key: cord-0733487-8l1vfsbc
authors: Liao, M.; Liu, Y.; Yuan, J.; Wen, Y.; Xu, G.; Zhao, J.; Chen, L.; Li, J.; Wang, X.; Wang, F.; Liu, L.; Zhang, S.; Zhang, Z.
title: The landscape of lung bronchoalveolar immune cells in COVID-19 revealed by single-cell RNA sequencing
date: 2020-02-26
journal: nan
DOI: 10.1101/2020.02.23.20026690
sha: 8f81392c0b84774a0907127a08737c94f780afb7
doc_id: 733487
cord_uid: 8l1vfsbc

The novel coronavirus SARS-CoV-2, etiological agent of recently named Coronavirus infected disease (COVID-19) by WHO, has caused more than 2, 000 deaths worldwide since its emergency in Wuhan City, Hubei province, China, in December, 2019. The symptoms of COVID-19 varied from modest, mild to acute respiratory distress syndrome (ARDS), and the latter of which is generally associated with deregulated immune cytokine production; however, we currently know little as to the interplay between the extent of clinical symptoms and the compositions of lung immune microenvironment. Here, we comprehensively characterized the lung immune microenvironment with the bronchoalveolar lavage fluid (BALF) from 3 severe and 3 mild COVID-19 patients and 8 previously reported healthy lung controls through single-cell RNA sequence (scRNA-seq) combined with TCR-seq. Our data shows that monocyte-derived FCN1+ macrophages, whereas notFABP4+ alveolar macrophages that represent a predominant macrophage subset in BALF from patients with mild diseases, overwhelm in the severely damaged lungs from patients with ARDS. These cells are highly inflammatory and enormous chemokine producers implicated in cytokine storm. Furthermore, the formation of tissue resident, highly expanded clonal CD8+ T cells in the lung microenvironment of mild symptom patients suggests a robust adaptive immune response connected to a better control of COVID-19. This study first reported the cellular atlas of lung bronchoalveolar immune microenvironment in COVID-19 patients at the single-cell resolution, and unveiled the potential immune mechanisms underlying disease progression and protection in COVID-19.

Since December, 2019, a disease outbreak caused by a novel coronavirus (now given the name SARS-CoV-2) started in Wuhan, and has quickly spread in China and subsequently inmany other countries [1] . The WHO has declared the new coronavirus (CoV) infection as a global public health emergency and named the coronavirus infected disease-19 . The genome analysis showed that the SARS-CoV-2 is a SARS-CoV like β-lineage coronavirus [2] and likely originates from bats [3] . SARS-CoV-2 uses human ACE2 protein as their receptors [4] , which explains its similarly high transmissibility. The clinical data indicated that the COVID-19 varied from asymptomatic to acute respiratory distress syndrome (ARDS), similar as SARS-CoV infection [5] [6] [7] . Although researchers are racing against time to develop vaccines and test anti-viral drugs in clinical trials [8] , there is no effective prophylactic and clinical treatment for COVID-19 yet.

Generally, the COVID-19 is less severe and less fatal than the SARS, however, some patients, especially aged populations with co-morbidities are prone to develop more severe symptoms and require emergent medical interventions [9] . It is not completely understood why some patients develop severe but others have mild or even asymptomatic diseases by the same SARS-CoV-2

infections. The immunopathogenesis of hCoVs-induced respiratory distress syndrome may involve deranged interferon production, hyper-inflammatory response and cytokine storms, inefficient or delayed induction of neutralizing antibody and specific T cell responses [10, 11] .

However, due to biosafety and ethics issues and technical limitations, most of the current knowledge was generated from animal model studies, and not directly from human subjects.

More investigations using patient samples are needed to explore the relevant protective or pathogenic mechanisms in clinical settings.

