Vaccine 17 (1999) 2951±2958
ww.elsevier.com/locate/vaccine
Neisseria meningitidis serogroup C polysaccharide and serogroup B outer membrane vesicle conjugate as a bivalent meningococcus vaccine candidate Lucila O. Fukasawa a, Maria CecõÂ lia O. Gorla c, Rocilda P.F. Schenkman a, Ligiane R. Garcia a, Sylvia M. Carneiro b, Isaias Raw a, Martha M. Tanizaki a,* a
Centro de Biotecnologia, Instituto Butantan, Av. Vital Brasil 1500, CEP 05503-900, SaÄo Paulo, Brazil LaboratoÂrio de Biologia Celular, Instituto Butantan, Av. Vital Brasil 1500, CEP 05503-900, SaÄo Paulo, Brazil c Servic° o de Bacteriologia, Instituto Adolfo Lutz, Av. Dr. Arnaldo 355, CEP 01246-912, SaÄo Paulo, Brazil
b
Received 20 March 1998; received in revised form 24 February 1999; accepted 13 April 1999
Abstract Neisseria meningitidis serogroup C polysaccharide (PS C) was conjugated to serogroup B outer membrane vesicles (OMV) in order to test the possibility of obtaining a bivalent group B and C meningococcus vaccine. The conjugate and controls were injected intraperitoneally into groups of ten mice with boosters on days 14 and 28 after the primary immunization. The following groups were used as control: (i) PS C; (ii) PS C plus OMV; (iii) OMV; and (iv) saline. The serum collected on days 0, 14, 28 and 42 were tested by enzyme-linked immunosorbent assay (ELISA) for PS C and OMV, and by complement mediated bactericidal assay against serogroups B and C. ELISA for PS C as well as bactericidal titres against serogroup C meningococci of the conjugated vaccine increased eight-fold (ELISA) and 32 fold (bactericidal) after 42 days in comparison with the PS C control group. ELISA for OMV and bactericidal titre against serogroup B meningococci of the conjugate showed no signi®cant dierence in comparison with the OMV containing controls. Furthermore, Western Blot assay of the conjugate immune serum did not bind OMV class four protein which is related to the complement dependent antibody suppressor. The results indicate that the PS C-OMV conjugate could be a candidate for a bivalent vaccine toward serogroups B and C meningococci. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Polysaccharide±protein conjugate vaccine; Neisseria meningitidis serogroup C polysaccharide; Neisseria meningitidis serogroup B outer membrane vesicle
1. Introduction Capsular polysaccharides from Neisseria meningitidis groups A, C, W135 and Y induce protective bactericidal antibodies in humans [1,2], having been used as vaccines for almost two decades. Capsular polysaccharide from group B was also tested as a vaccine but only limited success has been achieved with its native form [3]. Therefore, the candidate for group B vaccine * Corresponding author. Tel.: +55-11-813-7222; fax: +55-11-8151505. E-mail address:
[email protected] (Martha M. Tanizaki)
that has been most extensively tested in humans is based on outer membrane vesicles (OMV) extracted directly from cells with negatively-charged detergents. The ability of such vaccines to induce protective immunity against serogroup B meningococcal disease has now been tested in several large ®eld trials showing variable ecacy [4]. It is well known that no thymus-derived helper cells are involved in the immune response to polysaccharide antigens. The practical consequence of this immunological phenomenon is that polysaccharide vaccines are poorly immunogenic in young children. In the wake of the highly successful Haemophilus in¯uenzae type b (Hib) conjugate protein±polysaccharide vaccines [5],
0264-410X/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 4 - 4 1 0 X ( 9 9 ) 0 0 1 7 7 - 2
2952
L.O. Fukasawa et al. / Vaccine 17 (1999) 2951±2958
