Abstract
A serum neutralization assay (SN) was compared with the official plaque reduction neutralization test for the quantitation of West Nile virus antibodies. A total of 1,348 samples from equid sera and 38 from human sera were tested by these two methods. Statistically significant differences were not observed, thus supporting the use of SN for routine purposes.
TEXT
According to the International Committee on Taxonomy of Viruses (1), 53 virus species are included in the genus Flavivirus within the family Flaviviridae. Some of them are important mosquito-borne pathogens that have severe impacts on human and animal health. Together with important human flaviviruses, the West Nile virus (WNV) and Usutu virus (USUV)—the most widespread mosquito-borne flaviviruses in Europe (2, 3)—form the Japanese encephalitis virus serogroup (4, 5). Accordingly, flaviviruses may also cross-react and affect the specificities of many serological tests. The plaque reduction neutralization test (PRNT) is the most specific serological test for the proper serological identification of flaviviruses (4, 6, 7). The most stringent PRNT, which is represented by the 90% endpoint PRNT (PRNT90), is considered the gold standard protocol for the serodiagnosis of flavivirus infections. In differential PRNT, in which specimens are tested simultaneously for two or more viruses, a neutralizing titer of ≥4-fold for one virus compared with those of the others is considered virus specific (6). However, the possibility of coinfections cannot be discarded in cases of titers that are similar against two or more flaviviruses. PRNT still presents some drawbacks. First, it takes around 1 week to complete. Second, it is expensive and requires a highly trained staff. Third, PRNT with wild-type flaviviruses can be technically difficult, as many flaviviruses grow slowly and have pinpoint-sized plaque phenotypes. Indeed, viruses may comprise mixed populations, or quasispecies, that grow at different rates and produce plaques of various sizes. Last, PRNT requires a quantity of serum that is difficult to obtain when dealing with wild animals, particularly with some species of wild birds.
In order to test the consistency of serum neutralization in quantitating specific WNV antibodies, we tested by serum neutralization assay (SN) (8) and PRNT90 (9) 1,348 serum samples from equids (i.e., horses and donkeys) collected during the surveillance program for WNV neuroinvasive disease (WNND) organized by the Italian Ministry of Health and 38 serum samples from human patients collected during the first outbreak of WNND in humans in Croatia in 2012 (10). Collection of animal and human serum samples was approved by the Italian Ministry of Health and by the Croatian National Institute of Public Health, respectively. Furthermore, 25 serum samples from horses previously identified as positive for USUV were tested at the same time by SN and PRNT90 with USUV and WNV as antigens. The serum samples were inactivated at 56°C for 30 min. Starting from a dilution titer of 1:5, serial 2-fold dilutions were made in microtiter plates, and 100 tissue culture infective doses (TCID) of antigen were added to each dilution. Thereafter, the mixtures were incubated at 37°C for 1 h, and 105 Vero cells were added to each well. The plates were incubated at 37°C for 5 days. Starting from the third day after incubation, the plates were checked for cytopathic effect (CPE), and the antibody titer was defined as the reciprocal of the highest dilution of the serum that showed 100% neutralization. Positive and negative control sera were included in each plate. Sera with a titer of 1:10 were considered positive. Prior to SN and PRNT90, USUV and WNV antigens (strains 939/01 and Eg-101, respectively) were titrated by 50% TCID (TCID50) with Vero cells. After 4 days, the titer was determined by using the Reed and Muench formula. To compare SN and PRNT90, two different comparisons were performed for the resulting titers and positive/negative result values; (i) titers were compared with the Wilcoxon nonparametric test for dependent samples, and (ii) positive/negative results were compared using the McNemar χ2 test for dependent samples, and the level of agreement was evaluated by the Cohen κ coefficient (11). All the analyses were performed with SPSS Statistics version 17.0.
For serum samples from equids, the two diagnostic tests were not statistically different from each other in terms of titers or positive/negative results. The equid serum contingency table (Table 1) shows the level of agreement between the two tests. The McNemar χ2 test gave a value of 0.83 (P > 0.05), which indicates a nonsignificant difference between the two methods, and the Cohen κ, with a value equal to 0.78 (P < 0.05), indicates a good level of agreement beyond chance. Also, the titers were not significantly different from one another (Wilcoxon test, −1.69; P > 0.05). Similar results were also obtained when human sera were tested; the two tests were not statistically different from each other in terms of the titers and positive/negative results. The human serum contingency table (Table 2) shows the level of agreement between the two tests with human samples. The McNemar χ2 test gave a value of 1.33 (P > 0.05), which indicates a nonsignificant difference between the two methods, and the Cohen κ, which was equal to 0.81 (P < 0.05), indicates a good level of agreement beyond chance. Also, the titers were not significantly different (Wilcoxon test, −1.67; P > 0.05). The 25 samples from horses previously positive for USUV were also positive for WNV and USUV in the two tests across a broad range of titers (Table 3). The results from serum samples from horses 6, 8, 18, 19, 21, and 25 suggest that the presence of antibodies for the two viruses was a result of double infections, whereas the serum from horse 10 showed specific antibodies for only USUV. The remaining serum samples seemed to be positive for only WNV. Moreover, horses 6 to 10, 12, 23, and 25 showed recent infection by WNV, as suggested by the IgM enzyme-linked immunosorbent assay (ELISA) (Idexx IgM WNV).
