Abstract
Antibodies to West Nile virus (WNV) were examined by plaque-reduction neutralization test in the blood sera of 1023 wild artiodactyls: 105 roe deer (Capreolus capreolus), 148 red deer (Cervus elaphus), 287 fallow deer (Dama dama), 71 mouflons (Ovis musimon), and 412 wild boars (Sus scrofa), sampled in South Moravian district of Breclav (Czech Republic) in the years 1990–2008. Neutralizing antibodies were detected in 5.9% of wild ruminants (4.8% roe deer, 4.1% red deer, 6.3% fallow deer, 9.9% mouflons) and 4.1% of wild boars, with titers ranging between 1:20 and 1:320. The results indicate that WNV has circulated in wild artiodactyls at a variable frequency during the years in the area.
Introduction
T
In the Czech Republic, WNV was isolated from Culex pipiens mosquitoes in the district of Breclav (southern Moravia) after a big flood in 1997 (Hubálek et al. 1998) and this isolate was later identified as a new lineage 3 (“Rabensburg”) of WNV (Bakonyi et al. 2005). Neutralizing antibodies to WNV were detected in 2.1% of local human population in 1997, and five cases of WNV fever were recorded—two confirmed and three probable (Hubálek et al. 1999b). More recently, four strains of WNV lineage 2 were isolated from Culex modestus mosquitoes collected in reed beds on the South Moravian fishponds during August 2013 (Rudolf et al. 2014).
WNV lineage 3 (“Rabensburg”) was also isolated from Aedes rossicus mosquito during virological examination for arboviruses between 2006 and 2008 (Hubálek et al. 2010). This mosquito species feeds preferably on mammals including humans (Becker et al. 2010), whereas a majority of competent mosquito vectors of WNV are ornithophilic.
The aim of this study was to evaluate retrospectively the activity of WNV in southern Moravia in wild mammals before the flood year 1997 and later using serological survey of archived wildlife sera.
Materials and Methods
Hunting areas
Several hunting areas in the district of Breclav, South Moravian region, were surveyed. “Soutok” is an extensive area of the floodplain forest ecosystem with meadows at the confluence of the rivers Dyje and Morava (48.63–48.74 N, 16.90–16.98 E)—a game reserve with wild deer (roe, red, and fallow: Capreolus capreolus, Cervus elaphus, and Dama dama, respectively) and wild boars (Sus scrofa). “Palava” (Pavlovské vrchy hills and foothills) is a protected landscape area (48.80–48.87 N, 16.65–16.74 E) that also includes a game reserve with deer (roe, red, and fallow), mouflons (Ovis musimon), and wild boars. In addition, a number of smaller, dispersed other hunting grounds were sampled in the district of Breclav (marked as “BV-other”). Most of the sampling sites were situated in habitats with abundant mosquito populations (Sebesta et al. 2010).
Blood sampling
The wild animals were shot by hunters during hunting seasons, and blood samples were collected from the heart or from the thoracic cavity. After clotting, the samples were centrifuged in the laboratory and the sera were stored at −20°C until use.
Plaque-reduction neutralization microtest
Plaque reduction neutralization microtest (PRNT), originally proposed by Madrid and Porterfield (1969), was later adopted to a microtechnique on 96-well (flat-bottomed) microplates for cell culture (Hubálek et al. 1979). In brief, tested sera were thermally inactivated and diluted 1:10; 35 μL of the diluted serum was mixed in a microplate well with 35 μL test dose of the virus (containing 20–30 PFU (plaque-forming units) of WNV strain Eg-101) in L-15 medium, and incubated at 37°C for 60 min; 60 μL of Vero E6 cell suspension (20,000–30,000 cells) was then added to each test well, and 120 μL of carboxymethylcellulose sodium salt was overlayed after an incubation at 37°C for 4 h. The microplates were incubated at 37°C for 4–5 days and stained with naphthalene blue black. Sera were tested in duplicate, and controls included the virus test dose and its titration, immune WNV reference serum, control negative serum, and cells without virus. Cytotoxic sera were excluded from the analysis. Sera, revealing 80% or greater reduction in the number of WNV plaques at the 1:20 dilution during the screening, were titrated by twofold dilutions, and those dilutions corresponding to 80% reduction of plaque counts were regarded as the serum titers (PRNT80). Reciprocal titers ≥20 were considered positive. In addition, those sera reacting with WNV were then also tested in PRNT against two other flaviviruses occurring in Central Europe, that is, tick-borne encephalitis virus (TBEV) strain Hypr and Usutu virus (USUV) strain 939, to exclude cross-reactions with these antigenically related viruses.
Confidence intervals of proportions in seropositivity rate were calculated with Epi Tools, based on Wilson score interval. Differences in proportions among positive animals were statistically evaluated by χ 2 test considering the following variables: mammalian species, hunting areas, and years.