dierent meningococcal A and C conjugate vaccines are being developed and some of them are now in clinical trial phase [6,7]. Besides the use as group B meningococcus vaccine, N. meningitidis OMV has also been used as the carrier protein in the H. in¯uenzae polysaccharide vaccine (PedvaxHIBTM) [8] and recently, in S. pneumoniae [9] polysaccharide conjugate vaccine, both of which are produced by Merck Research Laboratories. In spite of the use of OMV proteins from N. meningitidis, these vaccines are not indicated for use against meningococcal meningitis. In this study, group C meningococcal polysaccharide was conjugated to OMV from serogroup B strain prevalent in Brazil in order to test the possibility of obtaining a polysaccharide±protein conjugate that could induce an immune response in mice against group C and group B N. meningitidis. Results have shown that this conjugate induced complement dependent bactericidal antibodies against serogroup C and serogroup B Neisseria meningitidis. 2. Material and methods 2.1. Extraction and puri®cation of outer membrane vesicles (OMV) and capsular polysaccharide (PS) OMV was obtained from N. meningitidis Brazilian serogroup B, N44/89 (B:4:P1.15) strain. Bacteria were recovered from the spinal ¯uid of a patient and stored in the Culture Collection of the Instituto Adolfo Lutz (SaÄo Paulo, Brazil). Bacteria were grown in 60 l of complete Catlin's medium in a 80 l fermentor (New Brunswick Model MPP) for 22 h at 378C. The cells were inactivated at 548C and removed from the culture medium by tangential dia®ltration in a Pellicon Cassette using a membrane with a 0.45 mm cut-o (Millipore±HVMP 000.05). The ®ltrate was concentrated from 60 l to 1.0 l in a Pellicon Cassette, 300,000 NMWL cut-o (Millipore±PTMK 000.05) and ultracentrifuged at 100,000 g for 3 h in 20 mM Tris±HCl, 1 mM EDTA buer, containing 0.5% w/v of sodium deoxycholate (DOC), pH 8.5. The pellet was resuspended in water without pyrogen containing 0.01% v/ v thimerosal. PS from N. meningitidis serogroup C, was puri®ed according to a previously described method [10]. Brie¯y, serogroup C strain IMC2135 bacteria were grown in 40 l of Frantz medium in a 80 l fermentor (New Brunswick Model MPP) according to Ramos et al. [11]. The fermentor content was precipitated by addition of 0.1% w/v Cetavlon. The precipitate was recovered after centrifugation at 15,000 g for 30 min and resuspended in 500 ml 1 M CaCl2. The cell debris was discarded after centrifugation at 10,000 g for 30 min. Nucleic acid contaminants were then removed
by fractional ethyl alcohol (25% v/v) precipitation and removed by centrifugation at 15,000 g for 30 min. PS was precipitated by adjusting the ethyl alcohol concentration to 80%. The mixture was left overnight and the PS was recovered by centrifugation at 15,000 g for 30 min. The PS precipitate was resuspended in Tris± HCl buer 20 mM, 2 mM EDTA, pH 8.5. This solution was incubated overnight at room temperature with 5 mg of each of the following proteinases: proteinase K, trypsin and nagarse, and this treatment was repeated for further 4 h. Elimination of lipopolysaccharide and low molecular weight proteins were obtained by tangential ultra®ltration in a 100 kDa cuto hollow®ber in the same buer containing 0.5% w/v DOC. The solution was washed in the hollow®ber with ®ve separate volumes of 1 l of the same buer followed by ®ve washes with 1 l each of the same buffer without DOC and ®ve washes with 1 l each of water without pyrogen. 2.2. Conjugation of N. meningitidis group C PS to group B OMV The conjugate was prepared with adipic acid dihydrazide (ADHÐSigma) as a linker and 1-ethyl-3-(3dimethylamino-propyl)carbodiimide (EDACÐSigma) as a coupling reagent. ADH (0.5 M) and EDAC (0.1 M) were added to PS (5 mg.mlÿ1 dissolved in 0.2 M NaCl, pH 7.5). After mixing, the pH was adjusted to 6.5 and the solution was stirred overnight. The reaction mixture was dialyzed against 0.2 M NaCl overnight at 88C. Then, OMV (10 mg.mlÿ1) and EDAC (®nal concentration of 0.1 M) were added and the reaction was stirred for 5 h, pH 6.5 at 3±88C. The conjugated PS was puri®ed from free PS by ultracentrifugation at 100,000 g for 3 h. 2.3. Chemical and physical characterization of the conjugates Protein content of OMV was measured by the method of Lowry and carbohydrate content was measured by the thiobarbituric acid method [12]. The hydrazide linked to the carbohydrate was determined according to the 2,4,6-trinitrobenzenesulfonic acid method [13]. OMV integrity was checked by electron microscopy. Suspensions of control and conjugated vesicles were placed onto a para®lm sheet in a moisture chamber. Parlodium carbon (300 mesh) coated grids were allowed to ¯oat on the droplets for 30 min and the excess of liquid was blotted dry. Negative staining was done with 2% w/v potassium phosphotungastate and the grids were examined in a JEOL 1010 electron microscope.
