key: cord-0958354-m1oqdiw6
authors: Kenney, Devin J.; O’Connell, Aoife K.; Turcinovic, Jacquelyn; Montanaro, Paige; Hekman, Ryan M.; Tamura, Tomokazu; Berneshawi, Andrew R.; Cafiero, Thomas R.; Al Abdullatif, Salam; Blum, Benjamin; Goldstein, Stanley I.; Heller, Brigitte L.; Gertje, Hans P.; Bullitt, Esther; Trachtenberg, Alexander J.; Chavez, Elizabeth; Nono, Evans Tuekam; Morrison, Catherine; Tseng, Anna E.; Sheikh, Amira; Kurnick, Susanna; Grosz, Kyle; Bosmann, Markus; Ericsson, Maria; Huber, Bertrand R.; Saeed, Mohsan; Balazs, Alejandro B.; Francis, Kevin P.; Klose, Alexander; Paragas, Neal; Campbell, Joshua D.; Connor, John H.; Emili, Andrew; Crossland, Nicholas A.; Ploss, Alexander; Douam, Florian
title: Humanized mice reveal a macrophage-enriched gene signature defining human lung tissue protection during SARS-CoV-2 infection
date: 2022-04-04
journal: Cell Rep
DOI: 10.1016/j.celrep.2022.110714
sha: 07d8701344945d8fe317ae2698ee4c06e4bb9a9b
doc_id: 958354
cord_uid: m1oqdiw6

The human immunological mechanisms defining the clinical outcome of SARS-CoV-2 infection remain elusive. This knowledge gap is mostly driven by the lack of appropriate experimental platforms recapitulating human immune responses in a controlled human lung environment. Here, we report a mouse model (i.e. HNFL mice) co-engrafted with human fetal lung xenografts (fLX) and a myeloid-enhanced human immune system to identify cellular and molecular correlates of lung protection during SARS-CoV-2 infection. Unlike mice solely engrafted with human fLX, HNFL mice are protected against infection, severe inflammation, and histopathological phenotypes. Lung tissue protection from infection and severe histopathology associate with macrophage infiltration and differentiation, and the upregulation of a macrophage-enriched signature composed of eleven specific genes mainly associated with the type I interferon signaling pathway. Our work highlights the HNFL model as a transformative platform to investigate, in controlled experimental settings, human myeloid immune mechanisms governing lung tissue protection during SARS-CoV-2 infection.

The immunological mechanisms driving susceptibility to Coronavirus disease 19 76 (COVID19), a recent viral respiratory disease caused by severe acute respiratory syndrome 77 coronavirus 2 (SARS-CoV-2) ( possibility of fLX-to-fLX viral transfer through the peripheral blood, some animals were inoculated 178 only in a single fLX. Non-inoculated fLX from non-infected and infected animals are hereafter 179 referred to as "naïve" and "contralateral" respectively. While NRGL mice did not display any signs 180 of clinical disease over the course of infection (Figure S3A-C) , inoculated fLX displayed gross 181 abnormalities ( Figure S3D ) which contrasted with a white homogenous appearance of naïve and 182 contralateral fLX. 183

Immunoreactivity to SARS-CoV-2 nucleocapsid (N) in inoculated fLX was observed in a 184 dose dependent manner (Figure 2A (Figure 2L ), suggesting that a limited amount of viral RNA, but not infectious viral particles, 199 may transfer between fLX through the blood. However, no viral RNA was detected in the 200 peripheral blood at any time point (Figure S3N) , suggesting that very low levels of free viral RNA 201 may circulate between fLX. We found a positive non-linear regression between viral load and PFU 202 in inoculated fLX (R 2 =0.675) ( Figure 2M ) and identified that viral loads in excess of a threshold 203 J o u r n a l P r e -p r o o f of 10 7 RNA copies/mg tissue (i.e. productive infection threshold [PIT] ) were indicative of 204 productive infection. Finally, using bioluminescence imaging and a recombinant SARS-CoV-2 205 expressing NanoLuc luciferase (Xie et al., 2020) , we revealed that SARS-CoV-2 replication 206 persisted within single animals for up to 12DPI (Figure 2N ,O; Figure S3O ; Movie S1) while 207 remaining undetectable in contralateral grafts ( Figure 2P; Figure S3O ).. 208

