Rooting the phylogenetic tree of middle East respiratory syndrome coronavirus by characterization of a conspecific virus from an African bat

Abstract

The emerging Middle East respiratory syndrome coronavirus (MERS-CoV) causes lethal respiratory infections mainly on the Arabian Peninsula. The evolutionary origins of MERS-CoV are unknown. We determined the full genome sequence of a CoV directly from fecal material obtained from a South African Neoromicia capensis bat (NeoCoV). NeoCoV shared essential details of genome architecture with MERS-CoV. Eighty-five percent of the NeoCoV genome was identical to MERS-CoV at the nucleotide level. Based on taxonomic criteria, NeoCoV and MERS-CoV belonged to one viral species. The presence of a genetically divergent S1 subunit within the NeoCoV spike gene indicated that intraspike recombination events may have been involved in the emergence of MERS-CoV. NeoCoV constitutes a sister taxon of MERS-CoV, placing the MERS-CoV root between a recently described virus from African camels and all other viruses. This suggests a higher level of viral diversity in camels than in humans. Together with serologic evidence for widespread MERS-CoV infection in camelids sampled up to 20 years ago in Africa and the Arabian Peninsula, the genetic data indicate that camels act as sources of virus for humans rather than vice versa. The majority of camels on the Arabian Peninsula is imported from the Greater Horn of Africa, where several Neoromicia species occur. The acquisition of MERS-CoV by camels from bats might have taken place in sub-Saharan Africa. Camelids may represent mixing vessels for MERS-CoV and other mammalian CoVs.

Importance: It is unclear how, when, and where the highly pathogenic MERS-CoV emerged. We characterized the full genome of an African bat virus closely related to MERS-CoV and show that human, camel, and bat viruses belong to the same viral species. The bat virus roots the phylogenetic tree of MERS-CoV, providing evidence for an evolution of MERS-CoV in camels that preceded that in humans. The revised tree suggests that humans are infected by camels rather than vice versa. Although MERS-CoV cases occur mainly on the Arabian Peninsula, the data from this study together with serologic and molecular investigations of African camels indicate that the initial host switch from bats may have taken place in Africa. The emergence of MERS-CoV likely involved exchanges of genetic elements between different viral ancestors. These exchanges may have taken place either in bat ancestors or in camels acting as mixing vessels for viruses from different hosts.

FIG 1

Neoromicia capensis bat. The absence of a tiny upper premolar separates it from similarly sized Pipistrellus and Hypsugo bats. The presence of an occipital helmet separates it from Neoromicia zuluensis, the species to which it was assigned based on preliminary morphological criteria.

FIG 2

Genome organization of NeoCoV and sequence identity compared to other clade c betacoronaviruses. (A) Genome organization of NeoCoV. The NeoCoV genome is represented by a black line; ORFs are indicated by gray arrows. The ribosomal frameshift site (RFS) is marked with an arrowhead. The locations of transcription regulatory core sequences (TRSs) following the leader (L) are marked by labeled dots and numbered in their order of appearance from the genomic 5′ terminus. (B) Genomic sequence identity between NeoCoV and other clade c betacoronaviruses. Plots were generated by using SSE version 1.1 (25). The graph representing the comparison of the phylogenetically basal camel virus NRCE-HKU205 and NeoCoV is not shown due to a total overlap in the curve resulting from the comparison between NeoCoV and human MERS-CoV.

FIG 3

Bayesian phylogenies of clade c betacoronaviruses, including NeoCoV. (A) Phylogenies of ORF1a, ORF1b, and ORFs coding for structural proteins. (B) Phylogenies of the S1 and S2 subunits, corresponding to amino acid positions 1 to 747 and 748 to 1353, respectively, of MERS-CoV strain EMC/2012. NeoCoV is shown in red, camel MERS-CoV is shown in blue, and human MERS-CoV is shown in cyan. HCoV-OC43 was used as an outgroup. (C) Phylogeny of MERS-CoV full genomes. MERS-CoVs obtained from humans are shown in black, and MERS-CoVs from camels are shown in blue. NeoCoV was used for rooting the tree. For all trees, statistical support of grouping from Bayesian posterior probabilities is shown at deep nodes. Only values above 0.7 are shown. The bar represents genetic distance. GenBank accession numbers are KJ477102 for NRCE-HKU205, KJ156881 for Wadi-Ad-Dawasir 1 2013, JX869059 for EMC/2012, KJ650296 for KFU-HKU19D, KC776174 for Jordan-N3/2012, KJ650297 for KFU-HKU1, KJ156910 for Hafr-Al-Batin2 2013, KF600613 for Riyadh 3 2013, KF186567 for Al-Hasa 1 2013, KC164505 for England1, KF961221 for Qatar3, KJ713299 for KSA-CAMEL-376, KJ156949 for Taif1 2013, KJ556336 for Jeddah1 2013, KJ713297 for KSA-CAMEL-503, KJ713295 for KSA-CAMEL-505, KF192507 for Munich 2013, KJ650098 for Qatar 2 2014, KF745068 for FRA/UAE, KF600630 for Buraidah1 2013, KJ650295 for KFU-HKU13, KF600628 for Hafr-Al-Batin1 2013, KJ713298 for KSA-CAMEL-363, KJ713296 for KSA-CAMEL-378, KF600620 for Bisha1 2012, KC869678 for NeoCoV, NC_005147 for HCoV-OC43, EF065512 for HKU5-5, NC_009020 for HKU5-1, NC_008315 for BtCoV/133, NC_009019 for HKU4-1, KC545386 for EriCoV/2012-216, KC545383 for EriCoV/2012-174, and KM027259 for Jeddah 2014 C9055.