key: cord-0888999-du8e0mx9
authors: Wyler, Emanuel; Adler, Julia M.; Eschke, Kathrin; Alves, Gustavo Teixeira; Peidli, Stefan; Pott, Fabian; Kazmierski, Julia; Michalick, Laura; Kershaw, Olivia; Bushe, Judith; Andreotti, Sandro; Pennitz, Peter; Abdelgawad, Azza; Postmus, Dylan; Goffinet, Christine; Kreye, Jakob; Reincke, S Momsen; Prüss, Harald; Blüthgen, Nils; Gruber, Achim D.; Kuebler, Wolfgang M.; Witzenrath, Martin; Landthaler, Markus; Nouailles, Geraldine; Trimpert, Jakob
title: Key benefits of dexamethasone and antibody treatment in COVID-19 hamster models revealed by single cell transcriptomics
date: 2022-03-24
journal: Mol Ther
DOI: 10.1016/j.ymthe.2022.03.014
sha: 3e9b1db3d23b7afdee994fa604885913a0f5aed6
doc_id: 888999
cord_uid: du8e0mx9

For COVID-19, effective and well-understood treatment options are still scarce. Since vaccine efficacy is challenged by novel variants, short-lasting immunity and vaccine hesitancy, understanding and optimizing therapeutic options remains essential. We aimed at better understanding the effects of two standard-of-care drugs, dexamethasone and anti-SARS-CoV-2 antibodies, on infection and host responses. By using two COVID-19 hamster models, pulmonary immune responses were analyzed to characterize effects of single or combinatorial treatments. Pulmonary viral burden was reduced by anti-SARS-CoV-2 antibody treatment, and similar or increased by dexamethasone alone. Dexamethasone exhibited strong anti-inflammatory effects and prevented fulminant disease in a severe disease model. Combination therapy showed additive benefits with both anti-viral and anti-inflammatory potency. Bulk and single-cell transcriptomic analyses confirmed dampened inflammatory cell recruitment into lungs upon dexamethasone treatment, and identified a specifically responsive subpopulation of neutrophils, thereby indicating a potential mechanism of action. Our analyses confirm the anti-inflammatory properties of dexamethasone and suggest possible mechanisms, validate anti-viral effects of anti-SARS-CoV-2 antibody treatment, and reveal synergistic effects of a combination therapy, thus informing more effective COVID-19 therapies.

Purpose and Study design 125 This study aims to compare two widely used COVID-19 treatments, dexamethasone and 126 monoclonal antibodies as well as a combination thereof. To this end, we employed two 127 COVID-19 hamster models, the Syrian and the Roborovski hamster, representing moderate 128 and more severe COVID-19 like disease, respectively. 24 individuals of both species were 129 experimentally infected with 1x10 5 plaque forming units (pfu) of the ancestral SARS-CoV-2 130 variant B.1 [BetaCoV/Germany/BavPat1/2020] and divided into 4 groups of 6 animals each 131 that received either mAb (30 mg, single treatment), dexamethasone (2 mg/kg/d), mAb (30 mg, 132 single treatment) and dexamethasone (2 mg/kg/d) or mock treatment (PBS, daily). Since the 133 course of disease varies considerably between both species, we choose to apply treatment at 134 the onset of clinical signs for each species, which is 24 h for Roborovski or 48 h post infection 135 for Syrian hamsters. To further account for species specific differences, we scheduled 3 136 animals per group for sampling at 3 and 5 days post infection (dpi) in case of Roborovski 137 hamsters or 5 and 7 dpi for Syrian hamsters. Due to early onset of severe disease, 2 mAb 138 treated and 1 mock treated Roborovski hamster reached defined humane endpoints at day 2 139 post infection and had to be terminated ahead of schedule. Clinical and virological parameters 140 were determined for each animal in this study, furthermore, lungs of Roborovski hamsters 141 taken at day 3 were subjected to scRNA sequencing to determine transcriptional response to 142 infection and treatment on a single cell level. 143

Dexamethasone treatment prevents severe disease, while monoclonal antibodies 144 decrease viral burden. 145 Following SARS-CoV-2 infection, Syrian hamsters lost body weight. Irrespective of treatment, 146