In contrast to NRGL mice, intra-fLX inoculation of HNFL mice with SARS-CoV-2 (10 6 PFU) 209 resulted in viral RNA copies per mg of tissues significantly below PIT (RNA copies/mg<1x10 7 ; 210 Next, we aimed to probe whether NRGL mice recapitulate histologic phenotypes observed in 227 patients with severe COVID19, and if HNFL mice are protected against these phenotypes. We 228 developed a semi-quantitative ordinal scoring system based on phenotypes of severe COVID19 229 J o u r n a l P r e -p r o o f (see methods). In NRGL-LX, the mean cumulative histologic score in animals inoculated with 10 6 230 PFU was 1.76-fold higher at 2DPI and 2.22-fold higher at 7DPI compared to those inoculated with 231 10 4 PFU, indicating a positive correlation with viral load and SARS-CoV-2 N positivity. ( Figure  232 3A; Figure S4A ). Of note, tissue integrity and architecture between naïve and contralateral fLX 233 was similar, and for this reason both were pooled together to define the histopathological baseline 234 ( Figure 3B, Figure S4B ). Compared to naïve/contralateral fLX, 10 6 PFU viral inoculation resulted 235

in an increasing histopathological score over time ( Figure 3B; Figure S4B) . Notably, neutrophil 236 influx, intra-airspace necrosis, capillary fibrin thrombi, and presence of syncytial cells were the 237 most significant independent observations that contributed to the increased cumulative score 238 ( Figure S4B ). There were no histopathological differences between NRGL-LX and NRGFL-LX at 239 2DPI, underscoring that loss of Flk2 expression does not impact histopathological outcome 240 ( Figure S4C) . 241

Histopathologic findings were observed in all SARS-CoV-2 inoculated NRGL-LX ranging 242 from mild focal to severe generalized disease ( Figure features were observed at 7DPI, but airspaces were more frequently filled with abundant necrotic 251 cellular debris (Figure 3J) , and in one fLX, a distinctive hyaline membrane lining pneumocytes 252 could also be observed ( Figure 3K) . Altogether, these data demonstrate that SARS-CoV-2 253 infection of NRGL mice is associated with the development of cellular and histopathological 254 features that resemble those observed in the lungs of patients with severe COVID19. 255

Ultrastructural analysis of NRGL-LX inoculated with 10 6 PFU supported our virological and 256 histopathological findings. Virus particles were observed within AT2 pneumocytes at various 257 stages of maturation and were often confined to double membrane bound vesicles (DMVs), with 258 morphology and particle size consistent with previously described coronaviruses (range 80-259 130nm in diameter) (Laue et al., 2021) (Figure 3L-N) . Potential single-particle budding events 260 ( Figure S4D ) were also observed. At 7DPI, airspaces were filled with abundant necrotic cellular 261 debris including lamellar bodies, erythrocytes, neutrophils and denuded viral particle-containing 262 AT2 pneumocytes (Figure 3O ; Figure S4E -G), which were occasionally undergoing apoptosis 263 as indicated by the presence of pyknotic nuclei ( Figure S4E ). DMV-containing viral particles and 264 electron-dense viral replication centers were still observed at 7DPI, suggesting persistence of 265 active viral replication ( Figure 3P, Figure S4H ). Faint Spike protein coronal surface projections 266 were sometimes visible within DMVs ( Figure 3Q ). Blood vessels also contained aggregates of 267 platelets ( Figure S4I ) with several small to intermediate-sized arteries occasionally occluded by 268 fibrin thrombi ( Figure S4J ). 269

Consistent with decreased viral loads, cumulative histology scores were significantly 270 decreased in HNFL-LX mice at 2 (p=0.01) and 7DPI (p=0.0002) when compared to NRGL-LX 271 (Figure 3R-U; Figure S4K ). HNFL-LX showed decreased syncytial cells and intra-airway 272 necrosis at 2 and 7DPI, respectively when compared to NRGL-LX. Hemorrhage and influx of 273 neutrophils were also significantly decreased at both 2 and 7DPI in HNFL-LX but not NRGL-LX 274 ( Figure S4K ). Taken together, these findings suggest that HNFL mice mount an effective host 275 response in fLX that not only prevents persistent SARS-CoV-2 infection, but also protects from 276 severe histopathological manifestations observed in patients with severe COVID19. Table S1 ). The major subsets 285 mediating hematopoietic expansion were macrophages (Naïve, 1.76%; 2DPI, 18.39%; 7DPI, 286 20.49%) and B cells (Naïve, 0%; 2DPI 7.24%; 7DPI, 3.31%) ( Figure 4L) . B cell infiltration was 287 consistent with flow cytometry ( Figure 4E ) and 6-color imaging data (Figure 2V,W) . Notably, the 288 frequency of AT2 cells within the epithelial compartment was reduced at 2DPI (12.7%) but 289 restored at 7DPI (45.5% vs. Naïve, 50.2%) (Figure 4M) Figure 4M ) and with previous studies reporting a loss of AT2 program/compartment upon 334 SARS-CoV-2 infection (Delorey et al., 2021) . At 7DPI, SFTPC (-0.5 Log2FC; FDR=1e-109) and 335 many ISGs were found to be returning to naïve fLX expression levels ( Figure 5H) . 336

