(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Cell-autonomous requirement for ACE2 across organs in lethal mouse SARS-CoV-2 infection [1]

['Alan T. Tang', 'Department Of Medicine', 'Cardiovascular Institute', 'University Of Pennsylvania', 'Philadelphia', 'Pennsylvania', 'United States Of America', 'David W. Buchholz', 'Department Of Microbiology', 'Immunology']

Date: 2023-02

Angiotensin-converting enzyme 2 (ACE2) is the cell-surface receptor for Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). While its central role in Coronavirus Disease 2019 (COVID-19) pathogenesis is indisputable, there remains significant debate regarding the role of this transmembrane carboxypeptidase in the disease course. These include the role of soluble versus membrane-bound ACE2, as well as ACE2-independent mechanisms that may contribute to viral spread. Testing these roles requires in vivo models. Here, we report humanized ACE2-floxed mice in which hACE2 is expressed from the mouse Ace2 locus in a manner that confers lethal disease and permits cell-specific, Cre-mediated loss of function, and LSL-hACE2 mice in which hACE2 is expressed from the Rosa26 locus enabling cell-specific, Cre-mediated gain of function. Following exposure to SARS-CoV-2, hACE2-floxed mice experienced lethal cachexia, pulmonary infiltrates, intravascular thrombosis and hypoxemia—hallmarks of severe COVID-19. Cre-mediated loss and gain of hACE2 demonstrate that neuronal infection confers lethal cachexia, hypoxemia, and respiratory failure in the absence of lung epithelial infection. In this series of genetic experiments, we demonstrate that ACE2 is absolutely and cell-autonomously required for SARS-CoV-2 infection in the olfactory epithelium, brain, and lung across diverse cell types. Therapies inhibiting or blocking ACE2 at these different sites are likely to be an effective strategy towards preventing severe COVID-19.

Data Availability: All the data and reagents are provided in the manuscript and supporting files, or available commercially, except for the new transgenic mouse lines (hACE2fl, hACE2hypo, and R26-LSL-hACE2). These mouse lines are available through a material transfer agreement with the University of Pennsylvania; interested researchers should contact the Principal Investigator, Mark Kahn, at [email protected] and CC the Office of Research Services at [email protected] . The remaining mouse lines used in this manucript are available at public repositories.

The application of mouse models to investigate COVID-19 pathogenesis has been hindered by the fact that the wild-type SARS-CoV-2 spike protein is unable to bind the mouse ACE2 protein, a necessary first step in viral cellular entry and infection [ 1 , 2 ]. Several hACE2-expressing mouse models have been generated to investigate COVID-19 pathogenesis [ 19 – 22 ], but only the K18-hACE2 line confers severe illness like that observed in patients [ 4 ]. K18-hACE2 random transgenic mice express hACE2 in a nonendogenous fashion primarily restricted to epithelial cells and do not enable genetic dissection of viral pathogenesis. To identify cell-specific roles of ACE2 during SARS-CoV-2 infection, we generated new lines of mice in which hACE2 is expressed from the mouse Ace2 locus in a manner that enables cell-specific loss of function (hACE2-floxed, hACE2 fl ) and Rosa26 mice in which hACE2 expression and gain of function can be conferred in a cell-specific manner (loxP-stop-loxP-hACE2, LSL-hACE2). Our results demonstrate that acute lung injury can occur in the absence of pulmonary infection and identify the olfactory epithelium (OE) and cerebral neurons as critical cellular sites of infection during lethal COVID-19. In all examined sites of infection—over multiple organs and diverse cell types—our genetic experiments illustrate an absolute, cell-autonomous requirement for ACE2 in viral entry.

SARS-CoV-2 is a respiratory virus, and initial infection of epithelium in the nasal cavity and airways is thought to be ACE2-mediated. However, the role of ACE2 in mediating viral spread after initial infection remains untested. Multiple ACE2-independent mechanisms have been proposed including alternative receptors (e.g., DC-SIGN, L-SIGN, CD147, NRP1) as well as receptor-independent phenomena [ 14 – 18 ]. Testing the absolute requirement of ACE2 as infection progresses is critical to understanding of COVID-19 pathogenesis and would inform the design of targeted therapies.

ACE2 is a type 1 transmembrane protein that exists in vivo in two states: a cell-surface protein and a circulating cleaved “soluble” form primarily generated by the ADAM17 (aka TACE) sheddase [ 5 , 6 ]. While a central role for membrane-bound ACE2 is accepted, recent studies have suggested that soluble ACE2 can also mediate viral entry and infection, perhaps explaining the widespread tropism of SARS-CoV-2 in cell types that do not express detectable ACE2 such as endothelium, myocardium, and immune cells [ 7 , 8 ]. Clarifying and testing the in vivo significance of these mechanisms is translationally relevant given proposed COVID-19 therapies utilizing recombinant soluble ACE2 or blocking antibodies to competitively inhibit viral spread [ 9 – 13 ].

