Abstract
In 2021, bovine polyomavirus 1 (BoPyV1; Polyomaviridae, Epsilonpolyomavirus bovis) was associated with nephritis in an aborted bovine fetus in Uruguay, with renal lesions resembling those typical of polyomavirus-associated nephropathy of humans. Given that little is known about the epidemiology of BoPyV1 infection in cattle, we screened for BoPyV1 in urine samples collected in 2015–2017 from beef and dairy herds from 12 of the 19 departments in Uruguay. We tested for BoPyV1 by PCR in 156 urine pools and 249 individual urine samples from 42 herds. We detected BoPyV1 in 33 of 42 (79%) farms across 100% of the departments studied, in similar proportions in beef (22 of 30; 73%) and dairy (11 of 12; 92%) herds. At the animal level, BoPyV1 was detected in 80 of 249 (32%) animals; this frequency was significantly higher in dairy (51 of 119; 43%) than beef (29 of 130; 22%) cattle, and in cows (36 of 81; 44%) than heifers (32 of 121; 26%). BoPyV1 strains circulating in Uruguay have a high degree (98.7–100%) of sequence identity at the major capsid protein VP1, which is slightly lower (96.2–99.7%) than for strains from other countries. We conclude that shedding of BoPyV1 in the urine of dairy and beef cattle is prevalent and geographically widespread in Uruguay.
The family Polyomaviridae contains a diverse group of non-enveloped viruses with a small, circular, double-stranded DNA genome that infects a wide variety of vertebrate and invertebrate animal species. 15 According to the 2023 taxonomy release of the International Committee on Taxonomy of Viruses (ICTV; https://ictv.global/taxonomy), the family has 8 genera, namely Alphapolyomavirus, Betapolyomavirus, Deltapolyomavirus, Epsilonpolyomavirus, Etapolyomavirus, Gammapolyomavirus, Thetapolyomavirus, and Zetapolyomavirus; many polyomaviral strains are awaiting genus and species assignment by the ICTV.
Although most polyomavirus infections are subclinical, under certain poorly understood circumstances, some members of this family cause disease in mammals and birds. Spontaneous diseases caused by polyomaviruses include polyomavirus-associated nephropathy (PyVAN) caused by BK polyomavirus (BKPyV; Betapolyomavirus hominis) and to a lesser extent JC polyomavirus (JCPyV; Betapolyomavirus secuhominis) in humans 4 ; progressive multifocal leukoencephalopathy (PML) associated with JCPyV in humans 3 ; meningoencephalitis caused by simian virus 40 (SV40; Betapolyomavirus macacae) in non-human primates 21 ; budgerigar fledgling disease (BFD) caused by Aves polyomavirus 1 (Gammapolyomavirus avis) in psittacine birds 13 ; and cancer, such as Merkel cell (neuroendocrine) carcinoma caused by Merkel cell polyomavirus (MCPyV; Alphapolyomavirus quintihominis) in humans, 6 and neuroglial tumors of the brain and olfactory tract associated with raccoon polyomavirus (RacPyV; Alphapolyomavirus procyonis) in raccoons. 7
Little is known about polyomaviruses infecting ruminants. Two species of polyomavirus have been identified in bovids, namely: bovine polyomavirus 1 (BoPyV1; Polyomaviridae, Epsilonpolyomavirus bovis),1,8 and BoPyV2 (awaiting genus assignment by the ICTV). 10 A third species, named BoPyV3 (GenBank KM496326, awaiting genus assignment by the ICTV), was originally identified in beef samples collected at supermarkets in the United States 18 and then in Germany. 9 Historically, BoPyVs had not been causally associated with any disease.11,16,20 However, in their natural host, BoPyV1 and BoPyV2 have been recognized respectively as pathogens causing severe nephritis resembling PyVAN in an aborted fetus 8 and nonsuppurative encephalitis in postnatal life. 10 BoPyV3 has not been reported as a cause of disease to date.
Polyomavirus of cattle origin was first identified in 1976 and initially thought to be a primate virus 19 ; however, in 1983, cattle were recognized as the original hosts of this virus. 16 Shortly after, researchers found that people occupationally exposed to cattle had an increased risk of seropositivity to BoPyV, raising the hypothesis of possible zoonotic transmission. 17 BoPyV1 can infect bovine kidneys producing renal lesions, including intranuclear inclusion bodies with abundant viral particles in the renal tubular epithelial cells, 8 and can be shed in urine, 11 as can polyomaviruses that cause nephropathy in humans (i.e., BKPyV).