Collectively, our findings highlight that viral clearance and tissue protection from SARS-337

CoV-2 in HNFL mice is associated with the upregulation of a defined genetic signature composed 338 of 11 specific genes. Finally, we aimed to identify the cellular compartment(s) driving PDG upregulation. Using 371 scRNAseq, we found that our PDG signature was significantly enriched in activated macrophages 372 in comparison to all other cellular compartments at 2DPI (Figure 6A ,B; Table S1 ). Three PDG 373 (IFIT2, IFIT3 and IFIH1) were also categorized as activated macrophage-defining genes ( Figure  374 6A). Notably, even though it was not identified as a PDG, ISG15 expression was statistically 375 restricted to activated macrophage clusters ( Figure 6C ), highlighting these clusters as the 376 dominant source of USP18-ISG15 co-expression ( Figure 6D ). Additionally, we found that 377 activated macrophage clusters were the major carriers of viral RNA (p=0.001), suggesting a 378 potential association between a dominant macrophage-mediated antiviral response ( Figure 6E ) 379 and macrophage infection. One limitation of our study involves the lack of evidence that human patients protected from 462 severe COVID19 also display an upregulated PDG signature in their lung macrophages in 463 contrast to severe COVID19 cases. However, eventhough the comparison remains imperfect, 464 combined PDG expression in lung monocytes and macrophages derived from lung autopsy 465 samples of patients with severe COVID19 (using previously reported datasets: (Delorey et al., 466 2021)) was significantly lower than in activated macrophages from infected HNFL-LX (Figure S7) , 467

supporting PDG upregulation in macrophages as marker of lung tissue protection. 468

Another weakness of our study relates to the limited number of human donors used. As 469 this is a clear limitation in comparison to human studies, future humanized mouse studies will 470 have to make every effort to increase intra-individual diversity within cohorts. Finally, direct SARS-471

CoV-2 inoculation into subcutaneous fetal lung tissues, which do not recapitulate neither the 472 function or level of maturation of the human adult lung, also constitutes a limitation of our work. 473

While direct injection within subcutaneous xenograft is a required trade-off for working with human 474 lung tissue in vivo, the well-defined compartmentalization of the fLX from the rest of the mouse, 475 its easy access from live animals and its connection to the vascular system underscores how 476 amenable the NRGL and HNFL models are to investigate host-pathogen interactions during viral 477 respiratory infection. That being said, it is likely that engraftment of human adult lung tissues over 478 fetal tissues could enhance the potential and biological significance of the HNFL mouse model. inoculation. See also Figure S5 and Table S1 . 606

and 7DPI (flow cytometric analysis). Dotted line represents mean frequency of CD45+ cells in 608 naïve HNFL-LX ( Figure 1C) (n=3-7) . Mean±SEM, one-way ANOVA, **p≤0.01, ***p≤0.001 over 609 naïve HNFL-LX. Il2rg tm1Wjl /J) were generated as described previously (Douam et al., 2018) and are available at 734

The Jackson Laboratory (Bar Harbor, ME, USA) (catalog number 033127). NRG and NRGF mice 735 were maintained at the Laboratory Animal Resource Center at Princeton University prior to 736 engraftment with human tissues and shipment to the NEIDL. 737

In the NEIDL BSL-3 facility, mice were group-housed by sex in Tecniplast green line individually 738 ventilated cages (Tecniplast, Buguggiate, Italy). Mice were maintained on a 12:12 light cycle at 739 30-70% humidity and provided sulfatrim-containing water and standard chow diets (LabDiet, St. 740

Louis, MO, USA). 741

All mice in this study were inoculated with SARS-CoV-2 at an age of 20 to 30 weeks old. Both 742 male and female mice were used. 743 744

Human fetal livers and lungs were procured from Advanced Bioscience Resources (Alameda, 746 CA, USA). Donor list is available in Table S5 . Table S5 ). 783

Generation of human immune system-engrafted mice. 3-5 weeks post fLX engraftment, 784 NRGFL mice were irradiated with 300 cGy and 7-10x10 5 human CD34+ HSC were injected 785 intravenously 4-6 h after irradiation. Male and female mice transplanted with CD34+ HSC derived 786 from three different human donors were used in this study. Twelve weeks post HSC engraftment, 787 peripheral levels of humanization were checked. Mice with peripheral engraftment level >40% 788 were enrolled in the study. One-week prior SARS-CoV-2 infection, NRGFL mice were injected 789 intravenously (tail vein) with 2x10 11 copies of AAV-Flt3LG resuspended in 200 µl of 1X phosphate-790 buffered saline (PBS) containing 35nM NaCL, 0.002%pluronic F-68 and 5%glycerol. 791

