key: cord-276017-2375ipkk
authors: Chen, Dongsheng; Sun, Jian; Zhu, Jiacheng; Ding, Xiangning; Lan, Tianming; Zhu, Linnan; Xiang, Rong; Ding, Peiwen; Wang, Haoyu; Wang, Xiaoling; Wu, Weiying; Qiu, Jiaying; Wang, Shiyou; Li, Haimeng; An, Fuyu; Bao, Heng; Zhang, Le; Han, Lei; Zhu, Yixin; Wang, Xiran; Wang, Feiyue; Yuan, Yuting; Wu, Wendi; Sun, Chengcheng; Lu, Haorong; Wu, Jihong; Sun, Xinghuai; Zhang, Shenghai; Sahu, Sunil Kumar; Chen, Haixia; Fang, Dongming; Luo, Lihua; Zeng, Yuying; Wu, Yiquan; Cui, ZeHua; He, Qian; Jiang, Sanjie; Ma, Xiaoyan; Feng, Weimin; Xu, Yan; Li, Fang; Liu, Zhongmin; Chen, Lei; Chen, Fang; Jin, Xin; Qiu, Wei; Yang, Huanming; Wang, Jian; Hua, Yan; Liu, Yahong; Liu, Huan; Xu, Xun
title: Single-cell screening of SARS-CoV-2 target cells in pets, livestock, poultry and wildlife
date: 2020-06-14
journal: bioRxiv
DOI: 10.1101/2020.06.13.149690
sha: 
doc_id: 276017
cord_uid: 2375ipkk

A few animals have been suspected to be intermediate hosts of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). However, a large-scale single-cell screening of SARS-CoV-2 target cells on a wide variety of animals is missing. Here, we constructed the single-cell atlas for 11 representative species in pets, livestock, poultry, and wildlife. Notably, the proportion of SARS-CoV-2 target cells in cat was found considerably higher than other species we investigated and SARS-CoV-2 target cells were detected in multiple cell types of domestic pig, implying the necessity to carefully evaluate the risk of cats during the current COVID-19 pandemic and keep pigs under surveillance for the possibility of becoming intermediate hosts in future coronavirus outbreak. Furthermore, we screened the expression patterns of receptors for 144 viruses, resulting in a comprehensive atlas of virus target cells. Taken together, our work provides a novel and fundamental strategy to screen virus target cells and susceptible species, based on single-cell transcriptomes we generated for domesticated animals and wildlife, which could function as a valuable resource for controlling current pandemics and serve as an early warning system for coping with future infectious disease threats.

In the past two decades, the world has witnessed the outbreak and spread of SARS, Middle East respiratory syndrome (MERS) 1 , ZIKA 2 , avian influenza and swine influenza 3 , which have been posing an urgent challenge to our infectious disease prevention and control system. Recently, SARS-CoV-2 has caused a highly contagious pandemic disease named coronavirus disease 2019 (COVID-19), which is rapidly spreading all over the world and has triggered a severe public health emergency. As of 3rd June 2020, globally the total number of confirmed COVID-19 cases and deaths has uncontrollably reached 6,287,771 and 379,941, respectively 4 .

The bat has been proposed to be the original host of SARS-CoV-2 5 , however, the transmission from bats to humans requires some intermediate hosts. Several studies have linked pangolins, cats, dogs and hamsters with SARS-CoV-2 infection and transmission [6] [7] [8] [9] [10] [11] [12] , indicating the potential widespread prevalence across animals, which would post potential threats to humans. The identification of the origin of this virus and its path to becoming a deadly human pathogen is needed to understand how such processes occur in nature and identify ways we can prevent the onset of these types of global crises in the future.

Evaluating host susceptibility is critical for controlling the infectious disease. The screening of virus putative host is usually performed using in vivo assay or inoculation experiments, which helps to reveal the host susceptibility, however, there are several limitations: 1) Experiments for some dangerous and infectious viruses to cell receptor angiotensin-converting enzyme 2 (ACE2) 13 and the cleavage of S protein by transmembrane serine protease 2 (TMPRSS2) 14 . Although SARS-CoV-2 like corona virus has been isolated from pangolins and bats, their susceptible cell types for SARS-CoV-2 is not clear. Given that the species barrier of SARS-CoV-2 was estimated to be relatively low 13 and livestock, poultry and pets have very close contact with humans, it is crucial to evaluate animal susceptibility to SARS-CoV-2.

