Fig 1.
Validation of the MMIA using serum samples from humans or NHPs known to be infected with RVFV (1 NHP), EBOV (14 NHPs), MARV (3 NHPs), CCHFV (10 humans from Tajikistan or Nigeria), or DOBV (2 humans from Europe).
Previously characterised serum samples were used as a validation panel to confirm the ability of the MMIA to detect anti-NP antibodies generated in response to infection with these viruses. These serum samples were assayed by MMIA under the same conditions that were used to assay the serum panels from the German blood bank and Guinea, thus providing a range of MFI values for each viral NP coupled microsphere region for known positive serum samples. For EBOV and MARV validation, serum taken from each NHP prior to experimental infection (pre) is compared to serum taken at various time points post-infection. For the RVFV NHP serum and for CCHFV and DOBV human serum, samples taken at a single time point after infection are compared to presumed negative human serum samples.
Table 1.
Comparison between the MMIA and the commercial IIFT/ELISA or in-house ELISA results using the serum samples from the validation panels.
* Only validation panels where all samples did not test 100% positive were confirmed by commercial immunoassay. ** The same serum samples tested positive in both immunoassays.
Fig 2.
Histograms displaying the results of the MMIA for each viral NP.
The x-axis displays the average MFI value for each serum sample tested in triplicate and the y-axis displays the number of serum samples that resulted in each MFI value. The serum panel from the (A) German blood bank and (B) from Guinea are shown individually. A conservative cutoff value is indicated by a vertical line on each plot, and was chosen based on the 99th percentile for serum samples in the German blood bank panel. This is a population based cut-off and assumes that 99% of serum samples in the German blood bank are negative for the indicated virus. Samples were assigned a positive or negative status based on this cutoff, and in all cases, negative serum samples are shown in grey and positive serum samples are shown in red.
Table 2.
Comparison of the number of serum samples that tested positive and negative against each viral NP in the German blood bank and Guinean serum sample sets.
The median MFI readings from three technical replicates per sample were analysed in R, and a positive or negative status was assigned to each serum sample. Positivity or negativity was based on a cutoff MFI value that was determined for each viral NP based on the 99th percentile for MFI values in the German blood bank sample set. Statistically significant differences in seroprevalence rates between the two sample sets and odds ratios were calculated using Fisher’s exact test.
Fig 3.
Correlation between MMIA results and commercially available or in-house ELISAs for EBOV (A), MARV (B) and CCHFV (C). For both the MMIA and the ELISAs results were obtained from 3 technical replicates for each sample and the average MFI value for each NP (EBOV, MARV or CCHFV) was then plotted against the average absorbance values obtained from the same sample in the corresponding ELISA. In (A) and (B) all serum samples from the Guinea panel and the validation panel were tested by in-house EBOV or MARV ELISAs, in (C) a subset of samples from the Guinea panel and validation panel were tested by commercial CCHFV IgG ELISA.
Fig 4.
Analysis by MMIA of serum samples taken pre- and post-inoculation with EBOV- and MARV-VLPs from Mops condylurus micro bats.
Serum samples were taken from micro bats 24 days after inoculation with EBOV VLPs and MARV VLPs, and their reactivity to EBOV NP and MARV NP coupled microspheres was compared to sera taken pre-inoculation. The MMIA was adapted to detect bat antibodies by exchanging the SAPE-conjugated anti-human secondary antibody with SAPE-conjugated protein A and protein G. MFI readings for the same negative and EBOV-NP and MARV-NP positive NHP sera used as controls in all MMIAs are shown in the first 2 columns.