key: cord-0427495-cf143byx authors: Kumakamba, Charles; Niama, Fabien R.; Muyembe, Francisca; Mombouli, Jean-Vivien; Kingebeni, Placide Mbala; Nina, Rock Aime; Lukusa, Ipos Ngay; Bounga, Gerard; N’Kawa, Frida; Nkoua, Cynthia Goma; Losoma, Joseph Atibu; Mulembakani, Prime; Makuwa, Maria; Tamufe, Ubald; Gillis, Amethyst; LeBreton, Matthew; Olson, Sarah H.; Cameron, Kenneth; Reed, Patricia; Ondzie, Alain; Tremeau-Bravard, Alex; Smith, Brett R.; Pante, Jasmine; Schneider, Bradley S.; McIver, David J.; Ayukekbong, James A.; Hoff, Nicole A.; Rimoin, Anne W.; Laudisoit, Anne; Monagin, Corina; Goldstein, Tracey; Joly, Damien O.; Saylors, Karen; Wolfe, Nathan D.; Rubin, Edward M.; MPassi, Romain Bagamboula; Tamfum, Jean J. Muyembe; Lange, Christian E. title: Coronavirus surveillance in Congo basin wildlife detects RNA of multiple species circulating in bats and rodents date: 2020-07-20 journal: bioRxiv DOI: 10.1101/2020.07.20.211664 sha: 28d9cf195b569a2fbf7783598a5c35c2c6e6d5a0 doc_id: 427495 cord_uid: cf143byx Coronaviruses play an important role as pathogens of humans and animals, and the emergence of epidemics like SARS, MERS and COVID-19 is closely linked to zoonotic transmission events primarily from wild animals. Bats have been found to be an important source of coronaviruses with some of them having the potential to infect humans, with other animals serving as intermediate or alternate hosts or reservoirs. Host diversity may be an important contributor to viral diversity and thus the potential for zoonotic events. To date, limited research has been done in Africa on this topic, in particular in the Congo Basin despite frequent contact between humans and wildlife in this region. We sampled and, using consensus coronavirus PCR-primers, tested 3,561 wild animals for coronavirus RNA. The focus was on bats (38%), rodents (38%), and primates (23%) that posed an elevated risk for contact with people, and we found coronavirus RNA in 121 animals, of which all but two were bats. Depending on the taxonomic family, bats were significantly more likely to be coronavirus RNA-positive when sampled either in the wet (Pteropodidae and Rhinolophidae) or dry season (Hipposideridae, Miniopteridae, Molossidae, and Vespertilionidae). The detected RNA sequences correspond to 15 Alpha- and 6 Beta-coronaviruses, with some of them being very similar (>95% nucleotide identities) to known coronaviruses and others being more unique and potentially representing novel viruses. In seven of the bats, we detected RNA most closely related to sequences of the human common cold coronaviruses 229E or NL63 (>80% nucleotide identities). The findings highlight the potential for coronavirus spillover, especially in regions with a high diversity of bats and close human contact, and reinforces the need for ongoing surveillance. 10 231 most closely related to bat coronaviruses with some similarity (>80% nucleotide identities) to 232 human coronavirus NL63 (Figures 2 and 3) . 242 Our data suggest that coronavirus circulation in bats, at least in the Congo Basin, may indeed 243 depend to some extent on species and seasonality (Supplement 4). We observed a significant 244 difference in the number of bats testing positive depending on the local calendric season (p = 245 0.0176), with 10.5% of coronavirus RNA positive bats in the wet season but only 6.6% in the dry 246 season at similar sample sizes for both seasons (Table 1) . Interestingly, when looking at the family 247 and species level, this holds true only for the Pteropodidae and Rhinolophidae species (p < 248 0.0001) while Hipposideridae, Miniopteridae, Molossidae, and Vespertilionidae species are more 249 likely to be positive for coronavirus RNA in the dry season (p < 0.0001) ( Table 1 ). The latter, 250 though not for those specific families but for bats in general, has been proposed to be the 251 correlation on a global scale [15] . We can only speculate as to the reasons of the apparent 252 seasonality, but family and species seem to be important determinants. Due to the diverse set of 253 species in our sample set, individual