**3. Discussion**

To our knowledge, this is the first study on the prevalence of WNV antibodies in the Tuscany region. The results show that, although between 2016 and 2019 WNV prevalence in the area of Siena Province was as low as less than 1%, in 2016 and 2017 WNV was actively circulating, as shown by the finding of specific IgM suggestive of recent infection. Moreover, in some of the positive subjects the presence of antibodies with neutralizing may sugges<sup>t</sup> a potentially protective immunity.

To date, the presence of WNV in Tuscany has been reported in horses in 1998 [4], in 2009 [10], and in 2016 [11], while WNV infection in humans was retrospectively diagnosed in 2007 [6] and two cases of WNND were reported in 2017 [13].

This is the first study showing that cases of human infection by WNV have occurred in the Siena area and may be considered of some relevance as no human cases have been reported by routine surveillance in the study period 2016–2019. In addition, it appears that WNV circulation was not an occasional finding but was detected in both 2016–2017 and 2018–2019 periods studied. The reason why no cases of WNV infection were reported in the study period from the Siena area may be due to the fact that WNV is often the cause of mild or sub-clinical infection, which may lead to misdiagnosis and underreporting of the disease.

Other serological studies conducted on the general population or blood donors in other areas of Italy where the circulation of WNV has been registered (Lombardy, Emilia-Romagna, and Veneto regions) [20–24] have found similar prevalence to this study. The same median age of 57 years we found in our study was observed in WNV positive blood donors in the Veneto region [25], probably due to the fact that our study most likely includes asymptomatic WNV infections or mild symptomatic infections.

Similar WNV prevalence studies performed in other European countries and in the Mediterranean Basin detected neutralizing antibodies in 1.5% and 2.34% of the population in Greece [26] and in Hungary [27], respectively, while the results of this study are more in line with the prevalence observed in Bulgaria [28].

Historically, surveillance activities detected WNV infection cases in the Northern East areas of Italy. However, over the years, increasing numbers of West Nile fever cases have been reported from other areas, suggesting viral circulation expanding in the general population in areas previously considered naïve [3], as shown in our study. In fact, the population included in this study, although not selected for the purpose, better represents the general population than that included in other epidemiological studies performed in blood donors or in international travelers [21,22,24,29,30].

This study has some limitation. Samples were collected for purposes different from the aim of this study; thus, no information on clinical findings such as fever or neurological signs and symptoms was available. Our study population did not include subjects younger than 20 years of age.

Recently, among mosquito-borne flaviviruses, with birds as reservoir hosts, circulating in different areas of Europe, Usutu virus has been reported to possess serological cross-reactions with WNV [31]. In our study, WNV-positive samples were not tested for Usutu virus neutralizing antibodies; therefore, a possible cross-reactivity between the two viruses cannot be totally excluded. Usutu virus circulation in the Tuscany region has been reported only in mosquito pools in 2018 and 2019 from two provinces in the Northern area of the region, and no animal or human cases were reported from routine surveillance activities [16,32]. Moreover, IgM ELISA antibodies have been detected in some samples of this study. IgM ELISA is usually considered to be more specific than IgG, with a lower cross-reactivity with other flaviviruses [33,34]. Taking into account the data on the circulation of both viruses in the area and the results obtained from all the serological assays performed in this study, the specific reaction to WNV can be reasonably assumed.

This study shows for the first time the active circulation in humans of WNV that occurred between 2016 and 2019 in the Siena area, an area considered not at high risk until 2019. Although the prevalence of WNV is limited as compared to other neurotropic arboviruses, such as Toscana virus [35], it appears to have acquired an established transmission pattern between 2016 and 2019.

In conclusion, WNV infection appears to be more widespread in the area of Siena than has been detected so far, and it is possible that some cases of infection are underdiagnosed and underreported. Taking into consideration the trend of the expansion of WNV in Central Italy, the absence of reported WNV human cases in the Siena area should not limit the application of preventive measures and epidemiological surveillance, as the low prevalence of antibodies does not prevent outbreaks of WNV disease in the future.

### **4. Materials and Methods**

### *4.1. Study Population*

The study was performed with samples available at the sera bank of the Molecular Epidemiology Laboratory of the University of Siena, Italy. Human serum samples are residual samples collected from a local laboratory in the province of Siena between 2016 and 2019. Samples were anonymously collected and stored in compliance with Italian ethics law. For each serum sample, information only on age, sex, place and year of sampling was available.

A total of 1800 samples were randomly selected from the sera bank: 879 for the years 2016–2017 and 921 for the years 2018–2019.

### *4.2. ELISA and Immunofluorescence Assay*

All samples were tested for the presence of IgG antibodies against WNV by use of "West Nile Virus IgG" (DIA.PRO, Milano, Italy) commercial ELISA kit. Testing was performed according to manufacturer's instructions, and test results were calculated by means of a cut-off value determined with the following formula: Cut-off = optical density (OD) of the negative control + 0.250. Samples were considered positive when the ratio

between the OD of the sample and that of the cut-off was >1.1, and negative when the ratio between the OD of the sample and that of the cut-off was <0.9. Samples with a ratio between 0.9 and 1.1 were considered borderline.

