Microsphere-based antibody assays for human parvovirus B19V, CMV and T. gondii

B19V IgG SIA: serum samples from 80 children or adults (2 to 58 years of age, median 10 years) with symptomatic B19V infection [1, 5]. These included single samples from 16 subjects and 171 samples from 64 subjects followed up serologically for up to 700 days after primary infection. As control group we included serum samples from 104 constitutionally healthy medical students. The seroprevalences of B19V and several emerging viruses [6‐8] have been previously determined with these samples.

CMV and T. gondii IgG SIAs: serum samples from 72 subjects; single sera from 64 subjects and two consecutive sera from eight subjects. These samples have been tested for IgGs against T. gondii and CMV by the corresponding Vidas enzyme linked fluorescent assays (ELFAs, bioMérieux) and in-house B19V EIA.

The following antigens were coupled with microspheres: for B19V IgG, cloned and purified in-house recombinant VP2 virus-like particles (VLPs) [1, 5, 9]. For CMV IgG, we used purified viral lysate (strain AD 169; Advanced Biotechnology) and for T. gondii IgG, purified tachyzoite lysate (RH strain; Advanced Biotechnology). The antigens and conditions for each assay are presented in Table 1.

The B19V-magnetic microsphere coupling was performed according to manufacturer’s protocol (“Sample protocol for two step carbodiimide coupling of protein to magnetic microsphere”, Luminex based xMap technology). In each coupling, a total of 1.25 × 10⁶ microsphere (100 μl) were coated with the corresponding antigen/protein followed by blocking with PBS + 50 mM/L Tris + 0.5 mL/L Tween-20 [10]. The coupled microspheres were washed twice with StabilGuard buffer (SG, SurModics) and stored in 1 mL SG at 4 °C in the dark. The antigen concentration was optimized by titration ranging from 200 ug to 0.8 ug per coupling.

The CMV- and T. gondii- microsphere couplings were performed using non-magnetic microspheres as described by Elfaitouri et al. [11].

The B19V SIA was performed following the manufacturer’s protocol (“Sample protocol for indirect antibody capture immunoassay”). The B19V SIA was optimized with serial serum dilutions ranging from 1:5 to 1:320 in phosphate buffered saline with 0.05 % Tween-20 (PBST). The optimal serum dilution is presented in Table 1. Naked microspheres were included as background control. In brief, 50 μl of diluted sera were dispersed in each well of 96-well flat bottom plate. Then the diluted sera were incubated with B19V-coated and naked microspheres for 45 minutes in the dark. After 3 times washing cycles, 50 μl of 2 ~ 4 μg/ml biotinylated protein G (Thermo Scientific) was added in each well. The plate was thoroughly mixed and incubated for 30 minutes in the dark. After 3 times washing cycles, the specific signal was developed by incubation with 2 ~ 4 μg/ml streptavidin conjugated phycoerythrin (R-PE, Life Technologies) in PBST. Each well was resuspended in 120 μl of PBST and read on the Bio-Plex®200 instrument.

The CMV and T. gondii SIAs were performed as described [11]. They were also compared with 27 sera against the corresponding Siemens BEPIII IgG tests, with good concordance (courtesy of Bo Albinson, Uppsala; data not shown).

In multiplex and singleplex SIAs, the intra-assay variability was calculated with 8 replicates in the same run, and the inter-assay variability with 6 distinct runs.

The SIAs cutoffs were calculated by the means and standard deviations (SDs) of test values. For B19V IgG SIA, the cutoff was defined with 72 sera confirmed to be B19V seronegative by both Biotrin’s and in-house B19V IgG and IgM EIAs. The sera were from children with expiratory wheezing, studied earlier for human bocavirus 1 [7]. The CMV and T. gondii cut-offs were established with separate sets of 60 sera each. These samples were IgG seronegative as confirmed by the Abott assays (Table 1).

We used two-way contingency table analysis in ‘VassarStats’ for the calculation of kappa value, sensitivity, and specificity. Borderline values in reference assays were excluded from the calculations. The agreement between SIA and EIA was evaluated by kappa values and defined as: poor (<0.20), fair (0.21–0.40), moderate (0.41–0.60), good (0.61–0.80), and very good (0.81–1.00) [12]. Pearson correlation coeffients (R²) were calculated by GraphPad Prism 6 to determine the correlation of results between singleplex and multiplex IgG SIAs.

The net cost was calculated by the sum of all the reagents’ costs (see Additional File 1).

The Helsinki University Hospital Ethics Committee approved the use of all clinical samples included in this study (Dnro 553/E6/2001, § 66, 13.4.2011). The control serum samples in B19V SIA study were obtained from the medical students with informed consent. All other samples in this study were taken as part of standard care and were analyzed anonymously.