sample numbers for most species are too small to draw 254 definite conclusions, however the significant seasonal difference between Yinpterchiroptera and 255 Yangochiroptera are largely supported by respective trends in the individual species. We tested 256 if the results from any particular species might be responsible for the observed correlation 11 257 between season and the rate of positive coronavirus RNA animals. The only species that turned 258 out to have a strong influence on the outcome was Eidolon helvum. However, the effect of 259 dropping it from the analysis did only influence the outcome for bats in total, while it was not strong 260 enough to negate the observed statistical significances for season within the Pteropodidae family 261 or the Yinpterchiroptera suborder. 262 We did find Eidolon helvum, a bat usually roosting in large colonies, to be significantly 299 RNA of either Kenya bat coronavirus BtKY56 or Eidolon bat coronavirus/Kenya/KY24 was 300 detected in ~70% (83) of the positive bats in this study and in several hundred bats reported 301 previously (GenBank). Interestingly Kenya bat coronavirus BtKY56 appears to be a common virus 302 species in the Congo Basin, while elsewhere it appears to be Eidolon bat 303 coronavirus/Kenya/KY24 that is more common. These observations are undoubtedly susceptible 304 to a sampling bias, for example due to the species composition of sample sets, particularly with 305 Eidolon helvum, which can be sampled in large numbers when colonies are present or when they 306 are present in markets [41] . However, we do find evidence of these two viruses in a relative wide 307 array of bat hosts, indicating that species barriers may not be a limiting factor for sharing these 308 specific Beta coronaviruses (Figures 2 and 3, Supplement 3) . In contrast, most of the other 13 309 sequences that we detected with related sequences in GenBank were detected in bats of the 310 same genus by us and previously by others, supporting some degree of general species 311 specificity and virus host co-evolution despite the latent ability of at least some coronaviruses to 312 jump species barriers within and outside of the taxonomic order of hosts [14, 15, 17] . How often 313 these events occur is not fully understood, but it is generally assumed that bats serve as a 314 reservoir for coronaviruses [16] . With SARS-CoV-1 and SARS-CoV-2, the available evidence 315 suggests that they were successfully transmitted from bats into humans, either directly or 316 indirectly [20]. When we add to these two coronaviruses MERS that originated in bats and 317 established a sustained reservoir in camels with occasional spillover into humans, we have 318 witnessed three coronavirus spillover events with a bat origin in less than two decades. 337 We conclude overall, that bats and to a much smaller degree rodents in the Congo Basin harbor 338 diverse coronaviruses, of which some might have the molecular potential for spillover into 339 humans. Considering the close contact between wildlife and humans in the region, as part of the 340 value chain or in peri-domestic settings, there is an elevated and potentially increasing risk for 341 zoonotic events involving coronaviruses. Thus, continued work to understand the diversity, 342 distribution, molecular mechanisms, host ecology, as well as consistent surveillance of 343 coronaviruses at likely hotspots, are critical to help prevent future global pandemics. Our goal was to determine the degree of coronavirus circulation and diversity, 103 using a consensus Polymerase Chain reaction (PCR) approach, coupled with the collection of, 104 and coupling with while 109 samples from the (bushmeat) value chain were collected from freshly killed animals voluntarily 110 provided by local hunters upon their return to the village following hunting, or by vendors at 111 markets. Fecal samples were collected from free-ranging NHPs. Some NHP samples were also 112 collected during routine veterinary exams in zoos and wildlife sanctuaries. Hunters and vendors 113 were not compensated, to