Understanding the cellular mechanisms that underlie Coronavirus Disease 2019 (COVID-19) is necessary to determine how to best prevent infection and treat affected individuals [ 1 ]. Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) utilizes the transmembrane carboxypeptidase angiotensin-converting enzyme 2 (ACE2) as a host cell-surface receptor for viral entry. This interaction between virus and ACE2 has been defined by structural biochemical studies, cell culture, and transgenic mouse studies wherein the expression of human ACE2 (hACE2) is sufficient to confer infectability of parental SARS-CoV-2 strains [ 2 – 4 ].

Results

Epithelial cell hACE2 expression is required for pulmonary SARS-CoV-2 infection Following infection with SARS-CoV-2, hACE2fl/y animals exhibited labored breathing, indicative of respiratory distress and COVID-19 lung disease. Prior studies using single-cell RNAseq and immunohistochemical analysis of mouse and human lungs and our immunostaining of hACE2fl mouse lungs identified AT2 cells as ACE2-positive, suggesting that AT2 cell infection might play a key role in COVID-19 pneumonia [30–32,42]. Therefore, we used the SftpcCreERT2 knock-in allele to test whether ACE2-dependent infection of AT2 cells is required for lung infection in hACE2fl animals. Following tamoxifen induction to activate Cre recombination, SftpcCreERT2; hACE2fl/y mice exhibited loss of hACE2 protein in lung AT2 cells (Fig 3A). However, tamoxifen-treated SftpcCreERT2; hACE2fl/y males infected with SARS-CoV-2 virus exhibited levels of viral nucleocapsid protein and viral RNA in the lung indistinguishable from those in hACE2fl/y animals (Fig 3B and 3C). These findings suggested that cell types other than AT2 cells can support SARS-CoV-2 infection of the lung. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. SARS-CoV-2 infects lung alveolar type I cells in an ACE2-dependent manner. (A) Immunodetection of ACE2 using pan-ACE2 antibodies and DC-LAMP in wild-type hACE2fl/y, and SfptcCreERT2/+; hACE2fl/y mouse lungs 6 days after infection with 104 PFU of SARS-CoV-2. Arrowheads indicate DC-LAMP+ AT2 cells. n = 4 for both genotypes, one experiment. (B) Immunohistochemistry of hACE2fl/y and SfptcCreERT2/+; hACE2fl/y mouse lungs 6 days after infection with 104 PFU of SARS-CoV-2 virus was performed using antibodies that recognize SARS-CoV-2 nucleocapsid, DC-LAMP (AT2 cells), and PDPN (AT1 cells) as well as the nuclear stain DAPI. Arrows indicate nucleocapsid colocalized with PDPN. N = 4 for both genotypes, one experiment. (C) qPCR was performed on hACE2fl/y and SfptcCreERT2/+; hACE2fl/y mouse lungs harvested 6 days after infection with 104 PFU of SARS-CoV-2 to measure total viral load. N = 4 for both genotypes, one experiment. (D) Immunohistochemistry of hACE2fl/y and HopxCreERT2/+; hACE2fl/y mouse lungs 2 days after infection with 105 PFU of SARS-CoV-2 virus was performed using antibodies that recognize SARS-CoV-2 nucleocapsid, DC-LAMP, and PDPN as well as the nuclear stain DAPI. Arrows indicate nucleocapsid colocalized with PDPN in AT1 cells. Arrowheads indicate nucleocapsid colocalized with DC-LAMP in AT2 cells. The boxed regions show DC-LAMP staining (red) only in the indicated nucleocapsid positive (green) AT2 cells because the dim DC-LAMP signal is obscured by nucleocapsid signal when both are visible. N = 4–5 for both