Syrian hamsters failed to show significant differences in body weight development nor did they 147 present with severe signs of disease ( Figure 1A , B). Titers of replication competent virus of all 148 hamsters receiving mAb or combination treatment was below the detectable level at all 149 sampling time points. The use of dexamethasone alone increased viral titers in the lungs of 150 J o u r n a l P r e -p r o o f Syrian hamsters and delayed viral clearance with moderately increased titers on day 5 and 151 significantly increased titers at 7 dpi ( Figure 1C ). The same trend was also evident in virus 152 gRNA levels in the lungs ( Figure 1D ), but not in the upper respiratory tract ( Figure 1E ), which 153 is the common site of sampling in patients. 154

Contrary to Syrian hamsters, Roborovski hamsters which can develop fulminant disease early 155 after infection 27 , displayed marked differences in clinical parameters in response to specific 156 treatments. Specifically, both dexamethasone alone and in combination with mAb protected 157

Roborovski hamsters from severe disease progression. By contrast, hamsters assigned to 158 mAb treatment (2/6 on 2 dpi) and animals receiving mock treatment (2/6 on 2 dpi or 3 dpi) had 159 to be euthanized prior to the terminal time point as they reached human endpoint criteria 160 ( Figure 1F ). Hamsters that developed severe disease in respective groups presented with 161 drastic drops in body temperature at 2 dpi ( Figure 1G ). Until the end of the experiment, body 162 weights in the dexamethasone treated groups remained stable, animals in the mAb treatment 163 group recovered from initial weight losses, while mock treated animals continued to lose weight 164 throughout the experiment ( Figure 1H ). Similar to Syrian hamsters, replicating virus was below 165 the detectable level in the lungs of Roborovski hamsters treated with either mAb or 166 combinatorial therapy at day 3 and 5 post infection. Only Roborovski hamsters that had to be 167 terminated at 2 dpi showed high titers of replication competent virus despite mAb treatment 168 ( Figure 1I ). In contrast to the results obtained from Syrian hamsters, no boost of viral replication 169 was observed in the dexamethasone treated group of Roborovski hamsters compared to mock 170 treated animals. This result was evident for all time-points, on both replicating virus and virus 171 gRNA level in the lungs as well as in the upper respiratory tract ( Figure 1J , K). 172 Figure E1 ). Inflammation and bronchitis scores were reduced at from 5 dpi on in all groups 180 receiving dexamethasone, which corresponds to 3 or 4 days post treatment start for Syrian 181 and Roborovski hamster, respectively ( Figure 2C -F). mAb treatment alone reduced 182 pneumonia, however, to a lesser extent as compared to dexamethasone (Figures 2A -F, E1) . 183

Next, we investigated how anti-viral and inflammatory transcriptional responses were 184 influenced by treatment in Syrian ( Figure 2G , E2A) and Roborovski hamsters ( Figure 2H , E2B) 185 over time. Therefore, we analysed previously established viral infection related gene sets, 186 response to type I interferon (IFN) and interferon-gamma (IFN-γ) 28, 29 . In Syrian hamsters, the 187 amplitude of the type I IFN response genes decreased from 5 to 7 dpi in the absence of 188 treatment ( Figure 2G , E2A). mAb treatment alone or in combination with dexamethasone led 189 to further reduction in gene expression of the type I IFN response genes. In contrast, response set genes decreased more upon dexamethasone compared to mAb treatment 191 ( Figure 2G , E2A). Similar effects were observed in Roborovski hamsters ( Figure 2H , E2B). 192

The combination treatment led to a strong reduction of both gene sets, independent of hamster 193 species ( Figure 2G ,H). 194

Taken together, treatment related improvement in clinical parameters and histopathology 195 correlated with substantially altered gene expression profiles in general, and a reduced 196 expression of the response to IFN-γ gene set following dexamethasone treatment specifically. 197 Dexamethasone reduces influx of immune cells and stabilizes endothelial cells 198 As described above, both mAb and dexamethasone treatment, and in particular their 199 combination, attenuated inflammatory aspects of pneumonia following SARS-CoV-2 infection, 200 thereby mitigating the otherwise severe disease observed in Roborovski hamsters. 201

In order to investigate cellular mechanisms underlying these treatment effects, we next 202 performed pulmonary scRNA-seq of Roborovski hamsters for all treatment groups at 3 dpi. 203