avoid incentivizing hunting. Oral and rectal swab samples were 114 collected into individual 2.0 ml screw-top cryotubes containing 1.5 ml of either Universal Viral 115 Transport Medium (BD), RNA later, lysis buffer, or Trizol® (Invitrogen), while pea-sized tissue 116 samples were placed in 1.5ml screw-top cryotubes containing 500ul of either RNA later or lysis 117 buffer (Qiagen), or without medium. All samples were stored in liquid nitrogen as soon as 118 practical RNA was extracted either manually using Trizol® or with a Zymo Direct-zol RNA kit (swabs 122 collected after 2014) and stored at -80ºC. Afterwards RNA was converted into cDNA using a First Strand cDNA Synthesis Kit (Thermo Scientific) or GoScript™ Reverse 124 Transcription kit (Promega) and stored at -20°C until analysis. Two conventional nested broad 125 range PCR assays Polymerase gene (RdRp) were used to test the samples for coronavirus RNA. The first PCR 127 amplifies a product of approximately 286nt between the primer binding sites The 132 second PCR was used in two modified versions: one of them specifically targeting a broad range 133 of coronaviruses in bats, the second one broadly targeting coronaviruses of other hosts In the second 136 round, either CoV-FWD4/Bat DNA was extracted using the Qiagen QIAquick Gel 141 Extraction Kit and either sequenced by Sanger sequencing at the UC Davis DNA sequencing 142 facility or was sent for commercial Sanger sequencing (GATC or Macrogen). Extracts with low 143 DNA concentrations were cloned prior to sequencing Viral sequences were deposited in the GenBank database under submission numbers KX284927-147 KX284930, KX285070-KX285095, KX285097-KX285105, KX285499-KX285513, KX286248-148 KX286258, KX286264-KX286286, KX286295-KX286296, KX286298-KX286322, MT064119 MT082032, MT082059-150 MT082060, MT082072, MT082123-MT082136, MT082145, MT082299 Maximum likelihood phylogenetic trees were constructed including different genera (Alpha, Beta 152 and Gamma) and species of known coronaviruses, as well as species/sub-species detected in 153 DRC and ROC during the PREDICT project. Only a single sequence was included representing 7 154 sequences with nucleotide identities of more than 95%. Multiple sequence alignments were made 155 in Geneious (version 11.1.3, ClustalW Alignment). Bayesian phylogeny of the polymerase gene 156 fragment was inferred using MrBayes The sequence of an avian Gamma Coronavirus (NC_001451) served as outgroup 159 to root the trees, and trees were sampled after every 1,000 steps during the process to monitor 160 phylogenetic convergence [36]. The average standard deviation of split frequencies was below 161 0.006 for the Watanabe PCR amplicon based analysis and below 0.0029 for the Quan PCR 162 amplicon based analysis (MrBayes recommended final average <0.01) Ecological data related to the locality and the host animals was compiled and analyzed with 166 respect to a correlation with the frequency of virus detection. The data included sex, human 167 interface at which the animals were collected and sampled (value chain or other), and local 168 calendric season (wet/dry) based on the climate-data.org data set 630 from DRC and 931 from RoC) were 173 sampled and tested, of which 1,356 were bats (24 genera), 1,347 rodents (33 genera), 836 NHPs 174 (14 genera), and 22 shrews, (Figure 1, Supplements 1 and 2). The majority of the 5,586 collected 175 samples were oral (2,258) or rectal (2,238) swabs, with others being tissue samples, including 176 liver and spleen (385), lung (187) or intestinal tract (175), as well as feces (167), blood, serum or 177 plasma (140) and others (36). Coronavirus RNA was Two of the animals with detected coronavirus RNA were rodents (<1% of 181 sampled rodents), while 119 were bats (8.8% of sampled bats). Coronavirus RNA positive animals 182 were found in 25% (27/106) of bat sampling events (same location and same day) In 10 of the bat sampling events, a single 184 coronavirus RNA positive bat was among the tested animals, while in 17 events the number of of rodents, one Deomys ferrugineus (1/1) and one Malacomys longipes (1/38), and in at 187 least 14 different bat species Micropteropus pusillus, Miniopterus inflatus, Mops condylurus, Myonycteris sp., Rhinolophus sp Among the five bat species 191 from which more than 100 individuals were sampled and tested, Eidolon helvum had the highest 192 rate of coronavirus RNA positives (22.3%) With 10.2% Yinpterchiroptera bats had a significantly (N=0.015 C 2 Y) higher rate of 195 coronavirus RNA positive animals than Yangochiroptera bats with 5.0% (Table 1) Significant seasonal differences for the rate of coronavirus RNA positive animals were detected 198 across the bats with a 10.5% PCR positive rate in the wet season and a 6.6% rate in the dry 199 season (p = 0.0176) (Table 1, Supplement 4). Bats that were associated with the (bushmeat) human animal (peri-domestic) interfaces (5%) Male bats were significantly overrepresented among the Coronavirus RNA positives p while there was insufficient data to analyze an influence of age Upon phylogenetic analysis, the sequences fall into 13 separate clusters based on the Quan PCR clusters Alpha 5, 6, and 7 (Q7=W2), as well as Beta 1, 2, and 3 correspond to each 208 other. In one bat, RNA corresponding to two different alphacoronaviruses was detected in the oral 209 and the rectal sample by the same PCR assay (ZB12030), while in another bat one PCR assay 210 amplified RNA indicating an Alpha-and the other assay an RNA indicating a betacoronavirus 211 213 single type/strain of these coronaviruses was detected, two in two events, and three, five or eight 214 in one event each. Identical or very similar coronavirus sequences were found with a spatial 215 distance of up to 1975 km apart and a temporal distance of up to 1708 days (Supplement 5). 80% identities with either of them (Figure 2, Supplement 3). The detected bat coronavirus 219 sequences on the contrary mostly clustered closely with known ones Eidolon bat coronavirus/Kenya/KY24 were detected in 30 individual bats of 3 different species 225 sampled on 8 occasions (Q-/W-Beta 3) (Supplements 3 & 5) sequences were closest to coronaviruses found in 229 bats and camels with a high similarity (>90% nucleotide identities) to human coronavirus 229E 230 (Figures 2 and 3). Similarly, the viral sequences in clusters Sustainable Development (1102/MEFDD/DGEFDFAP-SPR) and the 362 Ministry of Scientific Research and Technical Innovation 018/MRSIT/DGRST/DMAST) in the RoC. All authors declare that Fields Virology, 369 6th edn. Lippincott Williams & Wikins, Philadelphia Virus taxonomy. Classifcation and 371 nomenclature of viruses, ninth report of the international committee on taxonomy of 372 viruses. International Union of Microbiological Societies, Virology Division Identification of a novel coronavirus in patients 375 with severe acute respiratory syndrome Isolation of a 377 novel coronavirus from a man with pneumonia in Saudi Arabia A pneumonia outbreak associated with a new 380 coronavirus of probable bat origin COVID-19 compared to other 382 epidemic coronavirus diseases and the flu Bats are natural reservoirs of SARS-like coronaviruses Molecular evolution analysis and geographic investigation 386 of severe acute respiratory syndrome coronavirus-like virus in palm civets at an animal 387 market and on farms Phylogenetic perspectives on the epidemiology and origins 389 of SARS and SARS-like coronaviruses Rooting the phylogenetic tree of middle East 391 respiratory syndrome coronavirus by characterization of a conspecific virus from an 392 African bat Co-circulation of three camel coronavirus species 394 and recombination of MERS-CoVs in Saudi Arabia Distant relatives of severe acute respiratory 396 syndrome coronavirus and close relatives of human coronavirus 229E in bats Evidence supporting a zoonotic origin of human coronavirus 399 strain NL63 Recent transmission of a novel alphacoronavirus, bat 401 coronavirus HKU10, from Leschenault's rousettes to pomona leaf-nosed bats: first 402 evidence of interspecies transmission of coronavirus between bats of different suborders Global patterns in