Evolutionary trends of European bat lyssavirus type 2 including genetic characterization of Finnish strains of human and bat origin 24 years apart

Rabies is a fatal and incurable zoonotic disease caused by RNA viruses belonging to the genus Lyssavirus within the family Rhabdoviridae. The negative-sense lyssavirus genome encodes five proteins: nucleoprotein (N), phosphoprotein (P), matrix protein (M), glycoprotein (G) and RNA polymerase (L) in the order 3′-N-P-M-G-L-5′. Phylogenetic analysis and the virus-host relationship suggest that all 15 currently known lyssaviruses probably originated in bats and can be divided into phylogroups. Phylogroup I comprises the classical rabies virus (RABV) and the majority of bat lyssaviruses, whereas Lagos bat virus (LBV) and Shimoni bat virus (SHIBV) form phylogroup II [1, 2]. West Caucasian bat virus (WCBV), Ikoma lyssavirus (IKOV) and Lleida bat lyssavirus (LLEBV) may be representatives of a possible new phylogroups [3‐5].

The first record of a rabid bat in Europe occurred in Germany in 1954 [6], but only with the advent of antigenic typing and molecular tools could European bat lyssavirus (EBLV) be characterised and distinguished from classical RABV and other lyssaviruses [7, 8]. After the death of a Swiss bat biologist in Finland in 1985 and subsequent characterisation of the causative agent, it became evident that bat rabies in Europe at the time was caused by two different lyssaviruses, EBLV-1 and 2 [7, 9]. Both of these viruses are genetically and antigenically related to RABV but are significantly different from each other. Early genomic sequencing also indicated that two distinct genetic lineages (a and b) of EBLVs have evolved, which appeared to cluster geographically [10, 11].

During 1977–2013, 1064 rabies cases were reported in 11 of the 45 known indigenous bat species in 16 European countries [12]. Although numerous human contacts with European bats, primarily from handling sick or injured animals, have been reported [13], only a few EBLV-induced human casualties have been conclusively demonstrated: in Voroshilovgrad, Ukraine (1977), in Belgorod, Russia (1985), in Finland (1985) and in the UK in Scotland (2002) [9, 14‐17]. The Russian case was genetically typed as being EBLV-1a, while the Ukrainian case is only assumed to be EBLV-1 from antigenic profiling [15, 18]. The virus variants responsible for human rabies cases in the UK and Finland were typed as being EBLV-2a and b, respectively [17]. Subsequently, two other human rabies cases after a bat bite have been described in Ukraine [15]. However, these cases were diagnosed clinically and on the basis of anamnesis, and no virological or pathological results are available.

EBLV-2 is known to associate with two closely related Myotis bat species, Daubenton’s (Myotis daubentonii) and pond bats (M. dascyneme). The first isolation of EBLV-2 from a bat was carried out in 1987 from a pond bat in the Netherlands [19]. Elsewhere in Europe, EBLV-2 has been isolated sporadically from Daubenton’s bats in Switzerland [20], the UK [21, 22], Germany [23] and Finland [24]. In 1986, EBLV was detected in one Danish Daubenton’s bat and in one pond bat, but the causative strains were not analysed further. EBLV-2 was, however, demonstrated later in 2013 in mouth swabs from two Danish Daubenton’s bats using a molecular diagnostic strategy [25]. The strains from the UK and from the Netherlands appear to be closely related, as are the strains from Germany and those isolated from Switzerland [26]. The Finnish EBLV-2 strains and a strain from Switzerland have been suggested to form a lineage designated as EBLV-2b, with rest of the isolates forming the lineage EBLV-2a [10, 11]. EBLV-2 strains appear to form clusters according to the geographical area and year of isolation.

Evolutionary studies on lyssaviruses have tended to focus on the N protein, a conserved structural protein, and the G protein, which contains domains responsible for host-cell receptor recognition and membrane fusion, and which is the major target for the host neutralizing antibody response [1]. Previously, comprehensive evolutionary analysis has demonstrated that EBLV-1 and EBLV-2 have differing population structures and dispersal patterns. Molecular-clock estimates have suggested that the current linage of EBLV-1 arose some 500 to 750 years ago [27]. Little is known about the forces shaping the evolution of EBLV-2, and only three strains of EBLV-2 have previously been fully sequenced (EU293114, EF157977 and KF155004).

Finland has been free of rabies in terrestrial mammals since 1991. Nevertheless, rabies can remain a residual risk to public health due to the natural circulation of EBLV-2 [24, 28]. Here, we describe the phylogenetic analysis of two Finnish EBLV-2 strains, the first isolated from a human in 1985 and the second from a diseased bat in 2009. Using these isolates as a calibrator, this study provides an estimate of the EBLV-2 molecular clock. Increased information on the genome sequences of lyssaviruses is fundamental to understanding their epidemiology and evolution, and to demonstrating the importance of continuous surveillance and molecular characterization of lyssaviruses circulating in animal populations.

Sequencing, phylogenetic analysis and evolutionary analysis of Finnish EBLV-2 strains isolated from a human case in 1985 and from a bat in 2009 were described. It was shown that the EBLV-2 isolates share high nucleotide and amino acid sequence identity, regardless of the year and origin of isolation, but they also cluster according to the host species and the geographical place of isolation.