L.O. Fukasawa et al. / Vaccine 17 (1999) 2951±2958
2953
2.4. Immunization scheme
2.6. Determination of in vitro bactericidal activity
Male C3H/Hepas mice weighing 18±20 g each were used for the study. Five groups of 10 mice were injected intraperitoneally with 0.5 ml of preparations in saline and 25 mg of aluminium hydroxide as adjuvant:
Anti-meningococcal B bactericidal assay was carried out using a Brazilian epidemic serogroup B strain N.44/89 (B4:P:1.15) in 96-well plates, as described by Frasch and Robbins [15], with modi®cations. Brie¯y, the ®nal reaction mixture (50 ml) contained: 25 ml of serial two fold dilutions of test serum heat inactivated at 568C for 30 min, 12.5 ml baby rabbit serum screened for absence of anti-meningococcal activity as the complement source, about 5 103 CFU.mlÿ1 of log phase meningococci grown in serum agar with ironlimitation. The reaction mixture was incubated at 378C for 30 min, and 130 ml of tryptic soy agar (TSA) cooled to approximately 458C was added to each well. The mixture was allowed to solidify and the plates were incubated for 18±24 h at 378C in 5% CO2. Quantitative cultures were performed at time 0 (T0) before incubation with complement, in order to set the CFU value for 100% survival, plating (12.5 ml) the bacteria stock solution onto a 150 150-mm TSA plate by the tilt method. Control wells included on each microtitre plate contained: (i) bacteria±complement±buer (complement-dependent control) and (ii) heat-inactivated complement±bacteria±buer (complement-independent control). In addition, a positive control was included in each assay consisting of serial dilutions of a known positive serum. Assays were made in duplicate. Anti-meningococcal serogroup C bactericidal assays were carried out using a Brazilian epidemic N.1002/90 (C:2b:P1.3) strain and performed as described above, except that the reaction was incubated at 378C for 60 min as standardized by Maslanka et al. [16]. The bactericidal serum titre was de®ned as the reciprocal serum dilution yielding r50% killing compared to the number of target cells (CFU per well) present at T0.
. . . . .
group group group group group
I: 2.5 mg PS from the PS-OMV conjugate. II: 2.5 mg of PS, 10 mg of OMV. III: 2.5 mg of PS. IV: 10 mg OMV. V: saline control.
Mice of all ®ve groups were boosted with the same preparation 14 and 28 days after the ®rst immunization. Mice were bled via the retro orbital plexus at 0, 14, 28 and 42 days after the ®rst injection. Sera were pooled and stored at ÿ208C.
2.5. Determination of anti-polysaccharide and antiOMV antibodies N. meningitidis group C polysaccharide antibodies were quanti®ed by enzyme-linked immunosorbent assay (ELISA) with goat anti-mouse IgG conjugated with horseradish peroxidase. ELISA plates (Nunc Immunoplate Maxisorp, Roskilde) were pre-coated by adding 5 mg.mlÿ1 polylysine (Sigma) in PBS, pH 7.2 for 30 min at 378C and washed with water. Plates were coated by adding 100 ml per well of a PS solution (20 mg.mlÿ1) in carbonate±bicarbonate buer pH 9.6 and incubated overnight at 4±88C. Plates were blocked by addition of 200 ml of skimmed milk and washed with PBS±0.01% v/v Tween solution. Mouse sera were tested at dierent dilutions, followed by the addition of 100 ml of an enzyme-labeled goat anti-mouse IgG at a dilution of 1 in 1000. Enzyme activity was measured with o-phenylenediamine (0.4 mg.mlÿ1) and hydrogen peroxide (0.5 ml.mlÿ1) as substrate. The antibody concentration is expressed as Units.mlÿ1. In these assays a value of 2800 units.mlÿ1 and 1000 units.mlÿ1 was arbitrarily assigned to a standard sera that contained a high level of antibodies against OMV proteins and against PS C respectively. The mean value of the observed absorbance at 492 nm was transformed to arbitrary units per millilitre by a sigmoid standard curve (logit±log transformation) calculated from the values of the reference serum using a computer program [14]. Anti-OMV antibodies were quanti®ed by ELISA, with an OMV solution (4 mg.mlÿ1) as the coating antigen in Tris±HCl buer pH 8.5. The ELISA procedure was the same as that for PS determination.
