Serum bactericidal assay for the evaluation of typhoid vaccine using a semi-automated colony-counting method

Abstract

Typhoid fever, mainly caused by Salmonella enterica serovar Typhi (S. Typhi), is a life-threatening disease, mostly in developing countries. Enzyme-linked immunosorbent assay (ELISA) is widely used to quantify antibodies against S. Typhi in serum but does not provide information about functional antibody titers. Although the serum bactericidal assay (SBA) using an agar plate is often used to measure functional antibody titers against various bacterial pathogens in clinical specimens, it has rarely been used for typhoid vaccines because it is time-consuming and labor-intensive. In the present study, we established an improved SBA against S. Typhi using a semi-automated colony-counting system with a square agar plate harboring 24 samples. The semi-automated SBA efficiently measured bactericidal titers of sera from individuals immunized with S. Typhi Vi polysaccharide vaccines. The assay specifically responded to S. Typhi Ty2 but not to other irrelevant enteric bacteria including Vibrio cholerae and Shigella flexneri. Baby rabbit complement was more appropriate source for the SBA against S. Typhi than complements from adult rabbit, guinea pig, and human. We also examined the correlation between SBA and ELISA for measuring antibody responses against S. Typhi using pre- and post-vaccination sera from 18 human volunteers. The SBA titer showed a good correlation with anti-Vi IgG quantity in the serum as determined by Spearman correlation coefficient of 0.737 (P < 0.001). Taken together, the semi-automated SBA might be efficient, accurate, sensitive, and specific enough to measure functional antibody titers against S. Typhi in sera from human subjects immunized with typhoid vaccines.

Keywords: Functional antibody; Salmonella enterica serovar Typhi; Serum bactericidal assay; Typhoid vaccine; anti-Vi IgG.

Fig. 1

Schematic diagram for the procedure of semi-automated SBA against S. Typhi. Human serum from volunteer immunized with Vi PS vaccine was tested for bactericidal activity against S. Typhi Ty2 using microtiter plate or semi-automated bactericidal assay. Optical density (O.D.) at 600 nm of a microtiter plate and the number of colony forming unit (CFU) of a square agar plate were measured in triplicate.

Fig. 2

Sera from volunteers immunized with typhoid vaccine specifically inhibit the growth of S. Typhi. Human sera from volunteers administered with Vi PS vaccine were tested against various enteric bacteria including (A) S. Typhi Ty2, (B) S. Paratyphi A, (C) S. Typhimurium, (D) S. Typhi Ty21a, (E) V. cholerae, and (F) S. flexneri by the semi-automated bactericidal assay. Human convalescent sera obtained from typhoid (G) and cholera (H) patients were tested against S. Typhi Ty2 as the positive and negative control, respectively. Baby rabbit complement (10%) was used in the assay against all bacteria except S. Typhi Ty21a (2.5%). Each value represents mean ± SEM of three independent experiments.

Fig. 3

Baby rabbit complement is appropriate for SBA against S. Typhi Ty2. (A–C) Sera from volunteers administered with Vi PS vaccine were tested with various mammalian complements. SBA was performed using (A) 10% of each complement, (B) 5% of each complement except baby rabbit (10%), or (C) 10% of baby rabbit complement from three different batches. Each value represents mean ± SEM of three independent experiments. (D) Anti-Vi IgG antibodies in each complement were measured using AP-conjugated anti-human, -rabbit or -guinea pig IgG. Values are mean ± SD of triplicates from each group.

Fig. 4

SBA titers are highly correlated with serum anti-Vi IgG titers. (A) Eighteen pairs of serum samples from individuals (pre- and post-vaccination) were analyzed for SBA titer and serum anti-Vi IgG. Geometric mean titer (GMT) of SBA and anti-Vi IgG were analyzed in pre- and post-immunized sera. P values were determined using two-tailed Student’s t-test for comparison between pre- and post-vaccinated groups. (B, C) Individual SBA titers were plotted against (B) anti-Vi IgG titers against Vi PS in 18 paired serum samples and (C) anti-Vi IgM titers in 10 paired samples. Spearman correlation coefficient (r) and P value were obtained to examine the relationship between two assays using GraphPad Prism 5 software.