Usutu Virus Antibodies in Blood Donors and Healthy Forestry Workers in the Lombardy Region,Northern Italy

Introduction

Usutu virus (USUV) belongs to the family Flaviviridae genus Flavivirus, part of the Japanese encephalitis virus (JEV) group and related to human pathogens West Nile virus (WNV), JEV, Murray Valley encephalitis virus, and Saint Louis encephalitis virus (Calisher and Gould 2003). The virus was first isolated from Culex univittatus mosquitoes in South Africa in 1959 (Williams et al. 1964). USUV appeared for the first time in Italy, in the Tuscany region, in 1996 (Weissenböck et al. 2013) and in 2001 in Austria, where it caused an increase in mortality in blackbirds (Weissenböck et al. 2002).

The virus is transmitted by mosquitoes, mainly of the genus Culex, that represent the main vectors in Europe (Weissenböck et al. 2010). USUV rarely causes human diseases, but in 2009, in Emilia Romagna, Northern Italy, the first human cases with neurological involvement were reported in two immunocompromised patients (Cavrini et al. 2009, Pecorari et al. 2009) and recently three cases of neuroinvasive disease were reported in Croatia (Santini et al. 2015).

Since 2008, the entomological surveillance systems for WNV have been implemented in different regions in Northern Italy (Calzolari et al. 2010a) and since the two viruses (USUV and WNV), which have shown simultaneous circulation (Calzolari et al. 2010b), have similar transmission cycles with birds as main amplifying hosts and mosquitoes as vectors (Pradier et al. 2012), the WNV surveillance also detected the USUV with a prevalence of 2.6% in mosquito pools (Tamba et al. 2011).

Two consecutive studies were performed on blood donors in the Emilia Romagna region, Northern Italy, showing a prevalence increase of USUV in the population. The first study, on samples from 2008 and 2009 (Pierro et al. 2011), detected no positives, while the second, collecting samples from 2010 and 2011, showed a prevalence of antibodies against USUV of 0.23% (Pierro et al. 2013), prompting the authors to indicate a possible increased circulation of the virus in the area. A recent retrospective study conducted on 609 serum samples from patients of the Modena area (Northern Italy) showed the presence of USUV neutralizing antibodies, with a seroprevalence of 6.57%. This value is much higher than what found in previous studies conducted in the same areas, however, this value cannot be properly compared to other studies as the selected population is composed of inpatients and outpatients, and not healthy blood donors (Grottola et al. 2017).

Moreover, WNV and USUV share high similarity in the amino acid sequence of their major surface protein, the envelope protein (E-protein), which is the main target of flavivirus antibody responses (Oliphant et al. 2006, Nikolay et al. 2014) and the major neutralizing determinants (Sanchez et al. 2005). The close antigenic relationship has been confirmed by seroneutralization experiments and suggests that the viruses may interact immunologically within the host (Calisher et al. 1989). Serological assay results should thus be interpreted with care and confirmed by comparative neutralization tests using a panel of viruses known to circulate in Europe (Beck et al. 2013).

It must be noted that cross-reactivity between antibodies against USUV has been reported also in the presence of antibodies against the tick-borne encephalitis virus (TBEV) (Mansfield et al. 2011), a tick-borne virus, etiological agent of a potentially fatal encephalitis, that is widespread in Eastern Europe (Zeman 2016) and is currently spreading in Northeastern Italy (Rezza et al. 2015).

Materials and Methods

In the present case–control study, healthy forestry workers of the Park of the Ticino, in the Po river valley, Italy, highly exposed to the bite of competent mosquitoes, were monitored every 6 months for the detection of USUV-, WNV-, and TBEV-specific antibodies.

The study group consisted of 33 workers (11 females and 22 males) mean age 50 years (range 37–64) living in the city of Pavia and surrounding areas. After informed consent, three blood samples were collected: the first in October 2014, the second and the third, respectively, in June and December 2015. No worker reported any symptoms when interviewed during blood sample collection. In addition, samples collected in September 2016 from 200 donors living in the same study area were included in the study as controls. This group consists of 145 (72.5%) males, with a mean age of 43.9 years (range 21–68) and 55 (27.5%) females, with a mean age of 45.7 years (range 18–69). Thus, the ratio between cases and controls is 1:6. Laboratory strains used in this study were the USUV strain Vienna 2001-blackbird (939/01), WNV 3B2 strain, and TBE VR103457/2009.

Serum samples collected from each patient were tested by endpoint titration for USUV, WNV, and TBEV IgM and IgG with an indirect immunofluorescence assay (IFA) designed in-house. Sera were tested with a serial twofold dilution series starting from 1:10.

A neutralization assay (NTA) was performed in parallel. USUV, WNV, and TBEV were titrated in a cytopathic effect (CPE) TCID₅₀ assay using African green monkey kidney Vero cells (ATCC CRL-1586 VERO C1008). A 10-fold dilution series of the antigen was obtained and, after 5 days, the virus titer was determined using the Reed and Muench formula. For all three viruses, NTA was run starting from 1:5 with a serial twofold dilution series in microtiter plates with the addition of 50 μL TCID₅₀ of virus for each dilution. After a 1-h incubation at 37°C, 5% CO₂, 50 μL VERO cells were added to each well. CPE was read after 5 days of incubation. Serum titer was defined as the highest dilution showing a 50% CPE reduction compared to the virus control. Positive and negative control sera were included in each test. Statistical analysis of the results was performed using the chi-squared test.