Ten to fifteen weeks post engraftment, NRGL and HNFL mice of both sexes were inoculated via 793 intra-fetal lung xenograft (intra-fLX) injection with 10 4 or 10 6 PFU of SARS-CoV-2 in 50 µL of sterile 794 1X PBS. Inoculations were performed under 1-3% isoflurane anesthesia. Either one or both 795 implants were inoculated by direct injection into the fLX. Animals were euthanized at day two or 796 day seven post inoculation. 797

Tissue collection and lung inflation for histology. At the indicated endpoints, mice were 798 anesthetized using 1-3% isoflurane, followed by euthanasia using an overdose of 799 ketamine/xylazine. to the flask. The next day, media was removed, the cell monolayer was rinsed with 1X PBS, pH 811 7.5 (ThermoFisher Scientific) and 25 ml of fresh DMEM containing 2% FBS was added. Two days 812 later, when the cytopathic effect of the virus was clearly visible, culture medium was collected, 813 filtered through a 0.22 µm filter, and stored at −80°C. Our P2 working stock of the virus was 814 prepared by infecting Vero E6 cells with the P1 stock, at a multiplicity of infection (MOI) of 0.1. 815

Cell culture media was harvested at 2DPI and 3DPI, and after the last harvest, ultracentrifuged 816 (Beckman Coulter Optima L-100k; SW32 Ti rotor) for 2 h at 25,000 rpm (80,000 x g) over a 20% 817 sucrose cushion (Sigma-Aldrich). Following centrifugation, the media and sucrose were 818 discarded, and pellets were left to dry for 5 min at room temperature. Pellets were then 819 resuspended overnight at 4°C in 500 µl of 1X PBS. The next day, concentrated virions were 820 aliquoted and stored at −80°C. were seeded into a 12-well plate at a density of 2.5x10 5 cells per well and infected the next day 837 with serial 10-fold dilutions of the virus stock for 1 h at 37°C. Following virus adsorption, each well 838 was supplemented with 1 ml of overlay media, consisting of 2X DMEM supplemented with 4% 839 FBS and mixed at a 1:1 ratio with 2.4% Avicel (DuPont, Wilmington, DE, USA; RC-581). Three 840 days later, the overlay media was removed, the cell monolayer was washed with 1X PBS and 841 fixed for 1 h at room temperature with 10% neutral buffered formalin (ThermoFisher Scientific). tissues were placed on a 60 mm dish and minced using a disposable scalpel. Tissue pieces were 881 transferred to a 15 mL conical tube with 3 mL of digestion buffer (HBSS minus Ca 2+ , Mg 2+ , and 882 phenol red, 0.5 mg/mL Liberase TM, 1 mg/mL DNase I) and incubated at 37°C for 30 min with 883 agitation every 10 min. Minced pieces were transferred to a 70 µm strainer on a 50 mL tube and 884 mashed through using the plunger of a 3 mL syringe plunger. The strainer was washed two times 885 with 1 mL of FACS buffer (1X PBS with 1% (v/v) FBS) and the cell suspension was centrifuged 886

at 300 x g for 5 min at 4°C. The cell pellet was resuspended in 1 mL of ACK lysing buffer 887 (ThermoFisher Scientific; #A1049201) and incubated for 2 min at room temperature. After 888 incubation, 9 mL of FACS buffer was added to quench the lysis, samples were centrifuged at 300 889

x g for 5 min at 4°C, and the cell pellet was resuspended in 1 mL of FACS buffer prior to antibody 890

Single cell suspension from spleen. Spleen was collected and placed in RPMI with 10% FBS. 892