Previous studies have proposed that animal tissues show high heterogeneity in terms of cellular composition and gene expression profiles 15 , and ACE2 is only expressed in a small proportion of specific cell populations 16 , making single cell analysis of SARS-CoV-2 target cells an attracting field to investigate. Here, we constructed the single cell atlas for livestock, poultry, pets and wildlife, then screened putative SARS-CoV-2 target cells (indicated by the co-expression patterns of SARS-CoV-2 entry receptor ACE2 and SARS-CoV-2 entry activator TMPRSS2) and systematically evaluated their susceptibility, with the aim to understand the virus transmission routes and provide clues to fight against COVID-19.

Both pangolin and cat are suspected to be SRAS-CoV-2 intermediate hosts.

SRAS-CoV-2 like coronavirus has been isolated from pangolin, and SARS-CoV-2 was proposed to origin from the recombination of a pangolin coronavirus with a bat coronavirus 6, 7 . Cat is also a suspected intermediate host, as human-to-cat and cat-to-cat transmission of SARS-CoV-2 have been reported 9, 10 . Domestic pig is an animal in close contact with human and have been reported to be susceptible to SARS coronavirus 17 . Although those animals have been linked with coronavirus, yet a comprehensive single-cell atlas for those species is missing. In this study, we generated the single nuclei libraries for various tissues of pangolin (heart, liver, spleen, lung, kidney, large intestine, duodenum, stomach and esophagus) , cat (heart, liver, lung, kidney, eyelid, esophagus, duodenum, colon and rectum), and pig (heart, liver, spleen, lung, kidney, hypothalamus, area postrema, vascular organ of lamina terminalis, subfomical organ and cerebellum) (Fig. 1a Table 2) .

In cat, ACE2 and TMPRSS2 co-expressing cells were detected in lung (ATI, ATII, secretory cells, mesothelial cells, ciliated cells, endothelial cells, fibroblasts, macrophages), kidney (endothelial cells, non-proximal tubule cells, proximal tubule cells, stromal cells); eyelid (endothelial cells, epithelium cells and immune cells), esophagus ( immune cells) and rectum (enterocytes). Notably, we observed over 40% co-expression of ACE2 and TMPRSS2 in proximal tubule cells of cat kidney, and around 30% in epithelium cells of cat eyelid (Fig. 1d, Supplementary Table 3 ).

In pangolin, SARS-CoV-2 target cells were found in lung endothelial cells, kidney (endothelial cells, podocytes and proximal tubule cells), liver (hepatocytes) and spleen (immune cells) (Fig. 1g, Supplementary Table 3 ).

In pig, ACE2 and TMPRSS2 were mainly co-expressed in lung (ATI, ATII, ciliated cells, secretory cells, endothelial cells, fibroblasts, macrophages) and kidney (non-proximal tubule cells, proximal tubule cells, endothelial cells, podocytes) (Fig.   1j, Supplementary Table 3) . Overall, the proportions of SARS-CoV-2 target cells in cat are much higher than the proportions in corresponding cell types of all other species studied here.

In consistent with previous report that SRAS-CoV-2 replicates poorly in chicken and duck 8 , no SARS-CoV-2 target cells were found in lung cells of poultry (chicken, duck, goose and pigeon) (Fig. 2a, b) . In cat, SARS-CoV-2 target cells were detected in eight out of eleven cell types investigated, with the top two cell types being ciliated cells and secretory cells. In pig, co-expression of ACE2 and TMPRSS2 was observed in seven out of eleven cell types. In pangolin, a small proportion of endothelial cells were found to co-express ACE2 and TMPRSS2. In hamster, ciliated cells were the cell type with most abundant SARS-CoV-2 target cells. In goat, we detected the co-expression of ACE2 and TMPRSS2 in ATI, fibroblasts, endothelial cells, ciliated cells and T cells. Goat share highly similar ACE2 amino acids sequence with pig and human 22 , implying that goat ACE2 might have similar capability for mediating virus entering into host cells. In lizard, we detected the co-expression of ACE2 and TMPRSS2, mainly in B cells. In dog, less than 0.5% ACE2 and TMPRSS2 co-expressing cells were detected in ciliated cells and ATII (Fig. 2a, b) .