First, we evaluated the absolute content and composition of cell types by measuring total cell 204 counts of the dissociated tissue ( Figure 3A ) and relative cell type distribution from scRNA-seq 205 data ( Figure 3B -D, E3A-J). Lungs from dexamethasone (alone or in combination with mAb) 206 treated hamsters yielded significantly lower total cell counts ( Figure 3A ). This reduction likely 207 originated from reduced infection-triggered pulmonary immune cell immigration. NK cell 208 numbers were significantly lower in dexamethasone treated groups compared to mock and 209 mAb treated hamsters; similarly, neutrophil, monocytic macrophage, Treml4 + monocyte, T and 210 B cell showed reduced numbers in hamsters receiving dexamethasone, although the 211 difference was not statistically significant ( Figure 3B , C). Notably, endothelial cells had 212 significantly higher counts in groups treated with a combination therapy of dexamethasone and 213 mAb ( Figure 3D ) as compared to mock treated animals. Higher endothelial cell counts were 214 likely caused by mechanisms governing endothelial protection, rather than cell proliferation, 215 since increased Mki67 and Top2 expression was not detectable in endothelial cells ( For an unbiased view of the data, we selected all genes that were at least four-fold upregulated 234 in all cell types ( Figure 4A ). Again, monocytic macrophages and neutrophils stood out with 235 several upregulated genes, including Saa3 and F13a1 as mentioned above. We identified a 236 dexamethasone-induced transcriptional program common to several cell types, whereas some 237 genes, for example Gal (coding for galanin and galanin message-associated peptides) in 238 endothelial cells were cell-type specific. In contrast, tissue cells, including endothelial cells, 239 alveolar epithelial cell type 2 (AT2) or smooth muscle cells did not show substantial 240 upregulation of gene expression in response to dexamethasone alone ( Figure 4A ). Notably, 241 the mRNA of the glucocorticoid receptor, encoded by the Nr3c1 gene, is ubiquitously present 242 in both Roborovski hamsters and Syrian hamsters, and not modulated by SARS-CoV-2 243 infection or the employed treatments ( Figure E4D ). 244

Next, we asked which disease-relevant changes in gene expression were influenced by 245 treatment in different cell types. We therefore assessed changes in gene expression between 246 treatments for each cell type in an unbiased manner ( Figure E4E ). We noticed consistent 247 downregulation of a group of interferon-induced genes (ISGs) such as Ifit2/3, Ifi27, Ifi209 in 248 animals treated with mAb alone or in combination with dexamethasone, but not with 249 dexamethasone alone. Conversely, some genes, such as Tnfsf10 (coding for the pro-250 inflammatory cytokine Trail) in neutrophils were more reduced in dexamethasone treated 251 compared to mAb treated animals. 252

In order to understand the changes in gene expression patterns caused by these treatments, 253

we defined, based on our Syrian hamster scRNA-seq data 28 , two groups of gene sets. The 254 first was viral PAMP dependent (identified as "NF-kB-dependent"), the second induced by the 255 infection in general ("interferon-dependent") ( Figure E4F ). Whereas the "interferon-dependent" 256 gene expression was reduced more by mAb compared to dexamethasone treatment, for the 257 "NF-kB-dependent" gene set we in tendency observed the opposite ( Figure 4B ). We 258 scrutinized this effect in detail in monocytic macrophages and neutrophils, and found that in 259 neutrophils, the downregulation of the NF-kB-driven cytokine genes Cxcl10 and Tnfsf10 in 260 J o u r n a l P r e -p r o o f tendency experience stronger downregulation by dexamethasone compared to the ISG Mx2 261 ( Figure 4C ). For all genes, the combination treatment showed an additive effect ( Figure 4 ). 262

Overall, these data suggest that the reduced viral load in mAb-treated animals leads to a 263 generally reduced antiviral/type 1 interferon signal, whereas dexamethasone treatment 264 downregulates specific genes in some cell types, such as the pro-inflammatory cytokines 265 Among the genes that were particularly prominent in cluster 6 were the cytokines and 278 macrophage and lymphocyte attractants Csf1 and Ccl3 37,38 ( Figure E5C ). We therefore plotted 279 the expression of these two genes along with the ISG/NF-kB targets Mx2/Tnfsf10/Cxcl10, 280 which showed that neutrophils in cluster 6 express Csf1 and Ccl3 at particularly high levels 281 ( Figure 5D ), in the same time, these cells become less abundant upon dexamethasone and 282 particularly combination treatment ( Figure 5E ). Concomitantly, by histopathology analysis, we 283 observed less neutrophils in the dexamethasone treated groups ( Figure 5F ). Of note, cells 284 expressing mRNAs of receptors (Csf1r, Ccr1, Ccr4, and Ccr5) corresponding to cytokines Csf1 285 and Ccl3 were less abundant in the lungs upon dexamethasone treatment ( Figure E5D , 286 compare with Figure 3B and E3B). In addition, neutrophil-cluster 6 showed particularly low and 287 high expression of Il1r2 and Isg20 ( Figure E5E ), respectively, thereby recapitulating the 288 J o u r n a l P r e -p r o o f phenotypes seen for immunosuppressive and IFN active neutrophils in the peripheral blood of 289 COVID-19 patients 13 . 290