genotypes, one experiment. (E) qPCR was performed on hACE2fl/y and HopxCreERT2/+; hACE2fl/y mouse lungs harvested 2 days after infection with 105 PFU of SARS-CoV-2 to measure total viral load. N = 4–6 for both genotypes, one experiment. (F) Immunohistochemistry of hACE2fl/y and ShhCre/+; hACE2fl/y mouse lungs 2 and 6 days after infection with 104 or 105 PFU of SARS-CoV-2 virus was performed using antibodies that recognize SARS-CoV-2 nucleocapsid, DC-LAMP (AT2 cells), and PDPN (AT1 cells) as well as the nuclear stain DAPI. N = 6 for all genotypes, two independent experiments. Arrows indicate nucleocapsid colocalized with PDPN. (G) qPCR was performed on hACE2fl/y and ShhCre/+; hACE2fl/y mouse lungs harvested 2 and 6 days after infection with 104 or 105 PFU of SARS-CoV-2 to measure total viral RNA load. Simultaneous measurement using whole blood was performed 6 days after infection (shown in red) to measure circulating levels. (H) Infectious viral load was measured by PFU from hACE2fl/y and ShhCre/+; hACE2fl/y mouse lungs 6 days after infection with 105 PFU of SARS-CoV-2. N > 6 for both genotypes. Note: hACE2fl/y 104 titer data for 3C and 3G are the same. Scale bars in all images 50 μm. ***p < 0.001; *p < 0.05; ns p > 0.05, significance determined by unpaired two-tailed t test. Numerical data in corresponding S1 Metadata tab. ACE2, angiotensin-converting enzyme 2; AT1, alveolar type 1; AT2, alveolar type 2; PDPN, Podoplanin; PFU, plaque forming assay; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2. https://doi.org/10.1371/journal.pbio.3001989.g003 Histologic examination of hACE2fl/y lungs 2 and 6 days after infection with SARS-CoV-2 revealed viral nucleocapsid staining primarily along the cell membrane of Podoplanin (PDPN)-positive alveolar type 1 (AT1) epithelial cells (Fig 3B, 3D, and 3F). To specifically test the role of ACE2 in AT1 cells, we next examined SARS-CoV-2 infection of HopxCreERT2; hACE2fl/y animals in which Cre is active in AT1 but not AT2 cells when Cre activity is activated by tamoxifen administration in mature mice [43,44]. SARS-CoV-2 nucleocapsid was detected in DC-LAMP-expressing AT2 cells but not in PDPN-expressing AT1 cells in the alveoli of infected HopxCre; hACE2fl/y animals (Fig 3D). HopxCreERT2; hACE2fl/y animals infected with SARS-CoV-2 displayed significant but incomplete reductions in viral nucleocapsid protein and viral RNA in the lung (Fig 3D and 3E), consistent with infection of AT2 (Fig 3D) and bronchial epithelial cells (S3 Fig). Finally, to test the role of ACE2 in AT1, AT2, and bronchiolar epithelial cell infection in the lung, we examined SARS-CoV-2 infection of ShhCre; hACE2fl/y animals in which Cre is active in all lower respiratory and gastrointestinal epithelium ([45]; S4 and S5 Figs). SARS-CoV-2 infection was blocked completely in ShhCre; hACE2fl/y lungs, as assessed by immunostaining for viral nucleocapsid (Fig 3F), measurement of lung viral load using both qPCR for viral RNA (Fig 3G) and culture of viral plaque-forming units (PFUs) from freshly harvested lung tissue (Fig 3H). On day 6 postinfection, a low level of virus was detected using qPCR that was equivalent to the level detected in circulating blood (Fig 3G). These findings demonstrate that lung infection by SARS-CoV-2 occurs in both AT1 and AT2 epithelial cells in a cell-autonomous, ACE2-dependent manner.

