Abstract
Serological tests are widely used for Rift Valley fever (RVF) surveillance, disease control, and monitoring of immunological responses after vaccination. In recent years, several enzyme-linked immunosorbent assay (ELISA) formats have been developed for the detection of antibodies to RVF virus, but limited comparisons of their diagnostic sensitivities are available for specifically-targeted animal populations. This article describes the comparison of the commercially-available RVF recombinant nucleocapsid IgG indirect ELISA with the IgG sandwich ELISA in field-collected serum samples from 1262 domestic small ruminants in Mozambique. The agreement between the two tests measured by Cohen's kappa value was high (0.99 in goats and 0.92 in sheep), but the IgG sandwich ELISA was slightly more sensitive than the recombinant nucleocapsid IgG indirect ELISA in detecting the IgG response in sheep and goats naturally exposed to RVF virus.
Introduction
Among the serological tests available for the detection of antibodies against RVFV, the virus neutralization test (VNT) is reported to have the highest sensitivity and specificity, and it is regarded as the gold standard serological test (Swanepoel et al. 1986). However, a number of practical constraints are related to the use of this assay. It is expensive, laborious, can be performed only when a standardized stock of live virus and tissue culture are available, and requires high-level biocontainment facilities. To overcome these problems, enzyme-linked immunosorbent assay (ELISA) tests have been developed in various formats for the detection of specific IgM and IgG antibodies. A sandwich IgG and an IgM capture ELISA based on β-propiolactone-inactivated and/or gamma-irradiated, sucrose-acetone-extracted RVFV derived from mouse brain have been developed, extensively validated, and reported to be highly sensitive and specific for the detection of anti-RVFV antibodies in humans and domestic livestock (Paweska et al. 2003). However, since the production of whole antigens from live viral stocks poses a health risk to laboratory workers prior to inactivation, this procedure is limited to a few laboratories in Africa housing high-level biocontainment facilities. The recombinant nucleocapsid protein (recNP) of RVFV was successfully applied in an indirect ELISA format for the detection of antibodies to RVFV in domestic ruminants, humans, and wildlife (Fafetine et al. 2007; Jansen Van Vuren et al. 2007; Paweska et al. 2008).
The aim of this study was to compare results of two commercially-available ELISAs for the detection of IgG antibody to RVF virus (i.e., the indirect ELISA based on recNP antigen with the sandwich ELISA based on a whole antigen), in sheep and goat sera collected in Mozambique.
Material and Methods
Test sera
A total of 1262 serum samples (657 from goats and 605 from sheep), collected during a field survey of RVF in 2010 and 2011 in Mopeia District, Zambézia Province, Mozambique, were tested. The sera were from non-vaccinated animals and were distributed as follows: 453 female and 204 male goats; and 400 female and 205 male sheep. Most animals were from 0–6 months of age (336 goats and 382 sheep), followed by animals older than 12 months (197 goats and 170 sheep), and by animals between 7 and 12 months of age (124 goats and 53 sheep).
Enzyme-linked immunosorbent assays
The recNP-based IgG indirect ELISA was performed according to the manufacturer's instructions (BDSL, Dreghorn, Ayrshire, Scotland, U.K.), and published procedure (Jansen van Vuren et al. 2007). Briefly, the recNP of RVFV was coated overnight at 4°C on ELISA plates. The plates were washed three times with 0.1% Tweeen-20 in phosphate-buffered saline (PBS) (washing buffer), and blocked with PBS with 10% skim milk. After washing, duplicate volumes of the test, positive, and negative control sera were added and the plates incubated for 1 h at 37°C. After washing, recombinant Protein G HRPO conjugate (Zymed Laboratories, San Francisco, CA) was incubated, followed by washing and incubation with 2,2′-azinodiethylbenzothiazoline sulfonic acid (ABTS; KPL Laboratories, Inc., Gaithersburg, MD). The stop reagent, 1% sodium dodecyl sulfate (SDS) was added and optical densities (OD) were measured at 405 nm. The results were subsequently expressed as percentages of a high positive control serum (PP) using the formula [(mean OD of test serum/mean OD of high positive control)]×100.