The infection with SARS-CoV-2 could lead to a severe pneumonia, and respiratory diseases caused by other respiratory viruses is also noteworthy. To reveal the putative target lung cells of other respiratory viruses, we screened the expression patterns of 32 virus receptors for a total of 29 virus species derived from 12 virus families (coronavirida, orthomyxoviridae, adenoviridae, hantaviridae, matonaviridae, paramyxoviridae, parvoviridae, phenuiviridae, picornaviridae, pneumoviridae, reoviridae and rhabdoviridae) (Fig. 2c, Supplementary Table 4) , which have been shown to be able to transmit via the respiratory system 22 . Generally, poultry lung cells express less types of virus receptors than mammalians and reptiles. For example, coronavirus receptors were generally not expressed in poultry lung cells, except that human coronavirus 229E receptor ANPEP was expressed in chicken and duck lung cells but not in goose and pigeon lung cells. MERS coronavirus receptor DPP4 was expressed in chicken and goose lung cells but was barely detected in corresponding cell types in duck and pigeon. Coronavirus SARS receptor CLEC4G was expressed in goat lung cells (ciliated cells, endothelial cells, fibroblasts and macrophages) but was found absent in lung cells of other species. Influenza A virus receptors UVRAG and EGFR was present in lung cells of every species we investigated, but was absent or exhibited meagre expression in pangolin and pigeon. Adenoviridae virus receptors showed a preferential expression in mammals, except adenovirus type C receptors which was also present in poultry lung cells. Rhinovirus C receptor displayed preferential expression in ciliated cells of human, cat, dog, hamster, pig, goat and goose while respiratory syncytial virus receptor CD209 was only present in human macrophage cells. Taken together, our work, for the first time, revealed the putative target cells for respiratory viruses in an important organ of respiratory system (lung), which lays the foundation for dissecting the infection and transmission of respiratory system viruses at the single-cell level.

While comparing the frequencies of SARS-CoV-2 target lung cells across different species, we noticed that cat clearly outweigh other species, with 13.19% in ciliated cells, compared to pig (3.35%) and hamster (3.87%) (Supplementary Table 3 ).

Moreover, when taking the proportions of SARS-CoV-2 target cells in distinct organs among cat, pangolin and pig into consideration, it further indicated that the proportion of ACE2 and TMPRSS2 target cells were much higher in cat. For example, the proportion of ACE2 and TMPRSS2 co-expressing cell was as high as 40% in cat kidney proximal tubular cells while the proportions were only around 3% and 2% in corresponding cell type in pangolin and pig, respectively (Supplementary Table 3 ).

We also noticed that SARS-CoV-2 target cells were widely distributed among organs within the digestive system (esophagus, rectum), respiratory system (lung) and urinatory system (kidney) of cat (Fig. 3a) , implying that cats could be infected by SARS-CoV-2 via multiple routes such as dietary infections of the digestive tract or airborne transmission through respiratory system. To highlight SARS-CoV-2 susceptible cell types, we summarized all cell types with proportion of ACE2 and TMPRSS2 co-expressing cells higher than 1%, which clears shows that cat, as well as pig, have more susceptible cell types than other species we investigated (Fig. 3a) .

Taken together, our data explains the observation that cats are highly permissive to SARS-CoV-2 8 and raise the necessity to carefully monitor and evaluate the possible roles of cat as intermediate hosts in current pandemic.

SARS-CoV-2 replicate poorly in dog, chicken and duck 8 . Our data suggests that the co-expression of ACE2 and TMPRSS2 is very rare in dog lung cells and absent in poultry lung cells (chicken, duck, goose and pigeon). Besides, it has been predicted that dog and poultry ACE2 cannot be utilized efficiently by SARS-CoV-2 spike glycoprotein because of mutations in critical amino acids of dog ACE2 5 . Therefore, our data, to some extent, explains why dogs and poultry are not as permissive for SARS-CoV-2 infection as cat.

In addition to cats, pangolins 6,7 and hamsters 12 have been reported to be permissive for SARS-CoV-2 infection, however, the target cells for virus infection and putative transmission routes are largely unknown. Our study identified the SARS-CoV-2 target cells in distinct tissues of cat, pangolin and hamster, indicated by the simultaneous expression of SARS-CoV-2 entry factors: ACE2 and TMPRSS2. A detailed comparative analysis among cat, pangolin and hamster deciphered that proportion of SARS-CoV-2 target cells in cat was much higher than pangolin and hamster, implying that cats are more susceptible to SARS-CoV-2. Besides, as a companion animal, cats interact with humans more frequently than pangolins, thus we proposed that cats should be closely monitored in the current COVID pandemic. In addition, we also detected the co-expression of ACE2 and TMPRSS2 in lung cells of goat and lizard. 