Lung epithelial infection is not required for lethal COVID-19 in hACE2fl/y animals Studies of human patients with severe COVID-19 pneumonia have revealed evidence of both lung infection by SARS-CoV-2 virus [32,42,46,47] and acute respiratory distress syndrome (ARDS), manifest by noncardiogenic pulmonary edema, severe hypoxemia, and a systemic inflammatory response [48,49]. The extent to which these two mechanisms of lung disease are interconnected, and their respective roles in COVID-19 mortality, have been difficult to define clinically in humans. Following exposure to 104 or 105 PFU per mouse of SARS-CoV-2, both hACE2fl/y and ShhCre; hACE2fl/y animals exhibited weight loss and lethality (Figs 4A, 4B, S6A and S6B). There was a slight delay in the onset of symptoms in ShhCre; hACE2fl/y animals, but both hACE2fl/y and ShhCre; hACE2fl/y animals were severely hypoxemic 5–6 days after infection, with oxygen (O 2 ) saturations (SpO2) ranging between 70% and 85%, levels like those of severely ill patients requiring mechanical ventilation (Fig 4C). Hematoxylin–eosin (HE) staining of lung sections from wild-type, hACE2fl/y, and ShhCre; hACE2fl/y animals 6 days after exposure to SARS-CoV-2 revealed alveolar consolidation, interstitial thickening, and the presence of focal infiltrates and hemorrhage in both hACE2fl/y and ShhCre; hACE2fl/y animals that were not observed in wild-type controls (Fig 4D). Virtually identical results were obtained following exposure to 105 PFU of SARS-CoV-2 (S6 Fig). SARS-CoV-2-infected hACE2fl/y and ShhCre; hACE2fl/y animals also exhibited thrombus-filled pulmonary vessels that were not observed in the lungs of wild-type animals exposed to virus (Figs 4D, 4E, and S6D), a prominent finding also observed in human lungs harvested from individuals with lethal COVID-19 disease [48]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. hACE2 mice develop acute lung injury and hypoxemia in the absence of SARS-CoV-2 lung infection. (A, B) Weight loss and survival of hACE2fl/y and ShhCre/+; hACE2fl/y mice after infection with 104 PFU of SARS-CoV-2. N = 11 (hACE2fl/y) and 4 (ShhCre/+; hACE2fl/y), two independent experiments. Note: Data for hACE2fl/y same as Figs 2H and S2 since ShhCre/+; hACE2fl/y animals were littermates. (C) Pulse oximetry measured in WT, hACE2fl/y, and ShhCre/+; hACE2fl/y mice 6 days after exposure to 104 PFU of SARS-CoV-2 virus. (D) HE staining of WT, hACE2fl/y, and ShhCre/+; hACE2fl/y lung tissue 6 days after exposure to 104 PFU of SARS-CoV-2 virus. Boxed regions at higher magnification in images below. Arrows, sites of focal consolidation. Asterisks, intravascular thrombi. Hashtag, acute emphysematous changes. Representative of N = 4 animals per genotype. (E, F) Immunohistochemistry of WT, hACE2fl/y, and ShhCre/+; hACE2fl/y lung tissue 6 days after exposure to 104 PFU of SARS-CoV-2 virus using antibodies against ICAM-1 and PDPN, or vWF and Endomucin (endothelial cells). Arrowheads in F identify vWF-positive microvasculature of the lung in hACE2fl/y and ShhCre/+; hACE2fl/y animals. Representative of N = 4 animals per genotype. Scale bars in all images, 50 μm. Note: Images in each panel were taken at lower or higher magnification from the same tissue section respective to genotype. *p < 0.05, **p < 0.001; ****p < 0.0001 by unpaired two-tailed t test, one-way ANOVA with Holm–Sidak correction for multiple comparisons, or log-rank Mantel Cox test. Numerical data in corresponding S1 Metadata tab. HE, hematoxylin–eosin; ICAM-1, intracellular adhesion marker 1; PDPN, Podoplanin; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2; vWF, von Willebrand’s Factor; WT, wild-type. https://doi.org/10.1371/journal.pbio.3001989.g004 Consistent with a global lung inflammatory state, we observed uniformly elevated expression of the inflammation-induced proteins ICAM1 and PDPN (aka RTI40) in the alveolar epithelial cells of infected hACE2fl/y and ShhCre; hACE2fl/y animals compared with SARS-CoV-2 inoculated wild-type controls (Fig 4E) [50]. Expression of the pro-coagulant, inflammation-induced protein von Willebrand’s Factor (vWF) was also up-regulated in the lung capillary endothelial cells of SARS-CoV-2-infected hACE2fl/y and ShhCre; hACE2fl/y animals compared with wild-type controls (Fig 4F), consistent with the presence of intravascular thrombi in those lungs. These studies and those shown in Fig 3 demonstrate that acute lung injury and hypoxemic respiratory failure may arise in the absence of primary SARS-CoV-2 lung infection and implicate an inflammatory state and ARDS due to extrapulmonary infection as an important mechanism of respiratory failure and lethality in hACE2fl/y mice.

