Epidemiology and biology of a herpesvirus in rabies endemic vampire bat populations

Abstract

Rabies is a viral zoonosis transmitted by vampire bats across Latin America. Substantial public health and agricultural burdens remain, despite decades of bats culls and livestock vaccinations. Virally vectored vaccines that spread autonomously through bat populations are a theoretically appealing solution to managing rabies in its reservoir host. We investigate the biological and epidemiological suitability of a vampire bat betaherpesvirus (DrBHV) to act as a vaccine vector. In 25 sites across Peru with serological and/or molecular evidence of rabies circulation, DrBHV infects 80-100% of bats, suggesting potential for high population-level vaccine coverage. Phylogenetic analysis reveals host specificity within neotropical bats, limiting risks to non-target species. Finally, deep sequencing illustrates DrBHV super-infections in individual bats, implying that DrBHV-vectored vaccines might invade despite the highly prevalent wild-type virus. These results indicate DrBHV as a promising candidate vector for a transmissible rabies vaccine, and provide a framework to discover and evaluate candidate viral vectors for vaccines against bat-borne zoonoses.

Fig. 1. Widespread BHV infection across the bat phylogeny.

The bat phylogeny (left) was extracted from the mammalian super-tree to contain species that were tested here or elsewhere for BHV infection^,. Labels on tree branches indicate bat families. Taxa in black text indicate bat species tested here for the first time. Stacked bar charts (right) indicate the number of saliva samples that tested either positive or negative for BHV by PCR of the terminase gene UL89. A binomial generalised linear mixed model (GLMM) followed by post-hoc pairwise comparisons of estimated prevalence by species (multiple comparisons of means: Tukey contrasts, two tailed) was used to test the significance of differences in prevalence. The pairwise significance of prevalence in each species compared to D. rotundus is shown on the bar chart with corrected P-values: > 0.1, * > 0.05, ** > 0.01, *** > 0.005. All other pairwise comparisons were non-significant. D. rotundus is indicated with a red arrow. Source data are provided as a Source Data file.

Fig. 2. Seroprevalence of VBRV in D. rotundus and geographic distribution of outbreaks in cattle.

a–f The seroprevalence of VBRV in D. rotundus, grouped by year and department within Peru, are shown in the surrounding plots (a–f), with the location of each department marked by the corresponding letter on the map. Seroprevalence within a department is displayed as the mean average with 95% confidence intervals based upon sample size. Sample sizes for A-F in ascending year order are as follows: A-5, 21; B-144, 126; C-158, 72, 116; D-107, 181, 218; E-146, 167, 78; F-45, 24, 111. g Locations of laboratory confirmed rabies mortality in livestock between 2015 and 2017 are shown coloured by year (2015=red, 2016=yellow, 2017=blue) on the map of Peru. The location from which the PCR positive bat saliva sample originated is shown by the black diamond. Source data are provided as a Source Data file.

Fig. 3. Bayesian phylogenetic tree of mammalian BHV shows clustering of virus diversity by host taxonomy.

Nucleotide-based phylogeny of UL89 sequences (304 bp) showing clusters of BHVs that were generally restricted to single species. Posterior probabilities are represented by node colour. Tip labels are coloured according to host taxonomy (family level for orders Chiroptera and Rodentia, and order level for Primates) and include the host species in which the BHV was detected and GenBank accession number where appropriate. The inset amino acid sequence-based phylogeny recovers the monophyly of bat BHVs and indicates bat family specific clustering of viral diversity, apart from a BHV from Di. ecaudata which groups separately from all other BHVs from Phyllostomid bats. Clades are collapsed into bat families, primate and rodent groups, represented by the diamond tips, proportional to the size of the collapsed clade. Individuals are represented by circular tips. Colours and family/order abbreviations correspond to those in the main figure. Both datasets include human herpesvirus 6A (HHV6, genus Roseolovirus) as an outgroup.

Fig. 4. Co-phylogenetic analysis of hosts and betaherpesviruses reveal co-evolution across mammals.

Co-phylogeny of host species (phylogeny extracted from the mammalian super tree) (left) with BHV, based on viral nucleotide (a) and amino acid (b) sequences. The lines between trees show links between virus and host phylogenies and are coloured by host family. Link weight is inversely proportional to the PACo squared residual for each species; heavier lines therefore reflect the likelihood that a pair represents a co-evolutionary link.

Fig. 5. Metagenomic sequence data reveal multi-strain DrBHV infections in individual bats.

The phylogenies (left), nucleotide sequences (centre), and read count (right) of DrBHV strains present in the alignment of samples 10148_KF29 and 10144_KF29 (where read depth was sufficient for reads to be included) with the DrBHV consensus genome sequence. Nucleotide locations: (a) 47620– 47671 (equivalent to part of Miniopterus schreibersii BHV B7D8 B36), (b) 121057–121108 (B70), (c) 141567–141618 (B85) and (d) 229640–229691 (B161). These locations were selected due to high coverage depth in sample 10148_KF29. The tree tips are coloured by the individual bat from which the sequence originates.