key: cord-0965303-4h2yv9sy
authors: Zheng, Jian; Roy Wong, Lok-Yin; Li, Kun; Verma, Abhishek K.; Ortiz, Miguel; Wohlford-Lenane, Christine; Leidinger, Mariah R.; Knudson, C. Michael; Meyerholz, David K.; McCray, Paul B.; Perlman, Stanley
title: K18-hACE2 Mice for Studies of COVID-19 Treatments and Pathogenesis Including Anosmia
date: 2020-08-10
journal: bioRxiv
DOI: 10.1101/2020.08.07.242073
sha: 2bf1de7b92c76a5f62a4195916d94a606e38d96b
doc_id: 965303
cord_uid: 4h2yv9sy

The ongoing COVID-19 pandemic is associated with substantial morbidity and mortality. While much has been learned in the first months of the pandemic, many features of COVID-19 pathogenesis remain to be determined. For example, anosmia is a common presentation and many patients with this finding show no or only minor respiratory signs. Studies in animals experimentally infected with SARS-CoV-2, the cause of COVID-19, provide opportunities to study aspects of the disease not easily investigated in human patients. COVID-19 severity ranges from asymptomatic to lethal. Most experimental infections provide insights into mild disease. Here, using K18-hACE2 mice that we originally developed for SARS studies, we show that infection with SARS-CoV-2 causes severe disease in the lung, and in some mice, the brain. Evidence of thrombosis and vasculitis was detected in mice with severe pneumonia. Further, we show that infusion of convalescent plasma (CP) from a recovered COVID-19 patient provided protection against lethal disease. Mice developed anosmia at early times after infection. Notably, while treatment with CP prevented significant clinical disease, it did not prevent anosmia. Thus K18-hACE2 mice provide a useful model for studying the pathological underpinnings of both mild and lethal COVID-19 and for assessing therapeutic interventions.

provide opportunities to study aspects of the disease not easily investigated in human 23 patients. COVID-19 severity ranges from asymptomatic to lethal. Most experimental 24 infections provide insights into mild disease. Here, using K18-hACE2 mice that we 25 originally developed for SARS studies, we show that infection with SARS-CoV-2 causes 26 severe disease in the lung, and in some mice, the brain. Evidence of thrombosis and 27 vasculitis was detected in mice with severe pneumonia. Further, we show that infusion of 28 convalescent plasma (CP) from a recovered COVID-19 patient provided protection 29 against lethal disease. Mice developed anosmia at early times after infection. Notably, 30 while treatment with CP prevented significant clinical disease, it did not prevent anosmia. 31

Thus K18-hACE2 mice provide a useful model for studying the pathological 32 underpinnings of both mild and lethal COVID-19 and for assessing therapeutic 33 The COVID-19 pandemic is associated with more than 18 million cases and more than 39 650,000 deaths worldwide since its emergence in December of 2019. SARS-CoV-2 uses 40 the same receptor as SARS-CoV, human angiotensin-converting enzyme 2, (hACE2) 1 . 41

Many of the animals infectable by SARS-CoV can also be experimentally infected with 42 SARS-CoV-2, but these animals generally develop mild disease 2-6 . Conversely, mice, 43 which are SARS-CoV-susceptible, are resistant to infection with SARS-CoV-2 because 44 of incompatibilities between mouse ACE2 and the viral spike protein 7 . Several 45 approaches have been or will be used to sensitize mice to infection, including providing 46 hACE2 by adenovirus transduction 8, 9 , mutating SARS-CoV-2 spike protein so that it binds 47 to mACE2 10 , and modifying mACE2 so that cells are susceptible to SARS-CoV-2 11 . 48

