Identification of Usutu Virus Africa 3 Lineage in a Survey of Mosquitoes and Birds from Urban Areas of Western Spain

Usutu virus (USUV) is an emerging zoonotic arbovirus that has caused an increasing number of animal and human cases in Europe in recent years. Understanding the vector species and avian hosts involved in the USUV enzootic cycle in an area of active circulation is vital to anticipate potential outbreaks. Mosquitoes were captured in 2020, while wild birds were sampled in both 2020 and 2021 in Extremadura, southwestern Spain. The presence of USUV in the mosquito vectors was assessed by a real-time reverse transcription-polymerase chain reaction (rRT-PCR) assay and confirmed by sequencing amplicons from two generic RT-PCR sets for flaviviruses. Sequences were analysed phylogenetically. Bird sera were screened for flavivirus antibodies with a blocking ELISA kit and subsequently tested for virus-specific antibodies with a micro-virus-neutralization test. Overall, 6,004 mosquitoes belonging to 13 species were captured, including some well-known flavivirus vectors (Culex pipiens, Cx. perexiguus, and Cx. univittatus). Of the 438 pools tested, USUV was detected in two pools of Cx. pipiens. Phylogenetic analysis using a fragment of the NS5 gene assigned the USUV detected the Africa 3 lineage. Out of 1,413 wild birds tested, USUV-specific antibodies were detected in 17 birds (1.2%, 10 males and 7 females) from eight species. The first detection of USUV Africa 3 lineage in mosquitoes from Spain, together with serologically positive resident wild birds in urban and rural areas, indicates active circulation and a possible risk of exposure for the human population, with necessity to establish specific surveillance plans.


Introduction
Usutu virus (USUV) is a mosquito-borne favivirus (family Flaviviridae), closely related antigenically and genetically to West Nile virus (WNV) and belonging to the Japanese encephalitis serocomplex.In nature, USUV is maintained in a bird-mosquito-bird enzootic cycle [1].However, the virus has also been detected sporadically in humans, horses, and other mammals, acting as incidental hosts (dead-end hosts, i.e., they can be infected but do not transmit the infection), and in a limited number of cases, they can sufer neuroinvasive disease and USUV fever [2,3].
Te frst detection of USUV was in Culex neavei mosquitoes from South Africa in 1959, in a close area to the Usutu River in Swaziland [4].Afterwards, the virus was retrospectively detected in archived tissue samples from birds from Italy collected in 1996 [5].In 2001, USUV was identifed as the causative agent of massive bird mortality in Austria [6].Since then, USUV has been detected in several European countries in mosquitoes, birds, various mammals (horses, bats, ruminants, dogs, and wild boars), and humans [1,3,[7][8][9].
In recent years, the understanding of the ecology, epidemiology, and distribution of USUV has improved due to an increase in its detection as a result of enhanced favivirus surveillance and research activities in Europe [3,10].So far, eight lineages of USUV have been described, fve European lineages (Europe 1-5) and three African lineages (Africa 1-3) [11].Te great diversity of lineages circulating in Europe suggests diferent introductions from Africa, as well as a continuous geographical spreading across the continent and colonization of new ecological niches [10,12].Te rapid European expansion of the virus, together with several reports of infection cases or neuroinvasive diseases in humans, has confrmed USUV as an emerging zoonotic virus of public health signifcance [1,3].
In Spain, USUV-specifc antibodies have been detected from diferent species of wild birds, horses, and red deer [13].In mosquitoes, USUV has been identifed in Culex pipiens in 2006 (northeastern Spain) and in Culex perexiguus in 2009 (southern Spain), both belonging to the USUV Africa 2 lineage [14,15].Another USUV sequence was also obtained in 2012 from a song thrush (Turdus philomelos) related to a Senegalese and Central European sequences, although without a clear classifcation [16,17].In the region of this study, Extremadura (southwestern Spain), USUV antibodies were detected for the frst time in 2017-2019 in horses and wild birds [18][19][20], but evidence from mosquitoes or humans has not been reported yet.
To date, feld studies focusing on USUV enzootic cycle are scarce.USUV shares important features with WNV, including a degree of overlapping ranges of vector species and avian hosts, as well as some cross-reactivity in diagnostic methods.Hence, combined surveillance and control programs of both viruses ofer important benefts, contributing to improve the understanding of their epidemiology and the potential interactions which may afect the transmission of both pathogens under natural conditions.Here, we report the concomitant detection of the USUV genome in mosquitoes and USUV antibodies in avian hosts, confrming the presence of an enzootic cycle in an area close to urban centers from southwestern Spain.(Figure 1).Badajoz is the largest urban nucleus (152,764 inhabitants) of the Autonomous Community of Extremadura.Most of the sampling sites were located close to the border with Portugal and the Guadiana River.Extremadura region has a Mediterranean climate classifed as Csa (hot dry summer) according to the Köppen climate classifcation.

