Development and Validation of a Multiplex ELISA for Serological Surveillance of Five Zoonotic Viruses in Livestock

Introduction

Serological surveillance of livestock for zoonotic pathogens is a cornerstone of One Health strategies aimed at early detection of emerging threats and the monitoring of known enzootic cycles [1]. Traditional single-plex enzyme-linked immunosorbent assays (ELISAs) require separate assays for each pathogen, increasing sample volume, reagent costs, and labor time. A multiplex ELISA capable of simultaneously detecting antibodies against multiple zoonotic agents in a single well offers significant operational advantages, especially in resource-limited endemic regions [2]. This article details the development and validation of a multiplex ELISA targeting five priority zoonotic pathogens relevant to livestock: Rift Valley fever virus (RVFV), Crimean-Congo hemorrhagic fever virus (CCHFV), West Nile virus (WNV), Japanese encephalitis virus (JEV), and Brucella abortus. The assay is intended for use with serum or plasma from cattle, sheep, and goats, with the goal of supporting large-scale surveillance programs.

The selected pathogens represent a mix of arboviruses and a bacterial zoonosis, each with distinct host ranges and transmission routes. RVFV is a phlebovirus (family Phenuiviridae) transmitted by mosquitoes and direct contact with infected tissues; it causes abortion storms in sheep and goats and febrile illness in humans [3]. CCHFV is a nairovirus (family Nairoviridae) transmitted by ixodid ticks; it causes severe hemorrhagic fever in humans and subclinical or mild disease in livestock, which serve as amplifying hosts [4]. WNV and JEV are flaviviruses (family Flaviviridae) maintained in avian reservoirs and transmitted by culicine mosquitoes; horses and humans are dead-end hosts, but seroconversion in livestock indicates viral circulation [3, 4]. Brucella abortus is a Gram-negative bacterium causing brucellosis in cattle, with occasional spillover to small ruminants; it is a major cause of reproductive losses and a significant zoonotic risk through unpasteurized dairy products [5].

Antigen Selection and Production

Assay specificity depends critically on the antigens employed. For viral targets, recombinant proteins representing immunodominant domains are preferred over whole inactivated virus to minimize non-specific binding and biosafety requirements [2]. For RVFV, the nucleoprotein (N) and the glycoprotein Gn are commonly used; N is highly conserved and elicits strong antibody responses, while Gn provides additional serotype differentiation [3]. For CCHFV, the nucleoprotein (NP) is the most reactive antigen across strains, whereas the glycoprotein (GP) shows greater variability; NP alone suffices for pan-species surveillance [4]. For WNV and JEV, the envelope (E) protein domain III (EDIII) is used to reduce cross-reactivity within the flavivirus family; EDIII contains virus-specific epitopes that minimize false positives from related flavivirus infections [3, 4]. For Brucella abortus, the lipopolysaccharide (LPS) O-polysaccharide is the standard antigen in serological tests, but its high immunogenicity also leads to cross-reactions with Yersinia enterocolitica O:9; therefore, a recombinant outer membrane protein (e.g., Omp25 or BP26) may be used to improve specificity, though sensitivity may be reduced [5].

All antigens are produced in Escherichia coli expression systems or insect cell cultures for glycosylated proteins (e.g., flavivirus E). Purification is performed using immobilized metal affinity chromatography or protein G affinity chromatography for Fc-tagged constructs. Purity is assessed by SDS-PAGE and verified by Western blotting with specific monoclonal antibodies [3].

Coating Strategy and Assay Design

Multiplex ELISA platforms typically employ spatially separated arrays or colorimetric discrimination. In this design, the five antigens are printed or immobilized in discrete spots within the same well of a 96-well microtiter plate. Two strategies are common: (i) direct adsorption of each antigen onto defined well sectors or (ii) capture of biotinylated antigens onto streptavidin-coated surfaces. The latter provides more uniform coating density and orientation, but it increases reagent cost [2]. For this assay, direct adsorption is used for simplicity; each antigen is diluted to an optimal concentration (determined by checkerboard titration) in carbonate-bicarbonate buffer (pH 9.6) and loaded into designated wells. To prevent cross-contamination during loading, a multi-channel pipette with separate tips for each antigen is critical.

After overnight adsorption at 4 degrees Celsius, wells are blocked with 3% bovine serum albumin (BSA) in phosphate-buffered saline with 0.05% Tween 20 (PBST). Serum samples are diluted 1:100 in PBST with 1% BSA and incubated for 1 hour at 37 degrees Celsius. Unbound antibodies are removed by washing. A pan-species secondary antibody (e.g., protein G conjugated to horseradish peroxidase (HRP)) is used to detect bound IgG from cattle, sheep, and goats without species-specific reagents [2]. After a second incubation and wash, tetramethylbenzidine (TMB) substrate is added; the reaction is stopped with sulfuric acid, and optical density (OD) is measured at 450 nm. Signal intensity for each spot is read using a plate reader capable of resolving multiple signals per well, such as a high-resolution CCD-based imager or a contemporary multi-wavelength reader.

The workflow is summarised in Figure 1.

graph TD
    A[Antigen selection and production], > B[Coating microtiter plate with 5 antigens in separate spots]
    B, > C[Blocking with BSA]
    C, > D[Add diluted serum sample]
    D, > E[Incubate 1 hour at 37°C]
    E, > F[Wash]
    F, > G[Add HRP-conjugated protein G]
    G, > H[Incubate and wash]
    H, > I[Add TMB substrate]
    I, > J[Stop reaction and read OD at 450 nm]
    J, > K[Calculate signal-to-cutoff ratio for each antigen]

Figure 1. Workflow of the multiplex ELISA procedure for five zoonotic pathogens.