During the 2003-2004 SARS epidemic, since mice developed only mild disease, we and 49 others engineered mice that transgenically displayed hACE2 12-14 . K18-hACE2 mice 50 express hACE2 driven by cytokeratin 18 promoter, predominantly in epithelial cells 15 . 51 SARS-CoV infected-K18-hACE2 mice developed an overwhelming encephalitis, with 52 lung disease characterized by aspiration pneumonia and evidence of mild virus-induced 53 pneumonia 12 . Both SARS-CoV-2 and SARS-CoV use hACE2 but show differences in 54 disease manifestations in patients. SARS-CoV-2 infects the upper airways to a much 55 greater extent than SARS-CoV and has been associated with clinical manifestations such 56 as anosmia, ageusia, thrombosis and endothelial damage in the lung vasculature, cardiac 57 and neurological disease, and a multisystem inflammatory disease in children and 58

adolescents [16] [17] [18] . Based on these observations, we reasoned that SARS-CoV-2 infection 59 of the K18-hACE2 mice might also show differences in pathogenesis. Here we 60 SARS-CoV-2 productively infects the sinonasal epithelium including sustentacular 152 cells. To assess virus replication in the upper respiratory tract following intranasal 153 inoculation, we quantified viral RNA in nasal secretions by qRT PCR at 3 dpi. We 154 observed that secretions from 5 of 7 mice showed evidence of SARS-CoV-2 genomic 155 RNA (Ct =21.5 +/-1.2 (mean +/-SEM)), and of these 5, 4 had evidence of subgenomic 156 RNA (36.7 +/-1.1 (mean +/-SEM), consistent with active virus replication. Viral antigen 157 was readily detected in both the respiratory and olfactory epithelium at 2 and 5 dpi. At 2 158 dpi virus antigen was present at multiple sites, often at the interface of the olfactory and 159 respiratory epithelium (Fig. 3e ). Viral antigen was also detected in nerve bundles 160 subjacent to the olfactory epithelium ( Fig. 3f ) and occasionally in vascular endothelia ( Fig.  161 3g, left arrow) and Bowman's glands (Fig. 3g , right arrow). At sites of antigen positivity in 162 the olfactory epithelium and maxillary sinus, we observed cell death and cellular debris at 163 day 2 ( Fig. 3h-j) , which progressed to cell sloughing and loss of cellularity by 5 dpi. (Fig.  164 3k, l). ACE2 has been detected in sustentacular cells in the olfactory epithelium 25 , but not 165 in olfactory sensory neurons, suggesting that these cells are a primary site of infection. 166

Consistent with this, we detected SARS-CoV-2 antigen in sustentacular cells . 167

Infection of sustentacular cells is not expected to result in spread to the olfactory bulb and 168 its connections, but could still contribute to anosmia. 169 170 SARS-CoV-2 infected K-18-hACE2 mice exhibit anosmia. To directly assess anosmia, 171

we performed two sets of behavioral tests, both of which require a normal sense of smell, 172 as described in Experimental Procedures. First, male mice were exposed to bedding 173 containing female or male dander in a 2 ml Eppendorf tube (Fig. 4a) . Mice identify the 174 tube visually, and then preferentially spend time with the female dander, if olfaction is 175 normal. In a second experiment, we used a buried food test, in which mice are attracted 176 to a food item hidden in the bedding that they were previously conditioned to detect (Fig.  177 4c). K18-hACE2 mice infected with SARS-CoV-2 did worse than controls in both tests so 178 that at days 2 and 3 p.i., they spent less time in the vicinity of the female dander (Fig. 4b ) 179

and took longer to find buried food (Fig. 4d, e) . At 2 and 3 dpi, brains were not infected 180 (Fig. 1b) . Additionally, mobility was largely normal since there were no differences in the 181 amount of time spent in exploring the tube containing male dander, when infected or 182 uninfected mice were compared (Fig. 4b ). This suggests that anosmia at these times 183 points is primarily caused by infection of the nasal epithelium and is not a consequence 184 of spread to the brain. hours before infection. Undiluted plasma administered 12 hours before infection protected 195 mice from death but not mild weight loss and reduced lung tissue titers while CP diluted 196 1:3 provided partial protection (Fig. 5a, c) . Convalescent plasma pretreatment markedly 197 inhibited spread of infection to the brain (Fig. 5c ). On examination of lung tissues, levels 198 of viral antigen and pathological changes were greatly decreased by CP treatment (Fig.  199 5d, e). Treatment did not seem to affect primary infection of lung cells but rather prevented 200 secondary spread within the lung. As shown in Fig. 5b , delivery of undiluted plasma at 24 201 hrs post infection (p.i.) demonstrated a partial protective effect. Finally, to assess the 202 effects of CP treatment on SARS-CoV-2-induced anosmia, we treated infected mice with 203 CP at 1 dpi and assessed mice for olfactory loss as described above (Fig. 4) . Even though 204 mice had minimal signs of clinical disease after CP treatment, by day 4 all mice exhibited 205 profound anosmia (Fig. 4b, d, e) . 206