To investigate the susceptibility of host cells to different kinds of viruses, we screened Table 5 ). Intriguingly, we found Rabies lyssavirus receptor NCAM1 was widely expressed in cell types of pig neural system but rarely expressed in non-neural tissues. Besides, we found that vesicular stomatitis virus receptor UVRAG was preferentially expressed in pig lung cells (Extended Data Fig. 6 ). To demonstrate the proportions of virus receptor expressing cells in an intuitive manner, we proposed a system named "traffic light system for virus entry" to assign each cell type to one of the following three status: red light, green light and yellow light. Briefly, if the receptor is expressed in less than 1% of that cell types, a "red light" status will be assigned. If the proportion is between 1% and 10%, "yellow light" status will be assigned. If the proportion is over 10%, then a "green light" status would be assigned (see methods). Based on this system, we assigned status to all the cell populations of all the investigated species (Supplementary Data Table 5 ). To demonstrate the potential usefulness of this system, we employed this analysis and Our project provides a novel strategy to find out putative susceptible hosts, based on the distribution of virus target cells among distinct cell populations, which makes it possible to screen the susceptibility of all existing viruses (using the available virus receptor information) on all existing species (using the available scRNAseq data) in an unbiased manner. With the development of single cell sequencing techniques and the progress of international single cell atlas collaborative projects, the atlas for more species could be generated at an accelerated speed. We anticipate that the information gained from the present study will certainly augment future research work, and provide some novel insights about the prevention and control strategies against SARS-CoV-2 along with many other harmful viruses.

Sample collection and research were performed with the approval of Institutional Review Board on Ethics Committee of BGI (Approval letter reference number BGI-NO. BGI-IRB A20008). All procedures were conducted according to the guidelines of Institutional Review Board on Ethics Committee of BGI.

A total of 11 samples were collected in this study, including four pets: Felis catus After the execution of animals in accordance with the ethics of animal experiment, dissection was carried out quickly to separate each organ. The collected tissues were rinsed by 1X PBS, then quick-frozen and stored in liquid nitrogen. The single cell nucleus of each tissue was separated by mechanical extraction method 26 . Briefly, the tissues were first thawed, infiltrated by 1X homogenization buffer (containing 30mM CaCl 2 , 18mM Mg(Ac) 2 , 60mM Tris-HCl (pH 7.8), 320mM sucrose, 0.1% NP-40, 0.1mM EDTA and 0.2U/µl RNase inhibitor), then cut into smaller pieces, and the single nucleus was isolated by 2ml Dounce homogenizer set. After filtration with 30µm strainer, the nuclei extraction was resuspended by 1%BSA containing 0.2U/µl RNase inhibitor and span down at the speed of 500g for 10 min at 4 degrees (to carefully discard the cellular impurities within the supernatant). This step was repeated twice, and finally the nucleus was recollected with 0.1% BSA containing 0.2U/µl RNase inhibitor. Subsequently, DAPI was used to stain the nucleus, and the nucleus density was calculated under a fluorescence microscope to prepare for the subsequent library construction.

The mRNA within the single nucleus samples of different organs of pig (heart, liver, spleen, lung, kidney, hypothalamus, area postrema, vascular organ of lamina terminalis, subfomical organ and cerebellum) were captured and the libraries were constructed using inhouse DNBelab C4 kit and sequenced using DNBSEQ-T1. The separated single nucleus of different organs (including the lungs for pig, dog, cat, goat, pangolin, chicken, pigeon, goose, duck, lizard, and hamster; pangolin organs: heart, liver, spleen, lung, kidney, large intestine, duodenum, stomach and esophagus; cat organs: heart, liver, lung, kidney, eyelid, esophagus, duodenum, colon and rectum) 

To facilitate the integration of cross-species lung single cell data set, we converted genes from other species to mouse homologs. We downloaded homolog gene lists using BioMart 27 . Then, if a 1:1 match existed between a non-mouse gene and a mouse gene, the non-mouse gene name was converted. As for pangolin, goose and pigeon, which lack homologs records on Ensemble, single-copy orthologs were identified from two species genomes by cluster analysis of gene families using OrthoFinder 28 (OrthoFinder version 2.3.3) with the default parameters.

