Serological Survey of Lyssaviruses in Polish Bats in the Frame of Passive Rabies Surveillance Using an Enzyme-Linked Immunosorbent Assay

Background: Bats are known to host a number of nonpathogenic viruses, as well as highly pathogenic viruses causing fatal diseases like rabies. Serological surveys as part of active and passive bat rabies surveillance mainly use seroneutralization assays, demonstrating the presence of lyssavirus-specific antibodies in a variety of European bats, particularly against European bat lyssaviruses type 1 (EBLV-1). Here, we present the first serological survey in European bats of this kind during which European bats from Poland collected in the frame of passive rabies surveillance between 2012 and 2018, as well as Serotine bats (Eptesicus serotinus) and North American Big Brown bats (Eptesicus fuscus) from previous experimental studies, were tested using a commercial ELISA kit for the detection of anti-lyssavirus antibodies. Results: Lyssavirus-specific antibodies were detected in 35 (30.4%) out of 115 Polish bats of both sexes, representing nine out of 13 identified bat species endemic mainly to Central Southern Europe and Western Asia, i.e., Eptesicus serotinus, Nyctalus noctula, Myotis daubentonii, Plecotus auritus, Vespertillo murinus,Pipistrellus pipistrellus, Pipistrellus pipilstrellus/Pipistrellus pygmaeus, Myotis brandtii, and Barbastella barbastellus. Seroprevalence was highest in bat species of Nyctalus noctula, Eptesicus serotinus, Plecotus auritus, and Myotis daubentonii. More than 60% of the ELISA seropositive bats originated from the voivodeships of Silesia, Lower-Silesian, Warmian-Mazurian, and Mazowian. Rabies-specific antibodies were also found in Eptesicus fuscus bats from North America. Conclusions: The study demonstrates the principal application of the BioPro Rabies ELISA Ab Kit for the detection of anti-lyssavirus specific antibodies in body fluids and serum samples of bats. However, results may only be reliable for North American bats, whereas interpretation of results for European bats per se is difficult because proper validation of the test is hampered by the protected status of these species.


Introduction
Bats (Chiroptera) are reservoirs of a large number of viruses, some of which can cause infections and diseases in animals and humans [1]. One example is rabies, an acute viral infection of the central nervous system caused by viruses belonging to the Mononegavirales order, Rhabdoviridae family, and Lyssavirus genus [2]. Based on demarcation criteria including antigenic and genetic distances, 16 lyssavirus species are accepted by the International Committee on the Taxonomy of In this study, 115 body fluids from Polish bats found dead and collected in the frame of passive rabies surveillance were included. The sampling was performed between 2012 and 2018 in different voivodships of Poland. Except for two bats, which tested positive by the Fluorescent Antibody Test (FAT) for detection of lyssavirus antigen [35], all other bats tested negative when examined by Polish regional veterinary laboratories. Bat carcasses were then submitted to the National Veterinary Research Institute (NVRI) for further research including molecular identification of bat species and detection of anti-lyssavirus antibodies. Some of the bat carcasses were kept frozen for several years before testing. After thawing of bat carcasses, body fluids were sampled during necropsy, i.e., upon availability, blood was sampled from the heart chambers and body fluids were taken from the thoracic cavity. Subsequently, serum was obtained after centrifugation of blood at 2500 rpm, at 4 • C for 10 min. All samples were kept frozen at −20 • C until ELISA examination.
For the two bats that tested positive in FAT, EBLV-1 was detected using a heminested RT-PCR method as described elsewhere [36]. Next, lyssavirus typing was performed based on the nucleotide sequence analysis of N gene fragment and multiple sequence alignment to the reference isolates. For each bat specimen, the sex was estimated, and body fluids or serum samples were taken from all bat individuals during bat necropsy.
Additionally, 87 bat sera were obtained from previous experimental studies in Eptesicus serotinus (n = 27) [18] and North American big brown bats (Eptesicus fuscus, n = 60) [19]. All blood samples were taken from the vena uropatagialis prior to infection, thus resembling active surveillance samples. Of note, for Eptesicus fuscus, animals were bled twice within an interval of 105 days.