Lethal COVID-19 in hACE2fl/y mice is associated with SARS-CoV-2 infection of the olfactory epithelium, olfactory bulb, and cerebrum The above results suggested that SARS-CoV-2 infection at cellular sites outside the lung that lack ShhCre activity were sufficient to confer severe disease in hACE2fl/y animals. To define the extrapulmonary sites of infection in ShhCre; hACE2fl/y mice that might be responsible for pulmonary inflammation and respiratory failure, we performed lineage tracing of the ShhCre transgene along the path of inhaled virus using the R26-LSL-Red Fluorescent Protein (LSL-RFP, aka Ai14) Cre reporter allele. ShhCre activity was detected uniformly in epithelial cells lining the trachea, lung airways, and alveoli, as well as in the transition zone epithelium that lies between the nasal cavity RE and OE (S4 Fig). In contrast, the RE and OE of the nasal passages failed to display ShhCre activity (S4B Fig), suggesting that these sites of high ACE2 expression may be responsible for the lethal response to SARS-CoV-2 infection in hACE2fl/y mice. Therefore, we compared SARS-CoV-2 nucleocapsid staining in the RE, OE, and adjacent olfactory bulb (OB) of the brain during infection in wild-type, hACE2fl/y, and ShhCre; hACE2fl/y animals. Two days after infection, abundant SARS-CoV-2 nucleocapsid was observed in the RE and OE of hACE2fl/y and ShhCre; hACE2fl/y animals, but none was detected in the adjacent OB and cerebral cortex of the brain (Fig 5A and 5C). In contrast, 5 to 6 days after infection, relatively little SARS-CoV-2 nucleocapsid was detected in the RE and OE of hACE2fl/y and ShhCre; hACE2fl/y animals, although abundant signal was detected in the neighboring OB, cerebral cortex, and hippocampus (Figs 5B, 5C, and S7). Costaining with the nuclear neuronal marker NeuN and the glial cell marker GFAP revealed SARS-CoV-2 nucleocapsid specifically in neurons and not associated glial cells and demonstrated the presence of a reactive gliosis 5 days postinfection (Fig 5D). Sparse neuronal cell SARS-CoV-2 nucleocapsid and a reactive gliosis was also detected in the brainstem, a site more distant from the OE, but not in the cerebellum, 5 days after infection (S7 Fig). Consistent with the immunostaining studies to detect virus described above, in situ hybridization of the nasal cavity and adjacent tissues using RNAscope probes for SARS-CoV-2 RNA revealed little or no virus in the OE 5 days after infection with abundant viral RNA detected in the adjacent OB and cerebrum of the brain (Fig 5E and 5F). Unlike the OE, SARS-CoV-2 nucleocapsid was not detected in the choroid plexus or adjacent neurons at an early time point, 2 days after infection (S8A Fig). However, patches of neuronal SARS-CoV-2 nucleocapsid staining were observed 5 days after viral infection in proximity to the choroid plexus and meninges in both hACE2fl/y and ShhCre; hACE2fl/y animals (S8B Fig). HE staining of the choroid plexus along the lateral ventricle and cerebral cortex vasculature revealed immune cell infiltrates at 5 (but not 2) days after SARS-CoV-2 infection (S8C and S8D Fig), indicative of an inflammatory response like that described in humans [51,52].These findings are consistent with prior studies demonstrating that olfactory epithelial cells may rapidly clear the virus after infection with SARS-CoV-2 in hamsters [53] and in humans [28] and demonstrate that early infection of the OE is followed by later infection of neurons in the brain of hACE2fl/y mice. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. SARS-CoV-2 infection of the OE and brain in hACE2fl mice. (A, B) Immunohistochemistry of SARS-CoV-2 nucleocapsid and epithelial cell E-cadherin in the RE, OE, and OB 2 and 5–6 days after infection of hACE2fl/y and ShhCre/+; hACE2fl/y mice. Arrowheads indicate sites of viral nucleocapsid detection. Representative of N = 4 animals per genotype and time point. Scale bars 100 μm. (C, D) Immunohistochemistry of SARS-CoV-2 nucleocapsid, neuronal NeuN, and glial cell GFAP in the cerebral cortex (Co) 2 and 5–6 days after infection. Arrowheads indicate sites of GFAP+ reactive gliosis. Arrows indicate nucleocapsid colocalization with NeuN staining. Representative of N = 4 animals per genotype and time point. Scale bars 100 μm top, 50 μm bottom. (E) Diagram of the mouse nasal cavity and cranial anatomy. (F) In situ hybridization detection of SARS-CoV-2 mRNA 5 days postinfection reveals virus in the OB and cerebral cortex of the brain, but not the OE of the nose. Scale bar 250 μm. Note: Images in each panel were taken at lower or higher magnification from the same tissue section respective to genotype and highlight different anatomical regions. OB, olfactory bulb of the brain; OE, olfactory epithelium; RE, respiratory epithelium; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2. https://doi.org/10.1371/journal.pbio.3001989.g005