Sampling of Mosquitoes and Wild Birds.
Mosquitoes were collected from May to November 2020.Entomological sampling was carried out every 40 days, and each trap was set from sunset to early morning, during a minimum of 12 hours.Te combination of diferent traps (i.e., BG-Sentinel traps, CDC miniature light-traps, both containing CO 2 as bait, and gravid traps baited with hay infusion) ensured a more complete sampling of the vector community present in the area, with a greater diversity of species and catch density [21].On the following morning, mosquitoes were frozen in the feld with dry ice and transported to the laboratory for their morphological identifcation and molecular analysis.
Morphological identifcations at species level were based on broadly used identifcation keys [22].Te whole process was carried out under binocular magnifying glass with frozen plates to ensure the maintenance of the cold chain.Mosquitoes were grouped by species, collection site, and dates in pools of 1-25 individuals and conserved in the MEM culture medium.Males were equally identifed and classifed to be considered in abundance estimates but not analysed for USUV detection.
Molecular analyses were performed to confrm the identifcation of epidemiologically important mosquito species that are difcult to identify morphologically.For this purpose, genomic DNA (gDNA) was extracted from the legs of individuals using GeneJET ™ Genomic DNA Purifcation Kit (Termo Scientifc Inc., reference #K0722) following the manufacturer's instructions.Te cytochrome oxidase 1 (cox1) gene was partially amplifed using the primer set LCO1490 and HCO2198 following the PCR protocol described previously [23].Wild birds were captured using mist-nets from February to December 2020 and from March to May 2021.Each bird was ringed with a numbered metal ring, and its age and sex were determined when possible, according to their plumage characteristics and skull ossifcation [24].For each individual, a blood sample was extracted from the jugular vein using sterile insulin syringes.Te volume of extracted blood varied depending on the body size of each bird and never exceeded 1% of their body mass.Blood samples were transferred to sterile Eppendorf tubes and preserved in cold boxes during the feld work.In the laboratory, the samples were kept at 4 °C and centrifuged within 24 h after sampling for 10 min at 11,000 rpm (11.2 g) to separate serum and cellular fractions, which were frozen at −80 or −20 °C, respectively.Birds were immediately released unharmed at the site of capture after manipulation.

Detection of the Flavivirus Genome in Mosquitoes and
Phylogenetic Analyses.Viral RNA was extracted from mosquito pools using a MagMax ™ Pathogen RNA/DNA kit (TermoFisher ® ), according to the manufacturer's protocol (Pub.no.4463379).At least one negative control and a triplex positive extraction control (for WNV-L1, WNV-L2, and USUV) from inactivated viral cultures were included in each nucleic acid extraction run.Before extraction, 500 μl of MEM, 1x prepared with antibiotics (penicillin and streptomycin), L-Glutamine, and inactivated foetal bovine serum, was added to each tube and mosquito pools were crushed.A volume of 200 μl of this mix was used for RNA extraction.For favivirus detection, a triplex real-time reverse transcription-polymerase chain reaction (rRT-PCR), which simultaneously detects and diferentiates WNV-L1, WNV-L2, and USUV, was employed [25].For the characterization of USUV-positive pools, we used two overlapping generic RT-nested-PCR to detect favivirus genome on the NS5 gene region [26,27].Te amplifed products were visualized by electrophoresis on a 1.5% agarose gel and were purifed using ExoSAP-IT kit (GE Healthcare).Te purifed DNA was Sanger-sequenced in duplicates in both directions using the same primer sets of the RT-nested-PCR assays.Analysis of the sequences and assembly was performed by using SeqMan software (DNASTAR LASERGENE Software).Te genome sequence obtained in this study was submitted to GenBank database (accession no.ON838179).Multiple alignments were performed using ClustalW program, and the best-ftting evolutionary model was based on those defned using JModeltest2 [28] on the basis of the Akaike information criterion.Tree reconstruction was carried out with Mega 11 [29].Two phylogenetic analyses were performed on the basis of 1030 and 238-nucleotide fragment of the NS5 gene using the maximum likelihood (ML) method, Kimura 2-parameter model, and the general time reversible model [30].Initial trees for the heuristic search were obtained automatically by applying the Neighbor-Join and BioNJ algorithms to a pairwise distances matrix estimated using the maximum composite likelihood approach and then selecting the topology with a higher log likelihood value.All positions containing gaps and missing data were eliminated.Bootstrap coefcients were calculated for 1000 replicates.Although this ELISA kit is highly sensitive for WNV, crossreacting antibodies to other closely related faviviruses such as USUV may also react [18,31].Tus, in order to specifcally detect USUV-neutralizing antibodies and diferentiate from other cross-reacting faviviruses, ELISA-positive and doubtful sera were subsequently analysed with a micro-virus-neutralization test (micro-VNT) in 96-well microtitre plates as described in [31].Ten bird serum samples lacked enough volume so that they could not be analysed by micro-VNT.

