Previous Article | Next Article ![]()
Clinical and Vaccine Immunology, May 2007, p. 556-561, Vol. 14, No. 5
1071-412X/07/$08.00+0 doi:10.1128/CVI.00452-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Vaccine Evaluation Unit, Health Protection Agency North West, Manchester Laboratory, Manchester Medical Microbiology Partnership, P.O. Box 209, Clinical Sciences Building II, Manchester Royal Infirmary, Manchester M13 9WZ, United Kingdom,1 Institute of Environmental Science & Research Ltd., Kenepuru Science Centre, Porirua, New Zealand,2 Novartis Vaccines, Via Fiorentina 1, 53100 Siena, Italy3
Received 4 December 2006/ Returned for modification 11 January 2007/ Accepted 27 February 2007
|
|
|---|
|
|
|---|
A surrogate of protection against MenB is the serum bactericidal antibody (SBA) assay, with SBA titers having been shown to correlate with the efficacies of outer membrane vesicle vaccines in Norwegian and Cuban teenagers (6, 15). The MenB SBA assay was recently standardized between four laboratories (4) and has been recommended as the primary measure for evaluating candidate vaccine responses without the need for efficacy studies (5).
Early studies using the SBA assay by Goldschneider and colleagues (8, 9) used an exogenous human complement source. However, problems of obtaining human complement of sufficient quantity and quality due to the presence of antimeningococcal or cross-reacting antibody resulted in the use of a commercially available source as an attractive alternative. Complement-preserved baby rabbit serum (subsequently referred to as baby rabbit complement) has since been utilized in the standardized serogroup A and C SBA assays (23), but its use in the MenB SBA assay has been associated with elevated SBA titers in comparison with those obtained using complement-preserved human serum/plasma (subsequently referred to as human complement) (21, 30). This discrepancy has been associated with the presence of a low-avidity anti-MenB capsular antibody in test sera (21, 30), and for this reason, human complement has remained the preferred complement source for the MenB SBA assay.
Due to the difficulties in obtaining suitable human complement, the possibility of absorbing anti-MenB capsular antibody from test sera (14, 30, 31) to enable the use of baby rabbit complement was investigated. This method would dispense with the need for human complement and have significant cost and standardization benefits. Colominic acid, a capsular homopolymer from Escherichia coli K1 which shares the same alpha-(2-8)-linked N-acetyl-D-neuraminic acid linkage residue as MenB polysaccharide, was used due to the commercial nonavailability of purified MenB polysaccharide and as a replacement for the whole cells used in earlier studies (21, 30, 31).
|
|
|---|
SBA assay.
The MenB SBA assay was performed as previously described (4), with minor modifications. Briefly, either human complement (plasma) or baby rabbit complement (Pel Freez, AZ) was used at a concentration of 25% as an exogenous source of complement. SBA titers were expressed as the reciprocals of the final dilutions giving
50% SBA killing at 60 min compared with the control group (inactive complement/no test sera) against the MenBvac and MeNZB vaccine strains 44/76-SL (B:15:P1.7,16) and NZ 98/254 (B:4:P1.7-2,4), respectively.
For absorption studies, colominic acid (Sigma-Aldrich, Dorset, United Kingdom) was added to the bactericidal buffer (BB) (4) to give final concentrations of between 400 and 2,000 µg/ml. Sera were incubated with BB-colominic acid for 1.5 to 3 h at room temperature prior to being assayed.
Monoclonal antibodies (MAbs) against the MenB capsule (NIBSC code 95/750) and the PorA subtypes P1.4 (NIBSC code 95/700), P1.7 (NIBSC code 01/514), and P1.16 (NIBSC code 01/538) were obtained from the National Institute of Biological Standards and Control (Hertfordshire, United Kingdom) and assayed as unknown samples in the SBA assay.
Determination of anti-colominic acid IgG and IgM concentrations.
Using a subset of samples (n = 36), anti-colominic acid immunoglobulin M (IgM) concentrations were determined using a bead-based assay (17). The methodology was adapted to detect anti-colominic acid IgG concentrations by the inclusion of R-phycoerythrin-conjugated goat anti-human IgG Fc
fragment (Jackson ImmunoResearch Laboratories, PA).
