Novel Betaherpesvirus in Bats

Because bats are associated with emerging zoonoses, identification and characterization of novel viruses from bats is needed. Using a modified rapid determination system for viral RNA/DNA sequences, we identified a novel bat betaherpesvirus 2 not detected by herpesvirus consensus PCR. This modified system is useful for detecting unknown viruses.

Before using the RDV method, we had attempted to detect herpesvirus by nested PCR with the consensus primer sets DFA, ILK, KG1, TGV, and IYG, which were designed according to the consensus-degenerate hybrid oligonucleotide primers program (7). These consensus degenerate primers are effective for detecting many herpesviruses from vertebrate hosts. However, in this study they failed to detect any herpesviruses.
We then attempted to detect herpesvirus by using RDV version 3.1, our modifi cation from version 3.0 (5). The adapters and primers for construction of the second cDNA library in RDV version 3.1 were newly designed and replaced those used in RDV version 3.0 (online Technical Appendix 1, www.cdc.gov/EID/content/16/6/986-Techapp1.pdf). Both adapters have sticky-end structures digested with Sau3AI or HpyCH4 IV. RDV version 3.1 can determine an unknown viral cDNA fragment with 64 primer pairs, which we used for constructing the second cDNA library.
With RDV version 3.1, we obtained 4 unknown cDNA fragments, which had no matches in a BLASTn (www.ncbi. nlm.nih.gov/blast/Blast.cgi) search. In a BLASTx search, 1 cDNA fragment (deduced sequence of 29 aa) was homologous to the glycoprotein B (gB) amino acid sequence of the tupaiid herpesvirus 1 (TuHV-1) (79% identity), which belongs to subfamily Betaherpesvirinae. We designed new consensus-degenerate hybrid oligonucleotide primers (http://blocks.fhcrc.org/codehop.html) selective for the betaherpesvirus gB and DNA polymerase (DPOL) genes, and we determined the complete gB sequence and the partial DPOL sequence of the isolated virus (5,029 bp, DNA Data Bank of Japan accession no. AB517983). BLAST search indicated that the complete gB sequence was novel and most similar to that of TuHV-1 (59% aa sequence identity) (online Appendix Figure, www.cdc.gov/EID/content/ 16/6/986-appF.htm). We named the isolated virus bat betaherpesvirus 2 (BatBHV-2).
We constructed a phylogenetic tree by using the neighbor-joining method with the gB amino acid sequence and the available sequences of known herpesviruses (Figure). The phylogenetic tree based on betaherpesvirus gB genes showed that BatBHV-2 is most closely related to TuHV-1 and caviid herpesvirus 2 (guinea pig cytomegalovirus). The subfamily Betaherpesvirinae consists of the genera Cytomegalovirus, Muromegalovirus, and Roseolovirus. TuHV-2 and caviid herpesvirus 2 are species unassigned to any genus in the subfamily Betaherpesvirinae.
In May 2009, we collected, again with permission, another 50 bats belonging to 1 species, M. fuliginosus, from the same location for an epizootologic study (online Technical Appendix 2, www.cdc.gov/EID/content/16/6/986-Techapp2.pdf). Spleens and blood were collected from all bats, and other organs (liver, kidney, lung, brain, intestine, trachea, and urinary bladder) were collected from 10 bats. Nested PCR was performed by using specifi c primers selective for the DPOL gene of BatBHV-2, and PCR products were subjected to direct sequencing. Viral nucleotide sequences were obtained from 4 of the 50 spleen samples. Each nucleotide sequence showed complete identity to the partial DPOL sequence of the BatBHV-2. Other organs and serum collected from 2 of the bats were also tested by nested PCR, and viral DNA was detected in the liver, kidneys, and lungs of both bats.

Conclusions
Although PCRs with consensus primers effectively detect known and unknown viruses, they failed to detect BatBHV-2, possibly because of minor mismatches between the sequences of BatBHV-2 and the primer sets (TGV, IYG, and KG1). The variety of virus sequences and gene mutations often prevents successful amplifi cation of virus genes. RDV, however, can detect viral cDNA fragments independent of virus species and thus is useful as a fi rstchoice tool for identifying emerging known and unknown viruses in animals and humans.
BLAST search showed that the complete gB sequence of the isolated virus was novel and most similar to that of TuHV-1. Recently, bats have been described as hosts for herpesviruses in several countries in Europe, America, Africa, and Asia (4,9,10). Wibbelt et al. reported that the partial DPOL sequence (175 bp) of a betaherpesvirus, bat betaherpesvirus 1 (BatBHV-1), was obtained from several insectivorous bat species (10). Although the length of the BatBHV-1 sequence was short, similarity between BatB-HV-1 and BatBHV-2 was relatively high (58%). BatBHV-1 is most similar to TuHV-1(61%). These fi ndings suggest that BatBHV-2 is a different species than BatBHV-1.
Our epizootologic study found relatively high (8%) prevalence of BatBHV-2 in insectivorous bats. Although the virus genome was detected in a few parenchymal organs by nested PCR, no amplifi cation was possible for serum, intestine, or urinary bladder samples, which may exclude apparent virus shedding by the bats. In addition, all 50 bats collected appeared clinically healthy. To understand the life cycle of this virus, the possibility of a latent infection in these insectivorous bats must be explored.  4. Direct sequencing. After electrophoresis of PCR products on agarose gels, bands >120 bp were excised, and DNA was extracted from the gel by using MonoFas.
Direct sequencing was performed by using the forward or reverse primer.