To generate single cell suspensions, a 70 µm strainer was placed into one well of a 6-well plate 893 with 4 mL of FACS buffer. Whole spleen was then placed onto the strainer and mashed through 894 the strainer using a 3 mL syringe plunger. After the strainer was washed twice with 1 mL of FACS 895 buffer, the resultant single cell suspension was transferred to a 15 mL conical tube and samples 896 were centrifuged at 300 x g for 5 min at 4°C. The cell pellet was resuspended in 1 mL of ACK 897 lysing buffer and incubated for 2 min at room temperature. After incubation, 9 mL of FACS buffer 898 was added to quench the lysis, samples were centrifuged at 300 x g for 5 min at 4°C, and the cell 899 pellet was resuspended in 1 mL of FACS buffer. Briefly, serum was mixed with RNA/DNA shield (Zymo) at a 1:1 ratio. RNA buffer was then added 916 to the serum (2:1 ratio) and passed through a column by centrifugation at 13,000 x g. The column 917 was then washed twice, and RNA was eluted with 15 µL of RNase/DNase free water. Quantification of peripheral human chimerism in HNFL mice. 2-4x10 6 PBMCs of human or 935 murine origin were isolated as described above and stained for 1 h at 4°C in the dark with an 936 antibody cocktail targeting human(h)CD45, mouse CD45, hCD3, hCD4, hCD8, hCD16, hCD19, 937 hCD11c, hCD56 and hCD14. Following washing with FACS Buffer, cells were fixed with fixation 938 buffer (1% (v/v) FBS, 4% (w/v) PFA in PBS) for 30 min at 4°C in the dark. Chimerism of all 939 humanized mice was assessed by quantifying the following human populations: Human CD45 + , 940 human CD45 + murine CD45 -; T-cells, CD45 + CD3 + ; CD4 + T cells, CD45 + CD3 + CD4 + ; CD8 + T 941 cells, CD45 + CD3 + CD8 + ; CD45 + CD16 + leukocytes; B-cells, CD45 + CD19 + ; conventional dendritic 942 cells, CD45 + CD11c + ; NK/NKT cells, CD45 + CD56 + ; Monocytes, CD45 + CD14 + . 943 Antibody staining and flow-cytometry analysis of HNFL fLX. After generation of single cell 944 suspension, 5x10 5 -1x10 6 cells were used for flow cytometry staining. Cells were centrifuged at 945 300 X g for 5 min at 4°C. The cell pellet was resuspended in a mix of 22.5 µL FACS buffer and 946 2.5 µL of FcX (Biolegend; #422302) and incubated for 10 min at room temperature. After blocking, 947 25 µL of antibody cocktail targeting hCD3, hCD20, hCD16, hHLA-DR, hCD45, hCD8, hCD4, 948 hCD33, hCD45RA, hCD56, hCD14, mCD45, and containing a LIVE/DEAD viability dye 949 (ThermoFisher Scientific) was added to each sample and incubated in the dark for 30 min at 4°C. 950

After staining, 1 mL of FACS buffer was added to each sample, samples were centrifuged at 300 951

x g for 5 min, washed with 1 mL FACS buffer, centrifuged at 300 x g for 5 min, and then fixed in 952 200 µL 4% PFA for 30 min. After fixation cells were washed twice with 1 mL FACS buffer, 953 resuspended in FACS buffer, and stored protected from light at 4°C until analysis. Human immune 954 cell subsets were gated as follows: human CD45 + , hCD45 + mCD45 -; human CD3 + , hCD45 + 955 hCD3 + ; human CD4 + , hCD45 + hCD3 + hCD4 + ; human CD8 + , hCD45 + hCD3 + hCD8 + ; CD20 + , 956 hCD45 + hCD3 -hCD20 + ; human CD56 + , hCD45 + hCD3 -hCD20 -hCD33 -hCD56+. + . One-Step RT-qPCR using Applied Biosystems QuantStudio 3 (ThermoFisher Scientific), with the 966 following cycling conditions; reverse transcription for 10 min at 55°C and denaturation at 94°C for 967 J o u r n a l P r e -p r o o f 3 min followed by 45 cycles of denaturation at 94°C for 15 sec and annealing/extension at 58°C 968 for 30 sec. Ct values were determined using QuantStudio TM Design and Analysis software V1.5.1. 969

Calculations for RNA copies/mL were determined using a SARS-CoV-2 E RNA as a standard. 970 Quantification of infectious particles by plaque assay. SARS-CoV-2 infectious particles in 971 inoculated fLX were quantified by plaque assay. After euthanizing mice, tissues were collected in 972 500 µL of RNAlater (MilliporeSigma: # R0901500ML) and stored at -80°C. The day prior to 973 experiments, 8x10 4 cells per well were seeded in a 24-well plate. Between 20 and 30 mg of tissue 974 was placed into 500 µL of OptiMEM (ThermoFisher Scientific). Tissues were homogenized using 975 a Qiagen TissueLyser II (Qiagen) by two dissociation cycles (two min at 1800 oscillations/min) 976 with one min rest in between. Samples were then subjected to centrifugation at 13,000 rpm for 10 977 min at room temperature, and supernatant was transferred to a new 1.5 mL tube. From this, 1:10 2 978 -1:10 7 dilutions were made in OptiMEM and 200 µL of each dilution were plated onto a 24-well 979 plate. Supernatant was incubated at 37°C for 1 h with gentle rocking of the plate every 10 min. 980

