Acute SARS-CoV-2 Infection is Highly Cytopathic, Elicits a Robust Innate Immune Response and is Efficiently Prevented by EIDD-2801

All known recently emerged human coronaviruses likely originated in bats. Here, we used a single experimental platform based on human lung-only mice (LoM) to demonstrate efficient in vivo replication of all recently emerged human coronaviruses (SARS-CoV, MERS-CoV, SARS-CoV-2) and two highly relevant endogenous pre-pandemic SARS-like bat coronaviruses. Virus replication in this model occurs in bona fide human lung tissue and does not require any type of adaptation of the virus or the host. Our results indicate that bats harbor endogenous coronaviruses capable of direct transmission into humans. Further detailed analysis of pandemic SARS-CoV-2 in vivo infection of LoM human lung tissue showed predominant infection of human lung epithelial cells, including type II pneumocytes present in alveoli and ciliated airway cells. Acute SARS-CoV-2 infection was highly cytopathic and induced a robust and sustained Type I interferon and inflammatory cytokine/chemokine response. Finally, we evaluated a pre-exposure prophylaxis strategy for coronavirus infection. Our results show that prophylactic administration of EIDD-2801, an oral broad spectrum antiviral currently in phase II clinical trials for the treatment of COVID-19, dramatically prevented SARS-CoV-2 infection in vivo and thus has significant potential for the prevention and treatment of COVID-19.


Main Text
The recently emerged human pandemic coronavirus Severe Acute Respiratory Syndrome coronavirus-2 (SARS-CoV-2), the causative agent of coronavirus disease (COVID)-19, has spread to six continents resulting in substantial morbidity and mortality worldwide 1 . Bats serve as a natural reservoir for coronaviruses and are the presumed source of SARS-CoV-2 and the highly pathogenic human coronaviruses SARS-CoV and Middle East Respiratory Syndrome (MERS)-CoV 2 . Transmission of coronaviruses from bats to other species is well-documented and adaptation in an intermediary host can facilitate their transmission to humans 2 . While it is possible that SARS-CoV-2 was transmitted to humans via an intermediate host such as pangolins, phylogenic analysis indicates that the SARS-CoV-2 lineage has circulated in bats for decades and evolved in bats into a virus capable of replicating in human cells 3 . Thus, bats are a potential reservoir for coronaviruses with human pandemic potential that can be directly transmitted to humans. Given the repeated and accelerating emergence of highly pathogenic coronaviruses, it has become increasingly important to monitor and characterize coronaviruses circulating in bats and to identify the viral determinants of human infection, disease, and global spread as well as to develop effective therapeutic interventions. Animal models are useful in studying highly pathogenic human coronaviruses, the emergence potential of zoonotic coronaviruses, and to evaluate novel inhibitors for their ability to control coronavirus infection [4][5][6][7][8][9][10][11][12][13][14][15] . However, human coronaviruses do not replicate in mice without either extensive adaptation of the virus or genetic modi cation of the host by genetic editing of the receptor or by introducing the individual human receptor genes for each virus [4][5][6][7][8][9][10][11]14,15 . Although existing rodent models have made several important contributions to our understanding of coronavirus infection and pathogenesis, none of these models possess the diverse set of primary human cells present in the human lung that can serve as targets for viral infection 16 . Here, we show that human lung-only mice (LoM), immune de cient mice implanted with authentic human lung tissue 17 , allow for the in vivo study of all recently emerged human coronaviruses (SARS-CoV, MERS-CoV, SARS-CoV-2) in a single platform that permits direct comparison of experimental outcomes. Using LoM, we also established that bats harbor novel coronaviruses capable of e cient replication in human lungs without prior adaptation. In addition, we performed an in-depth in vivo analysis of acute SARS-CoV-2 infection in the human lung. Our results revealed robust SARS-CoV-2 replication, pathogenesis and sustained activation of the innate host immune response. Finally, we used this platform for the in vivo evaluation of EIDD-2801, an orally administered broad spectrum antiviral currently in phase II clinical trials for COVID-19 treatment, to prevent SARS-CoV-2 infection. Our results show that EIDD-2801 e ciently prevented SARS-CoV-2 infection in vivo strongly supporting its progression in clinical development for COVID-19.

