INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Neisseria meningitidis causes bacterial meningitis and septicemia worldwide. Epidemiological studies of endemic or epidemic outbreaks of meningococcal disease rely on polyvalent and monoclonal antibodies (MAbs) to capsular polysaccharides for monitoring serogroup specificity (27). This organism can be further classified into serotypes and subtypes with MAbs directed toward two major outer membrane proteins (OMPs) (6). Serotyping is based on MAbs which bind to the class 2 or class 3 OMP (PorB). Subtyping MAbs bind one of two immunodominant variable regions (VR1 and VR2) of the class 1 OMP (PorA) (1, 9, 10, 11, 13, 22). Complete subtyping requires identification of epitopes found in both variable regions (5).

MAbs to about 15 subtype epitopes have been developed and characterized; however, nonsubtypeable (NST) strains are still found. NST strains represent one of three categories: (i) strains possessing class 1 epitopes to which MAbs have not been developed and characterized, (ii) strains which do not express PorA, and (iii) strains in which the PorA subtype epitopes differ only slightly from known subtype epitopes due to genetic modifications, such as point mutations or duplication or deletion events, which eliminate binding of subtyping MAbs.

A significant part of recent group B meningococcal vaccine development efforts has been focused on OMPs as principal components of a subtype-serotype-specific vaccine (26). This vaccination approach is based on the observations that PorA elicits a human bactericidal antibody response (24) and that subtype-specific MAbs passively protect infant rats against challenge with N. meningitidis (17, 18). An effective subtype-specific vaccine should include the most prevalent subtype epitopes associated with the strains causing disease in the population in which the vaccine will be used. Human bactericidal antibodies induced by vaccination with a vaccine of one subtype are not equally effective in killing other subtype strains. Even single amino acid changes in VR1 and VR2 and deletions in regions flanking the epitopes may result in loss of reactivity with subtype-specific MAbs (12, 23), as observed in several recent outbreaks. One such variant also showed increased resistance to bactericidal activity (16), suggesting a possible influence of such subtype variants on the level of protection induced by a subtype-specific vaccine.

We have defined three different point mutations in the subtype-specific epitope P1.14 of N. meningitidis, using a combination of molecular and immunological techniques. The variability within this epitope highlights the importance of defining subtypes molecularly as well as immunologically.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Growth and maintenance of bacterial strains. Strains of N. meningitidis (strains 7967, 8659, 8778, 8779, and 9304) were obtained from the culture collection at the Walter Reed Army Institute of Research (WRAIR), and one strain (S3446) was kindly provided by C. E. Frasch, Food and Drug Administration, Rockville, Md. Strains were maintained at 70°C in skim milk or were lyophilized and stored at 4°C. Cultures were grown on supplemented GC agar (19) for 16 to 18 h at 37°C in a candle extinction jar.

MAbs. MN21G3.17, the prototype P1.14 MAb, was kindly provided by J. T. Poolman, Rijsinstituut voor Volksgezondheit en Milieuhygiene, Bilthoven, The Netherlands (20), and is referred to as 1.14R in this paper. MAb BZ-1-P1.14 was produced in our laboratory and is referred to as 1.14W in this paper.

Dot blot analysis. Cell suspension (2 to 3 µl of fresh live bacteria in 0.9% NaCl, with the cell density adjusted to between a no. 3 and no. 5 McFarland standard) was dotted onto nitrocellulose membranes, and the membranes were dried for 10 min at room temperature (RT). The membranes were used immediately or stored at 4°C until needed. Membranes were blocked for 30 min with 1% casein buffer and then washed once with phosphate-buffered saline (PBS). Casein buffer contained 10 g of casein in 400 ml of 0.1 N NaOH added to 400 ml of water containing 1.2 g of Tris, 8.8 g of NaCl, 1 g of MgCl₂ · 6H₂O, 1 g of sodium azide (pH 7.5), and water for a total volume of 1 liter. MAb diluted 1:10,000 in blocking buffer was added. Membranes were incubated overnight at RT on a rotator and then washed three times with PBS. The membranes were placed in a solution containing 1 µg of phosphatase-labeled goat anti-mouse immunoglobulin G (IgG) plus IgM (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, Md.) per ml diluted in blocking buffer and allowed to react for 2 h at RT on a rotator. The membranes were washed three times with PBS and one time with 50 mM Tris (pH 8.0) and developed with 1 mg of Naphthol AS-MX phosphate per ml-2 mg of Fast Red TR salt per ml in 50 mM Tris (pH 8.0) (Sigma Chemical Co., St. Louis, Mo.).