Data analysis. In the SBA assay, titers of <2 were assigned a value of 1 for computational purposes. The relationships between the results gained using the different complement sources (all on a log10 scale) were explored by linear regression analysis, and the Pearson correlation coefficient was determined. Additionally, sample/strain-specific differences in titer levels between results for the different complement sources were determined, and the averages for all samples were calculated.
|
|
|---|
Comparison of anticapsular and anti-PorA MAb results gained using human complement and baby rabbit complement, with and without colominic acid. Individual SBA titers gained using human complement (hSBA), baby rabbit complement (rSBA), human complement with colominic acid at 400 µg/ml (hcSBA), and baby rabbit complement with colominic acid at 400 µg/ml (rcSBA) are presented in Table 1 for each of the MAbs. Results from a single run are presented, with almost identical results achieved with two repeat experiments (replicate results were within one SBA titer for each analyte) (data not presented). Inclusion of colominic acid reduced the hSBA and rSBA titers of the anti-MenB capsular MAb to a greater extent than that attributable to assay variation. However, no such reduction was demonstrated with the PorA MAb, demonstrating the inhibition of anti-MenB polysaccharide antibodies by colominic acid.
|
View this table: [in a new window] |
TABLE 1. SBA titers of anticapsular and PorA MAbs with human and baby rabbit complement, with and without colominic acid (400 µg/ml)
|
|
View this table: [in a new window] |
TABLE 2. Correlation trend line data from linear regression of all results gained using human complement, baby rabbit complement, and baby rabbit complement with colominic acid (400 µg/ml)
|
512), where rcSBA titers were similar to hSBA titers, and the second consisting of samples with high rSBA titers (>512), where rcSBA titers were reduced but not similar to hSBA titers. Differences in these populations are demonstrated in Fig. 1a to d, which present sample-specific differences in SBA titer steps (one SBA titer step = one well/doubling dilution difference) between rSBA and rcSBA titers in comparison to hSBA results for each of the two groups (rSBA titers of
512 and >512). The average differences in SBA titer steps between hSBA and rcSBA results for these two groups are presented in Table 3 and demonstrate that samples with rSBA titers of
512 produced rcSBA titers similar to hSBA titers, while samples with rSBA titers of >512 did not.
![]() View larger version (28K): [in a new window] |
FIG. 1. Differences in SBA titer steps between results gained using baby rabbit complement, with and without colominic acid (400 µg/ml), and results gained using human complement. (a) Samples with rSBA titers of >512 for NZ 98/254. (b) Samples with rSBA titers of 512 for NZ 98/254. (c) Samples with rSBA titers of >512 for 44/76-SL. (d) Samples with rSBA titers of 512 for 44/76-SL.
|
|
View this table: [in a new window] |
TABLE 3. Average differences in SBA titer steps gained using baby rabbit complement, with and without colominic acid (400 µg/ml), and human complementa
|
|
View this table: [in a new window] |
TABLE 4. SBA titers against NZ 98/254 with human complement, baby rabbit complement, and baby rabbit complement with increasing concentrations of colominic acid
|
|
View this table: [in a new window] |
TABLE 5. Differences in SBA titers gained using human complement with and without colominic acid (400 µg/ml)a
|
|
|
|---|
Inclusion of colominic acid (400 µg/ml) with baby rabbit complement resulted in a general reduction of SBA titers which was previously demonstrated by Hodge et al., who also demonstrated that this was not due to complement exhaustion or epitope masking (14). Although this observation is not completely understood, it is expected that colominic acid competes for anti-MenB capsular antibodies, which have increased bactericidal activity in conjunction with baby rabbit complement. This was corroborated by the demonstration of the ability of colominic acid to reduce the bactericidal activity of an anti-MenB capsular MAb, but not a PorA MAb, with both human and baby rabbit complement.
Inclusion of colominic acid (400 µg/ml) with baby rabbit complement generally reduced SBA titers to similar levels to those gained with human complement for samples with low rSBA titers (
512). However, for samples with high rSBA titers (>512), colominic acid resulted in a reduction of titers, but not to similar levels to those achieved with human complement. For samples with rSBA titers of >512, increasing concentrations of colominic acid with baby rabbit complement resulted in differing results. Some samples demonstrated decreases in SBA titers with increasing colominic acid concentrations, suggesting that for these samples colominic acid at 400 µg/ml was not sufficient to absorb all anticapsular antibodies. Other samples demonstrated either little or no reduction in SBA titer with increasing colominic acid concentrations. Previously, Jennings et al. (16) demonstrated the presence of two populations of antibodies in equine antiserum, namely, those which could be absorbed by MenB polysaccharide and colominic acid and those which could be absorbed only by MenB polysaccharide. It is possible that similar populations were present in the human study subjects, with samples where increased concentrations of colominic acid did not increase absorption being due to an antibody population which may only be removed by MenB capsular polysaccharide. This may be a result of colominic acid not being a complete antigen (16) and therefore not absorbing all anti-MenB polysaccharide antibodies. Confirmation that purified MenB polysaccharide would absorb greater concentrations of antibody for these subjects was not possible due to its commercial unavailability.
