Seroconversion in Wild Birds and Local Circulation of West Nile Virus, Spain

A serosurvey for neutralizing antibodies against West Nile virus (WNV) in common coots (Fulica atra) was conducted in Doñana, Spain. Antibody prevalence was highest in 2003, intermediate in 2004, and lowest in 2005. Some birds seroreverted <1 year after first capture. Seroconversion of birds suggests local circulation of the virus.

I n western Mediterranean countries, the frequency of outbreaks of West Nile virus (WNV) infection has increased in recent decades. Evidence for WNV circulation in Spain has remained elusive, although WNV foci have recently been identifi ed in 3 neighboring countries (Morocco, Portugal, and France) (1)(2)(3). Recent WNV activity in Spain has been shown by serologic screening in humans, with detection of WNV-specifi c immunoglobulin M (4) and identifi cation of the fi rst clinical case in 2004 (5). In avian hosts, WNV-neutralizing antibodies have been found in chicks of wild migratory birds in southern Spain (6). However, interpretation of serologic data is not straightforward because antibodies in chicks may be the result of maternal transmission through eggs (7). To ascertain local circulation of WNV in Spain, we designed a capture-recapture study in which serum samples from wild birds were obtained at different times.

The Study
We focused on the partially migratory common coot (Fulica atra) because of its high seroprevalence for WNV detected during a preliminary screening of 72 bird species (J. Figuerola et al., unpub. data). Reasons for this high seroprevalence remain unclear, although preference of this bird for mosquito-rich habitats and its relative size (weight ≈800 g) might be involved in this pattern. Birds were captured in Overall, 853 captures of 515 different birds were conducted (1-7 captures/bird).
Blood was obtained from the tarsal vein and allowed to clot, and serum was stored at -20°C. All birds were marked with numbered metal rings. Age was determined by plumage characteristics before the birds were released. Neutralizing antibody titers for WNV (strain Eg101) were determined by using a micro-virus neutralization test as described (6). Only birds that showed neutralization (absence of a cytopathic effect) at dilutions >1:20 were considered seropositive. Controls for cytotoxicity in the absence of virus were included for every sample at a 1:10 dilution. Cytotoxic samples were excluded from the analysis.
Seroconversion was defi ned as a bird that was seronegative when fi rst captured and became seropositive at recapture with an antibody titer that had increased 4-fold (8). Seroreversion was defi ned as a seropositive bird whose antibody titer decreased below the cut-off value of 20 at recapture. The interassay coeffi cient of variation of titers, expressed as log 10 (calculated using an internal control repeated in 5 different assays, mean 2.56, standard deviation 0.35) was 13.67%. This variation is similar to that observed in individual samples and repeated in different assays. In a series of 27 samples tested twice, the mean fl uctuation observed was 0.29 log 10 units (≈2-fold). To obtain accurate measurements of titers, particularly when assessing seroconversion/seroreversion, we analyzed samples at least twice, and when results differed, they were assayed again until a consistent result was obtained. Specifi city of the test was assessed by parallel neutralization against Usutu virus (strain SAAR 1776), a fl avivirus found in wild birds that belongs to the same serogroup as WNV, with a panel of sera positive for WNV by micro-virus neutralization test. All titers were higher for WNV than for Usutu virus; 93.6% were >4× higher (Table 1). These results suggested that the neutralizing antibody response was generated by WNV or an antigenically related WNV-like virus.
Of 95 birds captured in 2 consecutive years, 59% had no detectable antibodies in either year, 21% seroreverted, 6.3% seroconverted, and 13.7% had antibodies in both years. Seroconversion confi rms that WNV circulation is present in the study area, and seroreversion indicates that antibody titers decreased. Antibodies persisted for >1 year in some birds, although whether this was caused by reinfection, which would stimulate the antibody response, is uncertain.
Of 54 birds captured at least twice in 2004-2005, 16.7% seroconverted (  (Figure 1), and antibody titers decreased to values near the cut-off point (Figure 2), which made changes in antibody status diffi cult to interpret. However, the most likely reason for these changes were fl uctuations in titers (from undetectable to 10 to 20) (Figure 2) because 7% of the birds showed changes in titers (from 10 to 20) at recapture.

Conclusions
We provide evidence for local circulation of WNV in our study area during 2004-2005. This evidence was obtained just a few months after a reported outbreak of WNV that affected humans in Algarve, Portugal, ≈100 km west of our study area. However, no increase in clinical signs or mortality rates was observed in the common coot  Additional studies are needed to evaluate the role of these 3 mechanisms in the dynamics of WNV in Spain. Combining serologic results for common coots and vector sampling for virus detection may provide information needed to address these issues.