Human and bat coronavirus replication in LoM
LoM are constructed by subcutaneous implantation of a small piece of human lung tissue into the back of immune de cient mice (Fig. 1a). Previously, we demonstrated that the human lung tissue expands to form a highly vascularized palpable implant 17 (Fig. 1a). These lung implants contain human broblast, epithelial, endothelial, and mesenchymal cells that form highly relevant lung structures including cartilaginous and non-cartilaginous bronchial airways lined with ciliated and non-ciliated epithelium, alveolar sac structures, and extensive vasculature 17 (Extended Data Fig. 1a,b). We also showed that the human lung tissue of LoMs supports replication of a diverse set of emerging and clinically relevant human pathogens including Zika virus, human cytomegalovirus, respiratory syncytial virus and MERS-CoV 17 . Recently emerged human coronaviruses have used at least two different receptors to gain entry into host cells, human angiotensin converting enzyme-2 (ACE2) and dipeptidyl peptidase 4 (DPP4). SARS-CoV and SARS-CoV-2 use human ACE2 as a receptor while MERS-CoV uses DPP4 18-22 . These differences in receptor usage in uence viral tropism, pathogenesis and disease progression 23 . Infection of LoM with MERS-CoV resulted in robust virus replication and infection of human epithelial, endothelial and mesenchymal cells in vivo 17 . These ndings were consistent with the broad cellular distribution of its receptor, DPP4 24 .
Here, we rst evaluated the potential of LoM to serve as a single platform to study all known recently emerged human coronaviruses and the potential of endogenous bat coronaviruses for human emergence. We rst con rmed that ACE2, the receptor for SARS-CoV and SARS-CoV-2 in human cells, was present on the surface of human epithelial cells (cytokeratin 19+) in the human lung tissues of LoM (Extended Data Fig. 1c,d). Next, LoM were inoculated with recently emerged coronaviruses SARS-CoV, MERS-CoV, or SARS-CoV-2 (Extended Data Table 1). LoM supported replication of all these viruses in vivo.
Consistent with our previous results 17 , we also observed robust replication of MERS-CoV in the human lung tissues of LoM with mean titers of 4.79x10 8 PFU/g in LoM human lung tissues at 2 days postinfection (Fig. 1b) and abundant viral antigen (Extended Data Fig. 2a).
Pre-pandemic bat coronaviruses WIV1-CoV and SHC014-CoV have high sequence homology to SARS-CoV, use ACE2 to infect human cells, and grow modestly in primary human airway cultures on liquid interface 4,5 . LoM were inoculated with WIV1-CoV or SHC014-CoV and virus titers in human lung tissues measured 2 days post-infection (Extended Data Table 1). WIV1-CoV and SHC014-CoV e ciently replicated in the human lung tissue of LoM with mean titers of 1.58x10 7 and 1.48x10 7 PFU/g, respectively (Fig. 1b) and viral antigen was readily detected in human lung tissues (Extended Data Fig.  2b). Collectively, these results demonstrate that LoM serve as single platform where all newly emerged coronaviruses SARS-CoV, MERS-CoV, and SARS-CoV-2 replicated e ciently in human lung tissue. Importantly, the fact that SARS-like bat coronaviruses WIV1-CoV and SHC014-CoV also replicated e ciently in the LoM platform indicates that bats harbor coronaviruses that are potentially capable of direct transmission to humans, thus bypassing the need for further adaptation in an intermediary host.

