|Home | About | Journals | Submit | Contact Us | Français|
Haemophilus influenzae type a (Hia) is an important pathogen for some American Indian, Alaskan native, and Northern Canada aboriginal populations. Assays to measure serum bactericidal activity (SBA) to Hia have not been developed or validated. Here, we describe two methods for the measurement of SBA: SBA with a viability endpoint (CFU counts) and SBA with a fluorometric endpoint using alamarBlue as the metabolic indicator. Both SBA assays measure Hia-specific functional antibody and correlate with anti-Hia IgG enzyme-linked immunosorbent assay (ELISA) concentration of naturally acquired antibodies.
Haemophilus influenzae is responsible for a variety of clinical illnesses, including otitis media, pneumonia, and meningitis. Since the introduction of the H. influenzae serotype b (Hib) conjugate vaccine in 1988 in the United States, the incidence of disease and carriage of this pathogen has been reduced significantly (3). However, another encapsulated H. influenzae, serotype a (Hia), has been recognized as an important cause of invasive disease in certain populations, including American Indians, Alaskan natives, and Northern Canada aboriginal groups (2, 11, 13). With no vaccine against Hia, natural immunity is the only means of immune protection against the organism. Since infants produce a weak or undetectable response to polysaccharide capsule (10, 19), maternal antibodies are the major source of immunity for young infants, and quantification of natural immunity in cord blood sera could be important in high-risk populations. No threshold has been established as a minimal protective level; therefore, monitoring fold increases in antibodies in acute- versus convalescent-phase sera by enzyme-linked immunosorbent assay (ELISA) to Hia polysaccharide can be used to monitor responses to infection.
ELISA to Hia polysaccharide can be used to measure the amount of Hia-specific serum antibodies but does not differentiate between functional and nonfunctional Hia antibodies. When assessing the natural immunity against this pathogen, it is important to also measure functional antibodies, such as serum bactericidal antibody (SBA). Several SBA methods using either viability or fluorometric endpoints have been standardized for organisms such as Hib (15) but have not been evaluated for Hia. The objectives of this study are to evaluate the applicability, reproducibility, and specificity of the SBA assay for Hia by using different endpoints and to compare the titers of this assay to the results of ELISA, which measures both functional and nonfunctional antibodies.
A total of 19 serum samples and one immune globulin preparation were used in the different assays. Sera were obtained after informed consent and protocol exemption by the CDC. Seven sera from normal, healthy, adult donors (age range, 24 to 60 years) were obtained through Emory Donor Services, Atlanta, GA; 12 unlinked cord blood sera were provided by Alberto Villaseñor Sierra (CIBO, IMSS, Jalisco, México) and contained only IgG antibodies. These 12 sera were selected from a larger bank of 69 cord blood sera collected from healthy mothers. Sera were selected randomly if enough volume was available and if ELISA was positive (cutoff level of 0.09). An immune globulin preparation was also used (Gamunex 10%; Talecris Biotherapeutics, Inc., Clayton, NC). Sera were stored frozen at −70°C in 500-μl aliquots until use.
The complement source was sterile serum from 3- to 4-week-old baby rabbits (Pel-Freez, Brown Deer, WI) previously used in the Hib SBA (15). Although the concentration of cross-reactive antibodies was not measured in the complement lots, three different complement lots were qualified for use on both assays by testing active and heat-inactivated complements. The lot chosen (8028) had 14% killing (level of acceptability is less than 25% killing) compared to time zero inoculum, where the inactivated complement had no killing.
Twelve bacterial strains were used in the SBA assays (Table (Table1).1). Six strains were obtained from the CDC Active Bacterial Core surveillance (ABCs) from Georgia (1989 to 2007) (GA41512, GA41513, GA18491, GA11151, GA44497) (12) and Oregon (M8881). Five strains were obtained from the CDC Arctic Investigations Program (AIP) (2000 to 2007) (AK1435C3, AK137339, AK137341, AK137342, AK137865). The last strain (ATCC 9006) was obtained from the American Type Culture Collection (ATCC). Seven isolates were obtained from sterile sites and caused invasive disease, and five were obtained from the colonized nasopharynx. Three isolates (GA41512, GA41513 , and M8881 [Satola personal communication] had the partial IS1016-bexA deletion, a mutation possibly associated with increased virulence in Hia (1).
Each Hia strain was plated on chocolate II solid medium overnight. Few colonies were selected from the medium and added to brain heart infusion (BHI) broth, supplemented with 2% Fildes enrichment (BBL, Becton Dickinson and Co., Sparks, MD), and then incubated at 37°C and 5% CO2 until optical density at 600 nm reached an average of 0.4. Aliquots were flash frozen and stored at −70°C for future use. A frozen aliquot of each strain was diluted to yield 1,000 bacteria in a 20-μl volume (amount added to each well).