2.7. Immunoblotting assay Sodium dodecyl sulphate±polyacrylamide gel electrophoresis (SDS±PAGE) and the Western Blot detection of IgG antibodies against OMV from the N44/89 strain were performed as described by Wedege and Froholm [17] with modi®cations. OMV proteins (120 mg) were separated on SDS±PAGE (12%) and transferred electrophoretically to nitrocellulose paper (0.45 m, Schleicher&Schuell). The transfer was performed at 250 mA for 1 h in a Bio-Rad Electroblotter as speci®ed by the manufacturer. After transfer, the nitrocellulose was cut into 3 mm strips and blocked with 3.0% bovine albumin in PBS for 30 min. The strips were then incubated overnight with the serum samples diluted 1:200 in bovine albumin solution. After washing four times, the strips were incubated with a peroxidase conjugated goat anti-mouse IgG
2954
L.O. Fukasawa et al. / Vaccine 17 (1999) 2951±2958
Fig. 1. Electron microscopy of Outer Membrane Vesicle (OMV) stained with 2% potassium tungstate. Bar corresponds to 0.1 mm. (A) Control vesicles. (B) PS C conjugated vesicles.
(Sigma) diluted 1 in 3000 in 3% w/v albumin in PBS. The ®lms were washed three times with PBS and with 0.05 M sodium acetate, pH 5.0 and incubated for 30 min with peroxidase substrates: 0.05% v/v of hydrogen peroxide (Aldrich) and 4% w/v of 3-amine9-ethylcarbazol (Sigma) in sodium acetate buer. The peroxidase reaction was stopped by washing the strips with water. Monoclonal antibodies against class 1 and 4 were used as positive controls.
Fig. 2A, low levels of anti-polysaccharide IgG were detected in the controls: group II (mice injected with PS C) and in group III (mice injected with PS C plus OMV). However, in the group of mice injected with conjugated PS C±OMV, the increase of anti-PS IgG started 14 days after the ®rst injection and a booster
3. Results 3.1. Characterization of PS C±OMV conjugate Ten mg of both protein and PS were used in the conjugation reaction. According to the chemical analysis, the puri®ed conjugate contained 0.22 mg PS per mg protein and 0.23 mg hydrazide per mg PS or 2.4 mmol sialic acid per mmole adipic acid hydrazide. The hydrazide/PS ratio represents the amount of PS that was coupled to the ADH spacer. The conjugate yield measured as PS bound to OMV was approximately 12%. The PS/protein ratio and the conjugation yield was similar to the data obtained with the conjugation of OMV and H. in¯uenzae type b PS [18]. Electron microscopy was used order to verify the integrity of the OMV after conjugation and Fig. 1 shows that the OMV maintained its original conformation even when coupled with PS. 3.2. Immunogenicity of the PS C±OMV conjugate ELISA assay was done to quantify anti-PS IgG (Fig. 2A) and anti-OMV IgG (Fig. 2B). As shown in
Fig. 2. Antibody titre of mice sera against PS C (A) and OMV (B) after immunization with dierent vaccines: + PS C±OMV, q PS C plus OMV, Q PS C, K OMV and saline.