TABLE 1.
Contingency table between PRNT and SN for equid serum samples
| SN result | PRNT result (no.) |
Total no. | |
|---|---|---|---|
| Positive | Negative | ||
| Positive | 1,010 | 54 | 1,064 |
| Negative | 45 | 239 | 284 |
| Total | 1,055 | 293 | 1,348 |
TABLE 2.
Contingency table between PRNT and SN for human serum samples
| SN result | PRNT result (no.) |
Total no. | |
|---|---|---|---|
| Positive | Negative | ||
| Positive | 10 | 0 | 10 |
| Negative | 3 | 25 | 28 |
| Total | 13 | 25 | 38 |
TABLE 3.
Serum samples from horses tested by SN and PRNT using WNV and USUV as antigens
| Horse no. | USUV titer from: |
WNV titer from: |
WNV IgMa | ||
|---|---|---|---|---|---|
| SN | PRNT | SN | PRNT | ||
| 1 | 1:20 | 1:10 | 1:160 | 1:80 | − |
| 2 | 1:10 | 1:5 | 1:160 | 1:80 | − |
| 3 | 1:10 | 1:20 | 1:40 | 1:40 | − |
| 4 | 1:10 | 1:20 | 1:80 | 1:40 | − |
| 5 | 1:10 | 1:20 | 1:40 | 1:40 | − |
| 6 | 1:20 | 1:20 | 1:20 | 1:20 | + |
| 7 | 1:20 | 1:20 | 1:40 | 1:80 | + |
| 8 | 1:20 | 1:20 | 1:40 | 1:20 | + |
| 9 | 1:10 | 1:20 | 1:40 | 1:20 | + |
| 10 | 1:40 | 1:20 | 1:5 | 1:5 | + |
| 11 | 1:5 | 1:5 | 1:320 | 1:320 | − |
| 12 | 1:5 | 1:10 | 1:40 | 1:20 | + |
| 13 | 1:5 | 1:5 | 1:40 | 1:20 | − |
| 14 | 1:10 | 1:10 | 1:80 | 1:80 | − |
| 15 | 1:40 | 1:20 | 1:640 | 1:320 | − |
| 16 | 1:80 | 1:40 | 1:640 | 1:640 | − |
| 17 | 1:20 | 1:20 | 1:320 | 1:160 | − |
| 18 | 1:80 | 1:80 | 1:160 | 1:80 | − |
| 19 | 1:160 | 1:80 | 1:640 | 1:320 | − |
| 20 | 1:10 | 1:10 | 1:320 | 1:320 | − |
| 21 | 1:80 | 1:40 | 1:320 | 1:160 | − |
| 22 | 1:10 | 1:20 | 1:640 | 1:320 | − |
| 23 | 1:10 | 1:10 | 1:160 | 1:320 | − |
| 23 | 1:80 | 1:80 | 1:640 | 1:320 | + |
| 25 | 1:320 | 1:160 | 1:160 | 1:80 | + |
−, negative assay result; +, positive assay result.
In this study, we showed with a large number of serum samples that SN can efficiently replace PRNT for the quantitation of WNV antibodies. The two diagnostic tests were not statistically different in terms of titers or positive/negative results, thus supporting the use of SN during the daily activities of diagnostic laboratories that deal with serological WNV investigation. SN is definitely faster (by up to 5 days) and requires less labor than PRNT. The use of SN can reasonably be translated for serological investigation for other flaviviruses. Cross-reactivity exists between WNV and USUV when the two serological methods are used. However, specific antibodies can be attributed to one or the other virus, as there is a neutralizing titer of ≥4-fold for one virus compared with the other independently in both serological assays. Nevertheless, these two viruses contemporaneously circulate in the same areas (12), which leads to the presence of seroreactors for both viruses. This seemed to be the case for horses 6, 8, 18, 19, 21, and 25. However, horses 19 and 21 showed ≥4-fold differences in titers by SN and PRNT for WNV compared with those for USUV, although the titers were significantly high for USUV. In this case, it was difficult to exclude the presence of a double infection. An efficient and up-to-date epidemiological picture of the circulation of flaviviruses in a given area helps to address the confusing results that potentially occur with these two assays. Importantly, even in the case of a double infection, SN performed similarly to PRNT and confirmed the feasibility of SN for routine serological purposes.