Out of 879 samples collected in 2016–2017, 92 were also tested for the presence of IgM antibodies by use of "West Nile Virus IgM" (DIA.PRO, Milano, Italy) commercial ELISA kit. IgM ELISA testing was performed on all the ELISA IgG positive and borderline samples and on a subset of ELISA IgG negative samples. Testing was performed according to manufacturer's instructions and results were calculated as for ELISA IgG kit.

ELISA borderline and positive samples were further tested by "Anti-West Nile virus (IgG)" (EUROIMMUN, Lübeck, Germany) IFA commercial kit, following manufacturer's instructions. Samples were tested with 2-fold dilutions from 1:50 to 1:6400. The IFA titer was defined as the highest serum dilution showing fluorescence, as reported by manufacturer's instructions.

All IgG and IgM ELISA and IgG IFA positive samples were further tested by MN and PRN assays.

### *4.3. Micro Neutralization and Plaque Reduction Neutralization Assays*

The cell substrate used was Vero E6 (African green monkey kidney cell line; ATCC® CRL-1586™) propagated in Dulbecco's Modified Eagle's Medium (DMEM; Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% Fetal Bovine Serum (FBS; Sigma-Aldrich, St. Louis, MO, USA). The WNV strain (lineage 2) viral stock, consisting of cell-free supernatants of acutely infected Vero E6 cells, was stored at −80 ◦C until use. Prior to the MN and PRN test, WNV was titrated for 50% tissue culture infectious dose (TCID50) and plaque forming unit (PFU) using Vero E6 cells, and all serum samples were heat-inactivated at 56 ◦C for 30 min.

MN assay was performed by exposing (1:1) serial twofold dilutions of heat-inactivated serum in DMEM (1:10 to 1:320) to 100 TCID50 of WNV. After 1-h incubation at 37 ◦C in 5% CO2 atmosphere, 50 μL of the serum/virus mixture was plated on each well of a 96-well plate covered by Vero E6 cell monolayers (10<sup>4</sup> cells/well), and incubated for 1 h at 37 ◦C, 5% CO2. Then, 50 μL of DMEM was added on each well and the plate was incubated for 4 days up to the appearance of an easy detectable cytopathic effect in control cultures (cell monolayers exposed to WNV). Additionally, IgG serum negative to WNV was used as control. The antibody titer was defined as the reciprocal of the highest dilution of the test serum sample, which showed at least 50% neutralization.

PRN assay was performed on heat-inactivated serum samples by exposing (1:1) serial twofold dilutions of them in DMEM (1:10 to 1:320) to 100 PFU of WNV. After incubation for 1 h at 37 ◦C, 5% CO2 atmosphere, 300 μL of the serum/virus mixture was plated on each well of 6-well plates seeded with 2.5 × 10<sup>5</sup> Vero E6 cells and incubated 1 h at 37 ◦C. Then, the overlay medium composed of 0.5% Sea Plaque Agarose (Lonza, Basel, Switzerland) diluted in propagation medium was added to each well. After 4 days of incubation at 37 ◦C, the monolayers were fixed with methanol (Carlo Erba Chemicals, Milan, Italy) and stained with 0.1% crystal violet (Carlo Erba Chemicals, Milan, Italy) and the viral titers were calculated by PFU counting. Percent of PRN was calculated by dividing the average PFU of viral serum treated samples by the average of viral positive control. All experiments were repeated at least twice. All experimental procedures were conducted under biosafety level 3 containment.

### *4.4. Statistical Analysis*

Categorical dichotomous data (sex and age group) and discrete data (IgG ELISA, IFA, MN/PRN assays results) were defined as categorical dichotomous data, described as counts and percentages and evaluated by Chi-square test. The relations between the IgG positivity of each assay as a dependent categorical dichotomous variable defined as a dummy variable and independent factors (sex and age group) were evaluated by logistic regression model, and OR, 95% CI, and *p*-values were assessed. In the univariate

logistic regression model, all the factors related to IgG positivity were investigated as independent variables. The statistically significant independent variables were assessed in the multivariate logistic regression model using Wald test and stepwise method for the selection of *p*-value. Statistical significance was set at *p* < 0.05.

Data from statistical analyses were performed with the software GraphPad Prism v.6.0.0 (GraphPad Software, San Diego, CA, USA).

### *4.5. Geographic Methods*

The spatial distribution of WND human and equine reported cases was mapped using QGIS 3.6.0 [36]. The shapefile of Tuscany region (WGS84 UTM32N) was retrieved from the National Institute of Statistics (ISTAT) [37]. The national geographic map was used as basemap to relate the study area to the European region.

**Author Contributions:** Conceptualization, S.M. and S.V.; methodology, S.M., S.G., M.A.S.; formal analysis, S.M., G.L., M.C.; investigation, S.M., S.G., M.A.S., C.A., R.C.; writing—original draft preparation, S.M.; writing—review and editing, E.M., S.V., S.G., M.A.S., G.L., M.C., C.M.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.
