Diverse Beta- and Gammaherpesviruses in Neotropical Rodents from Costa Rica

Herpesviruses (order Herpesvirales, family Herpesviridae) are large, enveloped, doublestranded DNA viruses belonging to three subfamilies (Alphaherpesvirinae, Betaherpesvirinae, and Gammaherpesvirinae; Ehlers et al. 2007). They circulate among a wide range of hosts, causing infections along a spectrum from benign to fatal (Riaz et al. 2017). All herpesviruses are associated with specific hosts and can be established and remain latent in the host cells (Riaz et al. 2017). Novel strains and host-virus associations are being found in wild mammals regularly (Host and Damania 2016).

Rodents are the most diverse mammalian order and host a high diversity of viruses, including numerous species that are zoonotic (Han et al. 2015). The detection of herpesviruses has been regularly reported for wild and urban rodent species. Specifically, some rodent betaherpesviruses, such as murine cytomegalovirus, are reported worldwide in wild house mice. Herpesvirus DNA sequences are reported in European house mouse (Mus musculus; Cizkova et al. 2018). Gammaherpesvirus DNA has been detected in wild rodents, namely, in yellow-necked mouse (Apodemus flavicolis), wood mouse (Apodemus sylvaticus), bank vole (Myodes glareolus), and greater bandicoot rat (Bandicota indica), and in urban rodents, namely, in house rat (Rattus spp.) and house mouse (Mus spp.), from several European and Asian countries (Ehlers et al. 2007; Teterina et al. 2009; Ettinger et al. 2012; Zheng et al. 2016). Some herpesviruses are present in a high prevalence and may have detrimental effects on their hosts, depending on the immune status of the host (Tischer and Osterrieder 2010; Knowles et al. 2012). As such, they may impact rodent population dynamics (Knowles et al. 2012), and their potential involvement in the conservation of wild populations must be considered.

There are 49 native rodent species in Costa Rica (Villalobos-Chaves et al. 2016), and these diverse rodent communities might host equally diverse herpesviruses. We assessed the abundance and diversity of herpesviruses in Costa Rican rodents. We analyzed voucher samples (134 livers and 26 hearts) from 160 individual rodents belonging to 15 genera and 20 species, collected during 2010 and 2011. These samples were previously used for phylogeographic and taxonomic studies (Soley 2014; Villalobos-Chaves et al. 2016), following consensus guidelines for the use of wild mammals in research (Sikes and Gannon 2011). The tissues were kept in ethanol at ^_20 C prior to analyses.

We extracted DNA from samples using the DNeasy Blood & Tissue Kit (Qiagen^®, Hilden, Germany), according to the manufacturer's protocol. Next, we tested samples for the presence of herpesvirus DNA, following the protocol of VanDevanter et al. (1996). As a positive control, we used a plasmid containing herpesvirus DNA homologous to the PCR target, donated by Ecology of Diseases and One Health Laboratory, National Autonomous University of Mexico (Mexico City, Mexico). We employed nuclease-free water as a negative control. The PCR products were visualized using a 2% agarose electrophoresis gel stained with GelRed^TM (Biotium Inc., Hayward, California, USA). Amplicons containing bands with the appropriate length (215–315 base pairs) were reamplified and Sanger-sequenced using the primers from VanDevanter et al. (1996). Sequences obtained varied in length from 150 to 216 nucleotides.

We found eight positive samples (one heart and seven livers) belonging to five rodent species as follows: 11% (3/17) Mexican deer mouse (Peromyscus nudipes), 9% (2/23) shortnosed harvest mouse (Reithrodontomys creper), 25% (1/4) sprightly colilargo (Oligoryzomys vegetus), 11% (1/9) long-tailed singing mouse (Scotinomys xerampelinus), and 4% (1/ 25) Talamancan oryzomys (Nephelomys devius), all belonging to Cricetidae family (Table 1), yielding a total presence of 5% (8/160; 95% confidence interval 0.096–0.220).

To partially characterize their phylogenetic placement within the Herpesviridae family, the sequences identified herein (GenBank accession nos. MH257590–MH257592 and MH257594–MH257598) were aligned using the species viruses inside the Betaherpesvirinae and Gammaherpesvirinae subfamilies retrieved from the International Committee on Taxonomy of Viruses, as well as the few available sequences of herpesvirus in rodent hosts obtained from GenBank, using the multiple alignment using fast Fourier transform algorithm in Geneious V9 (Kearse et al. 2012). A maximum likelihood evolutionary tree was constructed in MEGA 6.0 (Tamura et al. 2013; Fig. 1).

Since the rodent herpesvirus sequences found were so diverse (Supplementary Table 1), constructing a plausible phylogeny with known sequences found in the databases was challenging. Therefore, the available herpesvirus sequences we used in phylogenetic analyses were from diverse hosts and countries and represented only a basic phylogeny for the designated subfamilies. Even though the tree had no support in posterior nodes hindering further phylogenetic conclusions, seven sequences clustered with representative species of gammaherpesviruses and one to betaherpesviruses. The putative gammaherpesviruses found in Talamancan oryzomys, Mexican deer mouse, and Chiriqui harvest mouse clustered separately from other gammaherpesviruses with high bootstrap support, suggesting that wild rodent herpesviruses could form a separate clade within the gammaherpesviruses. The dependence of this cluster on either rodent species or geographical location remains to be determined. Nevertheless, the sequences we found grouped with known rodent gammaherpesviruses found in a house mouse from Germany (Gertler et al. 2017), which suggested that rodent gammaherpesviruses could form a cluster within gammaherpesvirus.

For the sequence H129 from a long-tailed singing mouse, a gammaherpesvirus detected in China in a brown rat (Rattus norvegicus; KX987274) shared an 81% identity in a National Center for Biotechnology Information (2017) BLAST search but is not shown in Figure 1 because of the limited genomic sequence available. This sequence also clustered within the gammaherpesvirus subfamily; however, low support made it difficult to infer the phylogenetic placement within the gammaherpesvirus subfamily (Fig. 1). For the putative betaherpesvirus sequence H35, 78% identity to a cytomegalovirus (AY728179; not shown in Fig. 1 due to the limited genomic sequence available) found in a greater whitetoothed shrew (Crocidura russula) from Germany hinted that this sequence belonged to a diverse cytomegalovirus. However, the low bootstrap support observed with related betaherpesvirus sequences found in databases precluded further classification (Fig. 1). In summary, the sequences identified belonged to diverse gamma- and betaherpesviruses; however, further sequencing of genome segments is needed for further classification.

Being a country with great species diversity, the Costa Rican rodent community might harbor more diverse herpesviruses yet to be discovered. Because of the low presence encountered and the sampling strategy, we could not exclude the possibility that other species of Neotropical rodents might also harbor diverse herpesviruses. More research is needed to assess the host specificity of herpesviruses. Changing habitat quality and climatic conditions may modify viral cycles (Rico-Chávez et al. 2015), so we must regularly evaluate if the detection rate remains stable over time and if detection of other herpesviruses could be plausible in different seasons of the year. The finding of diverse herpesviruses may serve as a first step toward research regarding biology, ecology, and epidemiology of murine herpesviruses and guide ecological investigations targeting prevalence and host association of herpesviruses in Neotropical wild rodents.

We thank the grants and institutions that funded the original fieldwork: the American Society of Mammalogists (Grant-in-Aid of Research; Latin American Student Field Research Award), American Museum of Natural History (Theodore Roosevelt Memorial Grant), and City University of New York (PSC-CUNY Research Award). The current research was funded by the University of Costa Rica.

Supplementary material for this article is online at http://dx.doi.org/10.7589/2018-05-117.