Detection of Flavivirus
Micro-VNTs were performed in parallel against USUV, WNV, and Bagaza virus (BAGV) as described in [31].Viral strains used in this assay were USUV SAAR-1776 (accession no.AY453412), WNV E101 (accession no.AF260968), and BAGV Spain/RLP-Hcc1/2010 (accession no.KR108244).Samples yielding neutralization (complete absence of CPE) at 1 : 10 or higher was scored as positives.To confrm antibodies as specifc for a determined virus, antibody titre should be at least fourfold higher for a given favivirus over the rest [32].If that threshold was not reached, the specifc favivirus that caused the infection could not be determined.Sera were considered USUV-positive only if the positive/ doubtful result by ELISA was confrmed by VNT.In this study, we focused on the USUV results, as the WNV outcomes are part of a diferent survey (unpublished data).

Statistical Analysis.
Te estimated prevalence of USUV in mosquitoes and the corresponding 95% CI from variable pool size and perfect tests were calculated with the EpiTools epidemiological calculator method (https://epitools.ausvet.com.au/ppvariablepoolsize). Tis method estimates prevalence and confdence limits for variable pool sizes and assumes 100% test sensitivity and specifcity [33].
Te prevalence of favivirus antibodies detected by ELISA and USUV-specifc antibodies determined by micro-VNT were estimated as the ratio of positives from the total number of samples analysed, with the exact binomial confdence intervals (CI) of 95% based on the score method [34].Tis prevalence should be considered as underestimated, or at least as the minimum detectable prevalence, taking into account that the potential cross-reactions detected with the ELISA test could not detect all sera with USUV antibodies.Descriptive statistical analyses were conducted in IBM SPSS Statistics for Windows, Version 26.0 (Armonk, NY: IBM Corp.).All the female mosquitoes were grouped in 438 pools.Of these, two pools of Cx. pipiens mosquitoes were positive for USUV by rRT-PCR, one captured on 16th August and one on 25th October in the same sampling site (B) (Figure 1).Te vector infection rate for USUV in Cx. pipiens species was 4e-04% (95% CI: 1e −04 -0.0014).A fragment of 1,030 nucleotides of the USUV NS5 gene was obtained from one Cx.pipiens pool (accession no.ON838179).Te phylogenetic analysis showed the obtained sequence clusters with other sequences belonging to the USUV Africa 3 lineage, previously detected in Africa and several European countries (Figure 2).Indeed, the sequence obtained in Cx. pipiens mosquitoes from Extremadura difers from previous Spanish sequences obtained from mosquitoes captured in Catalonia and Andalusia (in 2006 and 2009 from Cx. pipiens and Cx.perexiguus, respectively).Besides, our sequence even difers in the analysis of the 238-nucleotides fragment of the overlapping sequence obtained from Spanish song thrushes in 2012 (accession no.KC437386) (S1).
Most of the positive birds were adults, except for four juvenile females captured in 2020: two Azure-winged magpies (Cyanopica cyanus), one great tit (Parus major), and one red avadavat (Amandava amandava) (Table 1).Seropositive birds were detected in all sampling areas (Figure 1).