Here, we applied the emerging single-cell RNA sequence (scRNA-seq) and single-cell TCR-seq to comprehensively characterize the lung bronchoalveolar lavage fluid (BALF) cells from 6 of COVID-19 patients, including 3 severe and 3 mild cases. Our study depicts a high-resolution transcriptome atlas of lung resident immune subsets in response to SARS-CoV-2 infections. It reveals that compared to the lung immune microenvironment of mild symptom patients, monocyte-derived FCN1 + macrophages replacing the FABP4 + alveolar macrophages predominate macrophage lineage compostions in the severely damaged lung, which are highly inflammatory and potent chemokine producers. Furthermore, the lung resident highly expanded clonal CD8 + T All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint cells formed in the mildly infected patients support the notion that a rapid and robust adaptive immune response is potentially critical for controlling COVID-19.

To characterize the immune microenvironment of the SARS-CoV-2-infected lung, we performed scRNA-seq analysis of single cells in the lung BALF (37, 820 cells) using the 10X Genomics platform, from 3 of recovered mild cases and 3 of severe cases ( Figure 1A , Table 1 ). We also collected public available scRNA-seq data (43, 627 cells) of 8 normal lungs as control [12] . This dataset passed stringent high-quality filtering ( Figure S1 ), yielding a mean of 188K reads/cell with median gene and unique molecular identifier (UMI) counts of 2, 070 and 6, 852, respectively (Table S1 ). Figure 1D and S2B). We were also able to detect low levels of SARS-CoV-2 transcripts in various cell populations from severe COVID-19 patients but not mild cases and controls ( Figure S2C ). We assumed that these viral transcripts were likely ambient contaminations of viral loads in the BALF.

We then compared the distribution of different cellular compartments among control, mild and severe group. There were higher proportions of T and NK cells in the COVID-19 patients than those in controls, while epithelial cells in patients are fewer. As compared to mild cases, severe patients contained more macrophages but less proportion of T and NK cells ( Figure 1E-1F ).

Together, our data showed an increased recruitment of immune cells to the lung in response to SARS-CoV-2 infection, and that the lung immune cell compartments differed between mild and severe COVID-19 patients.

All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint

The increased lung macrophage population was present in severe COVID-19 patients. To further understand the macrophage heterogeneity, we re-clustered the macrophages and showed 22 clusters (Figure 2A ). Recent studies have identified three distinct human lung macrophage subsets by FCN1 (monocyte derived), SPP1 (pro-fibrotic) and FABP4 (alveolar macrophage, AM) marker expression [13] . Referring to the classification criteria and other macrophage markers ( Figure S3A ), we found that most macrophage clusters (except for C15 as a dendritic cell subset) in our study could also be grouped by FCN1, SPP1 and FABP4 expression patterns. Group 1 includes C9 and C10 (FCN1 hi only), Group 2 includes C0, C6, C12, C16 and C19 (FCN1 lo SPP1 + ), Group 3 includes C7, C8 and C18 (FCN1 -SPP1 + ), Group 4 contains C1, C2, C3, C4, C5, C11, C13, C14, C17, C20, C21, C22 (FABP4 + ) ( Figure 2B ).

We found that the composition of the macrophage groups differed significantly among controls, mild versus severe COVID-19 patients ( Figure 2C ). This transformation was also reflected by differential expression of FCN1, SPP1 and FABP4 genes in the controls and patients ( Figure 2D ).

We further confirmed that FCN1 was preferentially expressed by individual controls and mild COVID-19 patients, while SPP1 and FABP4 were highly expressed by severe COVID-19 patients ( Figure 2E ). Together, these data showed a dysregulated balance of lung macrophage populations during the progression of severe COVID-19, manifested by substantially increased group 1 monocyte-derived macrophages (FCN1 + ), group 2 & 3 macrophages (SPP1 + ), and almost a complete loss of lung AMs (FABP4 + ).