ELISA. OMC antigen was prepared from strains 7967, S3446 (B:14:P1.14 prototype strain), and 8659 as previously described (25).

SDS-PAGE and immunoblotting. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was carried out in 10% acrylamide gels measuring 16 by 16 by 0.15 cm. OMC at a concentration of 2.5 mg of protein per ml or low-molecular-weight protein standards (Bio-Rad Laboratories, Hercules, Calif.) were mixed 1:1 with treatment buffer (0.125 M Tris-Cl [pH 6.8], 4% SDS, 20% glycerol, 10% 2-mercaptoethanol) and boiled for 10 min before being loaded onto the gels. Gels were run at a constant current of 30 mA per gel at 10°C for approximately 4 h. The gels were stained with Coomassie brilliant blue R-250 or blotted onto nitrocellulose filters (0.45-µm pore size) at 30 V overnight at 10°C. Immunoblotting was performed as described for dot blotting, except the primary antibody was diluted 1:500.

PCR. A fresh overnight culture of each strain was grown on supplemented GC agar. A single colony of each was suspended in 25 µl of 0.5 N NaOH for 15 min to lyse the bacteria. Twenty-five microliters of 1 N Tris (pH 8.0) was added for neutralization, followed by dilution with 450 µl of water. Ten microliters of this suspension was added to the PCR mix by hot start at 95°C. The PCR mix contained 10 µl of 10× reaction buffer, 200 µM of each deoxynucleoside triphosphate, 20 pmol each of primers (forward [GCGGCCGTTGCCGATGTCAGCC] and reverse [GCGGCATTAATTTGAGTGTAGTTGCC]), 3.0 U of Hot Tub Polymerase (Amersham Corp, Arlington Heights, Ill.), and water to 90 µl. Cycling was performed in two steps: 95°C for 1 min and 70°C for 2 min for 20 cycles in a thermocycler (Perkin-Elmer Cetus Corp., Norwalk, Conn.).

Partial sequence analysis. Direct sequencing of PCR products was performed manually as previously described (15). Alternately, DNA from four PCRs was concentrated in Ultrafee-MC filter units (Millipore Corp., Bedford, Mass.). Concentrated DNA was separated in 1.2% SeaPlaque agarose (FMC BioProducts, Rockland, Me.) and cut from the gel under UV visualization. The amplified DNA was eluted and processed with Wizard PCR Preps (Promega Corp., Madison, Wis.). Sequencing primers included the forward amplification primer upstream of VR1 and AACAGCCTTCAGCCCTTCG or CACTCCCCTTAAAGCCGAT downstream of VR1 and CGGTTTCAGCGGCAGCGTC upstream of VR2 and CCCACATTGGCGTGTCTCGC or GTTCCCGGCAAAACCGCC downstream of VR2. Automated sequencing reactions were performed by using the DyeDeoxy Terminator Cycle Sequencing Kit on a model 373 DNA sequencer (Applied Biosystems, Inc., Foster City, Calif.).

Solid-phase peptide synthesis (modified Geysen pin method). Synthetic octapeptides covalently linked to derivatized polyethylene pins were prepared according to the modified technique of Geysen et al. (7). Synthesis of the overlapping peptides was directed from computer software and hardware designed at WRAIR (4). 9-Fluorenylmethyoxycarbonyl (Fmoc)-amino acid pentafluorophenyl esters were purchased from Peninsula Laboratories (Belmont, Calif.), and reagent grade chemicals were obtained from Aldrich Chemical Co. (Milwaukee, Wis.) and Fisher Scientific Co. (Springfield, N.J.). The synthesis was performed in duplicate, and the peptide composition was confirmed by amino acid analysis of one control pin per plate by complete hydrolysis with 6 N HCl at 110°C followed by analysis in a Beckman 6300 analyzer. Two control pins encompassing a previously mapped epitope were also included.