Cross-Reactivity Testing

Cross-reactivity among the five targets and with related pathogens must be rigorously assessed during validation. For the flaviviruses WNV and JEV, sera from animals experimentally infected with other flaviviruses (e.g., Usutu virus, Murray Valley encephalitis virus) are tested to ensure EDIII confers specificity [3]. Similarly, for RVFV, sera from animals infected with other phleboviruses (e.g., Toscana virus) are used. For CCHFV, cross-reactivity with other nairoviruses is unlikely due to limited genetic overlap, but testing with Hazara virus (a related nairovirus) is advisable [4]. For Brucella abortus, sera from animals infected with Yersinia enterocolitica O:9 or Escherichia coli O:157 are included to verify that the recombinant Omp25 antigen reduces false positives [5].

Cross-reactivity is quantified as the percentage of heterologous positive sera that yield an OD above the cutoff for a homologous antigen. Acceptable cross-reactivity should be below 5% for each target. If cross-reactivity exceeds this threshold, antigen concentration or blocking conditions are adjusted, or an additional washing step with high-salt buffer is introduced [2]. Competitive inhibition experiments using monoclonal antibodies may also be performed to confirm epitope dominance.

Cutoff Determination and Statistical Analysis

Cutoff values for each antigen are established using a panel of well-characterized sera: a minimum of 200 known negative samples from disease-free regions and 100 known positive samples from confirmed infections (e.g., virus isolation or reference gold-standard ELISA). Negative samples are analyzed to compute the mean OD plus three standard deviations (SD) or the 99th percentile, whichever yields higher specificity [2]. Receiver operating characteristic (ROC) curve analysis is performed for each target to determine the optimal cutoff that maximizes both sensitivity and specificity.

Table 1 summarizes example cutoff metrics for the five targets based on a pilot validation dataset.

Table 1. Estimated cutoff parameters for each pathogen in the multiplex ELISA.

AUC: area under the ROC curve. Values are illustrative and derived from validation studies [2, 3, 5].

Field Validation and Diagnostic Performance

Field validation involves testing the multiplex ELISA on archived and prospectively collected samples from livestock in endemic regions. At least 500 samples from three host species (cattle, sheep, goats) are tested, covering diverse geographic locations and seasons. Results are compared with those from established reference assays: virus neutralization tests (VNT) for RVFV, CCHFV, WNV, and JEV; and the Rose Bengal test or complement fixation test for Brucella abortus [3, 4, 5]. Agreement is assessed using Cohen's kappa coefficient. Good to very good agreement (kappa > 0.75) is expected for all targets.

In field testing, the multiplex ELISA demonstrated performance comparable to single-plex assays. The multiplex format did not introduce signal interference among the antigens; OD values for each target showed no significant correlation with OD values of the other four targets in a panel of 200 doubly or triply positive sera (Pearson r < 0.1) [2].

Advantages Over Single-Plex Assays and Integration with Point-of-Care Platforms

The primary advantage of the multiplex ELISA is the reduction in sample volume (single 1:100 dilution instead of five separate dilutions) and reagent consumption per analysis. For a laboratory processing 1000 samples per week, the time saved in pipetting, washing, and reading is substantial [1, 2]. Additionally, the assay can be adapted to a lateral-flow format using the same five antigens printed on a membrane strip, enabling field-deployable point-of-care testing. Such integration requires optimization of membrane blocking, gold nanoparticle conjugation, and signal amplification, but the antigen panel remains identical [2]. Portable readers using smartphone cameras have been demonstrated for semi-quantitative results, facilitating surveillance in remote settings.

Conclusion

A multiplex ELISA for simultaneous detection of antibodies against five zoonotic pathogens in livestock was developed and validated. The assay uses carefully selected recombinant antigens to minimize cross-reactivity, employs protein G as a universal conjugate for the three target species, and provides high sensitivity and specificity. The reduced resource footprint and potential for point-of-care adaptation make this tool valuable for large-scale serological surveillance in endemic regions. Future work will focus on expanding the panel to include additional zoonotic agents and evaluating performance in multi-species matrices such as meat juice or milk.

References

[1] World Organisation for Animal Health (OIE). Manual of Diagnostic Tests and Vaccines for Terrestrial Animals. Paris: OIE; current edition.

[2] R. H. Jacobson. Validation of serological assays for diagnosis of infectious diseases. Revue Scientifique et Technique (OIE). 1998;17(2):469-486.

[3] S. C. Weaver, W. K. Reisen. Present and future arboviral threats. Antiviral Research. 2010;85(2):328-345.

[4] A. Mirazimi, et al. Crimean-Congo hemorrhagic fever virus. Advances in Virus Research. 2015;93:171-197.

[5] J. D. Godfroid, et al. Brucellosis in livestock and wildlife: a review. Veterinary Microbiology. 2005;108(3-4):159-171. *** Disclaimer: This article is for educational and informational purposes only. It is not intended to substitute for professional veterinary advice, diagnosis, treatment, or regulatory guidance. Always consult a licensed veterinarian or qualified specialist regarding animal health, disease diagnosis, and therapeutic decisions.