Presence of Apis Rhabdovirus-1 in Populations of Pollinators and Their Parasites from Two Continents

Sofia Levin, David Galbraith, Noa Sela, Tal Erez, Christina M. Grozinger, Nor Chejanovsky

The viral ecology of bee communities is complex, where viruses are readily shared among co-foraging bee species. Additionally, in honey bees (Apis mellifera), many viruses are transmitted - and their impacts exacerbated - by the parasitic Varroa destructor mite. Thus far, the viruses found to be shared across bee species and transmitted by V. destructor mites are positive-sense single-stranded RNA viruses. Recently, a negative-sense RNA enveloped virus, Apis rhabdovirus-1 (ARV-1), was found in A. mellifera honey bees in Africa, Europe, and islands in the Pacific. Here, we describe the identification - using a metagenomics approach - of ARV-1 in two bee species (A. mellifera and Bombus impatiens) and in V. destructor mites from populations collected in the United States and Israel. We confirmed the presence of ARV-1 in pools of A. mellifera, B. impatiens, and V. destructor from Israeli and U.S. populations by RT-PCR and found that it can reach high titers in individual honey bees and mites (107-108 viral genomic copies per individual). To estimate the prevalence of ARV-1 in honey bee populations, we screened 104 honey bee colonies across Israel, with 21 testing ARV-1-positive. Tagged-primer-mediated RT-PCR analysis detected the presence of the positive-sense ARV-1 RNA in A. mellifera and V. destructor, indicating that ARV-1 replicates in both hosts. This is the first report of the presence of ARV-1 in B. impatiens and of the replication of a rhabdovirus in A. mellifera and V. destructor. Our data suggest that Varroa mites could act as an ARV-1 vector; however, the presence of ARV-1 in B. impatiens (which are not parasitized by Varroa) suggests that it may not require the mite for transmission and ARV-1 may be shared among co-foraging bee species. Given that ARV-1 is found in non-Apis bee species, and because "ARV" is used for the Adelaide River virus, we propose that this virus should be called bee rhabdovirus 1 and abbreviated BRV-1. These results greatly expand our understanding of the diversity of viruses that can infect bee communities, though further analysis is required to determine how infection with this virus impacts these different hosts.