The IgG sandwich ELISA was performed following a procedure published by Paweska and colleagues (2003). Polyclonal mouse anti-RVFV antibody (NICD-SPU) was passively adsorbed onto ELISA plates overnight at 4°C. The plates were washed with 0.1% Tweeen-20 in PBS (washing buffer) and blocked with 10% skim milk in PBS. After washing, RVFV-infected mouse brain (NICD-SPU), and control mouse brain antigen (NICD-SPU), were added to the rows of the top half of the plate (rows A–D: 1–12), and to the rows of the bottom half of the plate (rows E–G: 1–12), respectively. After incubation, the plates were washed and subsequently duplicate volumes of each test, positive, and negative control sera were added to the wells of the top half and the bottom half of the plates. The plates were incubated and washed and anti-sheep IgG HRPO conjugate was added. After incubation the plates were washed and ABTS was added to each well. The plates were then incubated and the reaction stopped by the addition of 1% SDS. Optical densities were determined at 405 nm. The net OD values were first recorded for each serum as the value determined with the RVFV antigen minus the value determined with the control antigen, and subsequently converted into a percentage of the OD value of a high positive control serum (PP), using the same formula as in the indirect ELISA.
Estimation of diagnostic accuracy
Since the IgG sandwich ELISA was validated against the VNT, it was considered the gold standard in this study and was used to calculate the agreement between the tests using Cohen's kappa value (Henken et al. 1997). The cut-off values previously optimized for the assays (a sample was positive with PP values greater than 25 for the indirect ELISA, and PP values greater than 11.1 for the sandwich ELISA), and recommended by the manufacturer were used in this study (Paweska et al. 2003; Jansen Van Vuren et al. 2007).
Results
In 657 goat and 605 sheep sera, the IgG sandwich ELISA was able to detect more positive animals than the IgG indirect ELISA (Table 1), but the overall agreement between the tests in sera of small ruminants was high, with kappa values of 0.99 and 0.92 in goats and sheep, respectively. Using the IgG sandwich ELISA as a gold standard, the sensitivity of the IgG indirect ELISA was 88.04% (95% CI 79.61,93.88) and 98.46% (95% CI 91.72,99.96), and the specificity was 99.81% (95% CI 98.92,100.00) and 100% (95% CI 99.38,100.00), in sheep and goats, respectively.
ELISA, enzyme-linked immunosorbent assay.
Conclusion
Diagnostic assay validation is paramount in determining estimates of specificity and sensitivity of specific assays for specific populations to be tested. Both assays used in this study were previously validated, but their performance has not been compared directly. The excellent agreement between the RVF recombinant nucleocapsid IgG indirect ELISA and the IgG sandwich ELISA for the detection of IgG to RVFV in domestic small ruminants demonstrates that these ELISAs are suitable diagnostic tools for the detection of specific anti-RVF IgG. The higher sensitivity of the IgG sandwich ELISA might be explained by the use of a species-specific HRPO conjugate in this assay, which possibly has higher affinity to IgGs generated early after infection than Protein G. Although the IgG sandwich ELISA was more sensitive than the indirect ELISA, advantages of the latter assay include a much easier and shorter procedure, which makes it more suitable for disease surveillance programs requiring the testing of large amounts of sera. Another major advantage is cost-effective production of recombinant nucleocapsid protein as the antigen in the ELISA, as well as its safety and stability, and the fact that high-quality antigen could be produced in large amounts without the need for high-level biocontainment facilities (Jansen van Vuren et al. 2007), thus enabling active surveillance programs in poor African countries.
Footnotes
Acknowledgments
We are very grateful to the staff of the Center for Emerging and Zoonotic Diseases (National Institute for Communicable Diseases, South Africa), and the Veterinary Faculty (Eduardo Mondlane University, Mozambique) for their technical support.
This work was supported by funds from SIDA/SAREC, Eduardo Mondlane University, and by the Utrecht University, The Netherlands.
Author Disclosure statement
No competing financial interests exist.