Peptide synthesis. Solid-phase peptide synthesis was performed with a 430A peptide synthesizer (Applied Biosystems, Inc.) with Fmoc-protected amino acids and side chain-protecting groups supported by p-hydroxymethylphenoxymethylpolystyrene resin. All solvents and side chain-protected amino acids were also obtained from Applied Biosystems, Inc. Following synthesis, protecting groups were removed by acid hydrolysis with trifluoroacetic acid in the presence of thioanisole, 1,2-ethanedithiol, and water as scavengers. Amino acid analysis and reverse-phase high-pressure liquid chromatography (HPLC) were used to determine the amino acid content and peptide purity, respectively.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Determination of the specificity of MAb 1.14W. The specificity of MAb 1.14W was determined by testing a panel of prototype N. meningitidis strains, including strain 7967, by dot blot assay. The only strains binding MAb 1.14W were 7967 and S3446, the subtype P1.14 prototype strain. Strain 7967 was negative by dot blot assay with MAb 1.14R, the prototype P1.14 MAb (Table 1). To further test the specificity of 1.14W, a panel of 100 strains from the WRAIR culture collection (19 nonsubtypeable strains and 81 subtypeable strains) were tested by dot blot. Five of the six strains that were previously positive with 1.14R were positive with 1.14W; only strain 8659 was positive with 1.14R and not with 1.14W. Eleven additional strains were identified by dot blot with 1.14W. A total of 18 strains were identified as P1.14; 16 of these strains were positive with both antibodies.

Western blot analysis. The two MAbs 1.14W and 1.14R showed different patterns of binding to strains S3446, 7967, and 8659 analyzed by Western blotting. MAb 1.14W bound well to its homologous strain 7967 but not to the others, whereas 1.14R bound only to S3446 and 8659 and not to strain 7967 (Table 1).

PCR-directed sequencing. Six strains that were dot blot positive with 1.14W or 1.14R were selected for sequencing, including four strains which bound to both MAbs and two which showed different binding patterns. The relevant regions (VR1 and VR2) of porA were sequenced, and the data were analyzed (Table 2). The deduced amino acid sequences in VR1 differed significantly for several of the isolates and were not considered the probable binding site for either MAb. The previously described VR1 subtype epitope P1.7 (ASGQ) was present in two of the strains but not the others. The deduced amino acid sequences for VR2 were all similar, and the sequence for strain S3446 was consistent with published data (VDEKKM) (10). Strains 7967 and 8659 varied from strain S3446 by single base changes which would result in differences of one amino acid (MK and KN, respectively). It was predicted that the sequences of VR2 could represent the binding sites for 1.14W and 1.14R.

Epitope mapping. The deduced sequences of 39 amino acids encompassing VR2 of strains 7967, S3446, and 8659 were selected in order to identify the epitopes of 1.14W and 1.14R. Overlapping octamers representing these sequences were synthesized on Geysen pins. MAb 1.14R bound to three pins of S3446, representing the sequence VDEKKM (Fig. 1). MAb 1.14W did not bind to solid-phase peptides synthesized from the VR2 sequence of 7967, S3446, or 8659. The possibility existed that the epitope for 1.14W was within this region, but determination of that epitope required either a longer peptide or a unique conformational structure of the epitope. Therefore, longer peptides which could be used for quantitative analysis of MAb binding were prepared.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Reactivity of MAb 1.14R by ELISA with overlapping octapeptides representing region VR2 of meningococcal strain S3446.

Peptide analysis. Three peptides consisting of 20 amino acids deduced from the sequences of VR2 of strains S3446, 7967, and 8659 were synthesized for quantitative analysis by ELISA with both MAbs. The additional C residues at the termini of the peptide were synthesized to allow the formation of a disulfide bond, thus mimicking the predicted exposed loop formation of VR2.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Typing of N. meningitidis includes determination of group, serotype, and subtype by either whole-cell ELISA or dot blot assays using MAbs. The subtype specificity of a MAb is defined by its reactivity to the approximately 40-kDa OMP (PorA) of the immunizing strain by Western blot analysis. Subtyping MAbs usually bind to one of two VRs of PorA, and many of these epitopes are known. In recent years, several mutants of subtypeable strains carrying point mutations and deletions in VR1 and VR2 that fail to bind subtype-specific MAbs have been isolated (12, 16, 23). Subtype variants of this kind which have been identified are designated as such by the addition of a lowercase letter to the subtype. Suggested designations for the variants of subtype P1.14 are presented in Table 3.