L.O. Fukasawa et al. / Vaccine 17 (1999) 2951±2958
2955
3B). PS C±OMV conjugate induced high levels of bactericidal antibodies against the C strain after the second dose and a clear booster eect was seen after the third dose (Fig. 3A) in 42 days showing a very similar pattern with that shown in ELISA assay. The bactericidal titre against the serogroup C strain was insigni®cant in the group of mice injected with PS C or PS C plus OMV (Fig. 3A). Regarding serogroup B, the PS C±OMV conjugate as well as the controls, PS plus OMV and OMV induced bactericidal antibodies though in dierent amounts. As shown in Fig. 3B, similar high titres were achieved with the PS C±OMV conjugate and OMV control after two vaccine doses, higher than that of the PS C plus OMV group. However, after three doses the PS C±OMV conjugate and all the OMV containing controls showed the same titre. The bactericidal assay pattern against serogroup B did not show a good correlation with ELISA assay. Thus, in the PS C±OMV conjugate group, the ELISA titre was lower than that induced by OMV control groups, however, the bactericidal titre was the same. As found in the serogroup B ELISA assay, again, the PS injected mice showed a low, but signi®cant, bactericidal titre. Fig. 3. Anti-meningococcal serogroup C (N.1002/90 strain) (A) and serogroup B (N.44/89 strain) (B) bactericidal activity of mice sera after immunization with three doses of dierent vaccines: + PS C± OMV, q PS C plus OMV, Q PS C, K OMV and saline.
eect of a 21-fold increase was obtained at 42 days after the ®rst injection, which meant that the thymus dependence response was achieved. Fig. 2B shows that the PS C±OMV conjugate and all the OMV containing controls were able to induce OMV antibodies although in dierent amounts. The highest titre was found in mice injected with OMV alone. PS C plus OMV also induced a higher titre than the conjugate PS C±OMV group. The anti-OMV titre of PS C±OMV group of mice was 2.7-fold lower than that of OMV alone control and 1.4 times lower than the PS C plus OMV group at 42 days after the ®rst injection. The serum from the PS group showed a low but signi®cant OMV ELISA titre after 42 days (324 U mlÿ1) which could be due to the 1.6% of protein contaminant in the puri®ed PS. 3.3. Bactericidal assay of the group C PS±OMV conjugate The bactericidal assay was performed quantifying complement mediated bactericidal activity of mice sera against serogroup C N1002/90 (C:2b:P1.3 strain) (Fig. 3A) and serogroup B N44/89 (B:4: P1:15 strain) (Fig.
3.4. Immunoblot studies Fig. 4 shows the binding pro®les of IgG antibodies reacting with serogroup B, N44/89 OMV in the serum samples after the third dose of immunization (42 days). All the sera from OMV containing preparations, except the PS C±OMV conjugate, bound to class 1, 3, 4 and 5 OMV proteins although the binding to class 3 was very weak. The serum from the PS C±OMV conjugate did not bind the class 4 OMV protein. Furthermore, the serum from group of mice injected with PS alone bound strongly to class 1 and 4 OMV proteins as well as to a high molecular weight protein that is present in all other strips in Fig. 4. It was observed that OMV electrophoresis of serogroup B N44/89, and serogroup C IMC 2135, but not serogroup C N1002/90 strain, presents a high molecular weight protein with similar mobility in SDS electrophoresis (not shown). Since the PS C was extracted from the IMC 2135 strain, it could be possible that 1.6% protein contamination might have induced antibodies sucient to react in ELISA (Fig. 2B), bactericidal assay (Fig. 3B) as well as to bind in the Western Blot (Fig. 4). 4. Discussion PS vaccines for prevention of Neisseria meningitidis groups A and C were licensed in 1975 in the United States and in Europe by the Institute Merieux and
2956
L.O. Fukasawa et al. / Vaccine 17 (1999) 2951±2958
Fig. 4. Immunoblots showing IgG binding to OMV proteins of N. meningitidis serogroup B N.44/89. AÐOMV proteins stained with amido black. BÐbinding of monoclonal antibodies against class1 and 4 proteins. CÐbinding of PS C±OMV conjugated vaccine antiserum. DÐbinding of OMV vaccine antiserum. EÐbinding of OMV plus PS C vaccine antiserum. FÐbinding of PS C vaccine antiserum. GÐsaline. HÐbinding of pre-vaccine sera.