BoPyV1 was detected incidentally in 2011 in the retropharyngeal lymph node of a cow with tuberculosis in Spain 1 and, in 2014, the virus was identified in aborted bovine fetus(es) in Belgium. However, its causal association with gestational loss was thought to be unlikely, although the aborted fetuses were not subjected to gross or histologic examination. 24
Given that BoPyV1 seems to infect primarily the kidney and has been detected in urine samples, 11 we studied archived DNA extracted from bovine urine from a previous study 26 to determine the prevalence and geographic distribution of BoPyV1 in urine shed by beef and dairy cattle in Uruguay. We used archived DNA extracted from bovine urine from a study in which urine samples were obtained by veterinarians following a procedure approved by the Ethics Committee for the Use of Animals for Experimentation, of the Ministry of Livestock, Agriculture and Fishery of Uruguay. 26 The samples were collected in 2015–2017 from 42 dairy and beef herds (30 beef, 12 dairy) located in 12 of the 19 departments of Uruguay (https://en.wikipedia.org/wiki/Departments_of_Uruguay). We pooled 512 individual urine samples (10 mL each) into 156 sample pools (3 or 4 samples per pool), and processed another 249 samples (10 mL each) individually. Pooled and individual samples were centrifuged at 10,000 × g for 15 min to obtain a pellet from each pool or sample that was rinsed once with PBS pH 7.4. Total DNA was extracted from the pellets (Invitrogen PureLink genomic DNA mini kit; ThermoFisher), 26 and stored at −20°C until screened for BoPyV1 as described below.
The 156 pools were used to obtain herd-level information from 29 different dairy and beef farms (Table 1); the 249 individual samples were analyzed to obtain both herd- and individual-level information from another 13 dairy and beef herds (Table 2). Pooled samples were collected from herds in 12 departments and originated from calves, steers, heifers, and cows (Table 1). Individual samples originated from heifers, cows, and a bull from herds in 8 departments (Table 2). The proportion of BoPyV1 positivity (see below) was compared between beef versus dairy farms, beef versus dairy individuals, and cows versus heifers using a chi-squared test (https://www.socscistatistics.com/tests/).
Herds, number of samples, department in Uruguay, year of sampling, production system, and age class of pooled urine samples analyzed for bovine polyomavirus 1, and frequency of detection.
ND = not determined.
The number of samples per pool was 3 or 4.
Herds, number of samples, department in Uruguay, year of sampling, production system, and age class of individual urine samples analyzed for bovine polyomavirus 1, and frequency of detection.
ND = not determined.
For polyomavirus detection, a real-time PCR targeting a 77-bp fragment of the BoPyV VP1 gene was performed as described elsewhere. 12 Briefly, 10 µL of SensiFAST Probe No-ROX kit (Bioline), 4 µL of nuclease-free water, 0.8 µL of 10 µM forward primer (QB-F1-1), 0.8 µL of 10 µM reverse primer (QB-R1-1), 0.4 µL of 10 µM probe (QB-P1-2), and 4 µL of DNA were mixed in 0.2-mL PCR tubes. Negative controls were included in each run.
For polyomavirus classification, a partial, 419-bp, BoPyV VP1 gene sequence was amplified from 20 BoPyV-positive DNA samples by a conventional semi-nested PCR, 25 with minor modifications. For the first round of amplification, 12.5 µL of MangoMix (Bioline), 4.5 µL of nuclease-free water, 1.0 µL of 10 µM forward primer VP1-F (5′-GGTATTCGCCCTCTGCTGGTCAAG-3′), 1.0 µL of 10 µM reverse primer VP1-R (5′-GCTGGCAATGGGGTATGGGTTCT-3′), 1.0 µL of dimethyl sulfoxide, and 5 µL of DNA were mixed in 0.2-mL PCR tubes. The reaction conditions were 95°C for 5 min, 35 cycles of 94°C for 1 min, 58°C for 1 min, and 72°C for 1 min, followed by a final elongation at 72°C for 10 min. For the second round of amplification, 12.5 µL of MangoMix, 4.5 µL of nuclease-free water, 1.0 µL of 10 µM forward primer VP1-F2 (5′-ATTTCAAAGCCCCCTATCATC-3′), 1.0 µL of 10 µM reverse primer VP1-R (5′-GCTGGCAATGGGGTATGGGTTCT-3′), 1.0 µL of dimethyl sulfoxide, and 5 µL of the first-round amplification were mixed in 0.2-mL PCR tubes. The reaction conditions were 95°C for 5 min, 35 cycles of 94°C for 1 min, 58°C for 1 min, and 72°C for 1 min, followed by a final elongation at 72°C for 10 min. Negative controls were included in each run. The PCR products were visualized in 2% agarose gels. Seventeen samples were purified (Zymoclean gel DNA recovery kits; Zymo Research) and sequenced (Macrogen, Korea) with Sanger technology.