After viral adsorption, 1 mL a 1:1 mixture of 2X DMEM 4% FBS and 2.4% Avicel (Dupont) was 981 overlaid into each well. Plates were incubated for 72 h at 37°C with 5% CO2. Avicel was then 982 removed, cells were washed twice with 1X PBS, and then cells were fixed in 10% buffered 983 formalin (ThermoFisher Scientific) for 1 h. After fixation, formalin was removed, and cells were 984 stained with 0.1% crystalline violet in 10% ethanol/water for 1 h and washed with tap water. 985

Number of plaques were counted, and infectious particles (PFU/mg of tissue) were calculated. 986 987

Histologic processing and analysis. Tissue samples were fixed for 72 h in 10% neutral buffered 989 formalin, processed in a Tissue-Tek VIP-5 automated vacuum infiltration processor (Sakura 990

Finetek USA, Torrance, CA, USA), followed by paraffin embedding using a HistoCore Arcadia 991 paraffin embedding station (Leica, Wetzlar, Germany). Generated formalin-fixed, paraffin-992 embedded (FFPE) blocks were sectioned to 5 μm using a RM2255 rotary microtome (Leica), 993 Table S3 . Related to Figure 5 . Phospho-proteomics analysis Matrix: Naïve HNFL-LX vs. 2DPI 1256 HNFL0LX, Naïve NRGL-LX vs. 2DPI NRGL-LX (Excel file). 1257 Table S4 . Related to Figure 5 . List of differentially expressed genes and IPA scores from bulk 1258 RNA sequencing analysis of naïve and inoculated NRGL-LX (Excel file). 1259 Table S5 . Related to all figures. List of the mice and fetal donor ID used in this study (Excel file). 1260 1261

Vascular Disease and Thrombosis in SARS-CoV-2-Infected Rhesus Macaques

STAT2 is an essential adaptor in USP18-mediated suppression of type I interferon 1267 signaling

Autoantibodies against type I IFNs in patients with life-threatening 1270 COVID-19

USP18-a multifunctional component in the interferon 1272 response

Monocytes and Macrophages, Targets of Severe Acute Respiratory Syndrome 1275 Coronavirus 2: The Clue for Coronavirus Disease

The Pro-Inflammatory Chemokines CXCL9, CXCL10 1278 and CXCL11 Are Upregulated Following SARS-CoV-2 Infection in an AKT-Dependent Manner

Global absence and targeting of protective immune states in severe COVID-1281 19

COVID-19 tissue atlases reveal SARS-CoV-2 1284 pathology and cellular targets

The use of humanized mice for studies of viral pathogenesis and 1286 immunity

Selective expansion of myeloid and NK cells in humanized mice yields 1289 human-like vaccine responses

Rapid evolution of HIV-1 to functional CD8(+) T cell responses in 1292 humanized BLT mice

MAST: a flexible statistical framework for assessing transcriptional changes 1295 and characterizing heterogeneity in single-cell RNA sequencing data

Utility of humanized BLT mice for analysis of dengue virus infection and antiviral drug testing

Circuits between infected macrophages 1301 and T cells in SARS-CoV-2 pneumonia

Impaired type I interferon activity and inflammatory responses in 1304 severe COVID-19 patients

Multiple functions of USP18

Selective inactivation of USP18 isopeptidase activity in vivo 1309 enhances ISG15 conjugation and viral resistance

Infection and Rapid Transmission of SARS-CoV-2 in Ferrets

Causal analysis approaches in Ingenuity 1314 Pathway Analysis

Morphometry of SARS-1316

-2 particles in ultrathin plastic sections of infected Vero cell cultures

R2-P2 rapid-robotic 1318 phosphoproteomics enables multidimensional cell signaling studies

Single-cell landscape of bronchoalveolar immune cells in patients with COVID-19

Coagulation Abnormalities 1322 and Thrombosis in Patients Infected With SARS-CoV-2 and Other Pandemic Viruses

UBP43 is a novel regulator of interferon signaling independent of its ISG15 isopeptidase activity

Lethal infection of K18-hACE2 mice infected with severe acute 1329 respiratory syndrome coronavirus

The multiple faces of CCL13 in immunity and 1331 inflammation

Pathological inflammation in patients with COVID-19: a key role for 1333 monocytes and macrophages

Of mice and not men: differences between mouse and human 1335 immunology

Altered ISGylation drives aberrant macrophage-dependent immune responses 1338 during SARS-CoV-2 infection

The PRIDE database and related tools and resources in 1341 2019: improving support for quantification data

COVID-19 1343 immune features revealed by a large-scale single-cell transcriptome atlas

The spatial landscape of lung pathology during COVID-19 progression

Altered increase in STAT1 expression and phosphorylation in 1349 severe COVID-19. medRxiv : the preprint server for health sciences

The chemokine CCL18 causes 1351 maturation of cultured monocytes to macrophages in the M2 spectrum