these vaccines are still commercialized. Due to the Tcell-independent properties of polysaccharide vaccines and in the wake of the highly successful Haemophilus in¯uenzae type b polysaccharide±protein vaccines, eorts have been made to obtain conjugated vaccines against N. meningitidis. Clinical trials were performed with meningococcal serogroups A and C polysaccharide conjugated with a mutant diphtheria toxin, CRM197 (Sclavo R&D Vaccines) [19]. On the other hand, the candidate serogroup B vaccines which have been most extensively tested in humans are all based on OMV. Among the recent OMV vaccines available for trials are the Cuban group B vaccine, VAMENGO BC [17], a noncovalent complex of OMV and PS C and the NIPH NmB vaccine from Norway, a partially puri®ed OMV from strain 44/76 [20] and the hexavalent PorA±OMV from the Netherlands [21]. The VA-MENGO BC is recommended for vaccination against group B and group C meningococcus. Furthermore, N. meningitidis OMV has been used with success as a protein carrier for a H. in¯uenzae type b polysaccharide [22] vaccine and in a conjugated pneumococcal vaccine [8], and both have been commercialized by Merck & Co. Therefore, in a group B/C meningococcal vaccine, OMV could possibly be used as a group B antigen and as a polysaccharide carrier.
For this purpose, OMV from the most prevalent B strain in Brazil (N44/89) was chosen as the carrier of group C polysaccharide, and this conjugate was tested for the induction of an immune response against groups B and C meningococcus. It has been well demonstrated that the polysaccharide must be conjugated with a carrier protein to induce antibodies in an ecient way. This paper shows that when polysaccharide and OMV are in a covalent conjugate form, the complex is able to induce a high level of anti-PS IgG titre, as well as bactericidal antibodies against serogroup C bacteria and a booster eect in both assays. PS alone or a noncovalent complex of PS and OMV are not able to induce a high level of IgG or bactericidal antibodies. In agreement with our results, a low bactericidal titre against serogroup C was also found in the Cuban VA-MENGO BC, a non covalent PS C and OMV complex vaccine which was tested in humans in Brazil [23]. In a covalent reaction between a polysaccharide and a protein there is a risk in destroying essential epitopes either of the polysaccharide or specially of OMV proteins and as a consequence, antibodies against native polysaccharide or OMV proteins would not be induced. The covalent attachment of the PS to the OMV probably induced some modi®cations in the OMV protein epitopes through a directly covalent linkage or steric hindrance. Thus, the addition of PS to the vaccine preparation hiding some protein epitopes is probably the reason for the dierences observed in the ELISA assay, although these dierences in total antibodies measured by the ELISA assay did not aect bactericidal antibody induction. This hypothesis is supported by the Western Blot pro®le showing that antibodies against the class 4 protein were not induced by the PS±OMV conjugate. Class 4 protein is an undesired protein, since it is related to the complement dependent antibody suppressor [24,25]. Therefore, this result would be a positive aspect of this conjugate. Among OMV proteins, class 1, named porin A, is the main protein involved in the bactericidal antibody induction, both in the animal model [26] and in humans [27] and it seems that class 1 protein induction was not aected by the conjugation. Serogroup C PS and OMV ELISA of the N44/89 strain as well as the bactericidal assays against group C and group B meningococcal strains demonstrate that it may be possible to produce a B/C conjugated vaccine. The integrity of OMV seems to be essential to achieve bactericidal activity against the group B strain, since in a conjugate reaction in which the vesicles were damaged, the product failed to induce bactericidal activity against group B, although this conjugate kept a similar bactericidal titre against group C compared to undamaged OMV vaccine. An unexpected result