Partial VP1 sequences were used for sequence identity and phylogenetic analyses. Multiple alignments of nucleotide and amino acid sequences with lengths of 377 nucleotides and 124 amino acids, respectively, were performed with MEGA 11. 22 The p-distance estimation was also performed with MEGA 11 using the nucleotide alignment of 11 sequences obtained herein and 15 BoPyVs from GenBank. The phylogenetic analysis was performed with the amino acid alignment of 11 sequences obtained herein (with a length of 124 amino acids) and others downloaded after a BLAST search using as query BoPyV-1/28P-O17/2015/Uy (GenBank PP531469); sequences selected for download included BoPyV and other related PyVs from other hosts (in particular, sequences of BKPyV and JCPyV, which can cause PyVAN). The selection of the best-fitting evolutionary model according to the Bayesian information criteria (rtREV+G4) and the maximum-likelihood phylogenetic tree with SH-aLRT branches support were performed online with W-IQ-TREE. 23 We deposited our sequences in GenBank (PP531462–PP531478).
The prevalence of BoPyV1 infection was assessed at the farm level (pooled samples and individual samples from 42 farms) and animal level (249 individual samples). BoPyV1 was detected in 33 of the 42 (79%) farms (herd-level prevalence); this frequency did not differ significantly between beef (22 of 30; 73%) and dairy (11 of 12; 92%) herds (p = 0.25). Among the 156 pools analyzed, 50 (32%) were positive for BoPyV1. These pooled samples originated from 29 farms, 20 of which (69%) had at least one pool testing positive for BoPyV1 (Table 1). At the individual level, 80 of the 249 samples tested positive (Table 2), indicating an overall prevalence of 32%, which was significantly higher in dairy (51 of 119; 43%, 95% CI 34–52%) than beef (29 of 130; 22%, 95% CI: 16–30%) cattle (p = 0.0005).
Across different age groups of cattle, BoPyV1 was detected in all categories, including calves, steers, heifers, and cows. Animal-level analysis was primarily focused on heifers and cows (given that the virus has been associated with abortion), with examination of one bull, which tested positive. The frequency of detection was significantly higher (p = 0.008) in cows (36 of 81; 44%, 95% CI 34–55%) than heifers (32 of 121; 26%, 95% CI: 19–35%). Although we succeeded in detecting BoPyV1 from these samples, they had originally been collected and processed with the aim of detecting Leptospira 26 ; thus, the g-force achieved during centrifugation may not have been sufficient to efficiently pellet viral particles, which could have led to an underestimation of the BoPyV1 prevalence.
BoPyV1 was detected across all 12 departments studied, indicating an extensive geographic distribution. Additionally, BoPyV1 was consistently detected throughout the entire analyzed period of 2015–2017. All negative controls included in the PCR runs yielded negative results.
The results of our sequence identity analysis using a fragment of 377 nucleotides of the VP1 coding sequence were somewhat expected. Within the Uruguayan sequences, including the previously reported OM938033, 8 a high degree of genetic identity was observed (98.7–100%). Comparing these Uruguayan sequences to BoPyV1 strains detected in other countries, the genetic identity was 96.2–99.7%, indicating a relatively high level of similarity. In contrast, the genetic identity between the Uruguayan BoPyV1 sequences and BoPyV2 was substantially lower (59.1–61.3%). The genetic identity between the Uruguayan BoPyV1 sequences and BoPyV3 was even lower (32.1–32.8%).
Phylogenetic analysis using the fragment of 124 amino acids of the VP1 protein gave interesting results. We confirmed that the Uruguayan sequences indeed belonged to BoPyV1, and that these sequences had low variability compared to other sequences of BoPyV1 available in the database. At VP1, BoPyV1 clustered with Potamochoerus porcus polyomavirus 1 (PporPyV1) and Capra aegagrus polyomavirus 1 (CaegPyV1; both in the Epsilonpolyomavirus genus), and closer to BKPyV and JCPyV and other betapolyomaviruses, such as SV40, but distant to some other members of the Betapolyomavirus genus, such as Meles meles polyomavirus 1, California sea lion polyomavirus 1, and Betapolyomavirus lepweddellii (Leptonychotes weddellii polyomavirus 1; Fig. 1).

Phylogenetic analysis. Bovine polyomavirus 1 VP1 amino acid sequences obtained herein compared with other sequences in GenBank. Uruguayan sequences are highlighted with a rhombus if obtained in this study, or a triangle if obtained previously.
BoPyV1 has been used to track environmental contamination of bovine origin,2,11,20 but there are no reports in the literature of its prevalence in cattle, to our knowledge. In Uruguay, BoPyV was used as a host-specific indicator of bovine contamination in the Santa Lucía river basin and the Uruguay river, with 13 of 120 (11%) of the water samples analyzed from both rivers being positive. 2 Water samples were collected in 2015 and 2016 in the Santa Lucía river basin and the Uruguay river, 2 which overlaps the period in which the urine samples that we analyzed were collected. We confirmed a high herd-level prevalence of BoPyV1 urinary infection in cattle, which is a likely source of the BoPyV found in surface water.
