Discovery of Novel Alphacoronaviruses in European Rodents and Shrews

Tsoleridis, Theocharis; Onianwa, Okechukwu; Horncastle, Emma; Dayman, Emma; Zhu, Miaoran; Danjittrong, Taechasit; Wachtl, Marta; Behnke, Jerzy M.; Chapman, Sarah; Strong, Victoria; Dobbs, Phillipa; Ball, Jonathan K.; Tarlinton, Rachael E.; McClure, C. Patrick

Eleven of the 813 rodents and shrews tested were found to be positive for CoVs (Table 1) . Five positives were found in Rattus norvegicus, three in Microtus agrestis, one in Sorex araneus and two in Myodes glareolus. The phylogenetic analysis ( Figure 2 ) showed that the new rodent/shrew CoVs, together with the Lucheng CoV reference strain formed a single clade (with 99% bootstrap support) in the Alphacoronavirus genus. Within this cluster one lineage contained a single sequence that was present in five separate samples (UKRn1-5) obtained from Rattus norvegicus, which were most related to the Lucheng CoV (100% bootstrap support). A separate cluster with 100% bootstrap support contained three sequences obtained from Microtus agrestis (UKMa1&2) and Myodes glareolus (UKMg1). A third branch contained a sequence obtained from Myodes glareolus (PLMg1) living in the Mazury lakes region of Poland and a further distinct sequence obtained from Sorex araneus (UKSa1) and Microtus agrestis (UKMa3) (99% bootstrap support). Detailed analysis of the sample archive, including details of the day and method of capture revealed that samples UKSa1 and UKMa3, which were identical across the region of ORF1a1b analysed, were not only obtained from the same geographical area but also captured by the same predator (Felis catus) on the same day. This was also true for samples UKMa2 and UKMg1. Therefore, we are left with three plausible explanations for the CoV phylogeny observed. Either the same or similar rodent coronaviruses were co-circulating and freely infecting different rodent species living in the same geographical area at the time of sampling; cross-species transmission of coronaviruses has been described previously [15] [16] [17] [18] [19] . An alternative explanation is that these carcasses were cross-contaminated during or after predation with coronavirus present in either animal or through predator-contamination from a different infected animal altogether that was not sampled. We do not consider that this result was due to contamination during tissue processing as new sterile dissection equipment was used for each carcass and this effect was observed between animals that were not processed consecutively. The cytochrome b sequence data ( Figure 3 ) clearly showed that the specimens were of different species (arguing against sample mix-up). Importantly, re-extraction of the original rodent tissues gave identical results (data not shown), which strongly suggests that contamination did not occur during nucleic acid extraction or PCR preparation. Finally, it is also possible that these animals were infected with distinct strains of coronavirus but the PCR primers, which were designed to target highly conserved regions of the ORF1ab gene, were unable to discriminate between highly related viruses. The inability of ORF1ab PCR sequencing to resolve different isolates was evident in recent analyses of bat CoVs [20, 21] . Sequencing additional regions of the genome, as has been performed previously with bat CoVs [22] , together with analysis of larger numbers of rodents, will help elucidate the overall diversity and distribution of alphacoronaviruses in European rodents and shrews. Interestingly, the samples UKSa1, UKMa2&3 and UKMg1 were killed by the same predator in a relatively limited sample set (Site 1, Figure 1 ), suggesting coronavirus pathogenesis may have increased susceptibility to predation in these animals. Whether or not coronavirus infection migh