When infected with SARS-CoV-2, K18-hACE2 mice developed a dose-dependent lung 209 disease phenotype with features similar to severe human COVID-19. This includes diffuse 210 alveolar damage, an influx of immune effector cells, tissue injury, lung vascular damage, 211 and death. Remarkably, the mice also support SARS-CoV-2 replication in the sinonasal 212 epithelium and associated with this pathology develop anosmia, a common feature of 213 human disease. Furthermore, the uniformly fatal disease outcome with a 10 5 inoculum 214 was prevented by pre-treatment with CP from a COVID-19 patient. Notably, CP pre-215 treatment enhanced the kinetics of virus clearance but did not prevent initial infection of 216 the lungs, damage to nasal respiratory and olfactory epithelia, or anosmia. SARS-CoV-2 217 infection of K18-hACE2 mice treated with CP or, potentially, neutralizing monoclonal 218 antibodies will be especially useful for studies of anosmia because mice do not succumb 219 to the infection, but like many infected patients with mild disease, have olfactory loss as 220 a major manifestation 18 . Anosmia may result from damage to supporting sustentacular 221 cells and not to olfactory sensory neurons, suggesting that the resulting inflammatory 222 milieu, rather than direct neuronal damage is disease-causing. Therefore, we postulate 223 that the observed anosmia has at least two possible explanations. First, the infection of Although both SARS-CoV and SARS-CoV-2 use ACE2 to gain entry into cells and both 235 cause lung and brain disease, the disease manifestations are different. Most importantly, 236 SARS-CoV caused a brain infection when as little as 3.2 PFU were administered 237 intranasally 24 , while even 10 4 PFU SARS-CoV-2 only variably infected the brain. A 238 consequence of this difference in CNS susceptibility is that lung damage was obvious 239 after SARS-CoV-2 infection, while it was obscured by the aspiration pneumonia that 240 developed in SARS mice as a consequence of probable infection of the cardio-respiratory 241 center in the medulla. 242 243 All K18-hACE2 mice succumbed to 10 5 SARS-CoV-2, from lung and, in some cases, 244 brain disease. This uniform lethality makes these mice useful for evaluation of anti-viral 245 therapies and vaccines, and also sets a high standard for their efficacy. This is illustrated 246 in Fig. 5 , where mice treated with convalescent plasma were protected from lethal 247 disease, but not infection. This absence of complete protection could reflect use of human 248 convalescent plasma in mice, since some antibody effector functions are species-specific. 249

Of note, while most mice receiving 1:3 or 1:9 dilutions of convalescent plasma had no 250 evidence of brain infection (Fig. 5c ), they succumbed to progressive lung disease. 251

Infection with a range of virus inocula will be useful for evaluation of therapeutic 252 interventions in a variety of pathological settings, as well as assessment of upper airway 253 disease and anosmia. Therefore, K18-hACE2 mice, readily available from Jackson 254

Laboratories, provide a useful model to study the pathogenesis of SARS-CoV-2-mediated 255 disease and to evaluate interventions. COVID-19 more than 4 weeks prior to their donation. Following the donation, she tested 452 positive for HLA antibodies so the plasma was not eligible for administration to patients 453 and was diverted to research. Antibody testing (EUROIMMUN SARS-COV-2 ELISA 454 (IgG)) performed on this donor was 9.8, well above the cutoff of 1.1 for a positive result. 455