Discussion. Te detection of USUV in mosquitoes and
wild birds from urban and rural areas has proven to be efective both to assess the epidemiological situation after outbreaks and to forecast possible future risks [35][36][37].Te survey carried out in this study with the analysis of mosquitoes and birds confrmed active USUV circulation in areas close to urban centers from southwestern Spain, providing new evidence that adds up to previous studies that had detected WNV and USUV antibodies in horses and birds in the same area in 2017-2019 [19,20].
Mosquito pools positive for the presence of USUV belong to Cx. pipiens species, and they were captured in late summer-early autumn at the sampling site B (Figure 1).In other European countries, USUV-positive mosquitoes have usually been detected between June and October [12,15].
Diferent mosquitoes, mainly ornithophilic species of the Culex genus, are known to participate in the transmission of USUV to wild or captive avifauna [1].Cx. pipiens is a proven vector for USUV, and this virus has already been detected in specimens of this mosquito species in Europe [1], including Spain [14].Culex pipiens was the most abundant mosquito species captured in this study, mainly in urbanized areas, the mosquito species with the greatest distribution in Extremadura [38].It has been suggested to play a key role in the epizootic transmission of pathogens to humans [39].So, it should be taken into account in control programs due to its potential for transmitting zoonotic faviviruses in areas close to urban environments.In Spain, USUV sequences of Africa 2 lineage have been detected in mosquitoes from Catalonia (Cx.pipiens, 2006) and Andalusia (Cx.perexiguus, 2009) [14,15] and other unclassifed USUV lineage in T. philomelos in 2012 [17].Our study represents the frst clear identifcation of USUV Africa 3 lineage in Spain.USUV strains of Africa 3 lineages were previously detected in Europe in Cx. pipiens mosquitoes in southern France [7,12] and in common blackbirds (Turdus merula) from Germany (2014), Belgium (2016) [11], France (2018) [12], Austria (2017) [40], Czech Republic (2018) [35], the Netherlands [41,42], the United Kingdom (2020) [43], and Luxembourg in 2020 [44].
In November 2012, USUV was detected in a song thrush (Turdus philomelos) from a die-of of ≈10 birds on a hunting estate in southern Spain [17].Here, we found six seropositive common blackbirds.Considering that common blackbirds showed high morbidity and mortality rates due to the USUV infection in diferent European countries [45], possibly Turdus sp. may represent a key host in the USUV epidemiology.However, the high number of seropositive individuals from this species may raise other questions: has there been a trade-of between the adaptation of the virus to this species and its increased survival of the disease?Or is the Africa 3 lineage less pathogenic to blackbirds than other lineages circulating in Europe?Accordingly, the histological lesion severity in common blackbirds for the two identifed lineages (Europe 3 and Africa 3) was compared, showing no signifcant diferences [41].Tus, it is likely that both lineages may produce a similar pathogenetic efect in naturally infected animals.Spain is the southernmost country in Europe where USUV has been detected, and the sequence obtained is closely related to another sequence from France detected in 2018 [12].As both sequences difer from other variants circulating in northern countries, this clade may represent an independent introduction event followed by geographical spread.As new sequences are continuously published, future analyses will allow us to better understand the spread dynamics of the USUV Africa 3 lineage in Europe and further decipher its evolutionary history.
Outside Europe, the Africa 3 lineage of USUV has been detected in mosquito species of the Univittatus subgroup (Cx.perexiguus, Cx. univittatus, and Cx.neavei) from Senegal [46], Uganda [47], and Israel [48], in Cx. perfuscus from Senegal and the Central African Republic [46], and in Ae. albopictus from Israel (unpublished, GenBank: MG461308.1).Te presence of Cx. perexiguus and Cx.univittatus, important vectors of WNV and USUV, has already been reported in Extremadura [18].Here, we confrmed the presence of both species in the same areas, which may imply a higher risk of favivirus transmission considering that WNV lineage 1 was detected in Cx. univittatus in Portugal [49], and WNV lineage 1 and USUV Africa 2 lineage were detected in Cx. pipiens and Cx.perexiguus species in Spain, respectively [15,50].Te difcult morphological diferentiation of adult specimens of the Univittatus subgroup, together with their possible presence in the same habitats requires special care in their identifcation during entomological surveys, at least in the south of the Iberian Peninsula.
USUV antibodies were detected in birds from all sampling sites, including locality A, which is 41 km away from the city of Badajoz.In previous transmission seasons, USUV-specifc antibodies were found in birds and horses in several areas from the Extremadura region [18][19][20].Te data obtained in this study confrm the occurrence of local transmission in 2020, with probable overwintering from past transmission seasons after its introduction in the area, indicating the establishment of an USUV enzootic cycle in southwestern Spain.In Europe, evidence of USUV infection has been found in at least 93 bird species from 35 families [1].With this study, we enlarge the list of USUV hosts, including woodchat shrike (Lanius senator), common nightingale (Luscinia megarhynchos), and red avadavat (Amandava amandava).In the case of the woodchat shrike and common nightingale, both species breed in the Iberian Peninsula but are trans-Saharan migrants [51,52].Te red avadavat is a bird indigenous to Asia which was introduced in the south of the Iberian Peninsula and Extremadura more than 40 years ago [53].
Overall, more than 75% of the positive birds of this study are resident species (and some juvenile specimens) with limited geographical mobility, which strongly indicates endemicity and establishment of the USUV in the region, considering that specifc antibodies against this virus have been detected in diferent species since 2017 [18,20].Furthermore, the observed USUV seroprevalence in the area (1.2%) is similar to previous studies in wild birds in nearby sampling areas (0.96%) [20] and in birds from rehabilitation centres (1.4%) [18].
Nevertheless, we are aware of the methodological limitations of the present study and these fndings should be interpreted with caution.Te ELISA kit employed for screening was developed specifcally for the detection of antibodies against WNV.Nevertheless, and in the absence of a commercial ELISA kit for specifc USUV detection, its cross-reactivity with antibodies directed to other faviviruses led to think it may be useful in USUV surveillance studies [31], but with unknown and possibly underestimated efcacy.
USUV and WNV share environmental and ecological drivers.Tus, cocirculation is often observed in nature [9,19,54,55].Indeed, both pathogens circulate within the same vectors in Spain (i.e., Cx. pipiens and Cx.perexiguus) [14,15].Consequently, USUV should be included, together with WNV, in surveillance programs and diferential diagnoses in humans and animals.Tis recommendation is supported owing to the fact that 16 of 17 positive samples were obtained from birds captured at less than 5 km of the city centre and that 12 of these samples correspond to resident birds.Although some USUV acute neurological human cases have been reported (e.g., encephalitis or meningoencephalitis) [2], in most cases, USUV (and WNV) infections are asymptomatic or cause mild clinical symptoms [1].Terefore, the actual incidence of USUV infection in humans in the region is essentially unknown, and thus, its epidemiological status and actual burden in public health is likely underestimated.Further epidemiological studies about USUV and other faviviruses are warranted, especially considering the current scenario of increasing urbanisation, where human activities, such as animal husbandry and intensive agriculture, greatly infuence the dynamics of vector-borne diseases through their impact on the distribution of hosts [56] and mosquito species [57][58][59].