coronavirus diversity Coronavirus diversity, phylogeny and interspecies 407 jumping Interplay between co-divergence and cross-409 species transmission in the evolutionary history of bat coronaviruses Identification and characterization of a novel 412 alpaca respiratory coronavirus most closely related to the human coronavirus 229E Evidence for an Ancestral Association of 415 Human Coronavirus 229E with Bats Origin and evolution of pathogenic coronaviruses Global trends in emerging infectious diseases Bushmeat hunting, deforestation, and 421 prediction of zoonotic disease Global hotspots and correlates of 423 emerging zoonotic diseases Detection of Novel SARS-like and Other 425 Genomic characterization of seven distinct bat coronaviruses 427 in Kenya Detection of new genetic 431 variants of Betacoronaviruses in Endemic Frugivorous Bats of Madagascar A metagenomic viral discovery approach 434 identifies potential zoonotic and novel mammalian viruses in Neoromicia bats within South 435 Coronaviruses Detected in Bats in Close Contact 439 with Humans in Rwanda The bushmeat market in Kisangani Democratic Republic of Congo: Implications for conservation and food security Trends in bushmeat trade in a 444 postconflict forest town: implications for food security Human coronavirus OC43 outbreak in wild 446 chimpanzees, Côte d´Ivoire Identification of a severe acute respiratory syndrome 448 coronavirus-like virus in a leaf-nosed bat in Nigeria Bat coronaviruses and experimental 450 infection of bats, the Philippines MRBAYES 3.2: Efficient Bayesian 452 phylogenetic inference and model selection across a large model space BEAST 2.5: An advanced software 455 platform for Bayesian evolutionary analysis R: A Language and Environment for Statistical Computing Reproduction of East-African bats may 459 guide risk mitigation for coronavirus spillover Breeding cycle of straw-colored fruit bat, Eidolon helvum Uncovering the fruit bat bushmeat 463 commodity chain and the true extent of fruit bat hunting in Ghana Serological Evidence of Bat SARS-Related Coronavirus 466 Infection in Humans The bat meat chain and perceptions of the risk of contracting Ebola in the Mount Cameroon region Sampling sites without viral RNA detection are sharing more than 95% nucleotide identities in brackets. GenBank accession numbers 482 are listed for previously published sequences, while sequences obtained during the project are 483 identified by cluster names (compare Supplement 3). Black font indicates coronavirus sequences 484 obtained from bats, brown font indicates rodents, blue humans and gray other hosts. The host 485 species and country of sequence origin are indicated for bats and rodents if applicable. In case 486 of clusters W-Alpha-1 sequences were detected in Mops condylurus and Chaerephon sp., host 487 species in cluster W-Beta-1 were Megaloglossus woermanni and Epomops franqueti and in case 488 of cluster W-Beta-2 Micropteropus pusillus, Epomops franqueti the sequences detected during the project (red boxes) and indicates the number of sequences 496 sharing more than 95% nucleotide identities in brackets. GenBank accession numbers are listed 497 for previously published sequences, while sequences obtained during the project are identified 498 by cluster names (compare Supplement 2). Black font indicates coronavirus sequences obtained 499 from bats, brown font indicates rodents, blue humans and gray other hosts. The host species and 500 country of sequence origin are indicated for bats and rodents if applicable Alpha-4 sequences were detected in Mops condylurus and Chaerephon sp., host species in 502 cluster Q-Alpha-7 were Epomops franqueti and Chaerephon pumilus Micropteropus pusillus and Epomops franqueti, and for cluster Q-Beta-3 Megaloglossus 504 woermanni, Eidolon helvum, and Epomops franqueti Table 1: PCR results by species and season (Bats) Suborder, family and species Significant difference between calendric seasons P<0.05 (Chi-square with Yates correction) Highly significant difference between calendric seasons P<0.01 (Chi-square with Yates 512 correction)