Results
In total, neutralizing antibodies against WNV were detected in 53 out of the 1023 wild artiodactyls examined (5.2%). Table 1 shows WNV seropositivity: no marked cross-reactions with the two other flaviviruses (TBEV and USUV) were recorded in that titers to WNV were in all cases (except no. 1169) higher than those to the other viruses. Specific antibodies to WNV were thus detected in 5.9% of wild ruminants (4.8% of 105 roe deer C. capreolus, 4.1% of 148 red deer C. elaphus, 6.3% of 287 fallow deer D. dama, 9.9% of 71 mouflons O. musimon) and 4.1% of 412 wild boars S. scrofa (a nonruminant artiodactyl), with titers between 1:20 and 1:320. Table 2 compares the WNV seropositivity including titer ranges and mean titers in the five animal species.
PRNT titer is defined as the reciprocal dilution of serum that produced an 80% decrease in virus PFU (plaque-forming units) count as compared with a virus-only titration control. The titers 20 or higher were regarded positive, and printed in bold.
f, female; m, male; nt, not tested (not enough serum); ng, data not given by the hunters.
PRNT, plaque reduction neutralization microtest; TBEV, tick-borne encephalitis virus; USUV, Usutu virus; WNV, West Nile virus.
CI, confidence interval.
Significantly higher seropositivity rate was detected in mouflons (9.9%) than in wild boars (4.1%; χ 2 = 4.21; d.f. = 1; p = 0.040). All the other differences among pairs of the five species were insignificant.
Significant difference (χ 2 = 6.05; d.f. = 1; p = 0.014) in the frequency of WNV seropositive animals was found between the hunting areas “Palava” (7.6% of 340 animals examined) and “BV-other hunting areas” (2.3% of 175 animals), but not between “Palava” and “Soutok” (4.5% of 508 animals; χ 2 = 3.64; d.f. = 1; p = 0.056).
There was considerable fluctuation in WNV seropositivity rate among five sampling periods (Table 3): overall value of χ 2 = 19.80 (d.f. = 4; p = 0.0005). The high prevalence of antibodies in wild animals was found especially in the period 1994–1997, that is, before and during the big flood, and then in 2008 (12.5% rate in that year).
Discussion
A few WNV serosurveys were published that included wild ruminants and boars in Europe. For instance, Kozuch et al. (1976) detected WNV neutralizing antibodies in 6.2% of 65 roe deer, 3.0% of 101 red deer, 8.3% of 24 fallow deer, and 2.6% of 38 wild boars in western Slovakia during the years 1969–1972. Juricová and Hubálek (1999) found 13–15% of 400 wild ruminants and 150 boars with WNV hemagglutination-inhibiting (HI) antibodies in South Moravia, Czechland. However, HI test with flaviviruses (as well as ELISA) is considered less specific than neutralization test because of cross-reactions among these viruses (Calisher et al. 1989, Niedrig et al. 2007). In the South Moravian region, Halouzka et al. (2008) detected antibodies neutralizing WNV in 6.5% of 93 wild boars, the figure being much closer to the data in this study. In Spain, the proportion of WNV-seropositive (in ELISA) individuals was 4.0% among 742 wild boars, whereas only 0.2% of 887 juvenile wild red deer reacted between 2003 and 2011 (Boadella et al. 2012). The authors, therefore, suggest that wild boars are potentially useful as sentinel animals for the WNV surveillance in southern Europe. Other Spanish serosurveys on WNV in wild artiodactyls include those of Gutiérrez-Guzmán et al. (2012) and García-Bocanegra et al. (2016), with WNV antibody prevalence rates similar to our study. In Serbia, Escribano-Romero et al. (2015) used ELISA combined with a confirmatory neutralization test, and found WNV antibodies in 5.5% of 91 roe deer and in 10.4% of 318 wild boars.
Sampling of sera was not homogeneous in our survey because of the inability to collect regularly samples from particular animal species in different hunting areas and years for such a long period, and the results might thus be biased to certain extent. It is, therefore, difficult to discuss or explain adequately the differences in seroprevalence rates among individual species of artiodactyls, between hunting grounds or periods.
We cannot attribute the detected antibodies to either WNV-1 or WNV-3 lineage because although both lineages differ genomically, they are very closely related antigenically (Hubálek et al. 1999b) and the infecting viruses could not be differentiated by using parallel titration in PRNT.
In conclusion, data of this retrospective study indicate fluctuating WNV activity in southern Moravia between the years 1990 and 2008 (i.e., already prior the flooding year 1997) and they suggest a moderate contact of wild artiodactyls with WNV in the area.
Footnotes
Acknowledgments
Technical assistance of L. Ševčíková and J. Peško is gratefully acknowledged. The USUV strain 939 was generously supplied by Prof. N. Nowotny and Dr. T. Bakonyi from Vienna Veterinary University. The study was partially supported by the Czech Science Foundation (16-20054S).
Author Disclosure Statement
No competing financial interests exist.