To generalize the observation of this transcriptional dynamic, we applied diffusion map 291 analysis of neutrophils to identify their most prominent direction of variation 39,40 ( Figure E5F ). 292

For each treatment, we show the neutrophil density along the diffusion axis ( Figure E5G , upper 293 part), which we defined as the first non-trivial component of the diffusion map. The directional 294 progression towards the right on this axis (which is the same cellular state represented as 295 neutrophil-cluster 6 above) is present in all conditions, as shown by the average RNA velocity 296 projected onto the diffusion axis ( Figure E5G , lower part). However, most neutrophils derived 297 from hamsters treated with dexamethasone or combinatorial treatment are found at the 298 leftmost part of the axis, whereas neutrophils from hamsters with mAb and mock treatment are 299 split into a left and right part, confirming that with dexamethasone treatment an otherwise 300 directional progression of neutrophils is limited. In order to relate the diffusion axis to biological 301 effects, we scored hallmark signatures 41 for every neutrophil and linearly correlated each 302 hallmark with the diffusion axis ( Figure E5H , upper part). In addition, we correlated the 303 expression profiles of each gene with the diffusion axis ( Figure E5H , lower part). These 304 correlations revealed that the drive towards neutrophil-cluster 6 marked by high expression of 305

Csf1 and Ccl3 and elevated amounts of viral RNA is accompanied by an increase of 306 interferon/inflammatory response gene expression (such as Isg15 or Cd274), and a decrease 307 in the levels of classical neutrophil marker genes such as S100a8/9 or Pglyrp1. 308

Dexamethasone limits this dynamic, effectively keeping the neutrophils in a stationary 309 transcriptomic state at the left part of the diffusion axis. As we will discuss in detail, this 310 stagnation could be a reason for the reduced production of lymphocyte attractants and 311 consequently, the reduction of lung infiltrates. 312

In this study, we examined the effects of separate and combined anti-viral and anti-314 inflammatory treatments for COVID-19 in two hamster models reflecting a moderate (Syrian 315 hamster) and severe (Roborovski hamster) disease course, respectively. Using histopathology 316 J o u r n a l P r e -p r o o f and bulk and single-cell transcriptomic analysis of hamsters subjected to dexamethasone, 317 monoclonal antibody, and combination treatment, we demonstrate treatment efficacy, and 318 identified a subset of neutrophils that express macrophage/lymphocyte attracting cytokines 319 and can be impeded by dexamethasone. with SARS-CoV-2 (1 x 10 5 pfu Wildtype (WT)) and treated once at 1 dpi with 30 mg/kg mAb 484 CV07-209 (mAb, n = 6), daily starting at 1 dpi with 2 mg/kg Dexamethasone (Dex, n = 6) or 485 received combination treatment (Dex + mAb, n = 6). Survival rates (F) in percent of SARS-486

CoV-2 infected Roborovski hamsters, body temperature (G) in degree Celsius and body weight 487 (H) development in percent after virus challenge were measured until planned analysis time 488 point (3 dpi, and 5 dpi) or until termination due to score sheet criteria (non-survivors) according 489 to treatment group. Two hamsters from the mAb group and one hamsters from the mock-490 treated group were euthanized at 2 dpi (represented by orange squares (I -K) Roborovski hamsters were challenged with SARS-CoV-2 (1 x 10 5 pfu Wildtype (WT)), treated 518 once at 1 dpi with 30 mg/kg mAb CV07-209 (mAb), daily starting at 1 dpi with 2 mg/kg 519 Dexamethasone (Dex) or received combination treatment (Dex + mAb). At 3 dpi n = 3 520