To further understand the difference of the aforementioned 4 groups of macrophages, we conducted gene expression (GEX) analysis ( Figure 2F ). Group 1 expressed S100A8, S100A9, VCAN, Figure 2G and S3B). Using SCENIC, we deduced the transcription factors (TFs) activity in macrophage groups ( Figure 2H ). Consistent with their inflammatory phenotype and functions, we All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint found enhanced STAT1, STAT2 and multiple IFN regulatory factors (IRFs) activity among group 1, 2 and 3 macrophages. We also identified M2-like TFs, MAF and CREB1, specifically in group 3 macrophages, and AM-TFs, PPARG and BHLHE41, specifically in group 4 macrophages. Finally, we analyzed the differentiation trajectory of macrophage populations, and found a linear path starting from group 1, then 2 & 3, eventually to group 4 ( Figure 2I ).

Collectively, consistent with early report [13] , we identified the group 1 as the monocyte-like population, group 3 as the recently reported pro-fibrotic SPP1 + population and group 4 as the lung AM. Moreover, we identified a novel intermediate macrophage population (group 2), only from the severe COVID-19 patients. Our data indicated that the monocytes are recruited from circulation (FCN1 + ) to the lung to fuel the inflammation during severe diseases, and some monocytes may further go through the differentiation process into the SPP1 + populations and eventually the FABP4 + AMs.

robust and early T cell response played crucial roles in viral clearance during acute respiratory infections [14] . After exclusion of contaminated clusters, our clustering analysis of T and NK lymphocytes identified 5 major subsets ( Figure 3A ) based on the typical T and NK cell markers ( Figure 3B ). NK cells highly expressed KLRF1, KLRC1 and KLRD1. CD8 + T cells expressed CD8A and CD8B. The CCR7 + cells expressing CD4 and IL7R were likely CD4 + T cells, Treg cells were CTLA4 + FOXP3 + IL2RA + , and proliferating cells were TYMS + MKI67 + .

Comparing few numbers of NK and T lymphocytes in controls, the proportions of NK and T lymphocytes in the lung was largely increased in COVID-19 patients ( Figure 1F and 3C). Among different subsets, we observed lower proportion of CD8 + T cells, and increased proportion of proliferating cells in severe patients than the mild patients ( Figure 3C and 3D). GEX analysis showed that CD8 + T cells expressed high levels of effector molecules, including GZMA, GZMK, FASLG, CCL5, etc., and resident T cell markers ITGA1, CXCR6, JAML, etc., likely played crucial roles in immune defense ( Figure S4 ).

The higher number of lung CD8 + T cells in mild patients may indicate a role of CD8 + T cells in the clearance of the virus. In contrast, the higher level of T cell proliferation in severe disease group is All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint intriguing. We assumed that it might be related to bystander activation induced by hyper inflammation, or a delayed response to the infection in severe COVID-19. To further understand the relevant functional status of each NK and T cell subsets between mild and severe patients, we conducted GEX analysis. We found T cell activation, migration, calcium ion signaling molecules were upregulated in CD8 + T cells from mild cases, but the virus responses, metabolites and energy generation, and translation initiation molecules was upregulated in CD8 + T cells from severe cases ( Figure 3E and 3F). Thus, these data confirm our assumption that more effective CD8 + T cells in mild patients help to clear the viruses, while CD8 + T cells in severe patients are in preparation to proliferate.

When encountering antigens, specific B and T lymphocytes respond through clonal expansion to generate effector and memory cells with the same specificity to deal with the invaded pathogens.

Expanded anti-viral T cell clones may exert protective or pathogenic functions during viral infections dependent on different situations. Here, we took advantages of single-cell TCR-seq to assess the status of clonal expansions in the BALF of patients. As expected, CCR7 + T cells as naive or central memory cells, showed little clonal expansion, while CD8 + effector T cells showed the highest expansion levels among different T cell subsets ( Figure 4A and S5). Proliferating T cells and Treg cells also showed some expansion, but were less pervasive. Next, we examined the proportions of expanded clones between COVID-19 patients with mild and severe diseases. We found that the expansion levels in both total T and CD8 + T cells were significantly higher in mild disease group than the severe patients ( Figure 4B and 4C, S5). In average, more than 50% CD8 + T cells in mild group were expanded clones, which likely represented the SARS-CoV-2-specific T cells. They also showed higher amplification index than those from severe patients ( Figure 4C ). At the individual levels, we also found the higher T cell clonality was consistently remained in the 3 mild patients as compared to the 3 severe patients ( Figure 4D ), supporting that highly expanded CD8 + T cell participated in resolving the SARS-CoV-2 infection. To confirm the functional status of expanded T cell clones, we performed GEX analysis between expanded CD8 + T cells versus non-expanded cells. We found increased expression of signaling and tissue resident genes including XCL1, XCL2, ZNF683, HOPX, CXCR6 and ITGAE, etc., which further supports a superior All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint effector functions of those expanded CD8 + T cells ( Figure 4E ).

