The area of study included 2 wetlands: Marismas del Odiel (tidal marshes) and Doñana (freshwater marshes), both located in southwestern Spain. Mosquitoes were captured in 2006 with U.S. Centers for Disease Control and Prevention light traps supplied with CO₂ and with gravid traps, which were used in the field during the late afternoon and retrieved the following morning. Mosquitoes were pooled by species, sex, collecting site, and date. The number of mosquitoes per pool ranged from 1 to 50. Mosquitoes/pools were homogenized in a range of 500–700 μL of minimal essential medium supplemented with 200 U/mL of antimicrobial drugs (penicillin/streptomycin) and 10% of fetal bovine serum and then were stored at –80°C until they were tested for flavivirus. The homogenate was centrifuged at 13,000 rpm for 5 min at 4°C, and the screening was performed with a generic nested reverse transcription PCR (15) to detect flavivirus genome.

This study comprised 35,424 mosquito specimens grouped in 1,641 pools and representing 14 species (Table 1). Approximately 11% of the pools (191) showed a positive result for flavivirus amplification. However, WNV was identified in only 1 (pool HU2925/06). The pool contained 50 unfed Cx. pipiens complex females, captured in June 2006 in Palos de la Frontera (Huelva; latitude 37°12′41.76′′N; longitude 6°55′9.58′′W).

Figure

Figure. Phylogenetic tree of 79 WNV isolates by the neighbor-joining method and distance-p model on MEGA3.1 (www.megasoftware.net/mega_dos.html). Bootstrap values correspond to 1,000 replications. A) Analysis of a 1,813-nt fragment of the...

A fragment of 1,813 nucleotides (nt) from the nonstructural protein 5 (NS5) gene from this WNV genome (GenBank accession no. GU047875) was amplified by using 3 WNV-specific nested-PCRs designed in this study. The phylogenetic analysis resulted in a tree in which, as expected, this sequence fell under the branch of WNV, with a value of certainty of 100% (Figure, panel A). A common evolutionary branch between the Spanish strain and lineage 4 (99% certainty) can be observed, and both strains seem related to lineage 3. Sequence differences observed between HU2925/06 and other strains of WNV are shown in Table 2. The minor genetic distance was obtained for lineage 4 and the highest for lineage 5. To confirm that the sequence detected in Spain did not correspond to those of the isolates KUN MP502–66 and KOUV, we analyzed part of the genome sequence of both viruses, and partial sequences showed that these viruses cluster into a distinct genetic lineage (Figure, panel B). The sequence data for KOUV was retrieved from GenBank (strain Koutango DakArD1470, accession no. AF013384), and the partial sequence for KUN MP502–66 was obtained in this work amplifying part of the NS5 gene (GenBank accession no. GU047874).

To isolate the virus, the positive pool was diluted 1:20 in the minimal essential medium, and 200 µL were injected onto C6/36 (Aedes albopictus cells), RK-13 (rabbit kidney cells), and Vero (African green monkey cells) monolayer cells grown at a constant temperature for each cell line (33°C, 37°C, and 37°C, respectively). Cell cultures were incubated under the same conditions for 7 days, and 3 blind passages were carried out. Signs of cytopathic effect were checked daily, and the culture supernatants were tested by reverse transcription–PCR. Neither cytopathic effect nor amplification was obtained. No virus was isolated from any of the 3 cell cultures.

The phylogenetic analysis performed on a 1,813-nt fragment of the NS5 gene clearly shows that the sequence recovered in Spain grouped within the branch of WNV with high values of certainty (100%). The tree topology shows a common evolutionary branch between the Spanish WNV genome (HU2925/06) and lineage 4, which clusters close to lineage 3. The lineages 3 and 4 were detected recently in Europe (1997 and 1998, respectively), and they have not been previously associated with natural disease in vertebrates. In addition, the phylogenetic analysis performed on 800 nt fragments of the NS5 gene indicated that the Spanish strain was not the same that KUN MP502–66 and KOUV, and that KUN MP502–66 seems to be a different lineage.

This report and the recent description of WNV lineage 1 in wild birds (14) demonstrate the circulation of both WNV lineages in Spain. This finding should lead to the analysis of serologic evidence of WNV infections in birds, horses, and humans in Spain and surrounding countries, where the highly pathogenic WNV strains sporadically cause clinical infections. An explanation for the high WNV seroprevalence levels found in birds, horses, and humans in the absence of neurologic disease in Spain could be that this new lineage infects birds and protects them from most pathogenic strains of WNV.