseria meningitidis. J. Infect. Dis. 150:71–79. Guttormsen, H.-K., L. M. Wetzler, and A. Næss. 1993. Humoral immune response to the class 3 outer membrane protein during the course of meningococcal disease. Infect. Immun. 61:4734–4742. Guttormsen, H.-K., L. M. Wetzler, and C. O. Solberg. 1994. Humoral immune response to class 1 outer membrane protein during the course of meningococcal disease. Infect. Immun. 62:1437–1443. Høgåsen, K., T. Michaelsen, O. J. Mellbye, and G. Bjune. 1993. Low prevalence of complement deficiencies among patients with meningococcal disease in Norway. Scand. J. Immunol. 37:487–489. Høiby, E. A., E. Rosenqvist, L. O. Frøholm, G. Bjune, B. Feiring, H. Nøkleby, and E. Rønnild. 1991. Bactericidal antibodies after vaccination with the Norwegian meningococcal serogroup B outer membrane vesicle vaccine: a brief survey. NIPH Ann. 14:147–156. Høiby, E. A., G. Bjune, L. O. Frøholm, J. Eng, A. Halstensen, E. Rosenqvist, E. Rønnild, and E. Wedege. 1991. The Norwegian meningococcal serogroup B outer membrane vesicle vaccine protection trials: case tracing, meningococcal antigen detection and serological diagnosis. NIPH Ann. 14:107–123. Holten, E. 1978. Serotypes of Neisseria meningitidis isolated from patients in Norway during the first six months of 1978. J. Clin. Microbiol. 9:186–188. Ida ¨npa ¨a ¨n-Heikkila ¨, I., E. A. Høiby, P. Chattopadhyay, U. Airaksinen, T. E. Michaelsen, and E. Wedege. 1995. Antibodies to meningococcal class 1 outer-membrane protein and its variable regions in patients with systemic meningococcal disease. J. Med. Microbiol. 43:335–343. Jennings, H. J., K. J. Johnson, and L. Kenne. 1983. The structure of an R-type oligosaccharide core obtained from some lipopolysaccharides of Neisseria meningitidis. Carbohydr. Res. 121:233–241. Jones, D. M., R. Borrow, A. J. Fox, K. A. V. Cartwright, and T. Riordan. 1994. Human antibody response to the meningococcal class 1 protein, p. 848–852. In C. J. Conde-Glez, S. Morse, P. Rice, F. Sparling, and E. Calderon (ed.), Pathobiology and immunology of Neisseriaceae. Instituto Nacional de Salud Publica, Cuernavaca, Morelos, Mexico.
VOL. 66, 1998 37. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 277:680–685. 38. Mæland, J. A., and E. Wedege. 1989. Serum antibodies to cross-reactive Neisseria outer membrane antigens in healthy persons and patients with meningococcal disease. APMIS 97:774–780. 39. Mandrell, R. E., and W. D. Zollinger. 1989. Human immune response to meningococcal outer membrane protein epitopes after natural infection or vaccination. Infect. Immun. 57:1590–1598. 40. Mandrell, R. E., and W. D. Zollinger. 1984. Use of a zwitterionic detergent for the restoration of the antibody-binding capacity of electroblotted meningococcal outer membrane proteins. J. Immunol. Methods 67:1–11. 41. Milagres, L. G., S. R. Ramos, C. T. Sacchi, C. A. E. Melles, V. S. D. Vieira, H. Sato, G. S. Brito, J. C. Moraes, and C. E. Frasch. 1994. Immune response of Brazilian children to a Neisseria meningitidis serogroup B outer membrane protein vaccine: comparison with efficacy. Infect. Immun. 62:4419–4424. 42. Moran, E. E., B. L. Brandt, and W. D. Zollinger. 1994. Expression of the L8 lipopolysaccharide determinant increases the sensitivity of Neisseria meningitidis to serum bactericidal activity. Infect. Immun. 62:5290–5295. 43. Orren, A., R. E. Warren, P. C. Potter, A. M. Jones, P. J. Lachmann, and J. T. Poolman. 