Pharmacologic ablation of the OE or genetic loss of hACE2 in the OE and neurons prevents lethal COVID-19 in hACE2fl/y mice The studies described above indicated that SARS-CoV-2 infection of the OE rather than the lung is associated with weight loss and lethality. Furthermore, it afforded the opportunity to stringently test the requirement for ACE2 during initial infection and subsequent viral spread in vivo. Since genetic tools to drive Cre expression selectively in the OE are not available, we tested the requirement for OE infection using pharmacologic ablation of olfactory epithelial cells with methimazole (MMZ), a compound that is highly and specifically toxic for the OE in rodents [54]. HE staining of the OE 24 hours after treatment of hACE2fl/y mice with MMZ revealed loss of almost all olfactory epithelial cells (Fig 6A). In contrast, the underlying OB and lung did not show evidence of cell loss, cell damage, or inflammation (Fig 6A). We next tested whether ablation of the OE with MMZ impacts the clinical course of hACE2fl/y mice infected with SARS-CoV-2 virus. Remarkably, pretreatment with MMZ 24 hours prior to intranasal inoculation of SARS-CoV-2 prevented both weight loss and death in the majority of hACE2fl/y mice (Fig 6B and 6C). Immunostaining of the brain for SARS-CoV-2 nucleocapsid protein 6 days after infection revealed no detectable virus in the cerebral cortex of MMZ-treated hACE2fl/y mice compared with vehicle-treated hACE2fl/y mice (Fig 6D). In contrast, similar expression of SARS-CoV-2 nucleocapsid was detected in the lungs of vehicle-treated and MMZ-treated hACE2fl/y mice (Fig 6E), a finding consistent with the lack of lethality associated with isolated lung infection in ShhCre;LSL-hACE2+/0 mice (Fig 8, described below). Analysis of MMZ-treated hACE2fl/y mice 14 days after SARS-CoV-2 infection revealed no viral nucleocapsid in either the brain or lung (Fig 6F and 6G), consistent with sparing of the brain and resolution of lung infection in surviving MMZ-treated hACE2fl/y mice. Consistent with brain infection being a driver of lethality, analysis of a single hACE2fl/y mouse that became ill and was euthanized 8 days after infection revealed strong staining for nucleocapsid in the brain and lung (S9 Fig). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. MMZ ablation of the OE or genetic loss of hACE2 in OE and neurons prevents brain infection and lethal SARS-CoV-2 infection. (A) HE staining of the indicated mouse tissues was performed 24 hours after intraperitoneal injection of 100 mg/kg MMZ or vehicle control. Scale bars 100 μm for the OE, OB, and 50 μm for the lung. Representative of N = 3 per condition. (B, C) Weight loss and survival of hACE2fl/y mice treated with MMZ or vehicle prior to infection with 105 PFU of SARS-CoV-2 virus. Asterisks indicate significant differences in weight between vehicle and MMZ treated hACE2fl/y animals. N = 14 (Vehicle) and 15 (MMZ). (D, E) Immunohistochemistry of WT, vehicle-treated hACE2fl/y, and MMZ-treated hACE2fl/y mouse cerebral cortex and lung 5–6 days after infection with 105 PFU of SARS-CoV-2 virus using antibodies that recognize viral nucleocapsid, the neuronal marker NeuN, the glial cell marker GFAP, the alveolar type II cell marker DC-LAMP, the AT1 cell marker PDPN, and the nuclear stain DAPI. Arrows indicate nucleocapsid staining colocalized with NeuN+ neurons (brain) or PDPN+ AT1 cells (lung). Representative of N = 4 per condition. (F, G) Immunohistochemistry of WT and MMZ-treated hACE2fl/y mouse cerebral cortex and lung 14 days after infection with 105 PFU of SARS-CoV-2 virus was performed as described in D and E. Note: Day 5 hACE2fl/y samples (middle panel) were included on the same tissue slide as a positive control. Arrows indicate nucleocapsid staining colocalized with neurons and AT1 cells in vehicle-treated hACE2fl/y mice. Representative of N = 4 per condition. Scale bars D-G 50 μm. (H) The Foxg1Cre/+ allele drives Cre expression in the OE and neurons of the brain (shown in red). (I, J) Weight loss and survival of hACE2fl/y and Foxg1Cre/+; hACE2fl/y mice after infection with 105 viral titer per mouse. n = 4 and 3 mice, respectively. (K) Pulse oximetry measured in WT, hACE2fl/y, and Foxg1Cre/+; hACE2fl/y mice 5–6 and 12 days after exposure to SARS-CoV-2 virus, respectively. *p < 0.05, **p < 0.001; ****p < 0.0001 by unpaired two-tailed t test, one-way ANOVA with Holm–Sidak correction for multiple comparisons, or log-rank Mantel Cox test. Numerical data in corresponding S1 Metadata tab. hACE2, human ACE2; HE, hematoxylin–eosin; MMZ, methimazole; OB, olfactory bulb; OE, olfactory epithelium; PDPN, Podoplanin; PFU, plaque-forming unit; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2; WT, wild-type. https://doi.org/10.1371/journal.pbio.3001989.g006 To further test the requirement for SARS-CoV-2 infection of the OE and brain for lethal COVID-19 in hACE2fl/y mice, we generated Foxg1Cre/+; hACE2fl/y mice in which Cre recombinase is expressed in the RE, OE, and neurons of the forebrain, but not in cells of the lung [55] (Figs 6H and S10), and K14-Cre; hACE2fl/y mice in which Cre recombinase is expressed in RE and transition zone epithelial cells of the nasal cavity (Keratin-14; S11A Fig). Following exposure to SARS-CoV-2, both hACE2fl/y and K14-Cre; hACE2fl/y mice exhibited weight loss and lethality associated with severe hypoxemia (S11B–S11D Fig), although K14-Cre; hACE2fl/y mice demonstrated absence of SARS-CoV-2 nucleocapsid staining in the nasal RE, consistent with a lack of infection at that site (S11E Fig). In contrast, Foxg1Cre/+; hACE2fl/y mice survived and did not experience weight loss or hypoxemia (Fig 6I–6K). Consistent with these findings, Foxg1Cre/+; hACE2fl/y mice exhibited evidence of SARS-CoV-2 infection in the lung but not the OE or RE at 2 days postinfection, and no evidence of brain infection at 6 days postinfection (S12 Fig). These pharmacologic and genetic findings are highly concordant and identify SARS-CoV-2 infection of the OE and brain as required for lethal COVID-19 in hACE2fl/y mice.

