Prevalence of coronaviruses in European bison (Bison bonasus) in Poland

Coronaviruses have been confirmed to infect a variety of species, but only one case of associated winter dysentery of European bison has been described. The study aimed to analyze the prevalence, and define the impact on the species conservation, the source of coronavirus infection, and the role of the European bison in the transmission of the pathogen in Poland. Molecular and serological screening was performed on 409 European bison from 6 free-ranging and 14 captive herds over the period of 6 years (2017–2023). Presence of coronavirus was confirmed in one nasal swab by pancoronavirus RT-PCR and in 3 nasal swab samples by bovine coronavirus (BCoV) specific real time RT-PCR. The detected virus showed high (> 98%) homology in both RdRp and Spike genes to BCoV strains characterised recently in Polish cattle and strains isolated from wild cervids in Italy. Antibodies specific to BCoV were found in 6.4% of tested samples, all originating from free-ranging animals. Seroprevalence was higher in adult animals over 5 years of age (p = 0.0015) and in females (p = 0.09). Our results suggest that European bison play only a limited role as reservoirs of bovine-like coronaviruses. Although the most probable source of infections in the European bison population in Poland is cattle, other wild ruminants could also be involved. In addition, the zoonotic potential of bovine coronaviruses is quite low.

hosts and ecological niches.The importance of this potential of CoVs has become evident in the last decades as the spread and adaptation of three coronaviruses of animal origin to humans led to outbreaks of severe acute respiratory syndrome coronavirus (SARS-CoV), Middle East respiratory syndrome coronavirus (MERS-CoV) and the pandemic of SARS-CoV-2 2,3 .Susceptibility of various wild and domestic animals to SARS-CoV-2 infections has been evaluated in several studies based on analysis of Angiotensin-converting enzyme 2 (ACE2) surface receptor of the host, study of virus prevalence in collected field samples and experimental infections 4,5 .The aforementioned ACE-2 receptor is responsible for coronavirus jumping between different hosts 4 , however little is known about interspecies transmission, except for the epidemic strains like SARS-CoV-2 6 .Spill-over events of SARS-CoV-2 from humans affected far more species than were identified at the start of the outbreak (pangolins and bats).The susceptibility to the virus was confirmed in over twenty animal species including ferrates, felines, and canids [7][8][9][10] .While the deer-to-deer and deer-to-human transmission of SARS-CoV2 was confirmed in white-tailed deer (WTD; Odocoileus virginianus) in North America, the involvement of European cervids in the pandemic has, however, been dismissed due to the lack of evidence of their exposure [11][12][13][14][15][16][17][18] .Exposure to Alpha, Delta, and Omicron variants was only shown in some individual fallow and red deer in the suburban areas of Madrid, Spain 19 , suggesting some possible spill-over events, however, rather accidental, not relevant for maintaining the virus in the environment.The ability of intraspecies transmission was also suggested based on the high homology of Bovine Coronavirus (BCoV) strains from cattle and BCoV-like Embecoviruses described in non-domesticated wild ruminants (Fig. 1) 3,20 , although such events have not been proven in field studies yet.Transmission between farmed and wild ruminants and vice versa may facilitate the persistence of this virus in nature, the recurrent emergence of epizootics, and its continuous evolution 20 .
The available data on coronavirus infections in wildlife in Poland is limited, as the only wild mammalian species in which the presence of CoVs was confirmed were hedgehogs and bats 21,22 .A small study on SARS-CoV-2 exposure in red deer in North-Eastern Poland did not confirm any possible involvement of this species in the transmission of the pandemic virus, how it was shown for farmed mink 23 , and incidental exposure of pets 14,24 .
Figure 1.Phylogenetic relationship of selected members of Betacoronavirus genus infecting wild,related domestic mammals, and human.Country of origin and year of detection of selected strains are included in the brackets.Three subgenus are distinguished: Embecovirus (green), Sarbecovirus (red) and Marbecovirus (blue).Phylogenetic tree was constructed by neighbour-joining method based on the RdRp gene sequences of coronaviruses available in GenBank, using MEGA11 software.
The studies on European bison as a reservoir of pathogens threating domestic animals and humans suggest mostly that the wild species exposure is due to the spillover from the synanthropic counterparts, however, no direct evidence on intraspecies transmission exists 27 .Due to their size, European bison are an ideal food source first for blood-sucking insects and second for scavengers and top predators.European bison is also considered as umbrella-species contributing to the protection of flora and small fauna and enriching biodiversity.We have shown that they are an important reservoir or are responsible for the persistence of emerging and re-emerging arboviruses in the environment 28,29 , which similarly to coronavirus 30 are considered climate-sensitive infections, meaning that they may be more seasonal, and frequency, rather than density-dependent.Previously, some herds of European bison in the Bieszczady mountains infected with Mycobacterium caprae may have contributed to the contamination of the environment and the exposure of other wild species including deer, wild boar and wolves 31 , while tuberculosis was successfully eradicated by selective culling in European bison in Bieszczady.The relevance of climatic parameters to the incidence of BCoV infection in cattle is reflected by the name of the disease it causes-winter dysentery (WD).The risk associated with coronavirus infections should also be considered important for the protection of European bison, as the species face the new challenges of the modern world related to environmental and anthropogenic changes.Poland is the homeland of the largest free-living population of European bison, whose number has increased significantly in recent years.The population size has exceeded the limits of the habitat capacity of large forest complexes such as the Białowieża Primeval Forest, resulting in the expansion of European bison into agricultural areas.In addition to the conflicts and losses associated with the damage caused by these large mammals, the possibility of direct (pastures) and indirect (manure fertilization) contact with livestock and humans, and thus the potential transmission of pathogens, is also increasing.Potential transmission of viruses between cattle and European bison is supported by results of recent studies on the seroprevalences of multiple viral pathogens known to infect cattle including bovine herpesvirus 1 (BoHV-1), bovine respiratory syncytial virus (BRSV), bovine adenovirus (BAdV), bovine viral diarrhea virus (BVDV) or bovine parainfluenza virus-3 (BPIV-3) 27,32 .The European bison protection strategy involves creating more, smaller populations by translocations into new areas and countries such as Spain, Romania, Netherlands, and Portugal, thus creating potential risk of pathogen transmission.Despite that the species' status by International Union for Conservation of Nature (IUCN) Red List has been changed from vulnerable to near threatened, the species remains endangered, also due to the low genetic diversity of the entire global European bison population, which may lead to impaired resistance to health hazards and resilience to changing environment.It urges the necessity for surveillance and control strategies as a tool for the conservation and prevention of epidemiological risks.
The epidemiological status of European bison remains poorly recognized as a single case of winter dysentery associated with Bovine-like CoV was described in an individual European bison housed in a zoo in Korea 25 .Additionally, the susceptibility of European bison to SARS-CoV-2 remains unexplored as only one individual was included in a recent serosurvey of wild ruminant screening in Germany, which also proved to be a negative 18 .Therefore, the study aimed to evaluate the occurrence coronaviruses in the European bison as it can represent unrecognized health risks that could affect the conservation program.