1992. Antibodies to meningococcal class 1 outer membrane proteins in South African complement-deficient and complement-sufficient subjects. Infect. Immun. 60:4510–4516. 44. Pavliak, V., J. R. Brisson, F. Michon, D. Uhrin, and H. J. Jennings. 1993. Structure of the sialylated L3 lipopolysaccharide of Neisseria meningitidis. J. Biol. Chem. 268:14146–14152. 45. Peeters, C. C. A. M., M. Schuller, J. Westdijk, and J. T. Poolman. 1994. Specificity of bactericidal antibody responses in volunteers immunized with meningococcal B15P1.7,16 outer membrane proteins complexed to polysaccharide C and in patients infected with B15P1.7,16, p. 880–887. In C. J. Conde-Glez, S. Morse, P. Rice, F. Sparling, and E. Calderon (ed.), Pathobiology and immunology of Neisseriaceae. Instituto Nacional de Salud Publica, Cuernavaca, Morelos, Mexico. 46. Peeters, C. C. A. M., H. C. Ru ¨mke, L. C. Sunderman, E. M. Rouppe van der Voort, J. Meulenbelt, M. Schuller, A. I. Kuipers, P. van der Ley, and J. T. Poolman. 1996. Phase I clinical trial with a hexavalent PorA containing meningococcal outer membrane vesicle vaccine. Vaccine 14:1009–1015. 47. Perkins, B. A., K. Jonsdottir, H. Briem, E. Griffiths, B. D. Plikaytis, E. A. Høiby, E. Rosenqvist, J. Holst, H. Nøkleby, F. Sotolongo, G. Sierra, C. Huergo, G. M. Carlone, D. Williams, J. Dykes, D. Kapczynski, E. Tikhomirov, J. D. Wenger, and C. V. Broome. 1998. Immunogenicity of two efficacious outer membrane protein-based serogroup B meningococcal vaccines among young adults in Iceland. J. Infect. Dis. 177:683–691. 48. Poolman, J., C. T. P. Hopman, and H. C. Zanen. 1983. Immunogenicity of meningococcal antigens as detected in patient sera. Infect. Immun. 40:398– 406. 49. Rosenqvist, E., S. Hartug, L. O. Frøholm, E. A. Høiby, K. Bøvre, and W. D. Zollinger. 1988. Antibody responses to serogroup B meningococcal outer membrane antigens after vaccination and infection. J. Clin. Microbiol. 26: 1543–1548. 50. Rosenqvist, E., E. A. Høiby, B. Bjune, K. Bryn, D. A. Caugant, O. Closs, J. H. Fredriksen, J. K. Grønnesby, H. Nøkleby, E. Wedege, and L. O. Frøholm. 1992. Meningococcal group B outer membrane vesicle vaccine trial among teen-agers in Norway. Adv. Mod. Biotechnol. La Habana 1:5.2. 51. Rosenqvist, E., E. A. Høiby, E. Wedege, K. Bryn, J. Kolberg, A. Klem, E. Rønnild, G. Bjune, and H. Nøkleby. 1995. Human antibody responses to meningococcal outer membrane antigens after three doses of the Norwegian group B meningococcal vaccine. Infect. Immun. 63:4642–4652. 52. Rosenqvist, E., E. A. Høiby, E. Wedege, D. A. Caugant, L. O. Frøholm, B. T. McGuinness, J. Brooks, P. R. Lambden, and J. E. Heckels. 1993. A new variant of serosubtype P1.16 in Neisseria meningitidis from Norway, associated with increased resistance to bactericidal antibodies induced by a sero-
Editor: R. N. Moore
IMMUNE RESPONSES OF PATIENT SERA
group B outer membrane protein vaccine. Microb. Pathog. 15:197–205. 53. Rosenqvist, E., E. A. Høiby, E. Wedege, B. Kusceck, and M. Achtman. 1993. The 5C protein of Neisseria meningitidis is highly immunogenic in humans and induces bactericidal antibodies. J. Infect. Dis. 167:1065–1073. 54. Rouppe van der Voort, E. M., B. Kuipers, H. F. Brugghe, L. M. A. van Unen, M. Timmermans, P. Hoogerhout, and J. T. Poolman. 1997. Epitope specificity of murine and human bactericidal antibodies against PorA P1.7,16 induced with experimental meningococcal group B vaccines. FEMS Immunol. Med. Microbiol. 