Metagenomic Analysis Reveals Previously Undescribed Bat Coronavirus Strains in Eswatini

We investigated the prevalence of coronaviruses in 44 bats from four families in northeastern Eswatini using high-throughput sequencing of fecal samples. We found evidence of coronaviruses in 18% of the bats. We recovered full or near-full-length genomes from two bat species: Chaerephon pumilus and Afronycteris nana, as well as additional coronavirus genome fragments from C. pumilus, Epomophorus wahlbergi, Mops condylurus, and Scotophilus dinganii. All bats from which we detected coronaviruses were captured leaving buildings or near human settlements, demonstrating the importance of continued surveillance of coronaviruses in bats to better understand the prevalence, diversity, and potential risks for spillover. Supplementary Information The online version contains supplementary material available at 10.1007/s10393-021-01567-3.


INTRODUCTION
Coronaviruses are a family of zoonotic viruses comprised of four genera, two of which, alpha-and betacoronaviruses, have an evolutionary origin in bats, while gamma-and deltacoronaviruses, originate in birds (Graham et al. 2013).
Coronaviruses have since radiated to a variety of hosts (Drexler et al. 2014). Notably, in humans, coronaviruses have caused COVID-19 (Zhou et al. 2020;Gorbalenya et al. 2020), Severe Acute Respiratory Syndrome (SARS) (Marra et al. 2003;Li et al. 2005), and Middle East Respiratory Syndrome (MERS) . While recent studies have increased our knowledge of coronavirus diversity and ecology, large gaps in sampling mean there are probably still many undiscovered species and strains in bats (Anthony et al. 2013(Anthony et al. , 2017. Southern Africa has a diverse bat community (Monadjem et al. 2020b) that appears to host many coronaviruses, including strains phylogenetically close to MERS-CoV (Geldenhuys et al. 2013(Geldenhuys et al. , 2018Ithete et al. 2013), although studies are still limited (Markotter et al. 2020). Globally, the diversity and distribution of coronaviruses in bats makes it likely that future transmission of these pathogens to humans or other animal species will occur (Woo et al. 2009;Anthony et al. 2017). Although there are no known cases of coronavirus spillover in Africa thus far (Markotter et al. 2020), this could occur where bat species come into frequent, close contact with humans or domestic animals (Monadjem 1998;Fenton et al. 2004;Jacobs and Barclay 2009;Noer et al. 2012;Monadjem et al. 2020b).
Therefore, we investigated the prevalence of coronaviruses in bats belonging to eight species from four families (Pteropodidae: Epomophorus wahlbergi; Emballonuridae: Taphozous mauritianus; Molossidae: Chaerephon pumilus, Mops condylurus, and Mops midas; and Vespertilionidae: Afronycteris nana, Scotophilus dinganii, and Scotophilus viridis). These species are all widely distributed and abundant across southeastern Africa and are commonly found in or near human settlements in northeast Eswatini (Monadjem et al. 2020b(Monadjem et al. , 2021Shapiro et al. 2020). We subjected fecal samples to virion enrichment followed by RNA sequencing to noninvasively investigate the prevalence and types of coronavirus in the bats of this region. We used this approach to recover whole coronavirus genomes and thus more reliably characterize them (Drexler et al. 2014;De Sabato et al. 2019). This method also allowed us to detect both known and unknown coronaviruses regardless of the specific sequences or genomic region present in samples.
We captured bats at eight sites in northeast Eswatini ( Fig. 1) from December 2013-May 2014 using mist-nets and/or a harp trap. Taxonomy follows Monadjem et al. (2010Monadjem et al. ( , 2020bMonadjem et al. ( , 2020a. To aid in the identification of species, we measured forearm length of each captured bat with calipers to the nearest 0.1 mm and mass to the nearest 0.5 g with a spring balance. Captured bats were placed individually in cloth holding bags for the deposition of feces. We trapped, handled, and released bats in accordance with a permit from the Eswatini National Trust Commission and University of Florida Institutional Animal Care and Use Committee approval (Protocol #201,508,751).
Fecal samples from insectivorous species were desiccated and preserved with silica gel (Sigma-Aldrich), while samples from the frugivorous Epomophorus wahlbergi were placed in RNALater (Thermo Fisher Scientific) because due to their wet condition they could not be properly dried. Samples were stored at -10°C until the end of the field season (May 2014), then transferred to -80°C.
Frozen fecal samples were vortexed vigorously in 600 ll of PBS with beads from the PowerFecal kit (MoBio) for 1 min and incubated at room temperature for 10 min. Following incubation, samples were vortexed for 1 min, then centrifuged at 2500 9 g for 3 min. The supernatant was then filtered and the flow-through nuclease-treated following Jensen et al. (2015). Viral nucleic acids were subsequently extracted using Roche High Pure Viral RNA kit (Roche) according to the manufacturer's guidelines after which 1 ll RNase Out (Invitrogen) was added to the final RNA extract (Jensen et al. 2015;Hansen et al. 2015).