SARS-CoV-2 replication in LoM
Given the state of the current COVID-19 pandemic and the urgent need to develop therapeutic and preventative approaches to control and prevent infection, we evaluated replication of SARS-CoV-2 in LoM in detail. Human lung tissues of LoM were inoculated with SARS-CoV-2 and titers of replication competent virus determined 2, 6, and 14 days post-exposure (Fig. 1c, Extended Data Table 2). High titers of replication competent virus were noted at all time points although they were highest 2 days postinfection (Fig. 1d). The distribution of virus-infected cells was determined with RNAscope (viral RNA) and immuno uorescence microscopy (viral nucleoprotein). Virus infection was widely distributed throughout the tissue with large numbers of cells positive for viral RNA (Fig. 1e) and nucleoprotein (Fig. 1f). Costaining with a human cytokeratin 19 antibody demonstrated that SARS-CoV-2 predominantly infects human epithelial cells in the lung (Fig. 1g). Viral antigen was not detected in human CD34 expressing (endothelial) cells, and only a few human vimentin expressing (mesenchymal) cells were positive for viral nucleoprotein (Fig. 1g). To identify the epithelial cell types in the lung tissue that are susceptible to SARS-CoV-2 infection, we further identi ed infected cells by assessing co-localization of viral nucleoprotein with antibodies against acetylated alpha-tubulin IV (ciliated cells), CC10 (club cells), HT1-56 (alveolar type I [AT1] pneumocytes), and pro-SP-C (alveolar type II [AT2] pneumocytes) (Fig. 1h). We were able to clearly identify virus antigen in cells which expressed pro-SP-C or acetylated alpha-tubulin IV; we did not detect virus antigen in HT1-56 or CC10 positive cells (Fig. 1h). These results demonstrate that SARS-CoV-2 has limited tropism in the lung with AT2 pneumocytes and ciliated airway epithelial cells being the predominant lung cells infected by virus.

SARS-CoV-2 pathogenesis in LoM
To evaluate the cytopathogenic effects of SARS-CoV-2 during acute infection of human lung tissue in LoM, we used a combination of histological analysis and electron microscopy. Histopathologic analysis revealed several features of early diffuse alveolar damage that have been described in lung tissues of COVID-19 patients including the accumulation of proteinaceous exudate and brin in alveolar spaces, desquamation of pneumocytes, multi-nucleated cell formation, and the appearance of brin thrombi in small vessels (Fig. 2) 25-27 . Proteinaceous exudate, including large globules of protein material, was observed in alveolar spaces, which overlapped with areas of virus accumulation (Fig. 2a,b). As early as 2 days post-infection, desquamation of pneumocytes was also noted; there were a large number of virally infected cells fully detached or detaching from the alveolar basement membrane into the alveolar space ( Fig. 2c,d). Infected multi-nucleated cells were also observed (Fig. 2c). While the formation of hyaline membranes was not noted, brin was detected in alveolar spaces (Fig. 2e). Importantly, we observed multiple occluded vessels containing brin thrombi as reported in the lungs of COVID-19 patients ( Fig.   2f,g) [25][26][27] . Electron microscopy demonstrated the normal architecture and integrity of uninfected AT2 pneumocytes that were present in human lung tissue obtained from LoM two days post-infection ( Fig.  2h). In contrast, AT2 cells containing virus particles in the same sample had swollen mitochondria with loss of matrix and cristae as well as rough endoplasmic reticula with distended cisternae, protein accumulation, and virus particles (Fig. 2i). Degenerative SARS-CoV-2 infected AT2 cells detached from the alveolar basal membrane could also be observed in the alveolar luminal space (Fig. 2j). Higher magni cation revealed subcellular accumulation of virus containing vesicles indicative of virus replication and egress. Virions with electron dense nucleocapsids and distinctive crown-like spikes were observed (Fig. 2i,j). Consistent with previous reports in human airway epithelial cell cultures and portmortem lung samples 26,28 , virions produced by human lung cells were pleomorphic in size (69 to 112 nm) and shape. Despite the extensive damage in icted in the lung tissue by the virus, the endothelium in the majority of blood vessels was intact with tight junctions, numerous pinocytotic vesicles, and normal mitochondria and endoplasmic reticulum (Fig. 2k,l). Virions were not detected within endothelial cells in agreement with a lack of infection as per our immuno uorescence analysis ( Fig. 1g and Fig. 2k,l). However, pleomorphic virions were present in capillary lumen surrounded by brillar protein deposits and cell debris (Fig. 2k,l). Together, these results demonstrate that acute SARS-CoV-2 infection of LoM closely resembles infection of human lung in humans and is highly cytopathic resulting in signi cant injury to the fragile alveolar lung structures.
To determine the effect of SARS-CoV-2 infection on human gene transcription, we performed RNAsequencing analysis of human lung tissues collected from animals 2, 6 and 14 days post-infection. Abundant viral transcripts were detected in infected lung tissue, ranging from 0.55% to 3.6% of the total reads at 2 days post-infection (Extended Data Table 3). Viral transcripts were still abundant but lower at 6 days and 14 days post-infection (Extended Data 3). Sequencing data was consistent with previously identi ed canonical SARS-CoV-2 transcripts 29 and con rmed maintenance of the furin cleavage site throughout the course of infection in vivo. Analysis of human gene transcripts revealed 1,504 differentially expressed cellular genes between naïve and infected human lung tissue at 2 days postexposure, the peak of infection (Fig. 3a, Supplementary Tables 1 and 2) (adjusted p value <0.05 after correcting for multiple testing). Of these, 1,043 genes were up-regulated and 461 genes were down regulated in the infected human lung tissue relative to non-infected lung tissue (Fig. 3a, Supplementary  Tables 1 and 2). Notably, numerous interferon-stimulated genes (ISGs) and in ammatory cytokine genes, including pro-in ammatory cytokines genes IL6, CXCL8 (IL-8), CXCL10 (IP-10), TNF, and CCL5 (RANTES) were potently induced in infected lung tissue (Supplementary Tables 1 and 2). We also observed dramatic upregulation of IFNB1 expression (>1,000 fold) at 2 days post-exposure, suggesting that this cytokine plays a key role in the antiviral response to SARS-CoV-2 (Supplementary Tables 1 and 2). Gene set enrichment analysis (GSEA) showed over 840 gene pathways signi cantly upregulated (p<0.05) including response to type 1 interferon (p=0.0011), response to virus (p=0.0010), innate immune response (p=0.0010), cytokine mediated signaling (p=0.0010), cytokine production (p=0.0010), response to stress (p=0.0010), in ammatory response, (p=0.0010), NIK NF-KB signaling (p=0.0011), acute in ammatory response (p=0.0035), regulation of cell death (p=0.0030), and coagulation pathways (p=0.0453) (Fig. 3b).
Complement activation, which contributes to SARS-CoV pathogenesis in mouse models 14 , was also increased (p=0.0470) (Fig. 3b). Importantly, analysis of host gene expression at later time points demonstrated a sustained upregulation of antiviral and in ammatory genes that in some instances (e.g. ISG15, IFITM1, TNF, CXCL9) persisted for up to 14 days post-infection (last time analyzed) (Fig. 3c,d, Extended Data Table 4, Supplementary Tables 1 and 2). These results demonstrate that acute SARS-CoV-2 infection causes a potent and sustained upregulation of innate immune responses in virus-infected human lung tissue.