Coronavirus infections in European bison
The virus was detected in only one (0.3%) nasal swab sample collected from European bison originating from Borecka forest in 2019 in pancoronavirus RT-PCR.The result was confirmed by BCoV specific real-time RT-PCR, together with two other positive swab samples, collected from two individuals from the free-ranging Knyszyńska forest population in 2019 and 2020 (Fig. 2).Two of the animals the age of 7 and 13 years were healthy individuals chemically immobilized for other purposes.While one animal from Knyszyńska forest was 15-year-old bull sampled post-mortem after a traffic accident.The death of the above-mentioned animal occurred as a result of multi-organ damage resulting from severe trauma, no lesions indicative of a respiratory infection were found.All fecal samples were negative for the presence of coronavirus in both pancoronavirus RT-PCR and real-time BCoV tests.
Partial sequences of S and RdRp genes of one coronavirus strain isolated from European bison were aligned with sequences of previously identified coronaviruses isolated from cattle, human, and other species.Identified coronavirus strain named EB/PL/BF/1/2019 showed high homology of the nucleotide sequence in RdRp (98.6-99.5%)and Spike genes (98.3-99.3%)fragments to recently identified Polish strains of BCoV isolated from cattle 33 .High genetic homology in the RdRp gene was also found in coronavirus strains, isolated from roe deer in Italy in 2021 (99.7%) and in the Spike gene from strains isolated from cattle in Israel in 2018 (99.3%) and in Slovakia in 2017 (99.0%) 15,34,35 (Fig. 3, Fig. 4).