17:139–148. 55. Rouppe van der Voort, E. M., P. van der Ley, J. van der Biezen, S. George, O. Tunnela, H. van Dijken, B. Kuipers, and J. Poolman. 1996. Specificity of human bactericidal antibodies against PorA P1.7,16 induced with a hexavalent meningococcal outer membrane vesicle vaccine. Infect. Immun. 64: 2745–2751. 56. Sierra, G. V. G., H. C. Campa, N. M. Varacel., I. L. Garcia, P. L. Izquierdo, P. F. Sotolongo, G. V. Casanueva, C. O. Rico, C. R. Rodriguez, and M. H. Terry. 1991. Vaccines against group B Neisseria meningitidis: protection trial and mass vaccination results in Cuba. NIPH Ann. 14:195–210. 57. Skevakis, L., C. E. Frasch, J. M. Zahradnik, and R. Dolin. 1984. Classspecific human bactericidal antibodies to capsular and noncapsular surface antigens of Neisseria meningitidis. J. Infect. Dis. 149:387–396. 58. Toma ´ska, L., and J. Nosek. 1995. Several polymers enhance the sensitivity of the Southwestern assay. Anal. Biochem. 227:387–389. 59. Wedege, E., G. Bjune, L. O. Frøholm, E. A. Høiby, and E. Rosenqvist. 1991. Immunoblotting studies of vaccinee and patient sera from a Norwegian serogroup B meningococcal vaccination trial. NIPH Ann. 14:183–186. 60. Wedege, E., K. Bolstad, L. M. Wetzler, and H. K. Guttormsen. 1996. IgG antibody activity against meningococcal class 1 and 3 outer membrane proteins in patient sera: comparison between immunoblot and ELISA analyses, p. 190–191. In W. D. Zollinger, C. E. Frasch, and C. D. Deal (ed.), Abstracts of the 10th International Pathogenic Neisseria Conference. Baltimore, Md. 61. Wedege, E., K. Bryn, and L. O. Frøholm. 1988. Restoration of antibody binding to blotted meningococcal outer membrane proteins using various detergents. J. Immunol. Methods 113:51–59. 62. Wedege, E., R. Dalseg, D. A. Caugant, J. T. Poolman, and L. O. Frøholm. 1993. Expression of an inaccessible P1.7 subtype epitope on meningococcal class 1 proteins. J. Med. Microbiol. 38:23–28. 63. Wedege, E., and L. O. Frøholm. 1986. Human antibody response to a group B serotype 2a meningococcal vaccine determined by immunoblotting. Infect. Immun. 51:571–578. 64. Wedege, E., E. A. Høiby, E. Rosenqvist, and L. O. Frøholm. 1990. Serotyping and subtyping of Neisseria meningitidis isolates by co-agglutination, dotblotting and ELISA. J. Med. Microbiol. 31:195–201. 65. Wedege, E., J. Kolberg, A. Delvig, E. A. Høiby, E. Holten, E. Rosenqvist, and D. A. Caugant. 1995. Emergence of a virulent clone within the electrophoretic type 5 complex of serogroup B meningococci in Norway. Clin. Diagn. Lab. Immunol. 2:314–321. 66. Wyle, F. A., M. S. Artenstein, B. L. Brandt, E. C. Tramont, D. L. Kasper, P. L. Altieri, S. L. Berman, and J. P. Lowenthal. 1972. Immunological response of man to group B meningococcal polysaccharide vaccines. J. Infect. Dis. 126:514–522. 67. Zollinger, W. D., J. W. Boslego, E. E. Moran, B. L. Brandt, C. Cruz, M. Martinez, and J. Garcia. 1994. Effect of vaccination with meningococcal outer membrane protein vaccine on subsequent antibody reponses to carriage and natural infections, p. 954–960. In C. J. Conde-Glez, S. Morse, P. Rice, F. Sparling, and E. Calderon (ed.), Pathobiology and immunobiology of Neisseriaceae. Instituto Nacional de Salud Publica, Cuernavaca, Morelos, Mexico. 68. Zollinger, W. D., and E. Moran. 1991. Meningococcal vaccines—present and future. Trans. R. Soc. Trop. Med. Hyg. 85(Suppl. 1):37–43.