Interestingly, for some samples, colominic acid with baby rabbit complement resulted in lower SBA titers than those achieved with human complement. This was demonstrated with colominic acid at 400 µg/ml but was accentuated with increasing colominic acid concentrations. Assays of samples with colominic acid and human complement demonstrated slight reductions in SBA titer for some samples, leading to the suggestion of the presence of an anti-MenB capsular antibody which is bactericidal with human complement, in agreement with the findings of Mandrell et al. (21). However, these minor reductions may contribute to but probably do not wholly account for the large reductions in the presence of colominic acid and baby rabbit complement which were demonstrated for some samples. It is plausible that for certain samples, colominic acid with baby rabbit complement causes an inhibition greater than that attributable to absorption of anticapsular antibody. This was particularly demonstrated with increasing concentrations of colominic acid, with the mechanism of this effect remaining to be established.
The incubation conditions for colominic acid with test sera were investigated both during initial optimization and with increased concentrations of colominic acid. Incubation at 4°C was previously demonstrated to increase binding of anticapsular antibody to MenB polysaccharide (1, 30). However, we demonstrated that temperature did not affect inhibition with colominic acid (data not presented), in agreement with previous reports of Hayrinen and coworkers (11, 12, 13). Increased incubation periods and agitation conditions were also investigated but, again, showed no effect on inhibition or SBA titers (data not presented).
In this study, we used colominic acid due to its commercial availability and promising data from early absorption studies (14). Absorption of anti-MenB capsular antibody has also been achieved using purified MenB capsular polysaccharide, purified E. coli K1 capsular polysaccharide, and whole cells from both MenB strains and E. coli K1 cells (10, 14, 16, 30, 31). The use of whole MenB cells would result in the absorption of antibodies against other meningococcal antigens and is therefore not suitable for this purpose. Purified MenB capsular polysaccharide has been demonstrated to be a more effective inhibitor than colominic acid (16, 19) and may be more appropriate for future absorption studies. Similarly, purified E. coli K1 capsular polysaccharide may also warrant further evaluation. Anticapsular antibody has also been removed from sera by attaching the absorbent to microtiter plates (31), resulting in the removal of absorbent/antibody from sera prior to assay. Incorporating soluble colominic acid may have affected our results and may explain some of the observed rcSBA titers which were lower than hSBA titers. Absorption which results in the removal of both the absorbent and the antibody from the test sample may be preferential over the soluble methodology but may increase the complexity of the SBA assay by adding extra steps and costs. However, it should be noted that for the pneumococcal serotype-specific IgG enzyme-linked immunosorbent assay, two absorption steps are recommended to increase the specificity of the assay (28).
Additional factors other than anticapsular antibody may account for the differences between results given by the investigated complement sources. These may include certain complement down-regulatory proteins which may enhance serum resistance due to species specificity in binding to meningococci. Human C4b-binding protein, but not that of baby rabbits or rodents, binds to gonococci, resulting in resistance to complement-mediated lysis by the classical pathway (25). Similarly, human factor H, but not those of other species, binds to factor H binding protein (previously called genome-derived neisserial antigen 1870) (24), enhancing resistance to complement-mediated lysis by the alternative pathway (20, 27). Such differences may be partially responsible for the higher titers achieved with baby rabbit complement than with human complement. However, for some samples, results obtained with human complement were either identical to, similar to (within assay variation), or even greater than those given with baby rabbit complement (data not presented). Furthermore, anticapsular and PorA MAbs gave almost identical titers with human and baby rabbit complement. This may suggest a sample-specific effect or may be due to the inclusion of heparin in the BB (due to the use of plasma as a human complement source), resulting in inhibition factor H (2, 3). Nonetheless, it has been demonstrated that heparin at the concentration used does not affect titers in the MenB SBA assay (4, 22), suggesting that binding of factor H was not sufficient to affect the outcome of the MenB SBA assay.
Most current candidate vaccines designed to confer protection against MenB disease are based upon target antigens other than the MenB capsule, and therefore absorption of anticapsular antibody should not interfere with the evaluation of SBA responses to these vaccines. Conversely, for certain vaccines, the response against the capsule may be important, and therefore the absorption of anticapsular antibody is not recommended in their evaluation.
In summary, colominic acid reduced rSBA titers, but these were not comparable to those gained using human complement for all samples. These results suggest that colominic acid will not enable the use of baby rabbit complement in the MenB SBA assay, thus not alleviating the issues of acquiring suitable human complement. The use of purified MenB capsular polysaccharide was not investigated during this study and may be an alternative which warrants further investigation. Furthermore, alternative complement sources, such as bovine sources (5), may be more appropriate, and further evaluation is awaited. However, successful large-scale comparisons between complement sources will be required prior to the use of nonhuman complement and any absorbent in any future vaccine trials.
We thank Sarah Frankland from the Vaccine Evaluation Unit for determining anti-colominic acid IgG and IgM concentrations.
Published ahead of print on 7 March 2007. ![]()
|
|
|---|
8 N-acetylneuraminic acid, the capsular polysaccharide of Neisseria meningitidis serogroup B. J. Infect. Dis. 172:1279-1289.[Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»