L.O. Fukasawa et al. / Vaccine 17 (1999) 2951±2958
shown in this paper is the observation that the serum from mice injected with polysaccharide alone was able to induce a low but signi®cant level of bactericidal antibodies against serogroup B OMV which crossreacted strongly with class 1 and 4 proteins in Western blot. Since all the bacterial strains used were retyped to con®rm their exact serotype, the probable explanation would be a cross-reactivity achieved with the 1.6% contaminant protein present in the polysaccharide preparation. This hypothesis is now under study. Meningococcal disease has occurred worldwide often in large epidemic waves which have been superimposed on a background of endemic disease. Strains of serogroups B and C are the most prevalent during endemic periods and have also been responsible for numerous epidemics. However, PS vaccine should be improved through a covalent attachment with a protein. Therefore, the development of a B/C bivalent anti-meningococcal vaccine would be desirable and this PS C±OMV conjugate vaccine could be a good candidate. Acknowledgements This work was supported by Fundac° aÄo de Amparo a Pesquisa do Estado de SaÄo Paulo (FAPESP) and PADCT/FINEP. Lucila O. Fukasawa is a recipient of scholarship from FAPESP. We thank Neusa M. GuimaraÄes and Umbelina Vassoler for technical assistance and Dr Carl Frasch, Dr Jose L. DiFabio and Dr Robert Pon for helpful discussions. References [1] Gotschlich EC, Austrian B, Cvjetanovic B, Robbins JB. Prospects for the prevention of bacterial meningitis with polysaccharides vaccines. Bull WHO 1978;56:509±18. [2] Griss JM, Brandt BL, Altieri PL, Pier GB, Berman SL. Safety and immunogenicity of group Y and group W135 meningococcal capsular polysaccharide vaccines in adults. Infect Immun 1982;34:532±725. [3] Wyle FA, Arteinstein MS, Brandt BL, Tramont DL, Kasper DL, Altieri SL, Berman SL, Lowenthal JP. Immunological response of man to group B meningococcal polysaccharide antigens. J Infect Dis 1972;126:514±22. [4] Zollinger WD. New and improved vaccines against meningococcal disease. In: Woodrow GL, Kaper JB, Cobon GS, editors. New Generation Vaccines. New York: Marcel Dekker, 1998. p. 469±88. [5] Wenger JD, Booy R, Heath PT, Moxon R. Epidemiological impact of conjugate vaccines on invasive disease caused by Haemophilus in¯uenzae type b. In: Woodrow GL, Kaper JB, Cobon GS, editors. New Generation Vaccines. New York: Marcel Dekker, 1998. p. 489±501. [6] Constantino P, Viti S, Podda A, Velmonte MA, Nencioni L, Rappuoli R. Development and phase I clinical testing of a conjugate vaccine against meningococcus A and C. Vaccine 1992;10:691±8.
2957
[7] Twumasi Jr PA, Kumah S, Leach A, et al. A trial of a group A plus group C meningococcal polysaccharide±protein conjugate vaccine in African infants. J Infect Dis 1995;171:632±8. [8] Frasch CE. Regulatory perspectives in vaccine licensure. In: Ellis RW, Grano DM, editors. Development and clinical uses of Hemophilus b conjugate vaccines. New York: Marcel Dekker, 1994. p. 435±53. [9] Klein D, Ellis RW. Conjugate vaccine against Streptococcus pneumoniae. In: Woodrow GL, Kaper JB, Cobon GS, editors. New Generation Vacines. New York: Marcel Dekker, 1998. p. 503±25. [10] Tanizaki MM, Garcia RL, Ramos JB, Leite LCC, His H, Furuta JA, Cabrera-Crespo J, Raw I, Puri®cation of meningococcal group C polysaccharide by a procedure suitable for scale-up, J Microbiol Meth 27:19±23. [11] Ramos JB, Hiss H, Vincentin MA, Paz MF, Peixoto A, Leal MBB, Gonc° alves PT, Raw I. Batch cultivation kinetics of Neisseria meningitidis (serogroup C) in Frantz medium. I Growth and polysaccharide production. Arq Biol