Serological evidence of BCoV exposure in European bison
According to ELISA results, 26 out of 409 (6.8%)European bison had antibodies to BCoV.The antibody levels with regard to the semi-quantitative gradation of the test (+ to + + + +) were almost equally distributed among the positive serum.The highest seroprevalences were observed in the free-ranging populations, while none of the captive European bison had BCoV antibodies (Table 1).That could explain the higher seroprevalences in the selectively eliminated due to the poor health conditions and dead in accident animals, which mostly concern the free-ranging European bison.According to those results, the exposure to BCoV increased along with the age of the animals (Table 1, Fig. 5).

Discussion
The current study is the first one to identify and characterise coronavirus circulating in European bison, indigenous to Europe free-roaming species of the Bison genus, Bovidae family.As expected, the virus detected belonged to a BCoV-like Embecovirus group of the Betacoronavirus genus.The low prevalence (1%) associated with viral shedding by nose only suggests that the strain circulating in European bison has a respiratory rather than enteric tropism 36 , however, CoV infections in the European bison population are incidental.Coronaviruspositive animals were detected by both pancoronaviral (1 sample) and BCoVs specific (3 samples) PCR reactions.While the number of positive results were different this should probably be associated with higher sensitivity of real-time PCR, as positive samples with high Ct values (> 35) in BCoV specific test were negative in pancoronavirus specific PCR.The fact that all positive samples were detected by BCoV-specific assays pointed on BCoV-related virus as a source of infection.This was further confirmed by sequence analysis of the one coronavirus positive sample isolated from European bison originating from the free population of Borecka forest.It showed 98.6 to 99.3% nucleotide sequence homology of the Spike protein encoding fragment and 98.6 to 99.7% in the RdRp gene to Polish strains of BCoV isolated from cattle 33 .Very high homology (over 99.5%) was also observed with RdRp sequences of 29944_2021 Embecovirus strain isolated in 2021 from Italian roe deer.All these strains could be assigned as related European lineage (or GIIb group as assigned by Shin et al. 2022 37 ) of BCoV, composed of strains isolated from various ruminants such as cattle, fallow deer, roe deer, or buffalo, in the last two decades.These results are in line with previous studies on coronavirus infection in captive and non-captive wild ruminants such as sambar deer, white-tail deer, water deer, antelope, or giraffes.Betacoronaviruses isolated from those species shared a high degree of homology to local strains of Bovine coronavirus with no species-specific mutations observed 38,39 .A similar observation was made in Korea for the only coronavirus isolated till now from a European bison kept in a zoo 25 .Two possible explanations were proposed for the lack of species-specific CoVs among ruminants.First, that new host adaptation of CoVs might be defined by a combination of substitutions scattered throughout the whole genome that could not be easily distinguished based on phylogenetic analysis of partial genome sequence, and second-that no adaptation changes are needed for CoV interspecies transmission between ruminants 38 .The second explanation was additionally supported by a large-scale study on the evolution of the BCoV genome that showed only limited signs of host or tissue adaptation 40 .As the CoV isolate acquired in our study was equally similar to BCoV strains isolated from cattle and roe deer it is not possible to unequivocally establish the original source of infection.It is known that European bison and various species of wild cervids such as red deer are present in the same habitats in Poland 41 and previous studies showed that deer species could be a source of CoV infection for other free-living ruminants (ex.sambar deer for waterbuck) 40 .The evidence that BCoV is not endemic in the European bison population and that the infection spilled out from other ruminants (cattle or cervids) is supported by the fact that we solely found infected individuals in free-ranging herds, not kept in enclosures.Further, it is more probable that the transmission originated from cattle, since our recent, yet unpublished study, showed that BCoV prevalence in Polish wild cervids is close to 0%.Whereas, our previous studies have confirmed that BCoV is highly prevalent in Polish unvaccinated cattle, with 72.6% antibody positives and 10.5% virus shedding by respiratory route individuals sampled between 2014 and 2015 33 .BCoV infection is more frequent in younger cattle and it is associated with bovine respiratory complex (BRD) in calves 33 .Coronavirus was detected in one-third of calves affected by BRD in the country, indicating its importance in the complex aetiology of the disease 42 .This contrasts with low (6.8%)BCoV seroprevalence among European bison and a strong correlation between age and the serological status with no seropositives among the youngest animals.Moreover, the lack of maternal antibodies to European bison calves below 6 months of age supports the rather incidental infection in the species unlike Polish cattle, which showed over 80% seroprevalence of this group of calves 33 .BCoV seroprevalence differed between herd sizes, with the lowest 50% infected in the smallest barns under 77 head which