Molecular Classification of Polish Bats Based on Genetic Markers
Molecular identification of bat species was performed based on cytochrome b sequence analyses as described before [37]. Briefly, around 25-mm 2 fragments of the wing membrane were sampled from all bat specimens. Total DNA was isolated from wing samples using DNA Mini Kit (Qiagen) after overnight lysis using proteinase K. The lysate was loaded on DNA spin columns following the manufacturer's instructions, and the resulting DNA was used for amplification of the cytochrome b fragment as published before [37]. Amplicons were subjected to Sanger sequencing in both directions on the automated sequencer ABI PRISM 310 Genetic Analyzer (Applied Biosystem/Thermo Fisher Scientific, Waltham, MA, USA) using a BigDye Sequencing Kit (Applied Biosystem/Thermo Fisher Scientific, Waltham, MA, USA) with GeneScan Analysis Software and the same primers as used for PCR. Bat classification was performed based on sequence mapping with reference sequences using the Basic Local Alignment Search Tool (BLAST) algorithm.

Detection of Lyssavirus Antibodies with Bio-Pro Rabies ELISA Ab Kit
Antibodies against lyssaviruses were detected using the commercially available BioPro Rabies ELISA Ab kit (O.K. SERVIS BioPro, Prague, Czech Republic), a blocking ELISA for the serological detection of rabies antibodies in domesticated [38] and wild animals [39]. In brief, following manufacturer's instruction on the modification of the ELISA for bat sera, 10 µL of bat sample was diluted 10-fold to a total of 100 µL using the sample diluent buffer provided in the kit. Additionally, the kit's controls were also diluted two-fold, and all samples were distributed into the respective wells of the plate. Thereafter, the plate was sealed and incubated at 4 • C for 24 h with gently stirring. The subsequent workflow was performed according to the manufacturer's instructions, essentially as described before [40].
The percentage of blocking (PB) or inhibition was calculated for each sample according to the manufacturer's recommendations (i.e., PB = [(ODNC − ODsample)/(ODNC − ODPC)] × 100, where ODNC is the optical density of the negative control, ODPC is the optical density of the positive control, and ODsample is the optical density of the sample. For bat sera and body fluids, a cut-off of PB ≥ 30% was considered positive for lyssavirus antibodies according to the manufacturer's instructions.

Modified Rapid Fluorescent Foci Inhibition Test (RFFIT)
The presence of virus neutralizing antibodies against EBLV-1 in bat sera was determined with a modified RFFIT using EBLV-1 as a challenge virus [41]. Sera were tested in two-fold dilutions on mouse neuroblastoma cells (MNA 42/13) with a starting dilution of 1:10.

Serological Testing of Eptesicus Bats (Eptesicus Fuscus and Eptesicus Serotinus) from Experimental Studies
The serological status of Eptesicus bats with regard to lyssaviruses was estimated using both an immunoenzymatic method with the BioPro Rabies ELISA Ab Kit and a modified SNT with EBLV-1 as the test virus. All Eptesicus serotinus bats captured in Germany were negative for the presence of antibodies against lyssaviruses, both in ELISA (Figure 3a) and in RFFIT (Table S1, Supplementary Materials). In contrast, eight out of 35 American Eptesicus fuscus bats tested positive in ELISA (Figure 3a), although they had no EBLV-1-specific VNAs (<1:10). A comparison between two samplings revealed a slight decline in percentage inhibition values (Figure 3b), albeit with a high correlation (Figure 3c).

Serological Testing of Eptesicus Bats (Eptesicus Fuscus and Eptesicus Serotinus) from Experimental Studies
The serological status of Eptesicus bats with regard to lyssaviruses was estimated using both an immunoenzymatic method with the BioPro Rabies ELISA Ab Kit and a modified SNT with EBLV-1 as the test virus. All Eptesicus serotinus bats captured in Germany were negative for the presence of antibodies against lyssaviruses, both in ELISA ( Figure 3a) and in RFFIT (Table S1,   ELISA results for all sera tested stratified by species. For Eptesicus fuscus, serum was taken at two occasions prior to infection. The comparative results between sampling point one and two are presented on an individual level (b) and were analyzed by a linear regression model (c). The slope (solid line) and the 95% CI (gray zone) are indicated. The results for the second sampling can be calculated using the formula Y = 0.7205X + 0.07097, with a slope that is highly significant divergent from zero (p < 0,0001).