Technol 1996;39:215±20. [12] Warren L. The thiobarbituric acid assay for sialic acids. J Biol Chem 1959;234:1971±5. [13] Schneerson R, Barrera O, Sutton A, Robbins JB. Preparation, characterization and immunogenicity of Haemophilus in¯uenzae type b polysaccharide±protein conjugates. J Exp Med 1982;152:361±76. [14] Frasch CE, Zahradnik LD, Wang LY, Mocca LF, Tasi CM. Antibody responsse of adults to na aluminum hydroxide± adsorbed Neisseria meningitidis serotype 2b protein group B polysaccharide vaccine. J Infect Dis 1988;158:710±8. [15] Frasch CE, Robbins JD. Protection against group B meningococcal disease. III. Immunogenicity of serotype 2 vaccines and speci®city of protection in a guinea pig model. J Exp Med 1982;152:361±76. [16] Maslanka SE, Gleisling LL, Libutti DE, Donaldson KB, Harakeh HS, Dykes JK, Arkin FF, Devi SJ, Frasch CE, KrizKuzemenska P, Levinson RD, Lorange M, Peeters CC, Quataert S, Tai JY, Carlone GM. Standardization and a multilaboratory comparison of Neisseria meningitidis serogroup A and C serum bactericidal assays. The multilaboratory study group. Clin Dian Lab Immunol 1997;4:156±67. [17] Wedege E, Froholm LO. Human antibody response to a group B serotype 2a meningococcal vaccine determined by immunoblotting. Infect Immun 1986;51:571±8. [18] Tai JT, Vella PP, McLean AA, Woodhour AF, McAleer WJ, Sha A, Dennis-Sykes C, Hilleman MR. Hemophilus in¯uenzae type b polysaccharide±protein conjugate vaccine (42460). Proc Soc Expl Biol Med 1987;184:154±61. [19] Anderson EL, Bowers T, Mink CM, Kennedy DJ, Belshe RB, Harakeh H, Pais L, Holder P, Carlone GM. Safety and immunogenicity of meningococcal A and C polysaccharide conjugate vaccine in adults. Infect Immun 1994;62:3391±5. [20] Fredriksen JH, Rosenqvist E, Wedege E, Bryn K, Bjune G, Froholm LO, Lindbak AK, Mogster B, Namork E, Rye U, Stabbetorp G, Winsnes R, Aase B, Closs O. Production, characterization and control of MenB-Vaccine ``Folkehelsa'': na outer membrane vesicle vaccine against group B meningococcal disease. NIPH Ann 1991;14:67±80. [21] Peeters CCAM, Rumke HC, Sundermann LC, van der Voort EMR, Meulenbelt J, Schuller M, Kuipers AJ, van der Ley P, Poolman JT. Phase I clinical trial with a hexavalent PorA containing meningococcal outer membrane vesicle vaccine. Vaccine 1996;14:1009±15. [22] Wenger JD, Booy R, Heath PT, Moxon R. Epidemiological impact of conjugate vaccines on invasive disease cause by Haemophilus in¯uenzae type b. In: Woodrow GL, Kaper JB,
2958
L.O. Fukasawa et al. / Vaccine 17 (1999) 2951±2958
Cobon GS, editors. New Generation Vaccines. New York: Marcel Dekker, 1998. p. 489±501. [23] Milagres LC, Lemos APS, Meles CEA, Silva EL, Ferreira LHML, Sousa JAM, Carlone CM. Antibody response after immunization of Brazilian children with serogroup C meningococcal polysaccharide noncovalently complexed with outer membrane proteins. Braz J Med Biol Res 1995;28:981±9. [24] Klugman K, Gotschlich EC, Blake MS. Sequence of the structural gene (rmp M)for the class 4 outer membrane protein of Neisseria meningitidis, homology of the protein to gonococcal protein III and Escherichia coli OmpA, and construction of meningococcal strains that lack class 4 protein. Infect Immun 1989;57:2066±71.
[25] Munkley A, Tinsley CR, Virji M, Heckels JE. Blocking of bactericidal killing of Neisseria meningitidis by antibodies directed against class 4 outer membrane protein. Microb Pathog 1991;11:447±52. [26] Saukkonen K, Leininen M, Abdillahi H, Poolman JT. Comparative evaluation of potential components for group B meningococcal vaccine by pasisve protection in infant rat and in vitro bactericidal assay. Vaccine 1989;7:325±8. [27] Milagres LG, Ramos SR, Sacchi C, Melles CE, Vieira VS, Sato H, Brito GS, Moraes JC, Frasch CE. Immune response of brazilian children to a Neisseria meningitidis serogroup B outer membrane protein vaccine. Infect Immun 1994;62:4419±24.