Neuronal SARS-CoV-2 infection is required for lethal respiratory failure in hACE2fl mice The genetic and pharmacologic studies described above suggested that SARS-CoV-2 infection of OE and/or neurons was required for lethal respiratory disease in hACE2fl animals. To further identify the infected cell type responsible for lethal disease, we tested whether hACE2 was required selectively in neurons for SARS-CoV-2 infection using the Baf53b (aka Actl6b)-Cre transgenic line that expresses Cre specifically in neurons but not in olfactory or other epithelial cells [56]. Importantly, lineage tracing studies demonstrated highly specific activity of the Baf53b-Cre allele in neurons, including olfactory sensory neurons (OSNs), but not in sustentacular cells or respiratory epithelial cells (S13 Fig). In contrast to hACE2fl/y littermates, following exposure to SARS-CoV-2, Baf53b-Cre; hACE2fl/y mice exhibited no signs of distress, did not lose weight or become hypoxemic, and had normal survival (Fig 7A–7C). Immunoblot analysis of brain lysates demonstrated partial loss of hACE2 in Baf53b-Cre; hACE2fl/y animals compared with hACE2fl/y littermate controls (Fig 7D) confirming neuronal expression of hACE2 in hACE2fl/y mice, while also suggesting the existence of nonneuronal sources of brain hACE2. Immunostaining for SARS-CoV-2 nucleocapsid revealed evidence of viral infection in the OE, cerebral cortex, and lung in hACE2fl/y animals (Fig 7E–7G). However, in Baf53b-Cre; hACE2fl/y mice, SARS-CoV-2 infection was detected in both the OE and lung but not the brain (Fig 7E–7G). These genetic findings demonstrate that neuronal infection is cell-autonomously ACE2-dependent and required for lethality after SARS-CoV-2 infection in hACE2fl/y mice. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. Neuronal infection is required for lethal SARS-CoV-2 infection in hACE2fl mice. (A, B) Weight loss and survival of hACE2fl/y and Baf53b-Cre; hACE2fl/y mice after infection with 104 PFU of SARS-CoV-2. N = 10 (hACE2fl/y) and 12 (Baf53b-Cre; hACE2fl/y), three independent experiments. (C) Pulse oximetry measured in WT, hACE2fl/y, and Baf53b-Cre; hACE2fl/y mice 6 days after exposure to 104 PFU of SARS-CoV-2 virus. (D) Immunoblotting of whole brain lysates from hACE2fl/y and Baf53b-Cre; hACE2fl/y and WT mice was performed using anti-panACE2, anti-hACE2, and anti-β-actin antibodies. Each lane represents a single animal. n = 4, two independent experiments. (E) Immunohistochemistry of SARS-CoV-2 nucleocapsid, epithelial cell Krt8, and OSN OMP in the OE of WT, hACE2fl/y, and Baf53b-Cre; hACE2fl/y mice 6 DPI. Representative of n = 4 animals per genotype. (F) Immunohistochemistry of SARS-CoV-2 nucleocapsid, AT1 cell PDPN, and AT2 cell DC-LAMP in the lungs of WT, hACE2fl/y, and Baf53b-Cre; hACE2fl/y mice 6 DPI. (G) Immunohistochemistry of SARS-CoV-2 nucleocapsid, neuronal NeuN, and glial cell GFAP in the cerebral cortex at low and high magnification of the same tissue section of WT, hACE2fl/y, and Baf53b-Cre; hACE2fl/y mice 6 DPI. Representative of N = 5–6 animals per genotype and time point. Scale bars in all images 50 μm. ****p < 0.0001 by unpaired two-tailed t test, one-way ANOVA with Holm–Sidak correction for multiple comparisons, or log-rank Mantel Cox test. Numerical data in corresponding S1 Metadata tab. AT1, alveolar type 1; AT2, alveolar type 2; DPI, days postinfection; OE, olfactory epithelium; OMP, olfactory marker protein; OSN, olfactory sensory neuron; PDPN, Podoplanin; PFU, plaque-forming unit; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2; WT, wild-type. https://doi.org/10.1371/journal.pbio.3001989.g007 We previously observed that infection with the Omicron BA.1 SARS-CoV-2 variant failed to confer lethal disease in of hACE2fl/y mice, a marked contrast to the universal lethality observed following infection with the USA-WA1 SARS-CoV-2 variant (Fig 2J and 2K). The findings described above suggested that an explanation for this difference in lethality might be a difference in the ability of the two variants to confer neuronal infection. To test this hypothesis, hACE2fl/y mice were infected with 105 PFU of Omicron BA.1 SARS-CoV-2. Omicron BA.1 SARS-CoV-2 nucleocapsid was detected in the OE 6 days postinfection at a level similar to that observed following infection with the USA-WA1 SARS-CoV-2 variant (S14A Fig). Exposure to Omicron BA.1 SARS-CoV-2 resulted in ACE2-dependent lung infection that was evident at both 2 and 6 days postinfection (DPI), persistent pulmonary immune cell infiltrates, but was not associated with hypoxemia (S14B, S14C, and S14E-S14G Fig). However, in contrast to hACE2fl/y mice infected with the USA-WA1 SARS-CoV-2 variant, SARS-CoV-2 nucleocapsid was not detected in the brain 6 DPI with the Omicron BA.1 variant (S14D Fig). These findings are highly concordant with those described above and support the conclusion that neuronal infection is required for lethal COVID-19 in hACE2fl/y mice.