The clinical symptoms of COVID-19 varied from asymptomatic to ARDS, which has been similarly observed in SARS-CoV, MERS-CoV and influenza infections [7] . Host immune responses on some extent determine both protection and pathogenesis to the respiratory viral infections [11, 14] . A well co-ordinated innate and adaptive immune response may rapidly control of the virus, while a failed immune response leads to viral spreading, cytokine storm, and high mortality [15] .

However, the immune players and the detailed molecular mechanisms related to these different processes are incompletely understood yet, in SARS-CoV-2 infection.

Previous studies have shown the significantly higher levels of inflammatory cytokines, chemokines and interferon stimulated genes associated with disease severity in SARS, MERS and COVID-19 patients [10, 16] . The increased infiltrating macrophages into the lung are identified in the autopsy and animal models, and are thought to be responsible for fueling the inflammation [17] [18] [19] . Macrophages are well known to contain heterogeneous subsets, including tissue resident and monocyte derived populations, covering a broad spectrum from M1 to M2 like phenotype [20] . However, little is known regarding the macrophage subsets and their molecular features in SARS-CoV-2-infected human lungs. Here, we identified 4 groups of lung macrophages which can be classified by FCN1, SPP1 and FABP4 markers according to recent reports [13] . Group 1 & 2 macrophages are FCN1 + and highly inflammatory. They express higher levels of interferon stimulated genes and multiple chemokines, CCL2, CCL3, CCL5, IL-8, CXCL9, CXCL10 and CXCL11, thus identifying the FCN1 + macrophages as the culprit for the deranged hyper-inflammation. We identified the group 3, SPP1 + macrophages likely represented a repaired but also a pro-fibrotic subset [12] . They may counteract the FCN1 + macrophages to dampen the inflammation. Their roles and associations with patient outcomes should be further investigated. The group 4 macrophages are FABP4 + AMs showing higher PPARγ activity which play important physiological roles in metabolizing lipid surfactant [20] . Here, we found that AMs consisted the principle lung macrophages in both controls and mildly infected patients, but was almost completely lost in severely infected lungs. The loss of AMs likely contributed to the failed lung functions. Together, these data improved our understanding of tissue macrophage heterogeneity in the immunopathogenesis of SARS-CoV-2 infection. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02.23.20026690 doi: medRxiv preprint

Adaptive immune system is specific and memorizes the pathogens, including two arms, the antibody and T cell responses. Inducing adaptive immunity is also the aim of vaccination. SARS studies showed that binding and neutralizing antibodies are elicited in SARS-CoV infections, but their association with clinical outcomes is unclear [21] . Robust antibody responses were developed in severely infected SARS patients [22] . However, it is debated whether antibody-dependent enhancement played roles in disease exacerbation [23, 24] . Memory T cell responses are induced and maintained in SARS-CoV infected subjects [25] . T cell immunity was also confirmed to be necessary for resolving the viral infection in mouse models [26, 27] .

However, the roles of T cells, especially resident T cells, in controlling the disease severity of SARS-CoV infected patients are not proved yet. Here, we applied both scRNA-seq and TCR-seq to characterize the lung resident T cell populations. Our data not only revealed the presence of a larger population of CD8 + T effectors in mild group, but also showed that these CD8 + T cells contain highly expanded clones, indicating their SARS-CoV-2 specificity. Thus, we presented the fresh evidence that CD8 + T cell response likely holds the key for successful viral control in COVID-19 patients. This data will guide the future development of vaccine and potential T-cell immunotherapy for the severe viral respiratory infections.