EIDD-2801 pre-exposure prophylaxis
Currently, there is no vaccine to prevent SARS-CoV-2 infection or effective therapy to treat patients with COVID-19. The ribonucleoside analog β-D-N 4 -hydroxycytidine (NHC) has been shown to broadly inhibit coronavirus infection in vitro in human airway epithelial cell cultures, with potent activity against SARS-CoV-2 as well as SARS-CoV, MERS-CoV, and bat SARS-like and MERS-like coronaviruses 7 . We therefore tested the ability of prophylactic EIDD-2801 (also known as MK-4482), the oral pro-drug of NHC, to inhibit SARS-CoV-2 replication in vivo. For this purpose, LoM were administered EIDD-2801 starting 12 h prior to SARS-CoV-2 exposure and every 12 h thereafter (Fig. 4a, Extended Data Table 5). Our results show that EIDD-2801 had a dramatic effect on virus infection, signi cantly reducing the number of infectious particles in the human lung tissue of EIDD-2801 treated animals in two independent experiments (Fig.   4b,c) by over 100,000 fold (Fig. 4c,d). Furthermore, in contrast to EIDD-2801 treated mice, abundant cell debris and nucleoprotein positive cells could be readily observed in the alveolar lumen of vehicle control treated mice consistent with the extensive pathogenic effects in icted on the lung by SARS-CoV-2 (Fig.4e,f). These results demonstrate that prophylactic administration EIDD-2801 is highly effective at preventing SARS-CoV-2 infection and pathogenesis in vivo.

