In July 2011, a bat was found in the City of Lleida and taken to the Wildlife Care Center of Vallcalent (Lleida, Catalonia). The bat arrived lethargic and dehydrated, died soon after admission, and its carcass was frozen at −20°C. On March 12, 2012, as part of the rabies surveillance program in Spain, the bat carcass was received by the National Center of Microbiology, where rabies testing was conducted by 2 generic reverse transcription PCR (RT-PCR) methods for lyssavirus detection (9,10) and 2 commercial rabies antiserum assays (Bio-Rad Laboratories, Marnes La Coquette, France; and Fujirebio, Inc., Tokyo, Japan) for antigen detection by fluorescent antibody testing.

Brain smears were positive for lyssavirus by RT-PCR and fluorescent antibody testing, and an oropharyngeal swab sample was positive by RT-PCR. Further attempts to isolate the virus by tissue cultures were unsuccessful after 2 blind passages in BHK-21 and murine neuroblastoma cells. The negative results could be explained by the fact that the sample had been stored at −20°C for 8 months and had been frozen and thawed twice before cell culture testing; however, the possibility of the cell lines not being permissive for the virus cannot be excluded.

The bat was morphologically identified as a bent-winged bat (M. schreibersii) and genetically identified by cytochrome b sequencing (11). The genomic sequence of the corresponding fragment of the diagnostic RT-PCR on the conserved region of the nucleoprotein gene, determined by BLAST (http://blast.ncbi.nlm.nih.gov/), showed no substantial sequence similarity to previously known lyssaviruses.

To determine the identity of the lyssavirus, we sequenced a larger fragment (565 bp), including the variable codifying region of the nucleoprotein gene (GenBank accession number submitted). We reconstructed an overall phylogeny of lyssaviruses by using a Bayesian Inference with the first 405 nt of the N-gene and MrBayes version 3.1.2 (http://mrbayes.csit.fsu.edu/). Two simultaneous runs of 10⁶ generations were conducted, each with 4 Markov chains, and trees were sampled every 100 generations. The best-fit nucleotide model, GTR + I + G, available in MrBayes was selected according to the corrected Akaike information criterion. The phylogenetic reconstruction was based on a dataset representative of all known lyssaviruses, including the recently described IKOV. The topology obtained showed that this sequence is more closely related to IKOV and WCBV than to the lyssaviruses in phylogroups I and II (Figure). These results suggest that this sequence tentatively belongs to a new Lyssavirus species named after the location of collection, Lleida bat lyssavirus (LLEBV).

The highest nucleotide identity was with IKOV (71.6%), followed by SHIBV (68.6%), IRKV (68.1%), KHUV (67.6%), EBLV-2 (67%–68.2%), ARAV (67.3%), WCBV (67.4%), ABLV (66.6%–67.7%), BBLV (66.1%), LBV (65.7%–68.6%), MOKV (65.7%–67.2%), DUVV (65.5%–65.8%), RABV (64.7–66.4%), and EBLV-1 (63.7%–64%). The lowest nucleotide identity was with the only lyssavirus found in bats of the Iberian Peninsula, EBLV-1. The nucleotide identity among the previously known lyssaviruses was 63.5%–80.0% in this particular fragment, and the lowest identities among strains belonging to the same lyssavirus were 80.4% for ABLV and 79.9% for LBV (the most distant LBV strain has been suggested to be a different lyssavirus) (Technical Appendix).

The lyssavirus-specific antigen reactivity and association with a genomic sequence found in a bent-winged bat in northeastern Spain could be derived from the tentative new virus LLEBV. According to our phylogenetic reconstruction, the virus does not seem to belong to phylogroup I, which comprises most bat lyssaviruses, or to the African phylogroup II. The evolutionary relationships between the LLEBV sequence with WCBV and IKOV sequences need to be clarified before it can be determined whether they form >1 different phylogroups.

Of note, the new LLEBV was detected in M. schreibersii bats, as was WCBV, the other European lyssavirus outside phylogroup I. The genus Miniopterus has traditionally been considered to belong to the family Vespertilionidae as do other bat genera linked to lyssaviruses in Eurasia (Eptesicus, Myotis, and Murina). However, recent molecular analyses have confirmed that the genus Miniopterus belongs to the family Miniopteridae (12). M. schreibersii bats are migratory, widely distributed across southern Europe and Eurasia. Large numbers (thousands) of these bats overwinter in caves and move in the spring to different and sometimes distant summer roosts for reproduction (13). These ecologic features make it relatively easy for an infectious agent to quickly spread out within and among the populations.

Consequently, it is difficult to imagine that WCBV or LLEBV are locally restricted; both could be located far from where they were found. Neutralizing antibodies against WCBV have been found in bats in Africa (14). The cumulative description of new bat lyssaviruses in recent years shows the convenience of always using generic amplification primers for rabies diagnosis based on RT-PCR to complement antigen detection.

No human exposure to the new virus has been reported. However, because of the divergence exhibited by LLEBV and IKOV, and the growing evidence of inadequate protection/cross-neutralization against viruses outside phylogroup I, the effectiveness of current rabies vaccines remains for these viruses a concern (15).

Table of nucleotide identities (upper diagonal) and similarities (lower diagonal). Analysis was performed with the 405-nt fragment of the N-gene of all the known Lyssavirus species.