The shortcomings of our study is that only 6 samples are able to be investigated and our endpoint study may not determine the triggering factors leading to different clinical outcomes. We propose that characterizing the earlier immune events and longitudinal analysis of immune dynamics of the infected patients may help gain more insights. These emerging high dimensional techniques, like scRNA-seq, mass cytometry and multiplex imaging tools [28, 29] , will greatly improve human immunology studies using minimal precious tissue samples. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is 

The throat swab, sputum, nasal swab and BALF were collected from the respiratory tract of the patients at various time-points after hospitalization. Viral RNAs were extracted by Viral RNA was extracted from samples using the QIAamp RNA Viral Kit (Qiagen, Heiden, Germany). Then quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed using the primers and probes targeting ORF1ab and N genes of SARS-CoV-2 recommended by China CDC, and a commercial kit specific for SARS-CoV-2 detection (GeneoDX Co., Ltd., Shanghai, China). The specimens were considered tentatively positive if qRT-PCR Ct value lower than 37 is detected at any time-points during the hospitalization. Then the assay is repeated to confirm the positivity if the Ct value was similar or between 37~40. If the repeat Ct was above 40, then the specimen was considered negative.

About 20 ml of BALF was obtained and placed on ice. BALF was processed within 2 hours and all All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint operations were performed in BSL-3 laboratory. By passing BALF through a 100 µm nylon cell strainer to filter out lumps and then the supernatant was centrifuged and the cells were re-suspended in the cooled RPMI 1640 complete medium. Then the cells were counted in 0.4% trypan blued, centrifuged and re-suspended at the concentration of 2×10 6 /ml for further use. 

The lung scRNA-seq data from the healthy controls was acquired from the Gene Expression Omnibus (GEO) database under the series number GSE122960 [12] , which contains data of lung tissue from eight lung transplant donors generated using 3' V2 chemistry kit on Chromium Single cell controller (10xGenomics). Filtered feature-barcode matrix was used in the following analysis.

The Cell Ranger Software Suite (Version 3.1.0) was used to perform sample de-multiplexing, barcode processing and single-cell 5' UMI counting with human GRCh38 as the reference genome.

Specifically, splicing-aware aligner STAR [30] was used in FASTQs alignment. Cell barcodes were then determined based on distribution of UMI count automatically. Finally, gene-barcode matrix of all 6 donors and 8 healthy control was integrated with Seurat v3 [31] the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint patients) were left for following analysis.

The filtered gene-barcode matrix was normalized with LogNormalize methods in Seurat and analyzed by principal component analysis (PCA) using the top 2, 000 most variable genes. Then Uniform Manifold Approximation and Projection (UMAP) was performed on the top 50 principal components for visualizing the cells. Meanwhile, graph-based clustering was performed on the PCA-reduced data for clustering analysis with Seurat v3.

MAST [32] in Seurat v3 was used to perform differential analysis. For each cluster, differentially-expressed genes (DEGs) were generated relative to all of the other cells.

Slingshot [33] was used to perform pseudotime inference for the four myeloid cell groups.

Single cell regulatory network for 4 myeloid groups was constructed with SCENIC [34] . Specifically, GRNBoost2 (https://github.com/tmoerman/arboreto) in pySCENIC was applied to infer gene regulatory networks from expression data. The regulators-group heatmap was generated with R package AUCell.

For DEGs, Gene ontology (GO), KEGG pathway analyses and Gene Set Enrichment Analysis (GSEA) [35] were performed with clusterProfiler [36] , which supports statistical analysis and visualization of functional profiles for genes and gene clusters. In GSEA analysis, 50 hallmark gene sets in MSigDB [37] were used for annotation.