Data on BoPyV prevalence in cattle at the individual level is scarce. A small set of 26 bovine urine samples was analyzed when developing a quantitative real-time PCR (qPCR) for the detection and quantification of BoPyV in Spain, of which 8 samples (31%) were positive. 11 This frequency of positivity is similar to the 32% positivity that we found in our study of a larger set of 249 samples. The same Spanish study included samples from 3 farms, obtaining a farm-level BoPyV positivity of 2 of 3 (67%), 11 also similar to our study in which 33 of 42 (79%) farms had at least one individual sample or pool testing positive for BoPyV1. Again, the number of farms (n = 3) and the number of samples included from each farm (n = 22, 3, and 1) were small compared to those included in our study.
Regarding production system and year, we detected BoPyV1 in both dairy and beef cattle, over the 3 years of sampling. Our findings suggest widespread infection of BoPyV1 across geographic regions, production systems, age groups, and over time. At the farm level, BoPyV1 was detected similarly in both production systems but, at the animal level, the prevalence was higher in dairy than beef cattle. It should be noted that the animal-level prevalence varied widely (6–96% in dairy farms; 11–47% in beef farms). The higher overall prevalence in dairy cattle is skewed by a particularly high prevalence in one farm (farm 35; Table 2), although the median animal-level prevalence was also almost twice as high in dairy (29%) as in beef (16%) farms. Breeds, husbandry, and management practices differ substantially among these production systems and between farms and may affect the transmission of BoPyV1 within and between herds. In addition, in Uruguay, dairy and beef farms are largely located in different geographic regions, with dairy farming being concentrated mostly in the south and beef cattle ranches predominating in the north; thus, these production systems occupy different agroecosystems, which may also affect viral transmission and environmental survival. Factors influencing BoPyV1 transmission should be studied to develop strategies to mitigate the impact of this virus both at the herd and animal levels.
It is interesting to note that we detected BoPyV1 in the urine of animals of both sexes and different age groups, with an overall individual prevalence of 32%, and with very low genetic divergence among the 17 analyzed sequences. The higher prevalence found in cows than heifers indicates that the infection is probably largely acquired postnatally (horizontal transmission) despite the fact that vertical (transplacental) transmission also occurs.8,24 Our animal-level analysis was somewhat biased toward females in reproductive age (heifers and cows, which comprise most of the individuals in a herd), with younger animals (calves) and bulls underrepresented. Despite this bias, BoPyV1 was detected not only in calves but also in an adult bull, which makes us wonder whether the virus could be transmitted venereally (i.e., through direct contact with the preputial mucosa and/or semen contaminated with urine).
Our sequence analysis results were expected. First, we verified the primer specificity for BoPyV1, thereby confirming that the positive samples indeed corresponded to BoPyV1 and not to other BoPyV species. In addition, all of the Uruguayan sequences had very similar genetic identity (≥98.7%) and with other BoPyV1 sequences (≥96.2%), which indicates a close genetic relationship among the sequences from Uruguay and BoPyV1 sequences from other countries, as expected for a DNA virus.
Polyomavirus capsid is icosahedral and formed by 72 pentamers of the major capsid protein VP1. This protein is also implicated in the recognition by neutralizing antibodies (determining different serotypes) and interacts with cellular sialic acid receptors; VP1 is involved in PyV variability, structure, host immune response, infectivity, and viral replication. 5 In the phylogenetic analysis we conducted using partial sequences of VP1, epsilonpolyomaviruses clustered, including BoPyV1 and other artiodactyl-infecting viruses (CaegPyV1 and PporPyV1). Like BKPyV, JCPyV, and SV40, BoPyV1 has urinary tract tropism, 4 and mutations in VP1 may alter virus tropism, at least in SV40. 14 Here, we observed a closer phylogenetic relationship of BoPyV1 with the betapolyomaviruses BKPyV, JCPyV, and SV40 than with other betapolyomaviruses, within the VP1 region studied; the implications of this relationship in tissue tropism should be explored further.
Footnotes
Acknowledgements
We thank the consortium Grupo de Trabajo Interinstitutional en Leptospirosis (INIA/MGAP/UdelaR/Institut Pasteur de Montevideo, Uruguay) for kindly supplying DNA samples.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
Our work was partially funded by grants PL_27 from INIA, and ALI_1_2014_1_4982 and FSA_1_2013_1_12557 from the Uruguayan Agencia Nacional de Investigación e Innovación (ANII), and funds from the Laboratorio de Virología, Centro Universitario Regional (CENUR) Litoral Norte, Salto, Uruguay.