Discussion
The zoonotic transmission of the pathogenic SARS-CoV-2 resulted in a pandemic that has in icted signi cant morbidity and mortality as well as dire world-wide economic and social consequences. Herein, we describe a unique model for the in vivo study of human coronavirus infection particularly well suited to model distal human lung virus infection. Our results demonstrate replication of all known recently emerged human coronaviruses in LoM.
Importantly, our results demonstrate that WIV1-CoV and SHC014-CoV, two pre-pandemic endogenous bat viruses can replicate relatively e ciently in LoMs suggesting that coronaviruses circulating in bats have future pandemic potential without the need for further adaptation to the human host. We also show that Interestingly, an analysis of post-mortem COVID-19 patient lungs 12 did not reveal increased IFNB1 expression and it has been shown in vitro that its expression is blocked during SARS-CoV infection 33,34 . These results suggest that in human lung tissue IFNB1 gene expression is induced during the acute phase of SARS-CoV-2 infection. We also observed increased expression of several human cytokine genes in LoM infected with SARS-CoV-2. A substantial number of these cytokines were also increased in the serum of COVID-19 patients and post-mortem lung tissue samples further establishing the similarities between LoM and human infection with SARS-CoV-2 12,35 . Currently, there is no vaccine to prevent or therapeutics to treat COVID-19. Pre-exposure prophylaxis approaches to infectious diseases have proven to be highly e cacious and can contribute to reduce the risk of infection. The continued global spread of the virus and its associated morbidity and mortality are strong incentives to the development of prevention strategies for COVID-19. In this regard, NHC was shown to have broad activity against human and bat coronavirus infection in vitro 7 . In addition, prophylactic and therapeutic administration of its oral pro-drug EIDD-2801 reduced SARS-CoV and MERS-CoV replication and pathogenesis in mice 7 . Here we show that prophylactic administration of EIDD-2801 e ciently prevents SARS-CoV-2 infection in vivo highlighting its potential utility as an effective prophylactic agent against SARS-CoV-2 and other past and future zoonotic coronaviruses. There are some limitations of our study including the fact that LoM do not possess the human nasal airway structures that are thought to be early sites of SARS-CoV-2 replication in humans 36 . Since LoM do not have an autologous human adaptive immune system they re ect the direct effect of viruses on their targets and bystander cells as well as their innate immune response to infection. Collectively, our results indicate that LoM re ect the pathogenic effects in icted by SARS-CoV-2 on the human lung and demonstrate their utility as a single in vivo platform to evaluate and compare the replication and pathogenesis of past, present, and future pre-emergent, epidemic, and pandemic coronaviruses accelerating the development and testing of pre-exposure prophylaxis agents such as EIDD-2801.

Immunohistochemical analysis of coronavirus infection
Immunohistochemistry was performed as previously described 17 . Brie y, xed (10% formalin) human lung tissues collected from coronavirus-infected LoM were para n embedded and sectioned (5 um

Immuno uorescence analysis of SARS-CoV2 infection
Human lung tissues collected from mice were xed in 10% formalin and para n embedded.
Immuno uoresence staining of 5 um tissue sections was performed as previously described 17 . Brie y, following depara nization and antigen retrieval (Diva Decloaker), tissue sections were incubated with a 10% normal donkey serum solution with 0.1% Triton X-100 in 1x PBS to block non-speci c binding.
Tissue sections were then incubated overnight with primary antibodies at 4°C followed by incubation with uorescent conjugated secondary antibodies (Supplementary Table 7). Primary antibodies were directed against SARS nucleoprotein and human cytokeratin 19, CD34, vimentin, acetylated alpha-tubulin IV, CC10, HT1-56, and pro-SP-C (Supplementary Table 7). Background auto uorescence was then quenched using a 0.1% Sudan Black B solution in 80% ethanol prior to staining with DAPI. Slides were mounted and then imaged using an Olympus BX61 upright wide-eld microscope using Volocity software (version 6.3) with a Hamamatsu ORCA RC camera. Appropriate negative controls without primary antibodies were also imaged using the same exposure time as matching stained sections. Whole image contrast, brightness, and pseudocoloring were adjusted using ImageJ/Fiji (Version 2.0.0-rc-69/1.51w) and Adobe Photoshop Tissue pieces were washed in deionized water, dehydrated in ethanol, and placed through two exchanges of propylene oxide before in ltration and embedment in PolyBed 812 epoxy resin (Polysciences