Full-length TCR V(D)J segments were enriched from amplified cDNA from 5' libraries via PCR All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint amplification using a Chromium Single-Cell V(D)J Enrichment kit according to the manufacturer's protocol (10x Genomics). The TCR sequences for each single T cell were assembled by Cell Ranger vdj pipeline (v3.1.0), leading to the identification of CDR3 sequence and the rearranged TCR gene.

Cells with both TCR alpha and beta chains were kept and cells with only one TCR chain were discarded. The quality of TCR-seq was listed in Table S2 . Then the clonality of the different T cells defined by scRNA-seq data was explored.

All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint HC1-HC8: healthy controls, M1-M3: mild cases, S1-S3: severe cases. All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is Dimensionality reduction produced by UMAP was used in lineage inference.

All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint Only genes specifically upregulated in T cells compared to macrophage cells are kept. And a gene is considered significant with adjust p < 0.05.

F. The GO BP enrichment analysis of DEGs of CD8 + T cells between severe vs. mild COVID-19 patients with adjust p < 0.05 upregulated in severe (left) and mild (right).

All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02.23.20026690 doi: medRxiv preprint Table 1 Clinical data of the enrolled subjects   T141  T142  T143  T144  T145  T146   Severity  mild  mild  severe  mild  severe  severe   Age  36  37  66  35  62  63   Gender  male  female  male  male  male  male   Wuhan exposure history  yes  yes  yes  yes  yes  yes Symptom onset day Jan 9, 2020 Jan 11, 2020 Jan 3, 2020 Jan 9, 2020 Jan 11, 2020 Jan 8, 2020

First the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint 

A Novel Coronavirus from Patients with Pneumonia in China

Genome Composition and Divergence of the Novel Coronavirus (2019-nCoV) Originating in China

Receptor recognition by novel coronavirus from Wuhan: An analysis based on decade-long structural studies of SARS

A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster

Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus-Infected Pneumonia in

Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro

Pathogenic human coronavirus infections: causes and consequences of cytokine storm and immunopathology

Immunopathogenesis of coronavirus infections: implications for SARS

All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is

Single-Cell Transcriptomic Analysis of Human Lung Provides Insights into the Pathobiology of Pulmonary Fibrosis

Proliferating SPP1/MERTK-expressing macrophages in idiopathic pulmonary fibrosis

Immune responses in influenza A virus and human coronavirus infections: an ongoing battle between the virus and host

Overlapping and discrete aspects of the pathology and pathogenesis of the emerging human pathogenic coronaviruses SARS-CoV, MERS-CoV, and 2019-nCoV

Dysregulated Type I Interferon and Inflammatory Monocyte-Macrophage Responses Cause Lethal Pneumonia in SARS-CoV-Infected Mice

Multiple organ infection and the pathogenesis of SARS

Lung pathology of fatal severe acute respiratory syndrome

Origin and ontogeny of lung macrophages: from mice to humans

Antibody responses against SARS coronavirus are correlated with disease outcome of infected individuals

Recovery from the Middle East respiratory syndrome is associated with antibody and T-cell responses

Molecular Mechanism for Antibody-Dependent Enhancement of Coronavirus Entry

All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is

T-cell immunity of SARS-CoV: Implications for vaccine development against MERS-CoV

Virus-specific memory CD8 T cells provide substantial protection from lethal severe acute respiratory syndrome coronavirus infection

T cell responses are required for protection from clinical disease and for virus clearance in severe acute respiratory syndrome coronavirus-infected mice

Dissecting the human immune system with single cell RNA sequencing technology

A systems approach to infectious disease

STAR: ultrafast universal RNA-seq aligner

MAST: a flexible statistical framework for assessing transcriptional changes and characterizing heterogeneity in single-cell RNA sequencing data

Slingshot: cell lineage and pseudotime inference for single-cell transcriptomics

SCENIC: single-cell regulatory network inference and clustering

Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles

clusterProfiler: an R package for comparing biological themes among gene clusters

The Molecular Signatures Database (MSigDB) hallmark gene set collection

Single-cell reconstruction of the early maternal-fetal interface in humans

All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02. 23.20026690 doi: medRxiv preprint