Roborovski hamsters of each group were subjected to pulmonary single-cell RNA sequencing 521

analysis. Pulmonary single cell suspensions were generated, cells were microscopically 522 counted and total numbers per lung lobe calculated. (1 x 105 pfu Wildtype (WT)), treated once at 1 dpi with 30 mg/kg mAb CV07-209 (mAb), daily starting at 1 dpi with 2 mg/kg Dexamethasone (Dex) or received combination treatment (Dex + mAb). At 3 dpi n = 3 Roborovski hamsters of each group were subjected to pulmonary single-cell RNA sequencing analysis. Pulmonary single cell suspensions were generated, cells were microscopically counted and total numbers per lung lobe calculated. (A) Cell count of isolated cells per lung lobe according to treatment group. Calculated numbers of indicated innate immune cells (B), T and B lymphocytes (C) and AT2 and endothelial cells (D) based on scRNA-seq determined cell frequencies ( Figure E3 ) and according to treatment group. Data display means ± SD. n = 3 per group. (A -D) Two-way ANOVA, Tukey's multiple comparisons test. * P < 0.05, ** P < 0.01, ***P < 0.001, **** P < 0.0001. (E) Edema Score resulting from semi-quantitative assessment of alveolar and perivascular edema (F) H&E stained histopathology of pulmonary vascular endothelia (upper panel) and lung parenchyma (lower panel) from Roborovski hamsters at 3 dpi. Mock and mAb treated groups had moderate to marked endothelialitis with activation and loss of endothelial cells whereas the vascular endothelium remained mostly intact in Dex and Dex + mAb treated groups. The inflammatory response was more pronounced in mock and mAb treated hamsters compared to Dex and Dex + mAb treated animals. Differences were particularly observed for infiltrating neutrophils, macrophages and, lymphocytes as well as for the degree of alveolar epithelial cell necrosis. (A) Neutrophils from the scRNA-seq data were sub-clustered using the Louvain algorithm based on their individual transcriptomes, and two-dimensional projections performed using the UMAP algorithm. Cells were coloured by their cluster identity. (B) Projection as in (A), but cells are coloured by the log10-transformed percentage of viral RNA. Overlaid are the stream arrows derived from the RNA velocity analysis. Neutrophil cluster 6 is marked with a light blue oval. (C) Changes in cellular density on the UMAP projection were calculated, and cells coloured by fold changes of the indicated Dex vs. mock. Red indicates increased density, and blue indicates decreased density. Neutrophil cluster 6 is marked with a light blue oval. (D) Graph indicates the log2-transformed fold changes of the cell counts in the respective neutrophil clusters 1-10, with all three treatments compared to mock. For example, in cluster 6 there are about one third less cells (dark blue bar at -0.6, which corresponds to log2 of 0.66) upon dexamethasone treatment. (E) Dot plots show the expression of selected genes over all hamsters in the clusters as defined in (A). The dot size indicates the fraction of cells in the clusters as indicated on the left from mock-treated animals, with ≥ one mRNA count for the respective gene. The colour represents the average expression in those cells. (F) Histopathology of Roborovski hamsters 3 days after infection revealed moderate to marked alveolar and interstitial infiltration with viable and degenerate neutrophils (black arrowheads) in mock and mAb treated animals as well as elevated numbers of alveolar macrophages (gray arrowhead). Dex and Dex + mAb treated hamsters had lower numbers of neutrophils especially in their alveolar spaces and mild to moderate numbers of neutrophils in alveolar capillaries (black arrowheads). Activated alveolar macrophages phagocytized cellular debris and cleared the inflammatory response (gray arrowhead). Scale bar = 20 µm.

A Novel Coronavirus from Patients with Pneumonia in China

SARS-CoV-2 Vaccines

Molnupiravir, an Oral Antiviral Treatment 583 for COVID-19

Cytokine Storm in COVID-19: The 588 Current Evidence and Treatment Strategies

Characteristics of lymphocyte subsets and cytokines in peripheral blood of 123 592 hospitalized patients with 2019 novel coronavirus pneumonia (NCP). medRxiv

Plasma IP-10 and MCP-3 levels are highly associated with disease severity and 596 predict the progression of COVID-19

Dexamethasone-mediated inhibition of 599 human T cell growth factor and gamma-interferon messenger RNA

Scope, quality, and inclusivity of clinical guidelines produced early in the 602 covid-19 pandemic: rapid review

Anti-glucocorticoid 604 effects of dehydroepiandrosterone (DHEA)

Determinants of prolonged 608 viral RNA shedding in hospitalized patients with SARS-CoV-2 infection. Diagn Microbiol Infect 609 Dis 100

Dexamethasone in Hospitalized Patients 612 with Covid-19

Dexamethasone modulates immature neutrophils 615 and interferon programming in severe COVID-19