We produced a new subtype-specific MAb with NST strain Z:4:NST as the immunogen. MAb 1.14W bound to the prototype P1.14 strain as well as to the Z:4:NST strain. Although the MAb generated by strain Z:4:NST reacted with subtype P1.14 N. meningitidis, the P1.14 subtyping MAb did not bind to the Z:4:NST strain. However, both MAbs did bind to 16 of 18 strains tested. DNA sequence analysis of six of these strains suggested that the VR1 sequences differed among these strains, while the VR2 sequences were very similar. We concluded that neither antibody bound to the different VR1 epitopes present in the six strains. Therefore, we directed our attention to the VR2 epitope.

Four forms of the subtype-specific antigen were compared to determine the functional epitopes of the two antibodies: (i) native antigen (whole organism or OMC) as it is exposed on the surface of the microbe by dot blot or ELISA, (ii) denatured antigen in Western blot analysis, (iii) soluble 20-amino-acid peptides circularized to mimic the predicted exposed VR2 loop of PorA, and (iv) solid-phase peptides designed to measure continuous epitopes.

MAb 1.14R bound native and denatured antigen in 17 of 18 strains. MAb 1.14R bound to the VR2 epitope VDEKKM or VDEKNM. However, substitution of the nonpolar amino acid (M) with a charged amino acid (K) eliminated binding of the MAb to the native antigen by dot blotting or ELISA or to denatured antigen.

MAb 1.14W appeared to bind to an epitope in a conformation-dependent manner. This antibody bound to native antigen which contained either the sequence VDEKKM or the sequence VDEKKK. After denaturation in SDS-PAGE, MAb 1.14W bound only the antigen with the VDEKKK epitope, the sequence of the immunizing antigen. It did not bind any solid-phase octamer peptides, a conformationally constrained antigen, but it did bind to the longer peptide analogous to the sequence of the immunogen and very weakly to that which differed by one amino acid. As expected, the antibody did not bind a conformationally intact antigen which varied from the immunizing antigen by two amino acids (VDEKNM).

Any protein molecule displays multiple determinants or epitopes which vary in immunological availability. A denatured molecule may renature to a conformation different from the original, thus exposing a modified array of epitopes. This is particularly true of B-cell epitopes, which when characterized at the level of X-ray crystallography are usually discontinuous or conformational (2, 8). Because most of the binding energy for antigen-antibody interactions occurs within a few of the critical residues to which the MAb binds, the epitope often can be mapped to the region where those few residues are located. Linearization or unfolding-refolding of the core region of the epitope will modify the interactions of the amino acids and hence modify the binding site of the MAb. This result has been demonstrated by comparisons of the two MAbs described in this study.

At the appropriate concentrations, each MAb described herein effectively defines most strains with the P1.14 subtype in screening procedures but fails to detect at least one variant strain. It would be impractical to include multiple antibodies for each subtype variant which has been identified, since epidemiological screens already require about 15 MAbs. In the case of P1.14, MAb 1.14R is probably the more appropriate MAb to use, because its immunogen, strain S3446, is the prototype P1.14 strain and VDEKKM appears to be the dominant subtype epitope. The investigator must be aware that some subtype variants may not be detected.

Determination of the most appropriate screening methods and subtype analysis depends on the immediate requirements of an investigation. A panel of well-characterized subtype-specific MAbs proposed for use in screening a large number of isolates during an outbreak should be selected to identify the majority of strains. This method of epidemiological screening has proven effective in assessing outbreaks during the last 10 years.

Data collected in vaccine trials during this 10-year period have focused attention on the importance of the subtype antigen in stimulating a protective immune response. More-rigorous studies of this protein, in particular VR1 and VR2, have been performed, yielding detailed analyses of single base changes, deletions, and duplications that modify subtype specificity. This molecular approach to subtype identification highlights the variability which is more difficult to detect immunologically and offers a molecular basis for description of subtle differences within the immunogenic class 1 epitopes. Rapid advances in PCR and automated sequencing technology have made this molecular approach to subtype identification a viable alternative for future investigations focusing on the protective immune response directed toward this antigen. Data concerning clonal variation within epidemic populations as well as identification of immunologically nonsubtypeable variants can be assessed. The effect that this subtype variation may have on the efficacy of a subtype-specific vaccine is yet to be determined.