Discussion
Although rabies in bats is the oldest known zoonotic disease associated with bats [2], our knowledge into various aspects of the disease and its association with bat biology is still limited. Partly, this is a result of unknown natural history of each bat species [42], their protected status [43], and the heterogeneous surveillance in Europe, terms of existing networks of bat biologists, the number of bat species submitted, and individual bats investigated [16]. Moreover, the discovery of novel lyssaviruses in European Natterer's bats (BBLV, Myotis nattereri), Schreiber's bent-winged bats (LLEBV, Miniopterus schreibersii), and Brandt's bat (KBLV, Myotis brandtii) [5,9,12] highlights that additional bat species could serve as reservoirs for known or still unknown lyssaviruses in Europe. Therefore, the aim of this study was to enhance our knowledge on the epidemiology of the disease by analyzing the serological status of Polish bats against lyssaviruses, using available samples and a ELISA results for all sera tested stratified by species. For Eptesicus fuscus, serum was taken at two occasions prior to infection. The comparative results between sampling point one and two are presented on an individual level (b) and were analyzed by a linear regression model (c). The slope (solid line) and the 95% CI (gray zone) are indicated. The results for the second sampling can be calculated using the formula Y = 0.7205X + 0.07097, with a slope that is highly significant divergent from zero (p < 0,0001).