SARS-CoV-2 infection of the lung is sufficient to confer pneumonia and transient hypoxemia but is not lethal in hACE2fl/y mice The above studies demonstrated that ACE2-dependent SARS-CoV-2 infection is required in the brain but not the lung for lethal respiratory failure in hACE2fl/y mice. However, studies of human COVID-19 have documented significant SARS-CoV-2 infection of the lung [32,42,46,47], raising the possibility that lethal respiratory failure may also arise through direct lung infection. Moreover, hACE2fl/y mice express hACE2 at higher levels in the brain compared to lung, which may mask the significance of primary SARS-CoV-2 lung infection. To address this possibility and rigorously test the sufficiency of lung epithelial infection for severe disease, we generated a gain-of-function model in which Cre recombinase drives cell-specific expression of hACE2 from the Rosa26 locus (“LSL-hACE2 mice”; Fig 8A). LSL-hACE2 mice were crossed to ShhCre animals to generate ShhCre;LSL-hACE2+/0 mice in which hACE2 is expressed exclusively in the epithelial cells of the lung, upper respiratory tract, and gut. Immunoblotting demonstrated strong expression of hACE2 in the lysate of lungs harvested from ShhCre;LSL-hACE2+/0 but not littermate LSL-hACE2+/0 or hACE2fl/y mice (Fig 8B). Immunostaining using anti-hACE2 or anti-pan-ACE2 antibodies revealed robust expression of hACE2 in the pulmonary epithelium of ShhCre;LSL-hACE2+/0 animals but not LSL-hACE2+/0 littermates (Fig 8C), consistent with Cre-dependent hACE2 expression. Exposure to SARS-CoV-2 resulted in lung infection of ShhCre;LSL-hACE2+/0 mice, with greater nucleocapsid staining 2 DPI than 6 DPI (Fig 8D and 8E). In contrast, SARS-CoV-2 infection was not detected in the brain of ShhCre;LSL-hACE2+/0 mice (Fig 8F). HE staining of the lungs of SARS-CoV-2-infected ShhCre;LSL-hACE2+/0 mice revealed the presence of bronchovascular inflammatory infiltrates and acute emphysematous changes at 2 DPI (Fig 8G). By 6 DPI, the presence of alveolar infiltrates and hyalinosis was evident (Fig 8H). This isolated SARS-CoV-2 pneumonia was associated with transient hypoxemia 10 DPI that was accompanied by a small and statistically insignificant drop in weight (Fig 8I and 8J). However, in contrast to infection of hACE2fl/y mice, infection of ShhCre;LSL-hACE2+/0 animals did not result in prolonged weight loss or death (Fig 8J and 8K). These findings indicate that isolated SARS-CoV-2 lung infection may confer pneumonia and transient hypoxemia but is not sufficient to cause death in otherwise healthy hACE2fl/y mice, a finding remarkably consistent with recent studies of K18-hACE2 animals demonstrating that infection primarily in the lung is not associated with lethality [57]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 8. Selective hACE2 expression in lung epithelial cells confers nonlethal pneumonia after SARS-CoV-2 infection. (A) Generation of LSL-hACE2 mice using gene targeting of the mouse Rosa26 locus. WPRE, SV40, PGK-NeoR. (B) Immunoblotting of whole lung lysates from WT, LSL-hACE2+/0, and ShhCre/+;LSL-hACE2+/0 mice was performed using anti-hACE2 and anti-β-actin antibodies. Each lane represents a single animal. Representative of n = 3 per genotype and two independent experiments. (C) Immunohistochemistry of lung from LSL-hACE2+/0 and ShhCre/+;LSL-hACE2+/0 mice was performed using anti-panACE2 and anti-hACE2 antibodies. Representative of N = 3 per genotype and two independent experiments. (D) Immunohistochemistry of lung from LSL-hACE2+/0 and ShhCre/+;LSL-hACE2+/0 mice was performed using antibodies to detect SARS-CoV-2 nucleocapsid, AT1 cell PDPN, and AT2 cell DC-LAMP 2 days after infection with 104 PFU of SARS-CoV-2 virus. Representative of n = 4 per genotype. (E) Immunohistochemistry of lung from LSL-hACE2+/0 and ShhCre/+;LSL-hACE2+/0 mice was performed using antibodies to detect SARS-CoV-2 nucleocapsid, AT1 cell PDPN, and AT2 cell DC-LAMP 6 days after infection with 104 PFU of SARS-CoV-2 virus. Representative of n = 4 per genotype. (F) Immunohistochemistry of cerebral cortex from LSL-hACE2+/0 and ShhCre/+;LSL-hACE2+/0 mice was performed using antibodies to detect SARS-CoV-2 nucleocapsid, neuronal NeuN, and glial cell GFAP 6 days after infection with 104 PFU of SARS-CoV-2 virus. Representative of n = 4 per genotype. (G) HE staining of LSL-hACE2+/0 and ShhCre/+;LSL-hACE2+/0 lung tissue 2 days after exposure to 104 PFU of SARS-CoV-2 virus. Arrows, sites of inflammatory cell infiltrate. Hashtag, acute emphysematous changes. Representative of n = 4 animals per genotype. (H) HE staining of LSL-hACE2+/0 and ShhCre/+;LSL-hACE2+/0 lung tissue 6 days after exposure to 104 PFU of SARS-CoV-2 virus. Arrows, sites of alveolar inflammatory infiltrate and hyalinosis. Representative of n = 4 animals per genotype. (I) Pulse oximetry measured in LSL-hACE2+/0 and ShhCre/+;LSL-hACE2+/0 mice 6, 10, and 14 days after exposure to 105 PFU of SARS-CoV-2 virus. (J, K) Weight loss and survival of LSL-hACE2+/0 and ShhCre/+;LSL-hACE2+/0 mice after infection with 105 PFU of SARS-CoV-2. N = 6 (LSL-hACE2+/0) and 7 (ShhCre/+;LSL-hACE2+/0 mice), two independent experiments. Scale bars in all images 50 μm. Note: Images in each panel were taken at lower and/or higher magnification from the same tissue section respective to genotype and highlight different pathology. ns, not significant, p > 0.05 **, p < 0.01 by unpaired two-tailed t test, one-way ANOVA with Holm–Sidak correction for multiple comparisons, or log-rank Mantel Cox test. Numerical data in corresponding S1 Metadata tab. AT1, alveolar type 1; AT2, alveolar type 2; hACE2, human ACE2; HE, hematoxylin–eosin; PDPN, Podoplanin; PFU, plaque-forming unit; PGK-NeoR, Phosphoglycerate kinase promoter-driven Neomycin resistance cassette; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2; SV40, simian virus 40; WPRE, woodchuck hepatitis virus posttranscriptional regulatory element; WT, wild-type. https://doi.org/10.1371/journal.pbio.3001989.g008

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001989

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.