Neutralization titer using a luciferase-expressing SARS-CoV-2 S protein pseudovirus 456 assay showed that the neutralization IC50 titer was 1:1,480. Control plasma was obtained 457 from an expired plasma unit collected prior to COVID-19 spread in our area and this 458 product was collected under an IRB (#201402735) approved protocol that allows for 459 research use of these products. o. Nasal and sinus tissue were examined at days 2 (e-j) and 5 (k-o) p.i. e-h. Olfactory 555 epithelium (OE) was stained with anti-SARS-CoV-2 N antibody (e-g) or H&E stain (h). f. 556

Nerve bundles (NB) subjacent to OE had punctate immunostaining (brown, arrows). g. 557

Subjacent to sites of immunostaining for N protein (brown, black arrows), endothelial 558 lining of vessels (left arrow) and bowman glands (right arrow) were occasionally positive 559 for virus antigen (brown). h. Sites of N protein localization in OE had evidenced of cell 560 death and cellular debris (arrows). Bar = 100 µm (e) and 25 µm (f-h). i-j. Maxillary sinus 561 (MS) stained with anti-SARS-CoV-2 N antibody (i) or H&E (j). i. MS lining epithelium had 562 extensive immunostaining for N protein (brown, arrows). j. ME epithelium had common 563 sloughing and cellular debris (arrowheads). The lateral nasal glands (LNG) also had 564 multifocal cellular and karyorrhectic debris (arrows). Bar = 75 µm (i) and 19 µm (j), 565 respectively. k-l. Olfactory epithelium still had foci of SARS-CoV-2 N protein 566 immunostaining at d. 5 p.i. (k, arrows, left) that was often localized near interface with 567 respiratory epithelium. In these sites, there was cellular sloughing and loss of cellularity 568 b Fig. 1 . Clinical and pathological disease in K18-hACE2 transgenic mice following SARS-CoV-2 infection. a. Percentage of initial weight and survival of K18-hACE2 mice infected with 10 3 , 10 4 , and 10 5 PFU SARS-CoV-2/mouse. b. Viral RNA detected by qPCR targeting viral N gene with normalization to HPRT for the indicated organs at 2, 4, and 6 dpi (left panel). Infectious virus titers detected by plaque assay in different organs at 2, 4, and 6 dpi (right panel). LOD=limit of detection. c. Lungs from uninfected (n=3), and infected (day 4 (n=4) and day 6 (n=3) p.i.) were analyzed by immunohistochemistry using anti-SARS-CoV-2 N antibody. d. Sections of paraffin-embedded lungs were stained with H&E. Note airway edema and alveolar hyaline membranes (asterisks, middle bottom panel), vascular thrombosis (inset, top right, 6 dpi 20x panel), dying cells with pyknotic to karyorrhectic nuclei, and a proliferative alveolar epithelium with mitotic figures (arrowhead and inset, lower right, 6 dpi 20x panel). e. Summary of histological scoring in the lungs, as described in Experimental Procedures. 2 . Inflammatory mediators and immune effector cells contribute to lung disease phenotype. a. Cytokine and chemokine transcripts were measured by qPCR following reverse transcription of RNA isolated from the lungs of SARS-CoV-2 infected K18-hACE2 mice (n=3 or 4 for each time point). Statistical significance compared to results obtained at 0 dpi. b. Representative FACS plot of IFNγ+TNF+ CD8 and CD4 T cells after stimulated with indicated peptide pools) in the lungs of 10 5 PFU SARS-CoV-2 infected K18-hACE2 mice. c. Summary data are shown (n=3 mice/time point). d. Quantification of immune cells in the lungs (n=3 for uninfected group and n=4 for 4 and 6 dpi). e. Sera were collected from infected mice at the indicated time points and IC 50 values determined by neutralization of SARS-CoV-2 pseudoviruses expressing luciferase.

survival (right panel) of K18-hACE2 mice receiving control plasma (n=2) or undiluted 591

LOD=limit of 593 detection. d. Scoring for N protein abundance in CP-treated mice in the nasal cavity at 2 594 (left panel) and 5 (right panel) dpi. 0 -none; 1 -rare <1%; 2 -multifocal or localized <33% 595 cells; 3 -multifocal, coalescing, 33-66%