Sequencing data filtered using custom script and gene expression matrix were obtained using Cell Ranger 3.0.2 (10X Genomics). The genomes using for reads alignment were downloaded from NCBI Assembly (Supplementary Table 6 ). Single cell analysis was conducted using Seurat 19,29 . Briefly, quality control was performed based on the following criteria: cells with mapped number of genes less than 200 or with mitochondrial percentage higher than 10% were removed. Variable genes were determined using Seurat's FindVariableGenes function with default parameters.

Clusters were identified using Seurat's FindClusters function and visualized using Seurat's RunTSNE. All the DEGs for each Seurat Objects were identified using Seurat's FindAllMarkers function. Cell types were annotated according to the expression of canonical cell type markers.

The human lung single cell RNAseq data was obtained from literature 30 . Data sets of lungs from different species were integrated using Seurat's FindIntegrationAnchors and IntegrateData function with features after homolog conversion.

Virus receptor list were downloaded from a virus-host receptor interaction database 31 and manually collected from published literatures (Supplementary Table 7) .

Single nuclei RNAseq data sets for frontal lobe, occipital lobe, parietal lobe, temporal lobe and hypothalamus of domestic pig were obtained from literature 26 .

The proportions of virus receptor were calculated and "red, yellow, green" status was assigned to each cell type based on receptor proportions (less than 1%, red light; greater than or equals to 1% & less than 10%, yellow light; greater than or equals to 10%, green light). In case of viruses with multiple receptors, the receptor with the highest proportion was considered for status assignment.

The single cell atlas of all the investigated species in this study are available via http://120.79.46.200:81/SARS-CoV-2. The raw data supporting the findings of this study will be made available upon request. Raw transcriptome sequencing data has been deposited to the CNSA (CNGB Nucleotide Sequence Archive) with the accession number CNP0001085 (https://db.cngb.org/cnsa/) and will be released to the public after the manuscript is accepted for publication. Smooth muscle cells 

Duodenum • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • TMPRSS2 ACE2 Duodenum TMPRSS2 ACE2

• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • TMPRSS2 ACE2

C O L 3 A 1 C O L 1 A 1 F B L N 5 A C T A 2 C L D N 1 6 T M E M 7 2 S L C 1 2 A 1 P A X 2 M A L N T N 1 L R P 2 P C K 1 P D Z K 1 S L C 5 A 8 B T B D 1 1 A C M S D S L C 1 0 A 2 S L C 2 7 A 1 A S S 1 F B P 1 S L C 2 2 A 6 S L C 3 4 A 1 A L D O B A D G R L 4 C D 3 4 F L T 1 K D R M Y C T 1 P E C A M 1 P T P R B E S A M N R C A M T C F 2 1 D P P 4 M A G I2 E YR G S 5 P R K G 1 M R C 1 IT G A M C D 1 6 3 C D 7 4 C 1 Q A A D G R E 1 S O X 1 7 G J A 4 E M C N P E C A M 1 T H B S 1 P C S K 6 G S N D D R 2 G F P T 2 C R IS P L D 2 E N O 3 M Y O Z 2 N E X N 0

Recent insights into emerging coronaviruses

Origins and evolutionary genomics of the 2009 swine-origin H1N1 influenza a epidemic

COVID-19 situation reports

Possible Bat Origin of Severe Acute Respiratory Syndrome Coronavirus 2. Emerg

Identifying SARS-CoV-2 related coronaviruses in Malayan pangolins

Isolation of SARS-CoV-2-related coronavirus from Malayan pangolins

Susceptibility of ferrets, cats, dogs, and other domesticated animals to SARS-coronavirus 2. Science (80-. )

SARS-CoV-2 neutralizing serum antibodies in cats: a serological investigation

Transmission of SARS-CoV-2 in Domestic Cats

Infection of dogs with SARS-CoV-2

Pathogenesis and transmission of SARS-CoV-2 in golden hamsters

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Efficient Activation of the Severe Acute Respiratory Syndrome Coronavirus Spike Protein by the Transmembrane Protease TMPRSS2

The human cell atlas

SARS-CoV-2 entry factors are highly expressed in nasal epithelial cells together with innate immune genes

SARS-associated coronavirus transmitted from human to pig

The authors declare no competing interests.