may represent those with which free-ranging European bison populations come into contact on pastures 33 .However, our results do not corroborate the observations of the results of a questionnaire among cattle breeders, which were asked to assess the quality of their cattle's contact with European bison 43 .On the basis of their opinion the highest risk for virus transmission by direct or indirect contact or sharing pastures between European bison and cattle should be considered in in the area of Białowieża Forest, which was not supported by the relatively low BCoV seroprevalence among the free-ranging populations in our study 43 .However, the higher exposure of European bison may be explained by the higher density of cattle and their breeding in extensive conditions and more frequently grazed.It was supported by the number of seropositive ( 15Figure 3. Neighbour-joining phylogenetic tree was constructed using a 601-nucleotide-long fragment of the gene encoding the spike protein of Betacoronaviruses with Human coronavirus HCoV OC43 used as an outgroup 54 .Sequence acquired in this study is marked by black square.In brackets the European (GIIb) phylogenetic group as described by Shin et al. 2022 37 was distinguished.Data about the country of origin, date of sample collection and hosts are included next to the name of each strain.out of 26) and BCoV infected (2 out of 3) European bison, which originated from the Podlaskie region, a part of Poland with the highest cattle density per 100 ha of agricultural land (95.9 animals per 100 ha 44 ).Moreover, one of the highest (over 80%) BCoV seroprevalences was previously reported in the cattle from this area 33 .The direction of transmission from cattle to European bison, has already been concluded for several endemic bovine pathogens 27,32 , which may suggest that this recently booming in numbers and expanding geographically population could threaten the very conservation of the species itself.European bison are often considered as an additional epidemiological risk to livestock 43 .Therefore, efforts to monitor the health threats should be undertaken, including at the destinations prior to European bison translocations.Although respiratory lesions are the most commonly observed in European bison 45 , there is no evidence that they may be related to coronavirus infection, as we only found BCoV shedding in individuals not displaying them.However, despite this, the highest seroprevalence was observed for European bison eliminated for health reasons and those that died in accidents, thus possibly co-infections with coronaviruses contribute to their overall poorer health condition 46 .Similarly, there is little evidence of clinical BCoV infections in other wild ruminant species 39,46,47 .Previously, it was shown that ruminant species, such as wild-tailed-deer could not only transmit SARS-CoV-2 but also be clinically affected by COVID-19 11,48,49 .In that case this was confirmation of previous theoretical studies, suggesting WTD vulnerability to SARS-CoV-2 basing on high degree of homology of ACE2 receptor, crucial for viral entry, with the human one 50 .While similar analysis of ACE2 receptor has not been done for European bison, it was previously shown that degree of this homology is lower in case of ACE2 receptor of closely related American bison 51 .Therefore, the lack of SARS-CoV-2 infections among European bisons tested in our study should not be surprising even though some close encounters with human might have occurred, especially in enclosures.
The free-ranging European bison population in Poland lives mainly in forested areas.The constant increase in the population of free-ranging European bison leads to an increase in animal density and migration of some individuals outside the previously occupied forest habitats and staying in agricultural areas (arable fields, meadows).Factors such as the increasing density of free-ranging European bison in a given area and the increase in the frequency of direct or indirect contact between them and farm or wild animals may pose a potential risk of an increase in the number of CoV infections in European bison.Currently, it is impossible to explain whether the spread of the CoV virus in European bison encounters obstacles to the adaptation of the virus to a new host or whether the spread is limited by too low frequency of contact between animals.In summary, to fully elucidate whether other wild ruminants or cattle play a more important role in the spread of CoVs among European bison, further serological and virological studies on populations of those animals in Poland are needed.Nevertheless, coronavirus infection in Polish European bison seems to be rare and probably has little effect on the health status of the population.

Samples
Samples were collected from 409 European bison (Bison bonasus) including healthy (immobilized for other purposes such as translocation or GPS collar placement); selectively eliminated due to poor health condition; fallen (found dead); and dead in an accident between 2017 and 2023.Serum samples were collected from each animal, along with either fecal samples, nasal swabs, or both.As a result, in total 284 nasal swabs, 354 fecal samples and 409 serum samples were included in the study.Samples originated from 221 free-ranging and 188 captive animals from 20 main populations including six free-ranging from all over the country.Among those tested 222 were males and 187 females.The age of the tested animals ranged from 1 day to 22 years.