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

Responses to acute infection with SARS-CoV-2 in the lungs of 1357 rhesus macaques, baboons and marmosets

Ad26 vaccine protects against SARS-CoV-2 severe clinical disease in 1360 hamsters

BNT162b vaccines protect rhesus macaques from SARS-CoV-2

Precision mouse models with expanded tropism for human 1366 pathogens

SARS-CoV-2 infection is effectively treated and prevented by EIDD-1369 2801

Humanized-BLT 1371 mouse model of Kaposi's sarcoma-associated herpesvirus infection

Celda: A 1374 Bayesian model to perform bi-clustering of genes into modules and cells into subpopulations using 1375 single-cell RNA-seq data. bioRxiv

Discriminating mild from critical COVID-19 by innate 1378 and adaptive immune single-cell profiling of bronchoalveolar lavages

SARS-CoV-2 infection of human ACE2-transgenic mice causes severe lung 1381 inflammation and impaired function

A nanoluciferase SARS-CoV-2 for rapid neutralization testing 1384 and screening of anti-infective drugs for COVID-19

Attenuated Interferon and Proinflammatory Response in SARS-CoV-2-Infected Human Dendritic 1387

Cells Is Associated With Viral Antagonism of STAT1 Phosphorylation

Decontamination of ambient RNA in single-cell RNA-seq with DecontX

Inborn errors of type I IFN immunity in patients with life-threatening COVID-19

Human intracellular ISG15 prevents interferon-alpha/beta over-1395 amplification and auto-inflammation

Severe Acute 1397 Respiratory Syndrome Coronavirus 2-Induced Immune Activation and Death of Monocyte

