Detection and characterization of a novel bat-borne coronavirus in Singapore using multiple molecular approaches

Abstract

Bats are important reservoirs and vectors in the transmission of emerging infectious diseases. Many highly pathogenic viruses such as SARS-CoV and rabies-related lyssaviruses have crossed species barriers to infect humans and other animals. In this study we monitored the major roost sites of bats in Singapore, and performed surveillance for zoonotic pathogens in these bats. Screening of guano samples collected during the survey uncovered a bat coronavirus (Betacoronavirus) in Cynopterus brachyotis, commonly known as the lesser dog-faced fruit bat. Using a capture-enrichment sequencing platform, the full-length genome of the bat CoV was sequenced and found to be closely related to the bat coronavirus HKU9 species found in Leschenault's rousette discovered in the Guangdong and Yunnan provinces.

Keywords: Betacoronavirus; Singapore; bats; capture enrichment.

Fig. 1.

Geographical distribution of bat roosts in Singapore. Locations surveyed for the presence of bats include Bedok Town Park, Bukit Brown, Bukit Timah Caves, Clementi Woods, East Coast Park, Hort Park, Hougang, Lower Peirce Reservoir, Old Changi Hospital, Pasir Ris Park, Rifle Range Flyover, Singapore Botanic Garden, Singapore Zoological Garden and West Coast Park. Singapore basemaps available from the Singapore Land Authority at www.oneMap.sg.

Fig. 2.

Genome analysis of BtCov92. Genome organization of BtCoV92 (in bold) and a representative coronavirus from Alphacoronavirus (porcine epidemic diarrhoea virus, human CoV-299E and swine acute diarrhoea syndrome coronavirus), Betacoronavirus (murine coronavirus, severe acute respiratory syndrome-related coronavirus, Middle East respiratory syndrome-related coronavirus, Tylonycteris bat coronavirus HKU4 and Rousettus bat coronavirus HKU9) and Gammacoronavirus (avian coronavirus). ORF 1ab (orf1ab), the gene for haemagglutinin esterase (HE), the spike protein (S), the envelope protein (E), the membrane protein (M) and the nucleocapsid protein (N) are represented by grey boxes. The genes for nonstructural proteins are represented by white boxes.

Fig. 3.

Phylogenetic tree of the full-genome sequences of coronavirus. Multiple sequence alignment of the BtCoV92 coronavirus sequence and 2211 sequences present in NCBI database was carried out using a fast Fourier transformation method in MAFFT v6.940b. An approximately maximum-likelihood phylogenetic tree was generated using the generalized time-reversible model of nucleotide evolution in FastTree v2.1.7. FastTree uses SH-like local supports with 1000 resamples to estimate and validate the reliability of each split in the tree. From the tree, the branch containing the sequence from the Bt92 coronavirus sequence and 58 sequences from NCBI was selected and a more robust maximum-likelihood phylogenetic tree was created using RAXML with 1000 bootstrap replications. The trees were visualized using FigTree v1.4.2.

Fig. 4.

Phylogenetic tree of amino acids sequences encoding the coronavirus nucleocapsid protein (N). A partial N protein sequence of BtCoV22/29 and the complete N protein sequence of BtCoV92 were compared to the representative coronavirus species based on (a) the available partial N protein sequence of a betacoronavirus (KX452687) previously detected in E. spelaea and (b) full N protein sequences. Maximum-likelihood trees were constructed using the LG+F amino acid model (best-fit model determined by mega7) with 1000 bootstrap replications. Only bootstrap values >80 % are shown.