The closest relatives to EBLV-2 (Bokeloh and KHUV) are lyssavirus strains from Myotis bats from Germany and Tajikistan. An interesting case is ARAV (originating from Kyrgystan), which has an uncertain position in the phylogenetic analysis. The possibility of recombination of rabies viruses has been demonstrated in earlier studies [34‐36]. Lleida bat lyssavirus, detected in a bent-winged bat (M. schreibersii) in Spain in 2011 [5], is the most unique lyssavirus in Europe based on the N gene, although the whole genome sequence is not available. WCBLV has also been detected in M. schreibersii, and antibodies against WCBV have been detected in Africa in Miniopterus bats [37]. The tree topology was somewhat different in the phylogeny based on either the complete coding sequences or the commonly used partial N sequences, and it should be appreciated that for a detailed analysis of evolutionary history, complete genomes should be used. The limitation of this study, as in other studies, is the amount of data available for comparison.

The nucleotide sequence identity (deduced amino acid sequence identity) between FI-85 and FI-09 was 99.6 % (99.8 %). The N, M and L proteins were similar in structure and length between members of different lyssavirus species and strains, whereas the lengths of the P and cytoplasmic regions of the G protein were variable [38‐40]. No differences in structure were seen in the Finnish EBLV-2 isolates in comparison to other EBLV-2 complete genomes [40‐42]. The similarity plot in Figure 1 further demonstrates the similarities along the genomes of different lyssaviruses. Interestingly, the non-coding regions, which are highly divergent in other lyssaviruses, were rather similar for all EBLV-2 strains.

Based on this study, we estimate that the Finnish EBLV-2 strains and one Swiss strain from among the other EBLV-2 strains have evolved from a common ancestor during the last 1000 years. Earlier, the Finnish strains and the strain from Switzerland were suggested to form the lineage EBLV-2b, whereas the rest of the isolates form the lineage EBLV-2a [10, 11, 20, 27]. In our study, there was a division into subgroups EBLV-2a and EBLV-2b when full genomes were compared (Fig. 2), but when more strains were compared using shorter, 400-bp N gene sequences, we could not confirm the suggested division into subgroups (Fig. 3). In our comparison, the Finnish strains and the strain from Switzerland also clustered together based on N gene analysis. Interpretation of the molecular epidemiology of the strains is further complicated by the history of the Swiss bat biologist who died in Finland in 1985 of EBLV-2 infection (JX129233), who had been bitten 51 days before the onset of clinical signs by a Daubenton’s bat that was abnormal. The bat was freed before the patient developed symptoms and was not available for rabies analysis [16]. However, it is possible that the bat originated from Central Europe, possibly from Switzerland, even though the exposure took place in Finland.

The molecular clock estimates suggest that the current linage of EBLV-2 arose some 2000 years ago and that the most recent common ancestor (MRCA) occurred around 8000 years ago. In earlier molecular clock studies, Tao et al. [43] estimated based on G gene analysis that the time to the most common ancestor (TMRCA) of all lyssaviruses was approximately 5030 years. The West Caucasian bat virus divided first, and then phylogroup I and phylogroup II divided about 4000 years ago [43]. Davis et al. [27] estimated that the MRCA of EBLV-1 existed around 500 to 750 years ago based on G and N sequences, or 70 to 300 years ago, depending on the substitution rate [44]. The molecular clock estimates cannot be used to precisely describe events that occurred thousands, or even tens of thousands of years ago, since sequence data only exist from the last few decades. The overall rate of evolution appears somewhat slower for the bat-type lyssaviruses than for RABV. This is in line with bats being considered the true reservoir of lyssaviruses. The overall evolution rate for EBLV-2 in our study was 7.67 × 10⁻⁵ substitutions per site per year. For RABV, it has been suggested to be 1.56–1.78 × 10⁻⁴ [45], and 1 × 10⁻³ to 5.5 × 10⁻⁴ [7, 43, 46, 47], and for EBLV-1, 5–6 × 10⁻⁵ to 5–6 × 10⁻⁴ [27].

Behavioural and ecological traits of bats, such as hibernation, migration and coloniality, might influence viral evolution. Some rabies virus lineages evolve up to 22 times faster than others, depending on the reservoir host [48, 49]. Virus lineages in bat species in the temperate climate zone were found to have four times slower rates of evolution than lineages in bats from tropical or subtropical climates, and viral evolution was faster in bats that remain active year-round in comparison to bats that hibernate. This could be a major factor in a cold climate region such in Finland. Viral evolutionary rates were similarly unconstrained by host evolutionary relatedness such that lyssaviruses associated with closely related bat species or subspecies often had dissimilar evolutionary rates. Streicker et al. [48] demonstrated that the local host environment determines the evolutionary rates of lyssaviruses. Daubenton’s bats are facultatively seasonal migrants covering middle-range distances of 100–150 km [50], which supports the clustering of EBLV-2 according to the geographical place of isolation. In Finland, antibodies to EBLVs have only been detected in a restricted geographical area close to the location where the EBLV-2-positive bat was found in 2009 [28]. In a study of Daubenton’s bats from Finland, the UK, Switzerland and Spain, no population structuring was observed [51]. Daubenton’s bats migrate between the UK and the mainland, and the genetic structure of Daubenton’s bats is relatively homogeneous in western parts of Europe [52]. This could explain the relatedness within the proposed lineage 2a. As the migration pattern of Daubenton’s bats does not support the finding that the Finnish EBLV-2 Daubenton’s bat strain is closely related to the EBLV-2 strain from Switzerland, one can speculate whether there could have been contact between indigenous Finnish bats and the bats that were freed in Finland and whose origin is unknown [9, 16].

Increased information on the complete genome sequences of lyssaviruses is fundamental to understanding their epidemiology and evolution and demonstrates the importance of continuous monitoring and molecular characterization of lyssaviruses circulating in human and animal populations.