Conclusions
In summary, this is the frst identifcation USUV Africa 3 lineage in Spain.Te detection of USUV in mosquitoes by RT-PCR and USUV-specifc antibodies in wild birds (especially in juveniles) is indicative of an active circulation of the virus and the possible establishment of its enzootic cycle in southwestern Spain.Te circulation of USUV close to urban areas represents a public health threat that demands its inclusion in the diferential diagnoses in patients with compatible symptoms.Terefore, it is highly advisable to establish integrated bird and mosquito survey programs in the region, targeting urban and rural areas with the aim to collect relevant data on the epidemiological scenario represented by emerging faviviruses, including USUV, and its Transboundary and Emerging Diseases evolution and spread, which may reveal new foci, seasonality, and ecological niches of USUV.

Figure 2 :
Figure 2: Phylogenetic analysis of a 1,030-nucleotide fragment of the NS5 gene using the maximum likelihood method and Kimura 2parameter model.Tis analysis involved 52 nucleotide sequences and a total of 1,028 positions in the fnal dataset.Sequence obtained in this study is highlighted in red (accession no.ON838179), and the Spanish USUV sequence obtained from mosquitoes is marked with a black triangle.Te percentage of replicate trees in which the associated taxa are clustered together in the bootstrap test (1000 replicates) is shown next to the branches.Only bootstrap >75% is shown.Taxon information indicated in the branches includes the country of origin, isolation/ detection year, host, and GenBank accession number.USUV genetic sublineages are indicated on the right.
Antibodies in Wild Birds.Bird sera were screened to detect favivirus antibodies with the blocking ELISA kit INGEZIM West Nile COMPAC (INgenasa, Spain), following the manufacturer's protocol.

Table 1 :
Results obtained for positive/doubtful samples by ELISA and with Usutu virus-specifc antibodies by micro-VNT in wild birds from Extremadura, Spain.