Forty-four RNA libraries were produced, each one from an individual fecal sample, using ScriptSeq v2 RNAseq library preparation kit (Epicentre, Illumina), according to the manufacturer's guidelines. Samples were DNasetreated with Promega DNase for 30 min at 37°C and purified on RNeasy MinElute columns (Qiagen). Seven or eight individually and uniquely single-indexed sequencing libraries were pooled together in equimolar ratios for sequencing with paired-end reads of 100 bp (PE100) on an Illumina Hiseq 2000 platform. The library from one sample  Table 2. Sites from which coronaviruses were detected in bats are marked in red, while coronaviruses were not detected in bats captured from sites marked in black. The area shaded in gray is Hlane National Park. Solid lines indicate national borders and dotted lines indicate roads.
(Bat50) was resequenced individually on one lane of PE100 on an Illumina Hiseq 2000 platform.
Reads with overlapping sections of sequences were assembled into longer contiguous sequences (contigs) using Ray Meta v2.2.0 with default settings (Boisvert et al. 2012). The contigs were searched for coronaviruses using megablast and BLASTn on the NCBI Nucleotide collection (nt) database (Altschul et al. 1990(Altschul et al. , 1997 and by mapping against NCBI's nr database using DIAMOND (Buchfink et al. 2014).
De novo assembly with an alternative assembler was attempted on the eight coronavirus-positive samples using MEGAHIT v1.1.1 (Li et al. 2015) with the following parameters: minimum contig length = 100, minimum kmer size = 15, maximum kmer size = 101, increment of kmer size of each iteration = 2. To search for potential coronavirus genomes, the 20 longest contigs from each assembly were selected and analyzed using BLASTn on the nt/nr databases, which resulted in the identification of longer coronavirus contigs spanning and extending shorter contigs already identified. Further assembly was attempted on the combined set of contigs using Geneious v.11 software (https://www.geneious.com/), resulting in full or near-full genomes for four bats. Reads were mapped back to the genomes using bowtie2 (Langmead and Salzberg 2012) to correct ambiguous bases.
We also mapped all the sequenced reads from individual samples back to the coronavirus contigs using bowtie2 (v2.2.9) (Langmead and Salzberg 2012). We did this in order to confirm which samples the sequences came from and identify any potential cases of bleed over (the misidentification of the sample from which each sequence read originated) following Kircher et al. (2012) and Jensen et al. (2015).
We identified three full-length and one partial alphacoronavirus genomes from four individual bats (accession numbers OL807608, OL807609, OL807610, OL807611; Supplementary File 1). Three of these were isolated from the species Chaerephon pumilus: two of the full genomes (from Bat143 and Bat151; 27,956 nt and 28,061 nt respectively) and one partial genome (Bat180; 20,826 nt). The best hit using BLASTn for all three of these coronavirus genomes was Chaerephon bat coronavirus/Kenya/KY22/ 2006 from Kenya (Tong et al. 2009). When aligned in Geneious, all three were 86-87% identical to this species. Pairwise identity for the ORF1ab gene was 97.1-97.2%, indicating these coronaviruses likely belong to the same species as Chaerephon bat coronavirus/Kenya/KY22/2006 based on the coronavirus species demarcation criterion of the International Committee on Taxonomy of Viruses (Lefkowitz et al. 2018;ICTV 2019). In a bootstrapped maximum likelihood tree using RAxML based on full coronavirus genomes following De Sabato et al. (2019), all three Chaerephon pumilus coronavirus genomes clustered together as a sister clade to Chaerephon bat coronavirus/ Kenya/KY22/2006 (Fig. 2). This coronavirus may be widespread within the bat genus Chaerephon across Africa. Other coronaviruses have been found in bats of the same species or genera that are geographically distant, sometimes across continents (Drexler et al. 2014) and could indicate connectivity between bat populations across their distribution.
When aligned to each other in Geneious, the three Chaerephon pumilus coronavirus genomes were 98.8% identical. The full genomes from Bat143 and Bat151 were slightly more similar to each other (99.4%) than to the partial genome from Bat180 (98.3 -98.6%). All three genomes were confirmed using real-time PCR using strainspecific primers and fluorescently labeled TaqMan probe designed with Primer3 software in Geneious based on the coronavirus sequences from these three bats (Untergasser et al. 2012) (Supplementary Fig. 1, Supplementary File 2).
In conclusion, from a sample of 44 bats in Eswatini, we detected both alpha-and betacoronaviruses. All eight bats from which coronaviruses were detected were captured leaving roosts in houses, churches, or within human settlements. More research is necessary to determine whether any of these detected coronaviruses could be a concern for the health of humans or livestock. Limiting direct contact with these bats or their feces might possibly aid in preventing future emerging infectious diseases, while continued monitoring may shed light on the diversity and ecology of coronaviruses. (N.D.N). Initial phylogenetic analyses used the Extreme Science and Engineering Discovery Environment (XSEDE) resources, which is supported by National Science Foundation grant number ACI-1053575. XSEDE resources were provided by project allocation TG-ASC160034.

OPEN ACCESS
This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativec ommons.org/licenses/by/4.0/.