ACKNOWLEDGMENT
Funding was provided by the Italian Ministry of Health.
Footnotes
Published ahead of print 6 August 2014
REFERENCES
- 1.International Committee on Taxonomy of Viruses. 2013. Virus taxonomy: 2013 release. http://www.ictvonline.org/virusTaxonomy.asp.
- 2.Calistri P, Giovannini A, Hubalek Z, Ionescu A, Monaco F, Savini G, Lelli R. 2010. Epidemiology of West Nile in Europe and in the Mediterranean Basin. Open Virol. J. 4:29–37. 10.2174/1874357901004020029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Savini G, Monaco F, Terregino C, Di Gennaro A, Bano L, Pinoni C, De Nardi R, Bonilauri P, Pecorari M, Di Gialleonardo L, Bonfanti L, Polci A, Calistri P, Lelli R. 2011. Usutu virus in Italy: an emergence or a silent infection? Vet. Microbiol. 151:264–274. 10.1016/j.vetmic.2011.03.036. [DOI] [PubMed] [Google Scholar]
- 4.Kuno G. 2003. Serodiagnosis of flaviviral infections and vaccinations in humans. Adv. Virus Res. 61:3–65. 10.1016/S0065-3527(03)61001-8. [DOI] [PubMed] [Google Scholar]
- 5.Heinz FX, Collett MS, Purcell RH, Gould EA, Howard CR, Houghton M, Moormann RJM, Rice CM, Thiel HJ. 2000. Family Flaviviridae, p 859–878 In van Regenmortel MHV, Fauquet CM, Bishop DHL, Carstens EB, Estes MK, Lemon SM, Maniloff J, Mayo MA, McGeoch DJ, Pringle CR, Wickner RB. (ed), Virus taxonomy. Seventh report on International Committee on Taxonomy of Viruses. Academic Press, San Diego, CA. [Google Scholar]
- 6.Calisher CH, Karabatsos N, Dalrymple JM, Shope RE, Porterfield JS, Westaway EG, Brandt WE. 1989. Antigenic relationships between flaviviruses as determined by cross-neutralization tests with polyclonal antisera. J. Gen. Virol. 70:37–43. 10.1099/0022-1317-70-1-37. [DOI] [PubMed] [Google Scholar]
- 7.Johnson BW, Kosoy O, Hunsperger E, Beltran M, Delorey M, Guirakhoo F, Monath T. 2009. Evaluation of chimeric Japanese encephalitis and dengue viruses for use in diagnostic plaque reduction neutralization tests. Clin. Vaccine Immunol. 16:1052–1059. 10.1128/CVI.00095-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lelli R, Savini G, Teodori L, Filipponi G, Di Gennaro A, Leone A, Di Gialleonardo L, Venturi L, Caporale V. 2008. Serological evidence of Usutu virus occurrence in north-eastern Italy. Zoonoses Public Health 55:361–367. 10.1111/j.1863-2378.2008.01146.x. [DOI] [PubMed] [Google Scholar]
- 9.World Organisation for Animal Health (OIE). 2012. Manual of diagnostic tests and vaccines for terrestrial animals, 7th ed, vol 1, p 346–355 World Organisation for Animal Health, Paris, France. [Google Scholar]
- 10.Pem-Novosel I, Vilibic-Cavlek T, Gjenero-Margan I, Pandak N, Peric L, Barbic L, Listes E, Cvitkovic A, Stevanovic V, Savini G. 2014. First outbreak of West Nile virus neuroinvasive disease in humans, Croatia, 2012. Vector Borne Zoonotic Dis. 14:82–84. 10.1089/vbz.2012.1295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Siegel S, Castellan NJ. 1988. Nonparametric statistics for the behavioral sciences, 2nd ed. McGraw-Hill, New York, NY. [Google Scholar]
- 12.Savini G, Capelli G, Monaco F, Polci A, Russo F, Di Gennaro A, Marini V, Teodori L, Montarsi F, Pinoni C, Pisciella M, Terregino C, Marangon S, Capua I, Lelli R. 2012. Evidence of West Nile virus lineage 2 circulation in northern Italy Vet. Microbiol. 158:267–273. 10.1016/j.vetmic.2012.02.018. [DOI] [PubMed] [Google Scholar]