West Nile Virus surveillance in mosquitoes in Emilia-Romagna (Italy)
Mattia Calzolari, Paolo Bonilauri, Francesco Defilippo1, Giulia Maioli, Romeo Bellini2, Rodolfo Veronesi, Alessandro Albieri, Paola Angelini, Ilaria Barbieri, Davide Lelli, Antonio Lavazza,
Marco Tamba, Vittorio Sambri, Michele Dottori.
In the summer 2008 a large epidemic of West Nile Fever (WN) occurred in three different Regions of Northern Italy (Emilia-Romagna, Veneto, Lombardia), causing 32 diagnosed cases in horses and 4 in humans.
An active entomological surveillance plan was started by the Emilia-Romagna Surveillance Group on Vectorial Disease in 2007. In the 2008 season a total of 78 stations were activated by CO2 baited
traps in Bologna and Ferrara provinces, 40 stations historically operating for mosquito density monitoring and 38 stations specifically positioned after the first evidence of disease in equine.
Mosquitoes were pooled according to date, location and species, grinded manually and tested with Flavivirus genus RT-PCR and with WNV Real Time PCR. In total 38791 mosquitoes were
analyzed, most of them belonging to the species Culex pipiens, Ochlerotatus caspius, Aedes albopictus and Aedes vexans. Two pools of Cx. pipiens (one collected in Cona, Ferrara province, on
September 23th, and the other collected in Argelato, Bologna province, on September 30th) resulted positive in PCR for the presence of RNA belonging to the Flavivirus genus and also for the
presence of WNV. Virus isolation was attempted starting from the two PCR positive pools by using different cells culture (Vero, Bhk21, Rk13, C6/C36) and by inoculation of SPF chicken embryonated eggs but no WNV grown was obtained.
The sequence of the amplified fragments (part of NS5 gene) obtained from of the two positive pools were identical and BLAST analysis showed a highest similarity with two isolates from the same
outbreak in Emilia-Romagna – one from magpie (Pica pica) (100% homology, FJ472945) and one from human (99% homology, FJ472946). For a more accurate molecular characterization of WNV
the complete sequence of the viral genome was required and that was possible only with an appropriate amount of viral RNA. As the isolation of virus in PCR positive pooled mosquitoes failed the determination of the whole genome sequence of the virus was precluded. Nevertheless the partial sequences obtained supported the specificity WN-PCR detections and were sufficient to
preliminarily classify WNV strains as belonging to lineage 1.
Direct detection of WN from mosquito vector is a rare event and confirms the high viral activity in the survey area. The maximum likelihood estimation (MLE) per 1000 mosquitoes obtained by
grouping weekly homogeneous samples together results 0.69 (CI 0.04-3.37) for the week of first positivity (22/09-28/09) and 1.82 (CI 0.11-8.82) for the week of the second one (29/09-04/10).