Standard slide agglutination capsule serological testing was performed as described by the manufacturer of the antiserum (Bacto-Difco Diagnostic Systems), and PCR molecular capsule typing was performed as described by Falla et al. (5).
SBA assays (with a viability endpoint [CFU counts] and with a fluorometric endpoint) were performed as described by Romero-Steiner et al. (15) for Hib, which is a modification of the method described by Schlesinger and Granoff (17).
SBA assays were performed on four separate days for each strain by two independent researchers blinded to the assay results using either method.
Ten negative-control sera with negative SBA titers using the reference strain AK1435C3 were used but were not included in the comparison of strains or the final analysis.
Serum samples were tested with a starting dilution of 1:8. Two-fold serum serial dilutions were made in 10 μl of Hanks buffer with Ca2+ and Mg2+ (Life Technologies, Grand Island, NY) supplemented with 2% Fildes enrichment. Each bacterial strain was diluted to yield 1,000 bacteria in 20 μl inoculum per well. After 15 min of incubation of the serum and the bacterial strain at 37°C and in 5% CO2, 20 μl of baby rabbit complement was added to each well. An additional 30 μl of the dilution buffer was added to each well to bring the total volume of the reaction to 80 μl. After an incubation of 60 min at 37°C in 5% CO2, 5 μl from each well was plated onto chocolate agar plates using a tilt method for better definition of the CFU (ChocII; BBL, Becton Dickinson and Co., Sparks, MD). After 16 h of incubation at 37°C in 5% CO2, viability counts were performed and SBA titers were determined. SBA titers were defined as the reciprocal of the serum dilution that resulted in more than 50% killing compared to the growth in the complement controls.
alamarBlue (Trek Diagnostics, Westlake, OH) is a commercially available metabolic indicator. In the presence of viable bacteria, alamarBlue is reduced and a color change from blue to pink occurs. The reduced compound is also fluorescent with an emission wavelength of 590 nm if excited by a UV light source at a wavelength of 530 nm. Therefore, the SBA assay using alamarBlue could be used as an indirect method for the measurement of metabolism of the surviving bacteria after SBA assay, and the fluorescent units collected represent a viability endpoint.
The same protocol for the SBA assay using viability counts as endpoints applies except that 30 μl of alamarBlue buffer was used once 20 μl of the complement was added to each well. The alamarBlue buffer consisted of 16% alamarBlue, 64% Hanks buffer (containing Ca2+, Mg2+, and 2% Fildes enrichment), and 20% BHI broth (BBL). alamarBlue was found to be stable under SBA assay conditions and did not result in bacterial death in the absence of antibodies (15). After a 6-h incubation period at 37°C in 5% CO2, assay plates were read in a fluorometer (model FL 600 synergy; BIO-TEK Instruments Inc., Winooski, VT). Reagent blanks (without bacteria) were used in each plate. SBA titers were defined as the reciprocal of the serum dilution with 50% of the fluorescent units (FU) detected in the complement controls.
The reproducibility of the SBA assays is evaluated using the immune globulin preparation with the 12 bacterial strains in 46 different independent assay runs.
The specificity of SBA assay for the Hia polysaccharide capsule was determined by performing a competitive SBA assay in the presence of 100 μg/ml of purified reagent-grade (<1.2 μg/ml protein) Hia-specific capsular polysaccharide (11) (and comparing the results with SBA titers obtained in the absence of added polysaccharide). For this evaluation, we used five selected bacterial strains (GA41512, GA41513, GA18491, GA11151, AK1435C3) and five selected healthy adult donor sera. We also evaluated the effect of heterologous Hib polysaccharide (provided by Moon Nahm, University of Alabama at Birmingham) in the Hia SBA assay. For this evaluation, we used the reference strain AK1435C3 and the same adult sera. Hib polysaccharide was added also at 100 μg/ml.
Anti-Hia polysaccharide IgG concentrations (μg/ml) were determined on cord blood sera using an Hia ELISA. The Hia ELISA is based on an anti-Hib polysaccharide IgG ELISA method previously described (6, 16). This optimized ELISA used a standard reference material which was assigned an anti-Hia polysaccharide IgG concentration (4.13 μg/ml). This anti-Hia IgG concentration was determined through a heterologous ELISA using the Hib standard reference material (FDA1983) as a calibrator serum (Schmidt, unpublished) based on the cross-standardization method described by Concepcion and Frasch (4).
Sera with negative ELISA (levels less than 0.01) were used as negative controls but were not included in the analysis.
Correlations between SBA assays were determined by Pearson's product moment correlation coefficient by use of Sigma Plot and Sigma Stat software, version 2.0 (SPSS, Inc., Chicago, IL). Significant differences among assays were determined by Student's t test or the Mann-Whitney rank sum test for data not normally distributed. The significance level was set at a P value of less than 0.05. The concordance correlation coefficients (CCC) were calculated to estimate the degree of agreement between pairs of log2-transformed SBA titers. Paired comparisons were made for each strain and AK1435C3 (CDC reference strain).