Discussion
Although rabies in bats is the oldest known zoonotic disease associated with bats [2], our knowledge into various aspects of the disease and its association with bat biology is still limited. Partly, this is a result of unknown natural history of each bat species [42], their protected status [43], and the heterogeneous surveillance in Europe, terms of existing networks of bat biologists, the number of bat species submitted, and individual bats investigated [16]. Moreover, the discovery of novel lyssaviruses in European Natterer's bats (BBLV, Myotis nattereri), Schreiber's bent-winged bats (LLEBV, Miniopterus schreibersii), and Brandt's bat (KBLV, Myotis brandtii) [5,9,12] highlights that additional bat species could serve as reservoirs for known or still unknown lyssaviruses in Europe. Therefore, the aim of this study was to enhance our knowledge on the epidemiology of the disease by analyzing the serological status of Polish bats against lyssaviruses, using available samples and a commercialized ELISA [34].
The seropositivity among Polish bats (30.4%) against lyssaviruses corroborates other European studies which demonstrated seroprevalence in insectivorous bat species (for a review, see Reference [16]). Specifically, the largest numbers of antibody positives were detected in Nyctalus noctula and Eptesicus serotinus, followed by Plecotus auritus and Myotis daubentoni (Table 1). While Eptesicus serotinus is the reservoir of EBLV-1, and seropositivity was detected in various countries for this bat species, the positive results in Nyctalus noctula are quite exceptional. In Germany, among more than 600 sera of this particular bat species tested, none rested positive in an SNT [21]. In the latter study, samples were taken from roosting animals, which may limit their contact to the rather anthropophilic Eptesicus serotinus.
Daubenton's bat (Myotis daubentonii) is the main reservoir of EBLV-2 [4], and EBLV-2 cases in bats were reported from several European countries [7]. Thus far, lyssaviruses reported from Poland were characterized as EBLV-1 or BBLV [8]. The detection of antibodies against lyssaviruses in two out of three Polish Myotis daubentonii could indicate the presence of EBLV-2 also in Poland, but an interspecies transmission of EBLV-1 from Serotine bats cannot be ruled out either.
Generally, seasonality of seroprevalence in bats may result from differences in infectious status, behavioral ecology, and population dynamics among bat species [44]. In our study, the highest numbers of seropositive samples were found in August and September, as well as in March and April. The peak in seroprevalence in late summer seems to corroborate the assumption that the breeding period may favor virus transmission and subsequent infections, as demonstrated for Eptesicus fuscus [44]. Of note, significantly higher seroprevalences in this period were also demonstrated in France and Spain [25,45,46]. However, confounding effects cannot be excluded. For example, bat lyssavirus-specific antibodies were shown to persist up to four years in European bats, as exemplified for EBLV-1 [25]; thus, the antibodies detected in bats might not reflect recent exposure to lyssaviruses but might have been acquired much earlier. Additionally, the high degree of variation in seropositivity over time is partly influenced by a low sample size ( Figure 2).
In our study, antibodies against lyssaviruses were detected in nine different bat species, some of which were not identified as reservoirs for lyssaviruses. These findings support hypotheses on (i) inter-species transmission of lyssaviruses from known reservoirs, (ii) an expansion of the reservoir range for one specific lyssavirus, or (iii) the potential presence of other novel lyssaviruses of Phylogroup I. Of note, our results should be carefully interpreted, and both the variable number of submissions and the limited dataset (Table, 1, Figure 2) prevent formal statistical analyses and drawing of general conclusions on the infection processes in insectivorous bats from Poland.
The second objective of this study was to determine whether a commercially available BioPro ELISA Rabies Ab kit is suitable for bat samples. The seroprevalence in this study assessed with the ELISA is in the range of results from other European studies conducted with mFAVN or RFFIT, thus indicating the principle applicability of this approach for the detection of lyssavirus antibodies in bat samples in the frame of passive rabies surveillance and serological surveys (for a review, see Reference [16]). An unspecific binding of serum components that would lead to false positive results in the ELISA cannot be excluded; however, such a scenario was not reported during the broad usage of this ELISA in foxes or raccoon dogs [39]. This was corroborated by analyzing Eptesicus serotinus and Eptesicus fuscus sera from experimental studies. While all sera from Eptesicus serotinus that tested negative for EBLV-1-specific VNAs also tested negative in the ELISA (Figure 3a), eight sera from Eptesicus fuscus negative for EBLV-1 VNAs demonstrated a high concordance of results for subsequent samplings (Figure 3b,c). The fact that the ELISA-positive sera were SNT-negative is not surprising, as (i) the animals destined for infection studies with EBLV-1 were tested for homologous VNAs of the challenge virus and not for RABV-specific VNAs at the time [19], (ii) there is not necessarily a Viruses 2020, 12, 271 8 of 11 correlation between VNAs and binding antibodies [47], and (iii) cross-reactivity between EBLV-1 and RABV is less pronounced compared to other lyssaviruses [48].
The modified ELISA protocol as suggested by the manufacturer uses a smaller quantity of the sample (10 µL/sample as opposed to 50 µL/sample), which could allow for including small bat species into active bat rabies surveillance. Evidently, the higher dilution of samples results in a lower PB, as demonstrated by comparative analyses of ferrets experimentally infected with bat lyssaviruses ( Figure S1a, Supplementary Materials). Even with a reduced cut-off for positivity of 30% PB, the sensitivity was reduced compared to the standard approach using a modified RFFIT. Notably, there was a low test agreement between the modified RFFIT and the ELISA, and only sera with a higher titer tended to test positive in the ELISA ( Figure S1b,c, Supplementary Materials).
Obviously, a stringent validation, as was done in the development of the ELISA to be used for foxes and raccoon dogs for post-vaccination monitoring [39], is lacking for bats. Against the background of the genetic diversity of bats, the protected status of European bats, and the need for adaptation of cut-offs per species [47], this is rather illusionary. A balanced use of serial bleedings from an individual animal may represent an alternative to increase the robustness of sensitivity estimates of diagnostic tests, especially when sample availability is restricted [49]; however, in this case, it would require experimental vaccination or infection studies with different bat lyssaviruses.
Nonetheless, while negative results, therefore, do not exclude the presence of antibodies below the threshold of positivity, ELISA-positive sera appear to be an indicator for the presence of lyssavirus-specific antibodies.
Another advantage of the BioPro Rabies ELISA Ab kit is that it allows for the detection of antibodies in different matrices such as serum, plasma, or body fluids, and that it is not as sensitive to the quality of the samples as cell-based SNTs [34,50]. The BioPro ELISA Rabies Ab kit was specifically developed for the detection of antibodies against the RABV. However, because of cross-reactivity [51], ELISA may allow for the detection of antibodies against all lyssaviruses within phylogroup I. Further investigations and validations are necessary to standardize this method for detection of antibodies against RABV and non-RABV Lyssaviruses within phylogroup I in bat reservoirs.
Since RABV is the only lyssavirus present in the Americas [4], the approach from this study using an ELISA could ideally be used there, in the frame of serosurveillance in animals, particularly bats, thus providing highly reliable data. In the case of European and American bats, however, further studies are needed to verify the results obtained.