Extended Data Fig. 1. Histological analysis of extrapulmonary tissue in SARS-CoV-600 2-infected K18-hACE2 mice. Mice were sacrificed at days 0, 4 and 6 p.i. and tissues 601 prepared for histological examination (n=3/4 per group

Pathological changes were minor and only 603 observed in the liver. In the liver, all mice had some blood vessels filled with clear space 604 or aggregates variably composed of erythrocytes / platelets (insets)

Extended Data Fig. 2. Gating strategy for identification of immune cells in lungs is shown

Fig. 3. Some K18-hACE2 mice develop brain disease after SARS-CoV-2 infection. a. Brains were prepared from uninfected and infected mice and analyzed for SARS-CoV-2 by immunohistochemistry. b, c. Multiple sites of dead cells (arrows) characterized by cellular and karyorrhectic nuclear debris (b, arrows) and of thrombi (c, arrow and inset) in thalamus. Bar = 17 µm. d. Examination of meninges at day 6 p.i. revealed increased cellularity composed of degenerate cells and neutrophils and mononuclear cells. Cellular and karyorrhectic nuclei debris were also detected in the perivascular regions (200X). e-o. Nasal and sinus tissue were examined at days 2 (e-j) and 5 (k-o) p.i. eh. Olfactory epithelium (OE) was stained with anti-SARS-CoV-2 N antibody (e-g) or H&E stain (h). f. Nerve bundles (NB) subjacent to OE had punctate immunostaining (brown, arrows). g. Subjacent to sites of immunostaining for N protein (brown, black arrows), endothelial lining of vessels (left arrow) and bowman glands (right arrow) were occasionally positive for virus antigen (brown). h. Sites of N protein localization in OE had evidenced of cell death and cellular debris (arrows). Bar = 100 µm (e) and 25 µm (f-h). i-j. Maxillary sinus (MS) stained with anti-SARS-CoV-2 N antibody (i) or H&E (j). i. MS lining epithelium had extensive immunostaining for N protein (brown, arrows). j. ME epithelium had common sloughing and cellular debris (arrowheads). The lateral nasal glands (LNG) also had multifocal cellular and karyorrhectic debris (arrows). Bar = 75 µm (i) and 19 µm (j), respectively. k-l. Olfactory epithelium still had foci of SARS-CoV-2 N protein immunostaining at d. 5 p.i. (k, arrows, left) that was often localized near interface with respiratory epithelium. In these sites, there was cellular sloughing and loss of cellularity (arrows, right). 200x. m-o. Arrows point to "classic" morphology and strong labeling of sustentacular cells with expanding labeling of adjacent cells. Bar = 50 (m) and 25 µm (n, o). The dotted line shows the time limit of 4 minutes. Data were analyzed by 1-way ANOVA. ****p<0.0001. e. The percentage of mice that found the buried food within 4 minutes is shown. In some experiments, mice were pretreated with undiluted convalescent plasma (denoted by blue bars in d, e). and survival (right panel) of infected K18-hACE2 mice receiving control serum (n=4, black), undiluted (n=4, blue), 1:3 diluted (n=4, red) and 1:9 diluted (n=3, purple) human convalescent plasma at 24 hours prior to challenge of 10 5 PFU SARS-CoV-2. b. Percentage of initial weight (left panel), and survival (right panel) of K18-hACE2 mice receiving control plasma (n=2) or undiluted CP (n=4) at 24 hours after challenge with 10 5 PFU SARS-CoV-2. c. Viral titers of CP-treated mice in the lungs (left panel) and brains (right panel) at 2 and 5 dpi. LOD=limit of detection. d. Scoring for N protein abundance in CP-treated mice in the nasal cavity at 2 (left panel) and 5 (right panel) dpi. 0 -none; 1rare <1%; 2 -multifocal or localized <33% cells; 3 -multifocal, coalescing, 33-66%; 4 -extensive >67%. e. N protein staining in the lungs of control or convalescent plasma treated mice at 2 and 5 days p.i. Bar = 370 µm (top) and 75 µm (bottom).