Human Macrophages and Dendritic Cells

A 1400 Novel Coronavirus from Patients with Pneumonia in China

transferred to positively charged slides, deparaffinized in xylene, and dehydrated in graded 994 ethanol. Tissue sections were stained with hematoxylin and eosin for histologic examination, while 995 unstained slides were used for immunohistochemistry. Qualitative and semi-quantitative 996 histomorphological analyses were performed by a single board-certified veterinary pathologist 997 (N.A.C.). An ordinal scoring system was developed to capture the heterogeneity of histologic 998 findings in fLX. Individual histologic findings that were scored included: syncytial cells, hyaline 999 membrane, intra-airspace neutrophils and necrosis, hemorrhage, edema, denuded pneumocytes, 1000 capillary fibrin thrombi, intermediate vessel fibrin thrombi and coagulative necrosis. The entire fLX 1001 was examined at 200x with a DM2500 light microscope (Leica) using the following criteria: 0 = 1002 not present, 1 = found in <5% of fields, 2 = found in >5% but <25% of fields, or 3 = found in >25% 1003 of fields. Several criteria were also restricted to 'not observed' (0) or 'observed' (1). Scores were 1004 combined to generate a cumulative lung injury score. For mouse derived primary antibodies, a linker antibody (Abcam) was used prior to application of 1014 the secondary antibody to prevent non-specific binding. DAB and purple chromogens (Roche) 1015 and chromogens used for TSA-conjugated Opal 480, 520, 570, 620, and 690 fluorophores (Akoya 1016Biosciences, Marlborough, MA, USA) were utilized to develop immunohistochemical assays. The 1017following anti-SARS-CoV-2 antibodies were used for immunohistochemistry: rabbit polyclonal 1018 anti-SARS-CoV Nucleoprotein (Novus Biological, Littleton, CO, USA), mouse monoclonal anti-1019 SARS-CoV-2 Spike clone 2B3E5 (This antibody was used in this study as clone E7U60, which 1020 was the pre-production clone ID of clone 2B3E5; Cell Signaling Technology). 1021Multispectral fluorescent imaging. Fluorescently labeled slides were imaged using either a 1022 Mantra 2.0 TM or Vectra Polaris TM Quantitative Pathology Imaging System (Akoya Biosciences). To 1023 maximize signal-to-noise ratios, fluorescently acquired images were spectrally unmixed using a 1024 synthetic library specific for the Opal fluorophores used in each assay plus DAPI. An unstained 1025 fLX section was used to create an autofluorescence signature that was subsequently removed 1026 from multispectral images using InForm software version 2.4.8 (Akoya Biosciences). 1027 Image analysis of multiplex immunohistochemistry. Digitized whole slide scans were 1028 analyzed using the image analysis software HALO v3.2 (Indica Labs, Inc., Corrales, NM, USA). 1029Slides were manually annotated to select regions of interest, excluding accessory skin and 1030 adipose tissue, to ensure inclusion of only the fLX. Visualization thresholds were adjusted in 1031 viewer settings to minimize background signal identification and maximize specificity of signals 1032 for each sample. Quantitative positive pixel analysis outputs were obtained using the Area 1033 Quantification (AQ) module, which reports total area of immunoreactivity of a specified parameter 1034 within a region of annotated interest. Values are given as a percentage of total tissue area 1035 analyzed. Minimum dye intensity thresholds were established using the real-time tuning field of 1036 view module to accurately detect positive immunoreactivity. For quantitative cellular phenotyping, 1037 the fluorescence Highplex (HP) module was utilized. Cells are identified using DAPI to segment 1038 individual nuclei. Minimum cytoplasm and membrane thresholds are set for each dye to detect 1039 positive staining within a cell. Parameters are set using the real-time tuning mechanism that was 1040 tailored for each individual biopsy based on signal intensity. Phenotypes are determined by 1041 selecting inclusion and exclusion parameters relating to stains of interest. This algorithm produces 1042 a quantitative output for each cell phenotype standardized to the area analyzed (cells/µm 2 ). 1043 washed in MB and water, and dehydrated in grades of alcohol (10 min each: 50%, 70%, 90%, 1049 2x10 min 100%). The tissue samples were then put in propyleneoxide for 1 h and infiltrated 1050 overnight in a 1:1 mixture of propyleneoxide and TAAB Epon. The following day the samples were 1051 Acetonitrile and 01% TFA). Before being added to the peptides, the Fe-NTA beads were washed 1088 with binding buffer. Peptides were then incubated with the Fe-NTA bead slurry for 20 min in a 1089Kingfisher Apex magnetic bead transferring system (ThermoFisher Scientific) before being moved 1090 into wash wells. Beads with bound phosphopeptides were washed twice in binding buffer, after 1091 which phosphopeptides were serially eluted twice by moving the beads into wells containing 200 1092 µL of elution buffer (50% acetonitrile and 2.5% ammonium hydroxide). Both phosphopeptide 1093 eluates corresponding to an orthogonal fraction were combined prior to drying in a speedvac. 1094Mass spectrometry analysis. Multiplexed peptide fractions from each time point were 1095 resuspended in mobile phase A solvent (2% acetonitrile and 0.1% formic acid) to be analyzed on 1096 the Exploris 480 mass spectrometer equipped with FAIMS (ThermoFisher Scientific). The mass 1097 spectrometer was interfaced to the Easy nanoLC1200 HPLC system (ThermoFisher Scientific). 1098Briefly, the peptides were first loaded onto a reverse-phase nanotrap column in mobile phase A, 1099 (75 mm i.d. 3 2 cm, Acclaim PepMap100 C18 3 mm, 100 A°, ThermoFisher Scientific) and 1100 separated over an EASY-Spray column, (ES803A, ThermoFisher Scientific) using a gradient (6% 1101 to 19% over 58 min, then 19% to 36% over 34 min) of mobile phase B (0.1% formic acid, 80% Bulk RNA sequencing. Total RNA was processed from fLX as described above, and sent to BGI 1147 genomics (Hong Kong, China) for library preparation and sequencing (Pair-ends, 100 bp, 20M 1148 reads per sample). Raw FASTQ files were quality-checked with FastQC v0.11.7. Reads were 1149 found to be excellent quality and were aligned to the combined human (GRCh38, Ensembl 101) 1150 and mouse (GRCm38, Ensembl 101) genomes with STAR v2.7.1a followed by quantification with 1151 featureCounts v1.6.2. Quality was checked with MultiQC v1.6. All samples passed quality 1152 thresholds of >75% sequences aligned and >15 million aligned reads per sample. Significantly 1153 up-and downregulated genes were identified with DESeq2 v1.23.10 in R v3.6.0. Three treatment 1154 groups were compared to non-inoculated samples in turn: 2DPI, 7DPI and 7DPI contralateral. P-1155 values were FDR-adjusted, and log2 fold change was shrunk with the apeglm method. Table S1 . Analytics. 1227

For histopathological score, viral loads/titers comparison and Spike quantification, Kruskal-Wallis, 1230Kolmogorov-Smirnov non-parametric t-test, or a two-way ANOVA with Benjamini, Krieger, and 1231Yekutieli correction for multiple comparisons were applied given the non-continuous nature of the 1232 data (i.e., viral inoculation in areas displaying differential stage of tissue development, and/or 1233 differences in fLX engraftment, and/or donor/gestational age differences). A two-way ANOVA with 1234Benjamini, Krieger, and Yekutieli correction for multiple comparisons was also used to generate 1235 the ISG p.value heatmap given the non-parametric nature of these data. Significance between 1236

Kruskal-Wallis test between pooled viral RNA copies/mg tissue values from inoculated HNFL-LX 1238