Antigenicity of the SARS-CoV-2 Spike Glycoprotein

A Potent SARS-CoV-2 Neutralizing Human Monoclonal Antibody 621

That Reduces Viral Burden and Disease Severity in Syrian Hamsters

A Therapeutic Non-self-reactive SARS-CoV-2 625

Antibody Protects from Lung Pathology in a COVID-19 Hamster Model

Fruitful Neutralizing Antibody 628 Pipeline Brings Hope To Defeat SARS-Cov-2

CoV-2 Neutralizing Antibody Sotrovimab

Efficacy and safety of two neutralising monoclonal antibody 635 therapies, sotrovimab and BRII-196 plus BRII-198, for adults hospitalised with COVID-19 (TICO): 636 a randomised controlled trial

The monoclonal antibody combination REGEN-639 COV protects against SARS-CoV-2 mutational escape in preclinical and human studies

Studies in humanized mice and convalescent humans 643 yield a SARS-CoV-2 antibody cocktail

Omicron Variant by Vaccine Sera and Monoclonal Antibodies. medRxiv

Hamster models of COVID-19 649 pneumonia reviewed: How human can they be? Vet Pathol

Dexamethasone ameliorates severe pneumonia but slightly enhances viral replication 653 in the lungs of SARS-CoV-2-infected Syrian hamsters

Beneficial effect of combinational methylprednisolone and 657 remdesivir in hamster model of SARS-CoV-2 infection

The Roborovski Dwarf Hamster Is A Highly 664

Susceptible Model for a Rapid and Fatal Course of SARS-CoV-2 Infection

Temporal omics analysis 668 in Syrian hamsters unravel cellular effector responses to moderate COVID-19

SARS-CoV-2 infection of human ACE2-672 transgenic mice causes severe lung inflammation and impaired function

Dexamethasone may 675 improve severe COVID-19 via ameliorating endothelial injury and inflammation: A preliminary 676 pilot study

CBP (CREB binding protein) integrates NF-kappaB 678 (nuclear factor-kappaB) and glucocorticoid receptor physical interactions and antagonism

Enhancer Turnover Is 681 Associated with a Divergent Transcriptional Response to Glucocorticoid in Mouse and Human 682

Serum amyloid A -a review

Crosstalk between glucocorticoids and mitogen-activated 686 protein kinase signalling pathways

RNA velocity of single cells

Generalizing RNA velocity to 692 transient cell states through dynamical modeling

Control of macrophage lineage populations by CSF-1 receptor and GM-CSF in 696 homeostasis and inflammation

Macrophage inflammatory protein-1

Diffusion maps for high-dimensional single-700 cell analysis of differentiation data

SCANPY: large-scale single-cell gene expression 703 data analysis

The Molecular Signatures Database (MSigDB) hallmark gene set collection

Glucocorticosteroids enhance replication of respiratory viruses: effect of adjuvant interferon

Association Between SARS-CoV-2 712 Infection and Immune-Mediated Myopathy in Patients Who Have Died

Associations of clinical characteristics and treatment regimens with the 716 duration of viral RNA shedding in patients with COVID-19

Impact of dexamethasone 720 on SARS-CoV-2 concentration kinetics and antibody response in hospitalized COVID-19 721 patients: results from a prospective observational study

SARS-CoV-2 infection triggers 725 profibrotic macrophage responses and lung fibrosis

Effectiveness of corticosteroids 728 on chest high-resolution computed tomography features of COVID-19 pneumonia

Neutrophils and COVID-19: The road so far. Int 732 Immunopharmacol 90

Animal Models for COVID-19: Hamsters, Mouse, Ferret, Mink, Tree Shrew, and Non-human 735 Primates

Ultra-and micro-structural changes of respiratory tracts in SARS-CoV-2 infected 738 Syrian hamsters

Latent Class Analysis Reveals 741 COVID-19-related Acute Respiratory Distress Syndrome Subgroups with Differential Responses 742 to Corticosteroids

Virological assessment of 745 hospitalized patients with COVID-2019

Age-Dependent Progression of 748 SARS-CoV-2 Infection in Syrian Hamsters

SARS-CoV-2 infection of Chinese hamsters (Cricetulus griseus) reproduces 751 COVID-19 pneumonia in a well-established small animal model

<em>De 755 Novo</em> Whole Genome Assembly of the Roborovski Dwarf Hamster 756 (<em>Phodopus roborovskii</em>) Genome, an Animal Model for Severe/Critical 757 COVID-19. bioRxiv

QuasR: quantification and 759 annotation of short reads in R

Integrated analysis of multimodal single-763 cell data