RNA extraction
Collected fecal samples were diluted in PBS to 10% solution and homogenized using Fastprep-24 5G™ (MP Biomedicals, Irvine CA, USA) homogenizer.Viral RNA was extracted from 140 µL of nasal swab and fecal homogenate samples using a QIAamp Viral RNA Mini kit (Qiagen, Hilden, Germany) according to the manufacturer's guidelines.Ribonucleic acid was eluted in 50 µL of an elution buffer and stored at −70 °C.

Real-time PCR
A real-time RT-PCR for BCoV detection was performed using previously described primers specific to the gene encoding the M protein 53 .Additionally, for internal control, a 200 µL mix of primers and probes specific to β-actin was prepared consisting of 5 µL of 100 µM ACT-1005-F and ACT-1135-R primers, 3.75 µL of 100 µM ACT-1081-HEX probe and 186.25 µL of water.The reaction was run in 20 µL of reaction mix that comprised 6.3 µL of water, 4 µL of 5 × QuantiTect Virus Master Mix (Qiagen, Hilden, Germany), 2 µL of both BCoV-F forward (10 µM) and BCoV-R reverse (10 µM) primers, 2 µL of BCoV-Pb probe (5 µM), 1.5 µL of a mixture of primers and probes specific to bACT, 0.2 µL of 100 × QuantiTect Virus RT Mix (Qiagen, Hilden, Germany) and 2 µL of RNA sample.After 30 min of reverse transcription at 42 °C and a 10 min incubation at 95 °C, 40 cycles of amplification were run each consisting of 15 s of denaturation at 95 °C and 45 s of annealing/elongation at 58 °C.As a positive control, for each PCR reaction, BCoV S379 Riems strain was used.All real-time PCR amplifications were performed using a LightCycler 96 Instrument (Roche, Mannheim, Germany).Sequences of primers used for real-time RT-PCR are presented in Table 2.

Figure 2 .
Figure 2. Map of the free-ranging populations (light grey silhouette) and enclosures (dark grey silhouette), where tested European bison originated from.Two non-existing in 2022 captive herds: Kiermusy (M) and Smardzewice (N) were sampled between 2018 and 2021 (M) and in 2018 (N), that is, until they were closed down.Seropositive herds were red framed, while red virus sign indicated locations where coronavirus infected individuals were detected.The size of the silhouette on the map corresponds approximately to the herd / population size.

Figure 4 .
Figure 4. Neighbour-joining phylogenetic tree was constructed using a 432 nucleotide-long fragment of the gene encoding the RNA-dependent RNA polymerase (RdRp) of Betacoronaviruses with Human coronavirus SARS-CoV-2 used as an outgroup 54 .Sequence acquired in this study is marked by a black square.In brackets 3 subgena of the Betacoronavirus genus were distinguished.Data about the country of origin, date of sample collection, and hosts are included next to the name of each strain.

mix and 2
µL of DNA product from first reaction.Reaction conditions of first amplification included 30 min reverse transcription in 50 °C, 15 min incubation in 95 °C and 45 cycles of amplification, each consisting of 30 s denaturation in 95 °C, 30 s of annealing in 55 °C and 1 min elongation in 72 °C.The reaction was finished with 10 min incubation at 72 °C.For nested PCR, the reaction consisted of 10 min incubation at 95 °C, 40 cycles of amplification, with 30 s denaturation at 95 °C, 30 s of annealing at 48 °C and 45 s elongation at 72 °C and 4 min of final incubation in 72 °C.As a positive control, the BCoV S379 Riems strain was used.5 μL of PCR product of the second reaction was used for visualization by electrophoresis on the 1.5% agarose gel.A sample was regarded as positive if a band of the expected 440 bp size was visible on the gel.Sequences of primers used for each step of the reaction are presented in Table

Figure 5 .
Figure 5.The local polynomial smoothed line represents BCoV seroprevalence by European bison age.

Table 1 .
Association between individual and herd-level parameters and Bovine coronavirus (BCoV) seroprevalence in European bison.a binomial exact 95% confidence interval; b included free-ranging and captive European bison; c one-sided 97.5% confidence interval.