Results and Discussion

Among the 33 workers tested, 6 (3 females and 3 males) resulted positive for USUV IgG and neutralizing antibodies (Table 1), setting the seroprevalence among these workers at 18.1%. No IgM antibodies were detected in any of the samples tested. Three workers were immune at the time of all three serum samplings, and in one of them an increase in neutralizing antibodies was detected, probably due to a recent infection. The other three positive workers exhibited seroconversion for IgG and NTA antibodies during the course of the study: one between October 2014 and June 2015, the other two between June and December 2015. The detection of IgM antibodies was negative, probably due to the long period between the two serum samplings.

The presence of IgG and NT antibodies against WNV was also examined in the 33 workers, revealing positivity only in 2 of the 6 USUV-positive workers, but with very low titers (Table 1). No WNV positivity was detected in the 27 workers negative for USUV, even if mosquitoes from this area have been found positive to WNV (Rovida et al. 2015). The low WNV titers, in conjunction with the high USUV titers in the same samples, lead us to classify these as cross-reactions, and not a true positivity, as previously reported (Gaibani et al. 2010). The presence of antibodies against TBE was tested as well, showing negativity for all the sera samples of forestry workers.

Two hundred samples from blood donors were tested using the same protocols for USUV, WNV, and TBEV. All donor samples resulted negative for TBEV and WNV, while two were positive for USUV IgG by IFA and NT and IgM negative. These two USUV-positive donors were retrospectively investigated and an additional serum sample, collected before the ones used for the first screening (i.e., in January 2016), was retrieved for both. These additional samples both resulted positive for IgG, but IgM negative. No history of vaccination against Yellow fever virus was reported among positive workers or blood donors.

Since 2001, USUV circulation was reported in Central Europe because of the presence of susceptible birds and competent mosquito species (Weissenböck et al. 2002). Due to the spread of WNV in summer 2008 in the Po valley, the entomological surveillance was implemented in three regions of Northern Italy, Emilia Romagna, Veneto, and Lombardy. In the context of this surveillance program, USUV circulation was documented in these areas (Busani et al. 2011).

The first human case of USUV infection was reported in 2009, the Emilia Romagna region (Italy) (Pecorari et al. 2009), in a patient affected by a diffuse large B cell lymphoma with a diagnosis of meningoencephalitis. Cerebrospinal fluid and plasma resulted positive for USUV by PCR and sequencing. In the same period, another research reported a USUV-related illness in a patient who underwent an orthotropic liver transplant (Cavrini et al. 2009). In a study conducted in 2011 (Cavrini et al. 2011) on 40 patients affected by meningoencephalitis, USUV was detected in 7.5% of cerebrospinal fluids, confirming that positivity to this virus can be correlated with neurological symptoms. Surveillance studies on healthy blood donors from the Emilia Romagna region (Northern Italy), performed in 2008–2009 (Pierro et al. 2011) and 2010–2011 (Pierro et al. 2013), show an increasing prevalence of USUV positivity, from 0% to 0.23%.

Seroprevalence values reported here were statistically compared with previous results, using the chi-squared test. The seroprevalence here reported in forestry workers (18.1%) results higher compared to that found in this study in blood donors (1.0% p < 0.001) from the same area (Lombardy region), which is in turn also significantly higher than what found in the blood donor population from the Emilia Romagna region in 2008–2009 (0% p < 0.001) (Pierro et al. 2011) and 2010–2011 (0.23% p < 0.05) (Pierro et al. 2013).

Our first result, prevalence in forestry workers, can be explained by the specific population studied, indeed the examined forestry workers are constantly exposed to mosquitoes in an area where USUV has been persistently detected at fairly constant levels in Culex pipiens mosquitoes since 2009 (ranging from 2.6% to 4.5% prevalence in pools), showing a continuous circulation (Calzolari et al. 2015). Supporting this relatively obvious explanation is the prevalence value recently reported in a population of patients in the Modena province (Northern Italy), which, at 6.57%, is higher that those of blood donors (Tamba et al. 2011). So, as expected, populations at higher risk present higher seroprevalence values.

The result obtained for blood donors (1% prevalence of IgG against USUV) is also interesting, as it is higher than what previously reported in a bordering region (Emilia Romagna). This result could indicate that the prevalence in the region examined here (Lombardy) is higher, or that the virus prevalence is growing. This second hypothesis is in accordance with the conclusions drawn by the authors of the recent studies carried out in Emilia Romagna. Additional studies, possibly performed in parallel in the two regions, could allow to better evaluate this issue.

Conclusions

The high prevalence (18.1%) of antibodies against USUV in a high-risk group in the area of the Po river valley, inhabited by millions of people, emphasizes the need for monitoring the cocirculation for both USUV and WNV in this geographical area. In parallel, the prevalence found among blood donors (1%) from the same area is also higher than expected, providing additional evidence to the hypothesis that the prevalence in this area is growing.

The results of this study corroborate the hypothesis that USUV can cause clinically asymptomatic infection among humans (Gaibani et al. 2010) and pose a possible problem for blood transfusion monitoring in areas where the virus is active. In conclusion, public health authorities, blood transfusion services, and clinicians should be aware of the growing risk of USUV infection in humans, especially during the summer, and possibly consider the screening for the presence of USUV RNA in blood donor samples of this area.