The reproducibility of the SBA assays was evaluated using the immune globulin preparation with different bacterial strains in 46 different independent assay runs. The geometric mean titer (GMT) was 237.4, with 97.8% of SBA titers within 1 dilution of the GMT.
SBA titers obtained by both methods were highly correlated (r = 0.96) (Fig. (Fig.1).1). No statistically significant difference was observed for the two assays when using healthy adult sera (P = 0.68) or cord blood sera (P = 0.94) among the 12 strains. SBA titers measured by both methods using all sera were within 1 and 2 dilutions for 96.6% and 98.3% of the titers, respectively (Fig. (Fig.11).
The specificity of the SBA assays was assessed by competitive inhibition in which the SBA assay activity was evaluated after the addition of purified Hia polysaccharide. The relative decrease in SBA titers for the five different strains tested varied between 99.2% and 99.4%, with an average of 99.3% (Table (Table2).2). No effect or reduction in SBA titers was observed by the addition of Hib polysaccharide in the panel of five sera evaluated with AK1435C3.
The correlation coefficients (r value) for SBA titers with fluorometric endpoint and IgG antibody concentrations as determined by ELISA varied between 0.77 and 0.89 depending on the strain (Table (Table3).3). SBA assay performed with the reference strain AK1435C3 had a moderate correlation (r = 0.89) with ELISA (Fig. (Fig.22).
Although Hib and Hia capsular polysaccharides are similar, they are not cross-reactive; Hia contains glucose in a 4-β-linkage, and Hib contains ribose in a 3-β-linkage (11). Fully vaccinated infants with Hib conjugate vaccine are not protected against Hia invasive disease (9).
Bactericidal antibodies against H. influenzae type b capsular polysaccharide confer immunity to this pathogen (14, 18). SBA assay has been extensively used for decades as a correlate of protection against invasive Hib as well as meningococcal diseases (7, 8). The many similarities between H. influenzae type a and type b, including their epidemiology, capsular polysaccharides (type-specific or serotype antigens), and the pathology of the systemic infections they cause, are the basis for our notion that serum bactericidal antibodies against Hia will have a comparable protective activity to that of Hib. Accordingly, we describe two new methods of performing this assay more efficiently. Although our study cannot determine if SBA assay is a correlate of protection for Hia, it demonstrated that the assay is reproducible, is specific for Hia polysaccharides, and has a moderate correlation with ELISA values (r value between 0.77 and 0.89). The moderate correlation between ELISA and SBA assay could be explained by the fact that ELISA measures functional and nonfunctional antibodies, while SBA assay measures only functional antibodies.
SBA assay results can be characterized with viability, fluorometric, or colorimetric endpoints. Previous work in Hib demonstrated that the best correlation for Hib SBA assay was obtained between the viability count endpoint and the fluorometric endpoint for titer determination (15). Therefore, we chose to compare only SBA assay with viability colony counts and fluorometric endpoints and not to assess the colorimetric endpoint. SBA assay using viability counts as endpoints is highly reproducible but time-consuming, with a low sample throughput and overnight incubation for CFU growth, and since Hia and Hib are grown on chocolate agar, the use of the available automated programs for data acquisition is quite difficult due to lack of contrast from the opaque medium. SBA assay using alamarBlue as a metabolic indicator is a same-day assay, highly reproducible, easy to use, with higher throughput, automated results, and an available interface to computer software. In our study, 96.6% of our SBA titers using either method were within one dilution of each other, and the coefficient correlation between the two methods was high (r = 0.96). The assay was highly specific and showed high agreement regardless of the target strain used.
We conclude that the viability colony counts and fluorometric SBA assays are reproducible, measure Hia-specific antibody function, and correlate with anti-Hia IgG antibody concentrations (μg/ml). In addition, SBA assay with a fluorometric endpoint is fast, easy to use, and highly reproducible and should be considered the preferred method for measuring Hia SBA.
We thank Tom Hennessy and Jay Wenger, AIP, CDC, AK, for their helpful comments. Our thanks go to Wendy Baughman and Paul Malpiedi for coordination of the Georgia Emerging Infections Program and Alisa Reasonover, Julie Morris, Marcella Harker-Jones, and Kim Boyd-Hummel from the Arctic Investigations Program for collection of the specimens and identification of the strains. Our special thanks to Alberto Villaseñor-Sierra, CMO, IMSS, and Javier-Diaz-Garcia, UNAM, for their participation in obtaining and managing cord serum samples and to Sabrina Ward, James A. Ferguson fellow, class of 2006, Association of Minority Health Professions Schools, Inc., stationed at the CDC, Atlanta, GA, for SBA work. We also thank Charles Rose, CDC, Atlanta, GA, for statistical support.
Published ahead of print on 22 December 2010.